text
stringlengths
34
10.9k
page_idx
int64
9
2.25k
document_name
stringclasses
1 value
file_path
stringclasses
1 value
file_url
stringclasses
1 value
loader_name
stringclasses
1 value
Skin And i TS APPE ndAgES 150SECT iOn 1 are long, unbranched, tubular structures, each with a highly coiled, secretory portion up to 0.4 mm in diameter, situated deep in the dermis or hypodermis. From there, a narrower, straight or slightly helical ductal portion emerges (see Fig. 7.1). The walls of the duct fuse with the base of epidermal rete pegs and the lumen passes between the keratinocytes, often in a tight spiral, particularly in thick hairless skin (Fig. 7.17), and opens via a rounded aperture (pore) on to the skin surface (see Fig. 7.7). In thick hairless skin, sweat glands discharge along the centres of friction ridges, incidentally providing fingerprint patterns for forensic analysis. Sweat glands have an important ther - moregulatory function, and their secretion enhances grip and sensitiv - ity of the palms and soles. Eccrine sweat glands are absent from the tympanic membrane, margins of the lips, nail bed, nipple, inner preputial surface, labia minora, penis and clitoris, where apocrine glands are located. They are numerous elsewhere, their frequency ranging from 80 to over 600/cm 2, depending on position and genetic variation. The total number is esti-mated to be between 1.6 and 4.5 million, and is greatest on the plantar skin of the feet. There are many sweat glands on the face and flexor aspects of the hands, and least on the surfaces of the limbs. People indigenous to warmer climates tend to have more sweat glands than those indigenous to cooler regions.the hair follicle and then to the surface of the hair and the general epidermis. While the full range of functions in the skin remains to be clarified, sebum is a key component of the epidermal barrier, as well as having a key role in the skin’s immune system. Sebum also contributes to a characteristic body odour. When first formed, sebum is a complex mixture of over 50% di- and triglycerides, with smaller proportions of wax esters, squalene, cholesterol esters, cholesterol and free fatty acids. At birth, sebaceous glands are quite large, regressing later until stimu - lated again at puberty. At that time, sebaceous gland growth and secre - tory activity increase greatly in both males and females, under the influence of androgens (testicular and suprarenal), which act directly on the gland. Excessive amounts of sebum may become impacted within the duct and this, associated with hyperkeratinization, may lead to blockage and formation of a comedone. This may become infected and inflamed, and is the primary lesion of acne. Acne is the most common disease caused by dysfunction of the sebocyte. Oestrogens have an effect opposite to that of androgens, and sebum secretion is considerably lower in women, becoming greatly decreased after the age of 50 years. Apocrine glands Apocrine glands are particularly large glands found in the dermis or subcutaneous fat. They are classed as a type of sweat gland. In the adult, they are present in the axillae, perianal region, areolae, periumbilical skin, prepuce, scrotum, mons pubis and labia minora. An apocrine gland consists of a basal secretory coil and a straight duct that opens into either the pilary canal above the duct of the seba - ceous gland, or directly on to the skin surface if there is no associated hair. The secretory region may be as much as 2 mm wide and its coils often anastomose to form a labyrinthine network. Each coil is lined by cuboidal secretory cells whose apical cytoplasm projects into the lumen and basally is in contact with a layer of myoepithelial cells within a thick basal lamina. The secretory cells contain vacuoles, vesicles and dense granules of varying size and internal structure; the numbers and character vary with the cycle of synthesis and discharge. The mechanism of secretion is still not clear but may involve merocrine secretion of granules, apocrine secretion or complete holocrine disintegration of the cells (see Fig. 2.6). Apocrine activity is minimal before puberty, after which it is androgen-dependent and responsive to emotional stimuli. It is control - led by adrenergic nerves and is sensitive to adrenaline (epinephrine) and noradrenaline (norepinephrine). The secretion is initially sterile and odourless, but it undergoes bacterial decomposition to generate odorous and musky compounds, including short-chain fatty acids, and steroids such as 5 α-androstenone. In many animals these are potent pheromonal signals but their role in humans is less certain. Obstruction of apocrine sweat ducts and associated upper hair follicles in the axillae, breast areolae and pubic region, mainly in women, is thought to under - lie Fox–Fordyce disease. Chronic inflammation involving the skin bearing apocrine glands leads to a painful, occasionally debilitating condition called hidradenitis suppurativa. Arrector pili muscles The arrector pili muscles are small bundles of closely packed smooth muscle cells that form oblique links between the dermal sheaths of hair follicles and the papillary layer of the dermis (see Figs 7.1, 7.12). They show the typical features of smooth muscle cells and are separated by narrow spaces containing collagen fibres and unmyelinated noradren - ergic sympathetic axons. The muscles are attached to the bulge region of the follicles by elastin fibrils, and are directed obliquely and towards the side to which the hair slopes superficially. The sebaceous gland occupies the angle between the muscle and the hair follicle, and muscle contraction helps to expel the gland contents. Contraction tends to pull the hair into a more vertical position and to elevate the epidermis surrounding it into a small hillock (goose bump), dimpling the skin surface where the muscle is inserted superficially. Arrector pili muscles are absent from facial, axillary and pubic hairs, eyelashes and eyebrows, and from the hairs around the nostrils and the external auditory meati. SWEAT GLANDS The vast majority of sweat glands ( Fig. 7.16) are eccrine, although their mode of secretion includes typical merocrine mechanisms. The glands Fig. 7.16 The coiled secretory portion of a sweat gland (SG) in the reticular dermis. The deeper-stained profiles (above) are the origins of the duct. An autonomic nerve fibre (NF) and accompanying arteriole (A) and venule (V) are seen below. SG NF V ASG NF V A Fig. 7.17 A sweat duct in a trichrome-stained section of thick skin. The duct (SD) spirals through the dermis and epidermis, and is visible most clearly in the cornified superficial layer. SD
212
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Microstructure of skin and skin appendages 151 CHAPTER 7 bubbles may produce white flecks. These defects move distally with the growth of the nail plate. The nail plate arises from compacted cornified epithelial cells derived from the dorsal, intermediate and ventral nail matrices. It is densely adherent to the matrices on its undersurface but becomes a free struc - ture distal to the onychodermal band, where it separates from the nail bed. The dorsal aspect of the nail plate originates from the more proxi - mal regions of the germinal matrix, i.e. dorsal and intermediate matri - ces, whereas the deeper, volar aspect of the plate originates from the ventral matrix. Nail folds The sides of the nail plate are bordered by the lateral nail folds, which are continuous with the proximal fold (see Fig. 7.18). They are bounded by the attachment of the skin to the lateral aspect of the distal phalanx margin and the lateral nail. The proximal nail fold provides the visible proximal border to the nail apparatus. It consists of a superficial and a deep epidermal layer. These two layers are separated by a core of dermis. The epidermis of the superficial layer lacks hair follicles and epidermal ridges, and its cornified distal margin extends over the nail plate as the cuticle or eponychium. The deep layer merges with the nail matrix.Microscopically, the secretory coil consists of a pseudostratified epi - thelium enclosing a lumen. Three types of cell have been described: clear cells, from which most of the secretion is derived; dark cells, which share the same lumen; and myoepithelial cells. Clear cells are approxi - mately pyramidal in shape, and their bases rest on the basal lamina or are in contact with the myoepithelial cells. Their apical plasma mem - branes line lateral intercellular canaliculi, which connect with the main lumen. The basolateral plasma membranes are highly folded, interdigi - tating with apposed clear cells, and they have the basal membrane infoldings typical of cells involved in transcytotic fluid and ion trans - port from the interstitial fluid to the glandular lumen. Their cytoplasm contains glycogen granules, mitochondria, rough endoplasmic reticu - lum and a small Golgi complex, but few other organelles. The nucleus is round and moderately euchromatic. Dark cells are pyramidal and lie closer to the lumen, such that their broad ends form its lining. Their cytoplasm contains a well-developed Golgi complex, numerous vacu - oles and vesicles, and dense glycoprotein granules, which they secrete by a typical merocrine mechanism. Myoepithelial cells (see Figs 2.3, 2.4) resemble those associated with secretory acini of the salivary glands and breast, and contain abundant myofilaments. The intradermal sweat duct is composed of an outer basal layer and an inner layer of luminal cells connected by numerous desmosomes. The intraepidermal sweat duct (acrosyringium) is coiled and consists of two layers of cells, which, developmentally, are different from the sur - rounding keratinocytes and can be distinguished from them by the presence of keratin K19. The outer cells near the surface contain kerato - hyalin granules and lamellar granules, and undergo typical cornifica - tion. The inner cells, from a mid-epidermal level, contain numerous vesicles, undergo an incomplete form of cornification, and are largely shed into the lumen at the level of the cornified epidermal layer. Sweat is a clear, odourless fluid, hypotonic to tissue fluid, and con - tains mainly sodium and chloride ions, but also potassium, bicarbo - nate, calcium, urea, lactate, amino acids, immunoglobulins and other proteins. Excessive sweating can lead to salt depletion. Heavy metals and various organic compounds are eliminated in sweat. When first secreted, the fluid is similar in composition to interstitial fluid. It is modified as it passes along the duct by the action mainly of the basal cells, which resorb sodium, chloride and some water. The hormone aldosterone enhances this activity. The sweat glands are capable of producing up to 10 litres of sweat per day, in response to thermal, emotional and taste stimuli, mediated by unmyelinated sympathetic cholinergic fibres. The glands also respond to adrenaline (epinephrine). Thermoregulation is coordinated via a heat centre in the preoptic region, in and near the rostral hypothalamus, which reacts to changes in blood temperature and afferent stimuli from the skin by initiating appropriate responses such as controlling cutaneous blood supply, the rate and volume of sweat secretion (for evaporation at the surface) and shivering. NAIL APPARATUS Nails (Fig. 7.18) are homologous with the cornified layer of the general epidermis. They consist of compacted, anucleate, keratin-filled squames in two or three horizontal layers. Ultrastructurally, the squames contain closely packed filaments that lie transversely to the direction of proxi - modistal growth, and are embedded in a dense protein matrix. Unlike the general epidermis, squames are not shed from the nail plate surface. A variety of mineral elements including calcium are present in nail. The hardness of nail is determined by the arrangement and cohesion of the layers of squames and their internal fibres, and not by the calcium content. The water content of nail is low but nail is 10 times more permeable to water than the general epidermis. The softness and elastic- ity of the nail plate are related to its degree of hydration. The nail apparatus consists of the nail plate, proximal and lateral nail folds, nail matrix, nail bed and hyponychium. Nail plate The nail plate is embedded within the proximal and lateral nail folds. It is approximately rectangular in shape and is mostly convex in both longitudinal and transverse axes. There is considerable inter- and intra-individual variation (see Fig. 7.18). The thickness of the plate increases proximodistally from about 0.7 mm to 1.6 mm, although the terminal thickness varies between individuals. The surface of the nail plate may show fine longitudinal ridges, and its undersurface is grooved by cor - responding ridges in the nail bed. Disturbances of growth pattern or disease may lead to transverse ridging or grooves. Minute trapped air Fig. 7.18 The organization and terminology of the structures associated with a fingernail. (With permission from Paus R, Peker S 2003 Biology of hair and nails. In: Bolognia JL, Jorizzo JL, Rapini RP (eds) Dermatology. London: Mosby.) Distal nail edge Nail plateLunula Eponychium (cuticle)Lateral nail fold Proximal nail fold Proximalnail edge Distal groove HyponychiumProximal nail fold Nail bedDistal nail matrixProximal nail foldEponychium(cuticle) Eponychium (cuticle) Proximalnail matrix Joint Tendon Distal phalanxHyponychium Collagen bundle Distal phalanx Lateral nail foldLateral nail groove Subcutaneous fatDermis CapillaryFree edge of the nailNail plateGlomus organCollagen bundleEponychiumDermisProximal nail matrix Distal nail matrix Proximal nail fold Eponychium (cuticle) Lateral nail fold Epidermal ridge Nail bed Hyponychium
213
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Skin And i TS APPE ndAgES 152SECT iOn 1 RSPO4 gene, which encodes R-spondin 4, have been shown to cause recessive anonychia (absence of all fingernails and toenails). Further - more, mutations in the FZD6 gene, which encodes the Wnt receptor protein frizzled 6, have been shown to underlie the molecular basis of autosomal recessive nail dysplasia, as well as twenty-nail dystrophy (a condition that affects the nails of all the fingers and toes). VASCULAR SUPPLY, LYMPHATIC DRAINAGE AND INNERVATION VASCULAR SUPPLY AND LYMPHATIC DRAINAGE Cutaneous blood flow amounts to approximately 5% of the cardiac output. The cutaneous circulation has an important thermoregulatory function, and is organized so that its capacity can be rapidly increased or decreased by as much as 20 times, in response to a need to lose or conserve body heat. The ability to control skin blood flow decreases with ageing, making older adults less able to thermoregulate adequately. Lifestyle factors, such as physical activity, diet and smoking, might interact with the ageing process to modulate ‘normal’ age-associated changes in the cutaneous microcirculation. The blood supply to the skin originates from three main sources: the direct cutaneous, musculocutaneous and fasciocutaneous systems. The direct cutaneous system of vessels is derived from the main arterial trunks and accompanying veins. These vessels course in the subcutane - ous fat parallel to the skin surface and are confined to certain areas of the body, e.g. the supraorbital artery, the superficial circumflex iliac artery and the dorsalis pedis artery. The musculocutaneous perforators arise from the intramuscular vasculature, pass through the surface of the muscle, and pierce the deep fascia to reach the skin by spreading out in the subcutaneous tissues. The fasciocutaneous system consists of perforating branches from deeply located vessels (deep to the deep fascia) that pass along intermuscular septa and then fan out at the level of the deep fascia to reach the skin. Examples include the fasciocutane - ous perforating vessels from the radial and ulnar arteries. The direct cutaneous vessels, musculocutaneous perforators and fasciocutaneous perforators each contribute to six anastomosing hori - zontal reticular plexi of arterioles (Fig. 7.19), which have vascular con- nections between them and which ultimately provide the blood supply to the skin. Three plexi are located in the skin itself and supply all ele - ments including the sweat glands and pilosebaceous units. The subpap - illary plexus is located at the junction of the papillary and reticular layers of the dermis and gives off small branches that form perpendicu - lar capillary loops in the dermal papillae (usually one loop per papilla) (see Figs 7.1, 7.4; Fig. 7.20). The reticular dermal plexus is located in the middle portion of the dermis and is primarily venous. The deep dermal plexus is located in the deepest part of the reticular dermis and on the undersurface of the dermis. The close association between arte - riolar and venous plexi allows the countercurrent heat exchange between bloods at different temperatures. The remaining three plexi are the subcutaneous plexus, and two plexi associated with the deep fascia. The deep fascia has a plexus on its deep surface and a more extensive plexus superficially. This arrangement is much more pronounced in the limbs than it is in the trunk. The general structure and arrangement of the microvasculature is described in detail in Chapter 6 and so only features particular to skin will be considered here. In the deeper layers of the dermis, arterio - venous anastomoses are common, particularly in the extremities (hands, feet, ears, lips, nose), where vessels are surrounded by thick muscular coats. Under autonomic vasomotor control, these vascular shunts, when relaxed, divert blood away from the superficial plexus and so reduce heat loss, while at the same time ensuring maintenance of some deep cutaneous circulation and preventing anoxia of structures such as nerves. Extensive capillary anastomoses are present. Generally, cutaneous blood flow is regulated according to thermoregulatory need and also, in some areas of the body, according to emotional state. In very cold conditions, the peripheral circulation is greatly reduced by vasoconstriction, but intermittent spontaneous vasodilation results in periodic increases in temperature that prevent cooling to the level at which frostbite might occur. This is thought to be due to a direct effect of oxygen lack on the arteriolar constrictor muscle, rather than to a neural influence. The lymphatics of the skin, as elsewhere, are small terminal vessels that collect interstitial fluid and macromolecules for return to the cir - culation via larger vessels. They also convey lymphocytes, Langerhans cells and macrophages to regional lymph nodes. They begin as blind-ended, endothelial-lined tubes or loops just below the papillary dermis. The eponychium is bounded by the fascial attachment of the skin to the base of the distal phalanx, distal to the insertion of the extensor tendon, and its distal free edge. It adheres to the dorsal aspect of the nail plate and overlies the root of the nail. Nail matrix The nail matrix is the main source of the nail plate and it can be sub-divided into three parts. Proximally, the dorsal matrix is defined as the volar surface (undersurface) of the proximal nail fold. The intermediate matrix (germinal matrix) starts where the dorsal matrix folds back on itself and extends as far as the distal portion of the lunula. The ventral matrix (sterile matrix) is the remainder of the nail bed; it starts at the distal border of the lunula and ends at the hyponychium. The matrix epithelium consists of typical basal and prickle layer keratinocytes, among which are scattered melanocytes and Langerhans cells. Cornified cells of the dorsal and ventral aspects of the matrix are steadily extruded distally to form the nail plate. The proximal 50% of the nail matrix contributes approximately 80% of the nail plate. This process continues into the nail bed at the distal edge of the lunula, which is formed where the distal portion of the ventral matrix underlies the nail plate. The lunula is pale, opaque and convex, and is more prominent in the thumb than the other digits. It is not known why the lunula is so pale compared with the more distal translucent pink nail bed. The lack of colour may reflect the thickness of the epidermis in the lunula and/ or a paucity of capillaries in the dermis of the lunula. Nail bed The epidermis of the nail bed extends from the distal margin of the lunula to the hyponychium. The distal margin of the nail bed, which is the point where the nail plate becomes free of the nail bed, is called the onychodermal band. The surface of the nail bed is ridged and grooved longitudinally, corresponding to a similar pattern on the undersurface of the nail plate. This results in a tight interlocking between the nail plate and the underlying nail bed, protecting against microbial invasion and the collection of debris underneath the nail. The epidermis of the nail bed is thin and lacks a stratum granulosum. It consists of 2–3 layers of nucleated cells that lack keratohyalin gran - ules, and a thin cornified layer that moves distally with the growing nail plate. It contains an occasional sweat gland distally. The dermis of the nail bed is anchored to the periosteum of the distal phalanx without an underlying subcutaneous layer. It forms a tight compartment, which means that infections involving the nail bed or an increase in pressure, e.g. resulting from a haematoma, may cause severe pain, which may be relieved by the removal of part or all of the nail plate. The dermis of the nail bed is richly vascularized. The blood vessels are arranged longitudinally and display numerous glomus bodies, which are encapsulated arteriovenous anastomoses involved in the physiological control of peripheral blood flow in relation to tem - perature. The dermis is well innervated and contains numerous sensory nerve endings, including Merkel endings and Meissner’s corpuscles. Nail bed cells differentiate towards the nail plate and contribute to its thickness ventrally. Hyponychium The hyponychium is the area under the free nail between the onychoder - mal band proximally and the distal groove. It is an epidermal ridge that demarcates the junction between the finger pulp and the subungual structures. Nail growth Nail growth is determined by the turnover rate of the matrix cells, which varies with digit, age, environmental temperature and season, time of day, nutritional status, trauma and various diseases. Generally, the rate of growth is related to the length of the digit, being fastest (approxi - mately 0.1 mm per day) in the middle finger of the hand and slowest in the little finger. Fingernails grow 3–4 times faster than toenails, more quickly in summer than in winter, and faster in the young than in the old. A fingernail grows out in about 6 months, whereas a toenail is replaced, on average, in 18 months. Genetic keratin disorders (see Haines and Lane (2012)) may lead to nail dystrophies such as pachyonychia congenita, where the nails become grossly thickened. In addition, germline mutations in the
214
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Vascular supply, lymphatic drainage and innervation 153 CHAPTER 7 ceptors found in hairy and glabrous skin. They are located in dermal papillae close to the dermal–epidermal junction. They are rapidly adapting mechanoreceptors and are responsible for sensing light touch. These receptors are particularly suited to detecting shape and texture during active exploratory touch and are numerous in the finger pads. The primary input is transmitted by neurones whose cell bodies lie in the spinal and cranial ganglia, and whose myelinated or unmyeli - nated axons are terminally distributed, mainly within the dermis. Effer - ent autonomic fibres are unmyelinated and can be either noradrenergic or cholinergic. They innervate the arterioles, arrector pili muscles, and the myoepithelial cells of sweat and apocrine glands. In the scrotum, labia minora, perineal skin and nipples they also supply the smooth muscle fasciculi of the dermis and adjacent connective tissue. Except in the nipples and genital area, the activity of the autonomic efferent nerves is mainly concerned with regulation of heat loss by vasodilation and vasoconstriction, sweat production and pilo-erection (although this last is a minor function in humans). On reaching the dermis, the nerve fasciculi branch extensively to form a deep reticular plexus that innervates much of the dermis, includ - ing most sweat glands, hair follicles and the larger arterioles. Many small fasciculi pass from this plexus to ramify in another superficial papillary plexus at the junction between the reticular and papillary layers of the dermis. Branches from this pass superficially into the papil - lary layer, ramifying horizontally and vertically, and terminate either in relation to encapsulated receptors, or as terminals reaching the level of the basal lamina. In some instances, they enter the epidermis as free endings that are responsive to light pressure and touch sensation or to nociceptive stimuli. As these fasciculi terminate, they lose their epineu - rial and perineurial sheaths, leaving the Schwann cell–axonal com - plexes or naked axons enveloped only by basal lamina, in direct contact with the matrix. These naked distal axonal terminals may be vulnerable to pathogens entering via a skin abrasion. The structure and classifica - tion of sensory endings are described in detail on page 59. The segmental arrangement of the spinal nerves is reflected in the sensory supply of the skin: a dermatome is the area supplied by all the cutaneous branches of an individual spinal nerve through its dorsal and ventral rami (p. 233 and Fig. 16.10). Typically, dermatomes extend round the body from the posterior to the anterior median line. The upper half of each zone is supplemented by the nerve above, the lower half by the nerve below. Dermatomes of adjacent spinal nerves overlap markedly, particularly in the segments least affected by development of the limbs (Ladak et al 2014). The interrelationship between the skin and the nervous system is important in the pain seen in certain polyneu - ropathies and after injury to nerve or skin.These drain into a superficial plexus below the subpapillary venous plexus, which drains via collecting vessels into a deeper plexus at the junction of the reticular dermis and subcutis, and this, in turn, drains into larger subcutaneous channels. INNERVATION Skin is a major sensory organ, with regional variations in its sensitivity to different stimuli. It is richly innervated and is also involved with autonomic functions such as thermoregulation. Information about the external environment is relayed through receptors that are responsive to various stimuli, which may be mechanical (rapid or sustained touch, pressure, vibration, stretching, bending of hairs, etc.), thermal (hot and cold) or noxious (perceived as itching, discomfort or pain). Pacinian corpuscles subserve deep pressure and vibrational sensation, and are located deep in the dermis or in the hypodermis, particularly of the digits. Meissner’s corpuscles are highly specialized mechanore -Fig. 7.19 Vascular supply to the skin. A, Note the various horizontal plexuses fed by direct cutaneous, fasciocutaneous and musculocutaneous arteries. Compare with Figure 79.6. B, Higher magnification of vascular supply. (A, Redrawn from McCarthy JG (ed) 1990 Chapter 9 in Plastic Surgery, Vol 1. Philadelphia: Saunders. B, Redrawn from Cormack GC, Lamberty BGH 1994 The Arterial Anatomy of Skin Flaps, 2nd edition. Edinburgh: Churchill Livingstone.)Subcutaneous tissue MuscleDirect cutaneous artery Skin Fascia Internal artery Fasciocutaneous arteryMusculocutaneous arterySubfascialPrefascialSubcutaneousSubdermalMid-dermal (primarily venous plexus)SubpapillaryPlexuses Subcutaneous tissueHair Reticular dermisPapillary dermisEpidermisSubpapillary plexusPapillary loopMid-dermal (primarily venous plexus) Deep subdermal plexusSubcutanous vesselsA B Fig. 7.20 A thick vertical section through palmar skin: the arteries, arterioles and capillaries have been injected with red gelatin to demonstrate the pattern of dermal vascularization. At the base of the dermis a broad, flat arterial plexus supplies a more superficial papillary plexus, which in turn gives off capillary loops that enter the dermal papillae.
215
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Skin And i TS APPE ndAgES 154SECT iOn 1 coat of glycosaminoglycans, and cytoplasmic vesicles become promi - nent deep to it. These developments reach a peak between 12 and 18 weeks of gestation, at which time the periderm is a major source of the amniotic fluid to which it may contribute glucose; it also has an absorp - tive function. From about 20 weeks onwards, the globular protrusions become undermined and pinched off to float free in the amniotic fluid. The now-flattened periderm cells undergo terminal differentiation to form a temporary protective layer for the underlying developing epider - mis, against an amniotic fluid of changing composition as a result of the accumulation of products of fetal renal excretion. Up to parturition, the periderm squames continue to be cast off into the amniotic fluid and they contribute to the vernix caseosa, a layer of cellular debris that covers the fetal skin at birth. Proliferation in the germinative layer leads to a stratified appearance with successive layers of intermediate cells between it and the periderm. From an early stage, cells of all layers are packed with glycogen granules. Differentiation of these layers is not synchronous throughout all regions of the developing skin, being more advanced cranially than caudally, and progressing on the body from the mid-axillary line ventrally. Reduc - tion in glycogen content of the cells is associated with a shift towards biosynthetic activity connected with terminal (cornifying) differentia - tion, manifested by the presence of different enzymes and expression of keratins. Simple epithelial keratins present from before implantation (K8 and K18) are replaced by typical keratinocyte basal cell keratins (K5 and K14), followed in the first suprabasal cell layer by those of higher molecular weight associated with differentiation (K1 and K10) at 10–12 weeks. This is soon followed by the expression of profilaggrin and filag - grin, and the appearance of keratohyalin granules among filamentous bundles of the uppermost intermediate layer cells at approximately 20 weeks. The first fully differentiated keratinocytes appear shortly after - wards. By 24–26 weeks a definite cornified layer exists in some areas, and by approximately 30 weeks, apart from some residual glycogen in intermediate cells, the interfollicular epidermis is essentially similar to its postnatal counterpart. Non-keratinocytes are present in developing epidermis from about 8 weeks’ gestation. Langerhans cells can be seen in the epidermis by 5–6 weeks and are fully differentiated by 12–14 weeks. Their numbers increase at least partially by mitotic division in situ, but at 6 months are only 10–20% of those in the adult. It is not known if Langerhans cells function in immune surveillance in fetal skin. Melanocytes are present in the bilaminar epidermis of cephalic regions as early as 8 weeks. By 12–14 weeks they can reach a density of 2300 per mm 2, reducing to DEVELOPMENT OF SKIN AND SKIN APPENDAGES The skin develops from the surface ectoderm and its underlying mes - enchyme. The surface ectoderm gives rise to the epidermis and its appendages such as the pilosebaceous units, sweat glands and nail units. Interactions between ectoderm and mesenchyme give rise to the oral and nasal epithelia, as well as teeth. Ectodermal cells differentiate mainly into the keratinocytes and probably Merkel cells. Melanocytes are derived from the neural crest while Langerhans cells and lym - phocytes originate from the bone marrow. The dermis is derived from the somatopleuric mesenchyme (in the limbs and trunk), and possibly the somitic mesenchyme (covering the epaxial musculature), and from the neural crest (in the head). Ang - iogenic mesenchyme gives rise to the blood vessels of the dermis. Nerves and associated Schwann cells, which are derived from the neural crest, enter and traverse the dermis during development. EPIDERMIS AND APPENDAGES General (interfollicular) epidermis In the first 4–5 weeks, embryonic skin consists of a single layer of ecto - dermal cells overlying a mesenchyme containing cells of stellate den - dritic appearance interconnected by slender processes and sparsely distributed in a loosely arranged microfibrillar matrix ( Fig. 7.21). The interface between ectoderm and mesenchyme, known as the basement membrane zone (BMZ), is an important site of mutual interactions on which the maintenance of the two tissues depends, both in prenatal and postnatal life. Ectodermal cells, which characteristically contain glycogen deposits, contact each other at gap and tight junctions (occlud - ing junctions, zonae occludentes). The layer so formed soon develops into a bilaminar epithelium, and desmosomes also appear. The basal germinative layer gives rise to the postnatal epidermis, and the superfi - cial layer to the periderm, a transient layer confined to fetal life. The periderm maintains itself, expresses different keratin polypeptides, and grows by the mitotic activity of its own cells, independent of those of the germinative layer. Originally flattened, the periderm cells increase in depth; the central area containing the nucleus becomes elevated and projects as a globular elevation towards the amniotic cavity. The plasma membrane develops numerous surface microvilli with an extraneous Fig. 7.21 Development of the skin. Shedding of periderm globular elevations Langerhans cellMerkel cell MelanocyteLangerhans cell Melanocyte Merkel cell Early embryonic phaseMiddle Peridermal phaseEarlyDefinitive epidermal organization with mature-type keratinizationTransformed 0 4 8 12 16 20 24 28 32 36 Development in weeksTermEctoderm Basal lamina Basement membrane zoneIntermediate cellsPeriderm DermisEpidermisSecretion AbsorptionPeridermal squamesKeratinocyte squames
216
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
development of skin and skin appendages 155 CHAPTER 7 at approximately 9 weeks, a flattened nail field limited by proximal, distal and lateral nail grooves is apparent. The nail field ultimately forms the nail bed, and the primordium of the nail is formed of a wedge of cells that grows diagonally, proximally and deeply into the mesen - chyme from the proximal groove towards the underlying terminal phalanx. The deeper cells of this wedge form the primordium of the matrix, which gives rise to the nail plate. The latter emerges from under the now proximal nail fold at about 14 weeks and grows distally over an already keratinized nail bed. The nail matrix is usually considered to have dorsal and ventral (intermediate) components, but there are conflicting opinions as to the extent to which each contributes to the nail, both in ontogeny and postnatally. It is generally agreed that the ventral matrix contributes the major part. At 20 weeks, the nail plate entirely covers the nail field (nail bed), now limited distally by a distal ridge, which, when the plate projects beyond the tip, becomes the hyponychium beneath it. At birth, the microstructure of the main nail unit components is similar to that postnatally. DERMIS The embryonic dermis is more cellular than the adult dermis, and many of these mesenchymal cells are involved in signalling pathways that regulate ectodermal differentiation. The mesenchymal cells underlying the surface ectoderm and early bi- and trilaminar epidermis contact each other by slender processes to form an intercommunicating network. They secrete a matrix that is rich in ions, water and macro - molecules, proteoglycan/glycosaminoglycans, fibronectin, collagenous proteins of various types and elastin. Further development of these intrinsic components involves the differentiation of individual cell types, fibroblasts, endothelial cells, mast cells and the assembly of matrix components into organized fibrillar collagen fibres and elastic fibres. During embryogenesis, the matrix is heterogeneous with regard to its biochemical and macromolecular components. The main gly - cosaminoglycans of embryonic and fetal skin are glycuronic acid and dermatan sulphate. Types I, III, V and VI collagens are distributed more or less uniformly, regardless of fetal age, with some local concentrations of type III and V collagens, the levels of which are higher than in post - natal skin. Collagens of types IV and VII are found predominantly in the BMZ. The progressive morphological differentiation of the dermis in - volves its separation from the subcutis at about the third month; changes in composition and size of collagen fibrils and their organiza - tion into bundles, amongst which cells become relatively fewer; down - growth of epidermal appendages; the organization of nervous and vascular plexuses; and the relatively late appearance of elastic net - works. The papillary and reticular regions are present as early as 14 weeks but the overall organization of the dermis continues to develop postnatally. Vascular supply and lymphatic drainage The dermal vasculature is generally thought to be developed in situ by the transformation of angiogenic mesenchymal cells. Closed endothelium-lined channels containing nucleated red cells are present by 6 weeks underneath the ectoderm, and by the eighth week are arranged in a single plane parallel to the epidermis; these ultimately form the subpapillary plexus. A second, deeper horizontal plexus is evident by 50–70 days. Both plexuses extend by budding and give rise to the final patterns of arterioles, venules and capillaries, which are established shortly after birth. Pericytes also develop from mesenchy - mal cells. Lymphatic vessels are formed by mesenchymal cells, which become organized to enclose pools of proteinaceous fluid leaking from the developing capillaries. Innervation Sensory cutaneous nerves (axons and Schwann cells) are derived from the neural crest (via dorsal root and cranial sensory ganglia). Motor fibres to vessels and glands arise from neurones in sympathetic ganglia. As individual parts of the embryo grow, the nerves grow and lengthen with them. Small axons are present superficially at a stage when the epidermis is bilaminar, and by 8 weeks of gestation a functioning cuta - neous plexus is already present. By the fourth gestational month, the dermal plexuses are richly developed and Meissner and Pacinian cor- puscles have appeared.800 per mm2 just before birth. Keratinocytes regulate the final ratio between themselves and melanocytes via growth factors, cell-surface molecules and other signals. Fetal melanocytes produce melanized melanosomes and transfer them to keratinocytes. This is an intrinsic activity independent of UV irradiation and suggests that melanin has functions other than photoprotection. Merkel cells begin to appear in the epidermis of the palm and sole of the foot between 8 and 12 weeks, and later in association with some hairs and with dermal axonal–Schwann cell complexes. Pilosebaceous unit Pilosebaceous units develop in a cranio–caudal direction at about 9 weeks’ gestation, first in the regions of the eyebrows, lips and chin. The hair placode is a collection of cells in the basal layer of the epidermis and develops adjacent to a local concentration of mesenchymal cells, which will eventually become the dermal papilla. Further proliferation and elongation of the cells lead to a hair germ, which protrudes down - wards into the mesenchyme in association with the primitive dermal papilla during weeks 13–15. The hair germ becomes a hair peg as it migrates downwards within the developing skin, and when its bulbous lower portion envelops the dermal papilla it becomes known as a bulbous peg. Melanocytes are individually present at the hair peg stage, and abundantly so and quite active in the bulbous peg. At this stage (approximately week 15), two or three swellings appear on the posterior wall. The uppermost is the rudiment of the apocrine gland (present only in some follicles), the middle forms the sebaceous gland and the lower one is the bulb, to which the arrector pili muscle (arising from underlying mesenchyme) later becomes attached. The cells of the low - ermost region of the bulb, the matrix, divide actively and produce a pointed hair cone. This grows upwards to canalize a developing hair tract, along which the fully formed hair, derived by further differentia - tion of cells advancing from the matrix, reaches the surface at approxi - mately week 18 of gestation. Sebaceous glands develop independently of hair follicles in the nos - trils, eyelids (as tarsal glands) and the anal region. Apocrine sweat glands are formed at the same time as eccrine (merocrine) glands and are at first widely distributed over the body. Their number dimin - ishes from 5 months’ gestation, producing the distribution seen in adult skin. Hairs produced prenatally are called lanugo hairs, which are short and downy, lack a medulla, and in certain parts of the body are arranged in a vortex-like manner into tracts. Late in pregnancy, lanugo hairs are replaced by vellus hairs, and these in turn by intermediate hairs, which are the predominant type until puberty. New follicles do not develop in postnatal skin. Eccrine sweat glands Eccrine (merocrine) sweat glands are one type of sudoriferous gland. Sweat gland rudiments appear in the second and third months of gesta - tion as cell buds associated with the primary epidermal ridges of the finger and toe pads of terminal digits. They elongate into the dermis and by 16 weeks the lower portion begins to form the secretory coil, within which, by 22 weeks, secretory and myoepithelial cells are present. The solid cord of cells connecting the coil to the epidermis becomes the intradermal duct, and the lumina of both are formed by dissolution of desmosomal contacts between the cells. The intraepidermal duct is foreshadowed by a column of concentrically arranged inner and outer cells within which a lumen is formed and opens on the surface at 22 weeks. As with hair follicles, no new eccrine sweat glands develop post - natally. Emotional sweating occurs in preterm infants from 29 weeks’ gestational age. Epidermal ridges The primary epidermal ridges develop as regularly spaced, small down - growths of epidermal cells separated by corresponding dermal ridges during the second and third months of gestation. In the fifth month secondary ridges develop, with the pattern becoming evident on the surface; this pattern is finalized through further remodelling postnatally. Nails Fields of proliferative ectoderm appear on the tips of the terminal seg - ments of the digits. They progressively reach a dorsal position, where,
217
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Skin And i TS APPE ndAgES 156SECT iOn 1 e.g. on the cheek the primary creases radiate from the hair follicles, on the scalp they form hexagons, while on the calf and thigh they form parallelograms. There is a relationship between the type of pattern and local skin extensibility. Wrinkle lines Wrinkle lines are caused by contraction of underlying muscles and are usually perpendicular to their axis of shortening. On the face they are known as lines of expression, and with progressive loss of skin elasticity due to ageing, they become permanent. Occupational lines are creases produced by repeated muscular contractions associated with particular trades or skills. Contour lines are lines of division at junctions of body planes, e.g. the cheek with the nose, and lines of dependency are pro- duced by the effect of gravity on loose skin or fatty tissue, e.g. the creases associated with the pendulous fold beneath the chin in older age. Flexure (joint) lines Flexure (joint) lines are major markings found in the vicinity of syno - vial joints, where the skin is attached strongly to the underlying deep fascia (see Fig. 7.22). They are conspicuous on the flexor surfaces of the palms, soles and digits, and in combination with associated skin folds they facilitate movement. The skin lines do not necessarily coincide with the associated underlying joint line. For example, the flexure lines demarcating the extended fingers from the palm lie approximately 12 mm distal to the metacarpophalangeal joints, the positions of which are more closely related to the distal palmar crease. The patterns of EPITHELIAL–MESENCHYMAL INTERACTIONS IN DEVELOPING SKIN Epidermal–mesenchymal (dermal) interactions at the BMZ occur during development and throughout life. At the ectodermal stage, the BMZ consists of the basal plasma membrane of an ectodermal cell, paralleled on its cytoplasmic side by various cytoskeletal filaments, and beneath it by a layer (0.1–0.2 µm) of microfibrillar–amorphous mate- rial deposited by the cell. At the bilaminar stage, a continuous lamina densa is present, separated from the basal plasma membrane by a lamina lucida traversed by loosely fibrillar material; similar filaments extend from the lamina densa into the mesenchymal matrix. Hemidesmosomes (see Fig. 1.19) begin to appear at 8 weeks as stratification starts, and anchoring fibrils at 9–10 weeks. By the end of the third month, the basic morphology of the interfollicular BMZ is essentially similar to that of the postnatal BMZ. Laminin and type IV collagen are present in the developing basal lamina at 6 weeks; bullous pemphigoid antigen (BPAG, in hemidesmo - somes) and anchoring fibril proteins are expressed later. These immu - nocytochemical and morphological observations are of importance for the prenatal diagnosis of genetically determined diseases such as epi- dermolysis bullosa. The basal lamina provides a physical supporting substrate and attachment for the developing epidermis, and is thought to be selectively permeable to macromolecules and soluble factors regu - lating epidermal–dermal morphogenetic interactions. NEONATAL GROWTH The surface area of the skin increases with growth. It is estimated that the surface area of a premature neonate weighing 1505 g is approxi - mately 1266 cm2, whereas a neonate of 2980 g has a surface area of 2129 cm2. The skin of the neonate is thinner than that of older infants and children. It cornifies over a period of 2–3 weeks, which provides protection. In the premature infant the even thinner epidermal layer allows absorption of a variety of substances such as chlorhexidine and permits a significantly higher transepidermal water loss than occurs in full-term neonates. At birth the skin is richly vascularized by a dense subepidermal plexus. The mature pattern of capillary loops and of the subpapillary venous plexus is not present at birth but develops as a result of capillary budding with migration of endothelia at some sites and the absorption of vessels from other sites. Some regions mature faster than others. With the exceptions of the palms, soles and nail beds, the skin of the neonate has almost no papillary loops. It has a disor - dered capillary network, which becomes more orderly from the second week when papillary loops appear; defined loops are not present until the fourth or fifth week, and all areas possess loops by 14–17 weeks postnatally. Neonates exhibit a regional sequence of eccrine gland maturation. The earliest sweating occurs on the forehead, followed by the chest, upper arm and, later, more caudal areas. Acceleration of maturation of the sweating response occurs in premature babies after delivery. NATURAL SKIN CREASES AND WRINKLES SKIN LINES The surface of the skin and its deeper structures show various linear markings, seen as grooves, raised areas and preferred directions of stretching. Surface pattern lines, tension lines and skin creases Externally visible skin lines are related to various patterns of epidermal creasing, ridge formation, scarring and pigmentation. A simple lattice pattern of lines occurs on all major areas of the body other than the thick skin of volar and plantar surfaces. The lattice pattern typically consists of polygons formed by relatively deep primary creases visible to the naked eye, which are irregularly divided by finer secondary creases into triangular areas ( Figs 7.22–7.23). These, in turn, are further subdivided by tertiary creases limited to the cornified layer of the epi - dermis and, finally, at the microscopic level, by quaternary lines, which are simply the outlines of individual squames (see Fig. 7.7). Apart from the quaternary lines, all the others increase the surface area of the skin, permit considerable stretching and recoil, and distribute stresses more evenly. Details of the pattern vary according to the region of the body; Fig. 7.22 The surface of hairless skin from the palm of the hand, showing epidermal friction (papillary) ridges and larger flexure lines (left). Fig. 7.23 A scanning electron micrograph of the surface of thin skin of the back, showing an interlacing network of fine creases and predominantly triangular areas between them.
218
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
natural skin creases and wrinkles 157 CHAPTER 7 importance. Measurable parameters include the frequency of ridges in particular patterns and the disposition of tri-radii, which are junctional areas where three sets of parallel ridges meet. Fingerprint ridge patterns can be separated into three major types (see Fig. 7.24): arches (5%), loops (70%) and whorls (25%). Arches have no tri-radii, loops have one tri-radius, and whorls have two or more. Whorl finger patterns are more common on the right hand, and males generally have more whorls and fewer arches than females, in whom the ridges are relatively narrower. The frequency of individual patterns varies with particular fingers. Similar patterns are seen on the toes. Adermatoglyphia is an extremely rare autosomal dominant condition in which fingerprint ridges are absent, and is caused by mutations in the SMARCAD1 gene. Other genetic disorders may also have reduced or absent dermatoglyph - ics as part of the clinical features. If the mechanical demands placed on the skin are greater than the skin creases and the dermis can accommodate, the lateral cohesion of dermal collagen fibres becomes disrupted with associated haemorrhage and cellular reaction, and, eventually, the formation of poorly vascular- ized scar tissue. These changes are termed intrinsic to distinguish them from scars formed by external wounding. Sites of dermal rupture are visible externally as lines or striae. They are initially pink in colour, later widen and become a vivid purple or red (striae rubrae), and eventually fade, becoming paler than the surrounding intact skin (striae albae). They develop on the anterior abdominal wall of some women in preg - nancy, when they are termed striae gravidarum (stretch marks). Variation in pigmentation can also produce externally visible lines on the surface of the skin. Futcher’s or Voigt’s lines mark differences in pigmentation between the darker extensor and paler flexor surfaces of the arms, and occur along the anterior axial lines, extending from the sternum to the wrist. They are more common in darker-skinned people. Lines detectable after manipulation or incision In certain regions of the body, surgical wounds heal with a better and less conspicuous scar if they are orientated in a particular direction. This finding is related to a number of factors including skin tension and naturally formed wrinkle lines. Skin is normally under tension and the direction in which this is greatest varies regionally. Tension is dependent on the protrusion of underlying structures, the direction of underlying muscles, and on joint movements. Many anatomists and surgeons have therefore attempted to produce a body map to indicate the best direc - tion in which to make an elective incision to obtain the most aesthetic scar. These maps frequently differ, especially in the region of the face. Out of the multitude of described cleavage lines, the most commonly referred to are relaxed skin tension lines (RSTLs), Langer’s lines and Kraissl’s lines. Of these, the RSTLs and Kraissl’s lines are probably more appropriate lines for surgical incision. Relaxed skin tension lines Relaxed skin tension lines (RSTLs) are those that correspond to the directional pull (which forms furrows) when the skin is relaxed; they do not always correspond to wrinkle lines. The tension across the RSTL is constant even during sleep but can be altered (increased, decreased or abolished) by underlying muscle contraction. The direction of the RSTLs can be determined by pinching the skin in different directions. Pinching at right angles to the RSTLs will result in fewer and higher furrows than pinching parallel to these lines. Lines of Langer and Kraissl Langer punctured the skins of cadavers with a circular awl and noted the subsequent elliptical-shaped openings that this protocol produced. By connecting the long axes of the holes, he produced the cleavage lines named after him. These lines represent skin tension in rigor mortis but they frequently do not relate to the lines of choice in making elective incisions, e.g. Langer’s lines often run at right angles to the RSTLs on the face. Kraissl’s lines are essentially exaggerated wrinkle lines obtained by studying the loose skin of elderly faces whilst contracting the mimetic muscles of the face. For the most part, these lines do correspond to RSTLs but slight variation exists on the face, especially on the lateral side of the nose, the lateral aspect of the orbit, and the chin. Blaschko’s lines Blaschko’s lines represent a pattern of cutaneous mosaicism that can be observed in a range of congenital and acquired skin conditions. They do not appear to correspond to vascular or neural elements of the skin, and may be related to earlier developmental boundaries of a ‘mosaic’ nature.flexure lines on the palms and soles may vary and are, to some extent, genetically determined. Papillary ridges Papillary ridges are confined to the palms and soles and the flexor surfaces of the digits, where they form narrow parallel or curved arrays separated by narrow furrows ( Figs 7.24–7.25). The apertures of sweat ducts open at regular intervals along the summit of each ridge. The epidermal ridges correspond to an underlying interlocking pattern of dermal papillae, an arrangement that helps to anchor the two layers firmly together. The pattern of dermal papillae determines the early development of the epidermal ridges. This arrangement is stable throughout life, unique to the individual, and therefore significant as a means of identification. The ridge pattern can be affected by certain abnormalities of early development, including genetic disorders such as Down’s syndrome, and skeletal malformations such as polydactyly. Absence of epidermal ridges is extremely rare. Functionally, epidermal ridges increase the gripping ability of hands and feet, preventing slip - ping. The great density of tactile nerve endings beneath them means that they are also important sensory structures. The analysis of ridge patterns by studying prints of them (finger - prints) is known as dermatoglyphics and is of considerable forensic IPFL Fig. 7.25 A scanning electron micrograph of the surface of thick hairless skin from the volar surface of a human digit, showing friction ridges along which lines of sweat ducts open as pores (one pore is arrowed). (Courtesy of Professor Lawrence Bannister and Dr Caroline Wigley.) Fig. 7.24 The palmar aspect of a terminal phalanx to show fingerprint ridges. Note the interphalangeal flexure line (IPFL).
219
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Skin And i TS APPE ndAgES 158SECT iOn 1 division of the complex, interrelated processes that constitute the wound healing response. For further reading, see Miller and Nanchahal (2005) and Rozen et al (2009). HAEMOSTASIS Wounding causes vascular injury and initiates fibrin–fibronectin clot formation. The clot consists of platelets (promoted to aggregate by fibrillar collagens type I and III) embedded in a mesh of cross-linked fibrin and fibronectin fibres. It serves as a haemostatic plug, protects the denuded wound with a temporary physical shield, and forms a provi - sional matrix through and over which cells migrate during the repair process. The clot also acts as a reservoir of growth factors and cytokines, which are released as platelets degranulate, and contribute to inflam- matory cell recruitment. INFLAMMATION Neutrophils and monocytes are recruited to wound sites within minutes of injury by a variety of chemotactic signals, including complement factors, bacterial polysaccharides, cytokines and growth factors such as tumour necrosis factor alpha (TNF- α), interleukin (IL)-1, platelet- derived growth factor (PDGF), transforming growth factor alpha (TGF-α) and TGF-β1,2, basic fibroblast growth factors (bFGF/FGF-2), serotonin and monocyte chemotactic protein-1 (MCP-1). Neutrophils are active against contaminating bacteria. Monocytes differentiate into macrophages that phagocytose not only pathogenic organisms, but also expended neutrophils, cell and matrix debris. Both neutrophils and macrophages act as a further source of cytokines and growth factors to amplify stimuli already present in the wound site. Macrophages are also a source of additional factors such as TGF-α, FGF and vascular endothelial growth factor (VEGF) that activate resident fibroblasts, endothelial cells and keratinocytes, which in turn are central to the processes of re-epithelialization and formation of granulation tissue, as discussed below. PROLIFERATION The proliferation phase of wound healing involves re-epithelialization and granulation tissue formation, which take place more or less simultaneously. Re-epithelialization Re-epithelialization begins within hours of wounding as a result of keratinocyte migration and proliferation from the wound edges and skin appendages. Cytokines such as epidermal growth factor (EGF), FGF, keratinocyte growth factor (KGF/FGF-7), insulin-like growth factor-1 (IGF-1) and TGF- α are released by activated fibroblasts and keratinocytes, and stimulate the re-epithelialization process. The mech - anisms of keratinocyte migration are not fully understood, although disruption of desmosomes and hemidesmosomes, extracellular matrix contact, formation of cytoplasmic actin filaments and degradation of the fibrin matrix are all likely to be important. There is evidence that keratinocyte expression of matrix metalloproteinases (MMPs) is associ - ated with re-epithelialization. During this phase of cutaneous wound healing, keratinocyte proliferation is thought to be upregulated after migration has started; dividing and migrating cells are found in the first suprabasal as well as basal keratinocyte layers. MMP activity is regulated by the production of tissue inhibitors of MMPs (TIMPs), and a relative imbalance, with upregulation of certain MMPs, may be associated with impaired wound healing. When the denuded wound surface has been covered by a monolayer of keratinocytes, migration ceases. A stratified epidermis with an underlying basal lamina is re-established from the margins of the wound inwards. Anchoring fibrils linking the basal lamina to the underlying connective tissue mark neo-epidermal maturity. Granulation tissue formation The term granulation tissue refers to the macroscopic appearance of wound connective tissue, which appears pink and granular. It contains numerous capillaries that invade the initial wound clot and become organized into a microvascular network (angiogenesis), together with the cells and molecules necessary to stimulate neo-matrix deposition. AGE-RELATED SKIN CHANGES Two main factors, chronological and environmental, are involved in skin ageing. Chronological changes are physiological or intrinsic in origin. A major environmental factor is chronic exposure to the sun, referred to as photoageing, which is to some extent preventable. Intrinsic ageing From about the third decade onwards there are gradual changes in the appearance and mechanical properties of the skin, which reflect natural ageing processes. These become very marked in old age. Normal ageing is accompanied by epidermal and dermal atrophy, which result in changes in the appearance, microstructure and function of the skin. Alterations include wrinkling, dryness, loss of elasticity, thinning and easy bruising. Epidermal atrophy is manifested by general thinning and loss of the basal rete pegs and flattening of the dermal–epidermal junc - tion. Flattening of the junction decreases resistance to shear, leading to poor adhesion of epidermis and its separation following minor injury. The thickness of the cornified layer is not reduced in old age, and its permeability characteristics seem little affected. Epidermal proliferative activity and rate of cell replacement decline with age, being reduced by up to 50% in elderly skin. Synthesis of vitamin D is also reduced. After middle age, there is a 10–20% decline in the number of melanocytes, and Langerhans cells become sparser, a change which is associated with a reduction in immune responsiveness. Depigmentation and loss of hair, as well as some local increases (eyebrows, nose and ears in males, and face and upper lip in females), are commonly observed. Alterations in non-keratinocytes may be aggravated by chronic exposure to UV irradiation. Dermal changes are mainly responsible for the appearance of aged skin, its stiffness, flaccidity and wrinkling, as well as loss of extensibility and elasticity. Its general, thickness diminishes as a result of the decline in collagen synthesis by a reduced population of fibroblasts, though the relative proportion of type III collagen increases. Senile elastosis is a degenerative condition of collagen, which may be partly due to exces - sive exposure to sun. Vascularization of the skin is reduced, the capillary loops of the dermal papillae are particularly affected, and the tendency towards small spontaneous purpuric haemorrhages indicates a general fragility of the cutaneous microvasculature. A decrease in sensitivity of sensory perception, associated with some loss of specialized receptors, occurs. Ageing is also associated with fat redistribution, which may contribute to the physical characteristics of loose, sagging skin. Photoageing Skin ageing is also influenced by external factors, such as tobacco exposure, malnutrition, airborne particulate matter and UV radiation. Photoageing, also known as extrinsic skin ageing, occurs as a result of the hazardous effect of UV radiation on human skin. It is a major concern because of an association with skin cancer. Ultraviolet light B (UVB) is primarily absorbed in the epidermis and has been shown to induce angiogenesis and lymphatic dysfunction in skin. It causes DNA damage by generating cyclobutane pyrimidine dimers and 6,4- photoproducts, which are photocarcinogenic. Ultraviolet A is absorbed by cellular chromophores such as melanin, riboflavin and urocanic acid, resulting in the formation of reactive oxygen species, which in turn damage lipids, DNA and proteins (for review, see Kohl et al (201 1)). CUTANEOUS WOUND HEALING AND SCARRING The end-point of healing of mammalian skin wounds is usually scar formation. Cutaneous scars result from injury to both the epidermis and the underlying dermis. The epidermis largely regenerates, but dermal architecture is abnormal after repair and the undulating pattern of rete ridges at the dermal–epidermal junction is not reproduced. Scar tissue is biomechanically inferior to unwounded skin. Appendages such as hair follicles, sebaceous and sweat glands, which are derived from the epidermis, do not regenerate after wounding. The molecular biology of cutaneous repair involves the coordination of numerous cell types, signalling molecules and matrix proteins. Many of these signalling molecules have pleiotropic effects and it is the complex balance of these mediators, rather than their individual action, that determines events in wound repair. Wound healing is often con - sidered in four overlapping temporal phases: namely, haemostasis, inflammation, proliferation and remodelling ( Fig. 7.26). These events will be discussed separately for clarity, although this is an artificial
220
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Cutaneous wound healing and scarring 159 CHAPTER 7 contribute to tissue resilience. Increased wound strength coincides with new collagen deposition. Approximately 3 weeks after injury, collagen breakdown and synthesis equilibrate; subsequent, more gradual increases in wound strength reflect collagen remodelling, with the for-mation of intermolecular cross-links and larger collagen bundles. Scar maturation is associated with proportional increases in collagen type I relative to collagen type III. Collagen fibres of the dermis in scars are arranged in an irregular fashion and scarred skin only reaches a maximum tensile strength of about 70% that of unwounded skin. Wound contraction An important part of remodelling is wound contraction. Once granula - tion tissue has been laid down, a proportion of activated fibroblasts (or other mesenchymal cells) transform into myofibroblasts, which play a key role in wound contraction; they are thought to generate forces that pull normal dermal and adipose tissue into the wound defect. SCARLESS WOUND HEALING Wounds heal with reduced or absent scarring in certain circumstances, notably cutaneous wounds in the early fetus. The development of a major inflammatory response at a wound site appears to be a significant determinant of whether scarring occurs. Scarless fetal wound healing is characterized by fewer inflammatory cells (which are less differentiated than their adult counterparts and which remain in the wound for a shorter time), rapid re-epithelialization, reduced angiogenesis and res - toration of the connective tissue architecture in which collagen is arranged in the normal reticular pattern. An ontogenic transition period, during which cutaneous healing changes from scar-free to scar-forming, has been identified in the third trimester of gestation. Studies of this transition period have led to the identification of molecules of key importance in the scarring process; the most extensively characterized factor to date is TGF- β. Three mammalian TGF-β isoforms have been identified: TGF- β1, 2 and 3. Angiogenesis is a complex process, promoted by dynamic interactions between endothelial cells, angiogenic cytokines (including FGF-2, VEGF (mainly from keratinocytes), PDGF and TGF- β1,2) and the extracellular matrix environment. Electron microscopic studies have shown that the epidermis, basal lamina and papillary dermis all develop on the surface of the granula - tion tissue. Wounds that fail to granulate do not heal satisfactorily, suggesting that granulation tissue formation is a key aspect of wound repair. Excessive granulation is also associated with delayed re-epithelialization. Activated fibroblasts in the healing wound proliferate in response to growth factors, notably TGF- β1, IGF-1, PDGF, FGF and EGF. Within 72 hours of injury, these fibroblasts synthesize components of the new extracellular matrix, which are deposited in an orderly sequence. The neo-matrix initially includes fibronectin and hyaluronan, which form a provisional substratum for cellular migration. Fibronectin acts as an initiation site for collagen fibrillogenesis, and as anchorage for myofi - broblasts to effect wound contraction. Hyaluronan forms a highly hydrated matrix that is easily penetrated by migrating cells. Ultimately this and other neo-matrix components are replaced, first by collagen type III, and subsequently by collagen type I, which imparts strength to the mature scar (see remodelling, below). Non-structural proteins such as tenascin are also found in the neo-matrix of healing wounds and provide signals that orchestrate the repair process. Cellularity decreases during the evolution of granulation tissue into a mature scar (and during other phases of wound healing), mainly as a result of apoptosis. REMODELLING Remodelling of the extracellular matrix is important throughout the wound healing process and persists for some time after closure of the defect. Fibroblasts are responsible for matrix remodelling, as well as deposition. Initially they replace hyaluronan in the neo-matrix with sulphated proteoglycans, such as decorin, biglycan and versican, which Fig. 7.26 The processes involved in the healing of a normal cutaneous wound. Abbreviations: EGF, epidermal growth factor; FGF, fibroblast growth factor; KGF, keratinocyte growth factor; PDGF, platelet-derived growth factor: TGF, transforming growth factor; VEGF, vascular endothelial growth factor. Haemostasis Inflammation Proliferation RemodellingPDGF TGF-β1, 2Neutrophil Macrophage Platelet plugFibrin/fibronectin clot TGF-β1,2 FGFVEGF 3 20 Time (days) 100Relative cell numberMacrophage Neutrophil/polymorpho- nuclear leukocyteFibroblast Endothelial cellMyofibroblast Quiescent fibrocyteEGF KGF FGFTGF-αAdvancing epithelial layer New capillary Fibroblast Superficial scar‘Basket weave’ collagenRandomly orientated collagen Myofibroblast
221
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Skin And i TS APPE ndAgES 160SECT iOn 1 angiosome is an anatomical territory supplied by a source artery and drained by accompanying veins and lymphatic vessels (p. 132). It may include fascia, nerve, muscle and bone, as well as skin. The concept is important in analysis of patterns of ischaemia and in planning inci - sions, and enables the raising of free, composite, vascularized tissue flaps. (For further reading about the angiosome concept, see Taylor and Palmer (1987), Taylor et al (1994) and Yin et al (2013)). A free flap (free tissue transfer) is a specific type of flap in which the tissue, whether skin, fascia, muscle or bone, or a combination of these, is completely removed from its original location in the body along with a single identifiable artery and vein, and transferred to a remote site. The blood vessels in the flap are anastomosed to vessels located in the new site using microsurgical techniques. This often allows for greater flexibility in performing reconstructive surgery. SKIN STEM CELLS To maintain, repair and regenerate itself, the skin contains stem cells, which reside in various parts of the hair follicles, including the bulge area and isthmus, the basal layer of interfollicular epidermis, the base of sebaceous glands and within sweat glands. Stem cells are able to self-renew, as well as give rise to differentiating cells. However, it is not clear whether every basal keratinocyte or only a proportion of cells is a stem cell. Two possible hypotheses have emerged concerning epithelial renewal. One theory, known as symmetrical cell division, divides basal keratinocytes into epidermal proliferation units, which comprise one self-renewing stem cell and about 10 tightly packed transit amplifying cells, each of which is capable of dividing a limited number of times before exiting the basal layer to undergo terminal differentiation. This unit gives rise to a column of larger and flatter cells, which culminates in a single hexagonal surface. The alternative theory, known as asym - metrical cell division, is that some interfollicular stem cells (perhaps up to 70% of cells) can shift their spindle orientation from lateral to per - pendicular, such that one daughter cell is committed to terminal dif - ferentiation while the other maintains its proliferative capacity. Asymmetrical cell divisions can therefore bypass the need for transit amplifying cells. Hair follicle stem cells (see Hsu et al (201 1) for recent research) are found in the bulge regions below the sebaceous glands but also in other parts of the hair follicle. These stem cells undergo varying degrees of growth, degeneration and rest, which are governed, in part, by Wnt signalling and bone morphogenetic protein (BMP) inhibition. The bulge area stem cells generate cells of the outer root sheath, which drive the highly proliferative matrix cells next to the mesenchymal papillae. After proliferating, matrix cells differentiate to form the hair channel, the inner root sheath and the hair shaft. Hair follicle stem cells can also differentiate into sebocytes and interfollicular epidermis. Despite this multipotency, the follicle stem cells only function in pilosebaceous unit homeostasis and do not contribute to the interfollicular epidermis unless the skin is wounded. Stem cells are also found in the base of sebaceous glands; the progeny of these cells differentiate into lipid- filled sebocytes. Stem cells in sweat glands can differentiate into epithe - lial, luminal or myo-epithelial daughter cells to maintain skin integrity and homeostasis. Apart from epithelial stem cells, other cells in the dermis and sub - cutis may have stem cell properties. Some stem cells, termed dermal sheath cells, reside close to hair follicles; other stem cells in the dermis, known as MUSE (multi-lineage differentiating stress-enduring) cells, are pluripotent stem cells (see Commentary 1.2 ). Trauma and/or hypoxia in the skin, e.g. during wound healing or skin grafting, may induce recruitment of epithelial and endothelial progenitors from stem cell sources beyond the skin, such as bone marrow, predominantly from the mesenchymal stromal cell pool. For review of epidermal stem cells, see Beck and Blanpain (2012) and Ghadially (2012). Comparisons of fetal scar-free and adult scar-forming wounds show that TGF-β1 and 2 are not present in fetal wounds, suggesting that scar - less wound healing is associated with TGF- β3 activity, rather than TGF- β1 and 2. The β1 and 2 isoforms are dominant in fetal, neonatal and adult wounds that form scars. The main sources of TGF- β3 are fibro - blasts and keratinocytes, whilst TGF-β1 and 2 are produced from degranulating platelets and subsequently from monocytes and macro - phages; the low levels of TGF- β1 seen in fetal wounds have been attrib - uted to a relative lack of platelet degranulation and fibrin clot formation. SKIN GRAFTS AND FLAPS A graft is a piece of tissue that has been detached from its blood supply and therefore needs to regain a blood supply from the bed in which it is placed in order to survive. In contrast, a flap is a piece of tissue that is surgically raised and transferred from one location in the body to another while maintaining its blood supply, which enters the base (pedicle) of the flap when it is transplanted. GRAFTS Grafts can be composed of skin, fat, fascia or bone, either separately or together as a composite piece of tissue. Skin grafts can be either full- thickness grafts or split-thickness grafts. Full-thickness grafts consist of the epidermis and the full thickness of the dermis. Split-thickness grafts consist of the epidermis and a variable quantity of the dermis. An essential difference is that the donor site following the harvest of a full- thickness graft has no epidermal elements from which new skin can regenerate. These grafts therefore tend to be taken from sites of the body where the donor defect can be primarily closed. The donor site from split-thickness grafting contains adnexal remnants (hair follicles in par - ticular), which have the propensity to divide and regenerate new epi-dermis, and so resurface the donor defect. Revascularization of grafts is dependent on early and direct connec - tion between host and graft vessels (inosculation), before which graft survival depends on fluid absorption (imbibition). Revascularization, which occurs as early as 48 hours post-transplantation, is by both anas - tomosis, whereby the severed ends of pre-existing graft vessels link up with vessels of the underlying wound bed; and neovascularization, which involves the de novo angiogenic ingrowth of vessels from the wound bed into the graft. During the first 2 weeks, blood vessels from the recipient site invade the graft edges along previous vascular channels in the direction of an ischaemic stimulus, whereas native graft vascula - ture begins to regress. Endothelial progenitor cells appear to play an important role in blood vessel formation, attracted by ischaemic gradi - ents. Inosculation occurs, restoring blood flow to the graft microcircula - tion. By the start of the third week, complete blood flow in the graft vasculature has been established and, in the absence of a continuing ischaemic stimulus, neovascularization ends. FLAPS Flaps are named according to the type of tissue transferred, e.g. a fascio - cutaneous flap contains skin and fascia, and a musculocutaneous flap contains both muscle and the overlying skin, whereas a skin flap, fascial flap and muscle flap contain only the separate elements that their names imply. The blood supply to a skin flap can be randomly orien - tated, which limits the flap length-to-breadth proportions to no more than 2 : 1 (except on the face, where longer flaps can be performed). Much longer skin flaps can be raised elsewhere if the blood supply to the flap is a direct cutaneous artery and vein; these are called axial pattern flaps and are usually based on the local angiosome. An KEY REFERENCES Beck B, Blanpain C 2012 Mechanisms regulating epidermal stem cells. EMBO J 31:2067–75. A detailed review of mechanisms regulating epidermal stem cell renewal and differentiation. Brown SJ, McLean WH 2012 One remarkable molecule: filaggrin. J Invest Dermatol 132:751–62.An overview of the anatomical, biochemical and clinical relevance of the skin barrier protein filaggrin in health, atopy and allergy.Elias PM, Gruber R, Crumrine D et al 2014 Formation and functions of the corneocyte lipid envelope (CLE). Biochim Biophys Acta 1841: 314–18.An update on the structure, composition and functions of the corneocyte lipid envelope with new insights from inherited and acquired disorders of lipid metabolism. Fuchs E 2007 Scratching the surface of skin development. Nature 445: 834–42.
222
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
161 CHAPTER 7key references Kohl E, Steinbauer J, Landthaler M et al 201 1 Skin ageing. J Eur Acad Der - matol Venereol 25:873–84. An account of the key intrinsic and extrinsic factors that contribute to skin ageing. Ladak A, Tubbs RS, Spinner RJ 2014 Mapping sensory nerve communica - tions between peripheral nerve territories. Clin Anat 27:681–90. Lucarz A, Brand G 2007 Current considerations about Merkel cells. Eur J Cell Biol 86:243–51.A review of some of the controversies in Merkel cell biology, ontology and possible functions. Miller M-C, Nanchahal J 2005 Advances in the modulation of cutaneous wound healing and scarring. Biodrugs 19:363–81. An overview of wound healing and scarring mechanisms, and how recombinant growth factors and cytokines might be used therapeutically. Pan X, Hobbs RP, Coulombe PA 2013 The expanding significance of keratin intermediate filaments in normal and diseased epithelia. Curr Op Cell Biol 25:47–56.A comprehensive review of the cell biology of keratins in healthy skin and in diseases such as cancer.A well-illustrated review explaining how epidermal stem cells contribute to hair follicle regeneration and also wound healing. Ghadially R 2012 25 years of epidermal stem cell research. J Invest Dermatol 132:797–810.A concise summary of the recent progress in understanding stem cell niches in the skin and their potential clinical significance. Haines RL, Lane EB 2012 Keratins and disease at a glance. J Cell Sci 125: 3923–8. A compact description of keratin intermediate filament biology and diseases associated with keratin gene mutations. Hearing VJ 201 1 Milestones in melanocytes/melanogenesis. J Invest Derma - tol 131:E1. A short summary of landmarks in melanin biology with links to six other short reviews on key historical discoveries and insights germane to melanocytes in health and disease. Hsu Y-C, Pasolli HA, Fuchs E 201 1 Dynamics between stem cells, niche and progeny in the hair follicle. Cell 144:92–105.A detailed original study that defines the point at which stem cells in the hair follicle become irreversibly committed along a differentiation lineage.
223
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Skin and its appendages 161.e1 CHAPTER 7 REFERENCES Beck B, Blanpain C 2012 Mechanisms regulating epidermal stem cells. EMBO J 31:2067–75. A detailed review of mechanisms regulating epidermal stem cell renewal and differentiation. Brown SJ, McLean WH 2012 One remarkable molecule: filaggrin. J Invest Dermatol 132:751–62. An overview of the anatomical, biochemical and clinical relevance of the skin barrier protein filaggrin in health, atopy and allergy. Elias PM, Gruber R, Crumrine D et al 2014 Formation and functions of the corneocyte lipid envelope (CLE). Biochim Biophys Acta 1841: 314–18. An update on the structure, composition and functions of the corneocyte lipid envelope with new insights from inherited and acquired disorders of lipid metabolism. Fuchs E 2007 Scratching the surface of skin development. Nature 445:834–42.A well-illustrated review explaining how epidermal stem cells contribute to hair follicle regeneration and also wound healing. Ghadially R 2012 25 years of epidermal stem cell research. J Invest Dermatol 132:797–810. A concise summary of the recent progress in understanding stem cell niches in the skin and their potential clinical significance. Haines RL, Lane EB 2012 Keratins and disease at a glance. J Cell Sci 125: 3923–8. A compact description of keratin intermediate filament biology and diseases associated with keratin gene mutations. Hearing VJ 201 1 Milestones in melanocytes/melanogenesis. J Invest Derma - tol 131:E1.A short summary of landmarks in melanin biology with links to six other short reviews on key historical discoveries and insights germane to melanocytes in health and disease. Hsu Y-C, Pasolli HA, Fuchs E 201 1 Dynamics between stem cells, niche and progeny in the hair follicle. Cell 144:92–105.A detailed original study that defines the point at which stem cells in the hair follicle become irreversibly committed along a differentiation lineage.Kohl E, Steinbauer J, Landthaler M et al 201 1 Skin ageing. J Eur Acad Der - matol Venereol 25:873–84.An account of the key intrinsic and extrinsic factors that contribute to skin ageing. Ladak A, Tubbs RS, Spinner RJ 2014 Mapping sensory nerve communica - tions between peripheral nerve territories. Clin Anat 27:681–90. Lucarz A, Brand G 2007 Current considerations about Merkel cells. Eur J Cell Biol 86:243–51.A review of some of the controversies in Merkel cell biology, ontology and possible functions. Miller M-C, Nanchahal J 2005 Advances in the modulation of cutaneous wound healing and scarring. Biodrugs 19:363–81.An overview of wound healing and scarring mechanisms, and how recombinant growth factors and cytokines might be used therapeutically. Pan X, Hobbs RP, Coulombe PA 2013 The expanding significance of keratin intermediate filaments in normal and diseased epithelia. Curr Op Cell Biol 25:47–56.A comprehensive review of the cell biology of keratins in healthy skin and in diseases such as cancer. Rozen WM, Garcia-Tutor E, Alonso-Burgos A et al 2009 The effect of anterior abdominal wall scars on the vascular anatomy of the abdominal wall: a cadaveric and clinical study with clinical implications. Clin Anat 22:815–22. Taylor GI, Gianoutsos MP, Morris SF 1994 The neurovascular territories of the skin and muscles: anatomic study and clinical implications. Plast Reconstr Surg 94:1–36. Taylor GI, Palmer JH 1987 The vascular territories (angiosomes) of the body: experimental study and clinical applications. Br J Plast Surg 40: 1 13–41. Wu X, Hammer JA 2014 Melanosome transfer: it is best to give and receive. Curr Op Cell Biol 29:1–7. Yin Z-X, Peng T-H, Ding H-M 2013 Three dimensional visualization of the cutaneous angiosome using angiography. Clin Anat 26:282–7.
224
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
e10SECTION 1 COMMENTARY Fluorescence microscopy in cell biology today 1.1 (Fig. 1.1.1a). Model organisms, such as nematodes, zebrafish and fruit flies, can be studied in this way. The newest developments in in vivo imaging that are currently transforming the field are based on light sheets. Termed SPIM (selective plane illumination microscopy), a flat sheet of excitation light is projected through the sample and imaged by a perpendicularly arranged lens (Huisken et al 2004) ( Fig. 1.1.1b). This allows high-resolution three-dimensional imaging at high speed and with minimal photo-damage to the sample. In all the above techniques, the resolution of the final fluorescence image is limited by the diffraction of light to be 200–300 nm; two fluo - rescent objects separated by less than this distance would not be distin - guishable as two separate entities. In 2008, the emerging field of super-resolution far-field microscopy or nanoscopy was named Method of the Year by Nature Methods, and in 2014 the Nobel Prize for Chem - istry was awarded for the development of the techniques. By a combina - tion of new fluorophores, optics and image analysis, the diffraction limit was circumvented by three new methodologies. Broadly, these are techniques based on structured illumination microscopy (SIM), in which a grid pattern of excitation light is projected on to the sample (Gustafsson 2000) ( Fig. 1.1.2a); stimulated emission depletion (STED) microscopy, in which a ‘doughnut’-shaped depletion beam is used to de-excite fluorophores and narrow the excitation spot used in confocal microscopy (Hell and Wichmann 1994, Vicidomini et al 201 1) ( Fig. 1.1.2b); or single-molecule localization techniques such as photoacti-vated localization microscopy (PALM), in which individual molecules Fluorescence microscopy is one of the most widely used tools in cell biology today. Its major strength is that it allows the distributions of individual, specific protein species to be mapped at submicron resolu - tion in living cells and even whole organisms. There are several tech - nologies that came together in the 1990s and early years of the twenty-first century to make this possible. These are the development of fluorescent fusion protein constructs based on genetically encoded fluorescent proteins; advances in excitation and detector technology allowing new levels of specificity, speed and signal-to-noise; the devel - opment of multiphoton excitation permitting imaging deep within tissues; and new image analysis and computing techniques for the manipulation and quantification of the resulting data. These advances combined to make fluorescence a standard research tool, found within most laboratories working in the biological sciences. However, a new wave of technological development arrived in the last 5–10 years and is only now beginning to be adopted by the biological sciences com- munity. New fluorescent probes, optical technology and image- processing algorithms have led to the development of super-resolution microscopy in which fluorescence images can be obtained with resolu - tion approaching the molecular scale. In effect, this allows the mapping of all fluorescent molecules in a sample with nanometre accuracy, a possibility that is changing the way we think about what microscope images mean. Fluorescent fusion constructs allow a protein species of interest to be genetically fused to a fluorescent protein so that it can be visualized in a microscope (Chudakov et al 2010). Originally, this meant fusion to the green fluorescent protein (GFP) from the Pacific jellyfish Aequoria victoria (Tsien 1998). For the development of this technology, the 2008 Nobel Prize for Chemistry was shared between Osamu Shinomura, who identified and purified GFP; Martin Chalfie, who created and imaged the first fluorescent fusion (Chalfie et al 1994); and Roger Tsien, who mutated GFP to create a whole palate of possible colours – a palate that is now vast (Shaner et al 2005, Giepmans et al 2006). Unlike immu - nostaining, in which fluorescently tagged antibodies to the protein of interest are introduced into the cell, GFP is genetically encoded and is therefore compatible with live cell imaging, so allowing the investiga - tion of protein dynamics under physiological conditions. The newest fluorescent protein technology includes photoactivatable and colour- switchable proteins, which are useful for tracking intracellular events (Patterson and Lippincott-Schwartz 2002), and as timers (Subach et al 2009) and sensors of various environmental parameters, such as pH (Tantama et al 201 1). On the hardware side, advances in laser technol - ogy have allowed more specific excitation of the sample. Unlike mercury arc lamps, laser light is monochromatic: that is, it contains only a single wavelength. This specificity greatly expanded the scope for multichan - nel imaging and the visualization of several fluorescent species simul - taneously. The latest developments in this field centre on white light lasers and tuneable lasers for more flexible imaging (McConnell 2004). Simultaneously, detectors have become faster and more sensitive. The charge-coupled device (CCD) camera evolved into the electron- multiplying CCD, which was sensitive enough to detect individual photons and therefore image and track single molecules within cells. Currently, compound metal-oxide semiconductor (CMOS) camera technology is transforming microscopy with its very high frame rates and huge fields of view. For point detectors such as those used on con - focal and multiphoton systems, new hybrid detectors are being imple-mented, offering the Holy Grail of extremely high sensitivity and large dynamic range. While GFP advanced imaging into the domain of live cells, multi - photon excitation allowed imaging in living organisms (Denk et al 1990, Helmchen and Denk 2005). Here, two low-energy photons from a pulsed laser combine at the sample to excite the fluorophores, rather than a single high-energy photon as is normally used. Low-energy (red) photons are scattered less by complex biological tissue, which means that high resolution can be maintained at greater imaging depths Fig. 1.1.1 Illustration of in vivo imaging by multiphoton and selective- plane microscopy. A, In multiphoton microscopy, excitation only occurs where the excitation photons are most dense – at the focus. Not only does this generate intrinsic optical sectioning but also the low scattering of long-wavelength (red/infra-red) light means that focus can be deep into tissue. B, Selective-plane illumination creates a flat light sheet projected from the side by a cylindrical lens, with fluorescence collected by a standard objective and imaged using a camera. This allows extremely high-speed three-dimensional imaging. A BIllumination IlluminationSample SampleObjective ObjectiveFluorescence FluorescenceDylan M Owen
225
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Fluorescence microscopy in cell biology today e11 COMMENTAR Y 1.1 Fig. 1.1.2 The main techniques for super-resolution imaging. A, Structured illumination microscopy (SIM) projects a pattern of excitation light (blue lines) on to the sample (green). Multiple acquisitions with different pattern positions and orientations allow a super-resolution image to be reconstructed computationally. B, Stimulated emission depletion (STED) uses a red-wavelength depletion laser beam to cancel out fluorophore excitation at the periphery of a standard confocal excitation spot, leading to a narrower effective excitation area. C, Single-molecule localisation microscopy (SMLM) uses special fluorophores or chemicals to image a sparse set of molecules in any one frame exclusively, leading to individual point spread functions that can be imaged and centroided to find the true position of the molecule. Over many thousands of frames, the positions of all fluorophores in the sample can be mapped. + + =+ ++ = =A CB Fig. 1.1.3 Example data sets acquired with SIM, STED and STORM. A, Microtubules (green) and mitochondria (pink) imaged in U2OS cells using SIM imaging. B, STED image of the nuclear pore protein Nup153 in the nucleus of a fixed PtK2 cell. C, STORM image of actin fibres in Cos7 cells. A from York et al, Resolution doubling in live, multicellular organisms via multifocal structured illumination microscopy. Nature Methods 9(7):749–754 (2012); B from Wurm et al, Novel red fluorophores with superior performance in STED microscopy. Optical Nanoscopy 1(1) (2012); and C from Xu et al, Dual-objective STORM reveals three- dimensional filament organization in the actin cytoskeleton. Nature Methods 9(2):185–188 (2012). A B C 300 µm -200 µmz STED ConfocalRaw data 5 µmNuclear Pore complex/ Abberior STAR635 are imaged and localized in sequence (Betzig and Chichester 1993) (Fig. 1.1.2c). SIM achieves resolutions of around 100 nm and is compatible with conventional fluorophores and live cell imaging. STED typically achieves 50–100 nm in biological samples but is more challenging for live cell imaging because of the damaging laser powers used. PALM and the related technique of direct stochastic optical reconstruction microscopy (dSTORM) (Rust et al 2006, Heilemann et al 2008) deliver the highest resolution of 20–30 nm, but are relatively slow and therefore mainly used on fixed cells where new switchable fluorophores allow fluorescent molecules to be turned on stochastically. Individual molecules are then imaged and their coordinates recorded before the fluorophores are bleached and a new subset of molecules is activated. In this way, all molecules in the sample are imaged in sequence, circumventing the diffraction limit. Using PALM, the ability to acquire tables of the x, y and z coordinates of all individual fluorescent molecules, rather than images per se, requires new ways of thinking about the analysis and quantification of data sets; this challenge is only just beginning. Example images acquired with these three super-resolution methods are shown in Figure 1.1.3. While work on new improved fluorophores, laser technology, optical components and processing algorithms continues, more radical break - throughs in microscopic techniques and data analysis are likely. They will focus on further enhancements to resolution, imaging speeds and applicability to whole-organism imaging: for example, PALM is begin - ning to be applied to live cells when the biological structure is relatively stable (Shroff et al 2008) and the speed of STED has been greatly improved (Chmyrov et al 2013). There is no doubt that these advances will make fluorescence microscopy an even more valuable tool within the biological sciences. REFERENCES Betzig E, Chichester RJ 1993 Single molecules observed by near-field scan - ning optical microscopy. Science 262:1422–5. Chalfie, M, Tu Y, Euskirchen G et al 1994 Green fluorescent protein as a marker for gene expression. Science 263:802–5. Chmyrov A, Keller J, Grotjohann T et al 2013 Nanoscopy with more than 100,000 ‘doughnuts’ . Nat Meth 10:737–40.Chudakov DM, Matz MV, Lukyanov S et al 2010 Fluorescent proteins and their applications in imaging living cells and tissues. Physiol Rev 90:1 103–163. Denk W, Strickler JH, Webb WW 1990 Two-photon laser scanning fluores - cence microscopy. Science 249:72–6.
226
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
FluORESCENCE MICROSCO pY IN CEll b IOlOgY TOdAY e12SECTION 1 Shroff H, Galbraith CG, Galbraith JA et al 2008 Live-cell photoactivated localization microscopy of nanoscale adhesion dynamics. Nat Meth 5: 417–23. Subach FV, Subach OM, Gundorov IS et al 2009 Monomeric fluorescent timers that change color from blue to red report on cellular trafficking. Nat Chem Biol 5:1 18–26. Tantama M, Hung YP, Yellen G 201 1 Imaging intracellular pH in live cells with a genetically encoded red fluorescent protein sensor. J Am Chem Soc 133:10034–7. Tsien RY 1998 The green fluorescent protein. Annu Rev Biochem 67: 509–44. Vicidomini G, Moneron G, Han KY et al 201 1 Sharper low-power STED nanoscopy by time gating. Nat Meth 8:571–3. Wurm C, Kolmakov K, Gottfert F et al 2012 Novel red fluorophores with superior performance in STED microscopy. Opt Nanosc 1:7. Xu K, Babcock HP, Zhuang X 2012 Dual-objective STORM reveals three- dimensional filament organization in the actin cytoskeleton. Nat Meth 9:185–8. York AG, Parekh SH, Nogare DD et al 2012 Resolution doubling in live, multicellular organisms via multifocal structured illumination micros - copy. Nat Meth 9:749–54.Giepmans BNG, Adams SR, Ellisman MH et al 2006 The fluorescent toolbox for assessing protein location and function. Science 312:217–24. Gustafsson MGL 2000 Surpassing the lateral resolution limit by a factor of two using structured illumination microscopy. J Microsc 198:82–7. Heilemann M, van de Linde S, Schüttpelz M et al 2008 Subdiffraction- resolution fluorescence imaging with conventional fluorescent probes. Angew Chem Int Ed 47:6172–6. Hell SW, Wichmann J 1994 Breaking the diffraction resolution limit by stimulated emission: stimulated-emission-depletion fluorescence microscopy. Opt Lett 19:780–2. Helmchen F, Denk W 2005 Deep tissue two-photon microscopy. Nat Meth 2:932–40. Huisken J, Swoger J, Del Bene F et al 2004 Optical sectioning deep inside live embryos by selective plane illumination microscopy. Science 305: 1007–9. McConnell G 2004 Confocal laser scanning fluorescence microscopy with a visible continuum source. Opt Express 12:2844–50. Patterson GH, Lippincott-Schwartz J 2002 A photoactivatable GFP for selec - tive photolabeling of proteins and cells. Science 297:1873–7. Rust MJ, Bates M, Zhuang X 2006 Sub-diffraction-limit imaging by stochastic optical reconstruction microscopy (STORM). Nat Meth 3:793–6. Shaner NC, Steinbach PA, Tsien RY 2005 A guide to choosing fluorescent proteins. Nat Meth 2:905–9.
227
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
e13 COMMENTAR Y 1.2 COMMENTARY Stem cells in regenerative medicine 1.2 Introduction There is a great deal of hype, hope and optimism surrounding the use of stem cells in regenerative medicine. Historically and in recent years, stem cell successes have attracted widespread media attention, which has further fuelled the public perception of stem cells and what they are capable of achieving. Although significant breakthroughs have been made in the past decade in stem cell research and the resulting scientific output has increased exponentially, so far, only a small proportion of this research has been successfully translated into the clinical arena. Scientific effort is therefore now focusing on translating this research from ‘bench’ to ‘bedside’ and on testing the clinical efficacy and safety of stem cell therapies through clinical trials. (For further reading, see Dimmeler et al (2014).) What are stem cells? Stem cells may be defined as cells that exhibit properties of multilineage differentiation and self-renewal (Thomson et al 1998). The term ‘mul - tilineage differentiation’ means that cells have the potential to differen - tiate into any of the three embryonic germ layers – ectoderm, mesoderm or endoderm. By self-renewing, stem cells are able to generate further stem cells, thereby propagating themselves. Types of stem cell Stem cells can be broadly categorized into embryonic or adult stem cells. The different types of stem cell are depicted in Figure 1.2.1. From an immunological perspective, stem cells can also be syngeneic (from identical twins), autologous (from the same individual), allogeneic (from a different member of the same species) or xenogeneic (from a different species altogether). Syngeneic and autologous cells have obvious advantages since they are unlikely to be rejected following transplantation. The advantages and disadvantages of the different types of stem cell are outlined in Table 1.2.1. Embryonic stem cells (ESCs) Embryonic stem cells are the archetypal pluripotent stem cells, derived from the embryonic inner cell mass of the blastocyst and capable of differentiating into any cell type (Thomson et al 1998). However, current limitations on using ESCs include immunological rejection, safety concerns about the formation of tumours (teratomas) and ethical dilemmas concerning the utilization of cells derived from aborted fetuses. A recent trial of ESCs in spinal cord injury patients has been halted due to financial constraints, although clinical trials are currently under way in the UK and USA within the field of retinal research (Schwartz et al 2012, Watts 201 1). Evidence suggests that human ESCs may be generated through somatic cell nuclear transfer (cloning) tech - niques, a challenge previously believed to be insurmountable (Tachibana et al 2013). This has the potential of generating patient-specific (matched) ESCs in future that will not be rejected by the patient’s immune system. Preliminary data suggest that reprogramming by nuclear transfer may be slightly more effective than reprogramming by transcription factors (Krupalnik and Hanna 2014, Ma et al 2014). Amniotic fluid stem cells (AFSCs) The amniotic fluid that surrounds the developing fetus contains a rich stem cell population that was first discovered in 2007 (De Coppi et al 2007). Such cells are c-kit+ (CD1 17+) and fulfil the criteria of true stem cells, in that they are pluripotent and exhibit self-renewal. Although not in clinical trials as yet, these cells offer the prospect of correcting fetal defects either in utero or at the time of birth.Fig. 1.2.1 Different types of stem cell. Mesenchymal stem cellsEmbryonic stem cells Induced pluripotent stem cellsKIf4 Oct 3/4 c-Myc Sox2Amniotic fluid stem cells Umbilical cord stem cellsJonathan M Fishman, Paolo De Coppi, Martin A Birchall Umbilical cord stem cells Umbilical cord blood is a potential source of stem cells that may be used to treat a variety of different diseases, including haemopoietic and genetic diseases. Cord blood stem cells display embryonic stem cell markers but are negative for blood cell lineage markers. The main advantages they offer are ease of procurement with minimal risk to the donor; ease of cryopreservation and banking for future use; and minimal ethical concerns. Adult stem cells Mesenchymal stem cells (MSCs) Mesenchymal stem cells (MSCs) are broadly multipotent stem cells that are capable of differentiating into a variety of different tissue types, being fairly restricted towards differentiation along the mesodermal lineage. Their potential to differentiate into cartilage, bone or adipose tissue depends on the ability to create the appropriate microenviron-ment in which this might occur, a goal that continues to be the focus of intense study. MSCs are often derived from bone marrow but can also be enriched from a variety of other sources, including adipose tissue, synovium, skeletal muscle and placental tissues. They have the advantage of generating large numbers of cells and their application in
228
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
STEM CEllS iN REgENERAT ivE MEdiCiNE e14SECT iON 1 Fig. 1.2.2 Tissue-engineering an organ for transplantation. ImplantationStem cellsScaffold (may be natural or synthetic) Bioreactor Brain Ear Temporomandibular joint Heart valveAorta Pulmonary artery Heart muscleSpinal cord within vertebral canal Skin Finger KneeUrethraBladderUreterIntestinePancreasLiverChest wallTracheaJawSkull Eye vivo should encounter few ethical concerns when compared to the use of ESCs, given that their utilization is comparable to bone marrow transplantation, which is already in widespread clinical use. MSCs are currently in numerous clinical trials for a variety of different diseases, including the treatment of myocardial infarction, where initial results have been promising. induced pluripotent stem cells (iPSCs) Induced pluripotent stem cells are a relatively novel type of stem cell. Following their discovery in 2006 (Takahashi and Yamanaka 2006), Yamanaka was awarded the 2012 Nobel Prize in Physiology or Medi - cine. If several critical transcription factors are introduced into a cell – Oct3/4, Sox2, Klf4, c-Myc (so-called ‘Yamanaka’ or ‘stemness’ factors), any cell (e.g. a fibroblast) can be reverted to a pluripotent state. More recently, it has been demonstrated that such cells may be reprogrammed to totipotency (ability of a cell to differentiate into any cell type, includ - ing the extra-embryonic membranes and tissues) (Abad et al 2013). Limitations of the iPSC approach include the low efficiency of the process, safety concerns around viral transduction, immunogenicity of the iPSCs and the risk of tumour formation. Current research is aimed at generating iPSCs without the use of viruses, through the employment of drug molecules (Hou et al 2013). In addition, contrary to earlier opinion, it appears that iPSCs, being an autologous cell source, are indeed non-immunogenic (Araki et al 2013). Phase I/II clinical trials involving iPSCs are eagerly awaited to determine their overall efficacy and safety in disease states. Somatic cell reprogramming – stimulus- triggered acquisition of pluripotency (STAP): fact or fantasy? Two papers in Nature in 2014 reported that differentiated mouse somatic cells were able to revert to a pluripotent, or possibly even a totipotent, phenotype after transient exposure to low pH. The repro - gramming did not require either nuclear transfer or genetic manipula - tion and was remarkably rapid, unlike the time taken to prepare iPSCs (Obokata et al 2014a, Obokata et al 2014b). The reprogrammed cells were named STAP (stimulus-triggered acquisition of pluripotency) cells. Not surprisingly, the papers were headline news around the world. However, what seems to be too good to be true usually is just that; within weeks of publication, the methodology and the nature of the cells used in the studies were called into serious question and both papers were subsequently retracted ( Nature editorial 2014).Regenerative medicine through tissue engineering Tissue engineering is an interdisciplinary field that applies the princi - ples and methods of engineering and the life sciences towards the development of biological substitutes that can restore, maintain or improve tissue function (Langer and Vacanti 1993). The main approach to tissue engineering includes using the various types of stem cell mentioned above and seeding them, either on or within scaffolds, to create ‘off-the-shelf’ organs and tissues for transplantation ( Fig. 1.2.2). Seeding of stem cells on to scaffolds may be undertaken either in vitro Table 1.2.1 Comparison of the different types of stem cell Advantages Disadvantages Embryonic stem cellsPluripotency Fear of immunological rejection Safety concerns – tumour (teratoma) formation Ethical dilemmas surrounding aborted fetuses Oocytes required Amniotic fluid stem cellsPluripotencyNon-tumorigenicCan be harvested (through amniocentesis) and manipulated prenatally so that defect can be corrected either in utero or at the time of birthLow yield – 1% of amniotic fluid cells are c-kit + (CD117+) stem cells Further research concerning origin of cells and characterization required Umbilical cord stem cellsEase of procurement with minimal donor morbidity Ease of cryopreservation and banking for future use Minimal ethical concernsLow yield of stem cells and only finite number of cells available from donor Problems of storage – long-term storage may affect cell quality; cost implications; quality control issues Mesenchymal stem cellsMultipotentLarge numbers can be harvestedCan be enriched from a variety of different tissues including bone marrow, adipose tissue, etc Easily expanded in tissue culture for tissue engineering purposes Autologous and possess immunomodulatory properties Appear safe in clinical trialsMinimal ethical concernsDifferentiation dependent on an appropriate microenvironment Optimal mode of delivery unclearLimited long-term therapeutic potential Induced pluripotent stem cellsPluripotentCan be derived from any cell typeProbably non-immunogenicLow efficiency of reprogramming with current techniques (<1%) Safety concerns – viral vector transduction commonly required (risks of viral infection and genetic manipulation) and risk of tumours
229
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Stem cells in regenerative medicine e15 COMMENTAR Y 1.2 in a bioreactor, or in vivo using the host as a ‘living’ bioreactor. Tissue engineering has the potential to overcome existing organ shortages and generate organs and tissues that are not rejected by the patient’s immune system (Fishman et al 2013). The trachea was the first stem-cell- based, tissue-engineered organ successfully transplanted into humans (Fishman et al 201 1) ( Box 1.2.1). Using stem cells to replace organs and tissues through blastocyst complementation A complementary approach to organ and tissue replacement using interspecific blastocyst complementation has been reported recently. In this technique, iPSCs home to developmentally deficient niches, where they regenerate tissues of the donor cell species (Kobayashi et al 2010). Using two transgenic pig lines, normal fluorescently-labelled pig pluripotent stem cells have been transplanted into genetically altered pigs lacking pancreata, resulting in the development of chimeric pigs Box 1.2.1 Clinical highlights Several patients have now received a new trachea using tissue- engineering techniques. The patient’s own (autologous) stem cells were harvested from the bone marrow and respiratory epithelium, and seeded on to either decellularized (Elliott et al 2012, Macchiarini et al 2008) or synthetic (Jungebluth et al 2011) scaffolds before being transplanted back into the patient. Early results have been encouraging and clinical trials of full-scale partial laryngotracheal implants are anticipated in the near future.Fig. 1.2.3 Blastocyst complementation to generate human organs for transplantation. iPSCs Injection of human iPSCs into blastocysts genetically modified to lack specific organs Generation of a human organin livestock animals Organ transplantation Patient with fluorescent orange pancreata derived from donor cells (Matsunari et al 2013). This is an important step towards the generation of human organs in large animals by complementing pig embryos with human iPSCs (Fig. 1.2.3). The clinical possibilities of using blastocyst comple - mentation as a means of generating transplantable human organs may offer one solution to the current shortfall in organs for transplantation: by growing human organs in livestock animals such as pigs, an unlim - ited supply of immunologically matched human organs might be made available for transplantation. Concerns have been raised over possible immunological rejection (targeting porcine-derived vasculature sur- rounding the new organ), and ethical and safety issues associated with injecting human iPSCs, capable of differentiating into any cell and/or tissue type, into pigs (and so theoretically generating a new human–pig hybrid species, especially if such cells home to the brain or the germline). Conclusion At present, the jury is still out as to which stem cell is most effective and safest for tissue regeneration. For this reason, all of the above stem cells are being explored as viable options for future therapies without relying on a single stem cell at the present time. iPSCs probably offer most potential for the future, although it may be that different stem cells perform differently under different conditions (e.g. some stem cells may be better at generating one tissue than another). In addition, for tissue engineering purposes, a stem cell is required that can generate large numbers of cells in a relatively short period of time. Either way, well-conducted clinical trials will be required to determine the stem cell’s efficacy, safety and cell fate prior to its widespread use in medicine. REFERENCES Abad M, Mosteiro L, Pantoja C et al 2013 Reprogramming in vivo pro - duces teratomas and iPS cells with totipotency features. Nature 502: 340–5. Araki R, Uda M, Hoki Y et al 2013 Negligible immunogenicity of terminally differentiated cells derived from induced pluripotent or embryonic stem cells. Nature 494:100–4.De Coppi P, Bartsch G Jr, Siddiqui M et al 2007 Isolation of amniotic stem cell lines with potential for therapy. Nat Biotechnol 25:100–6. Dimmeler S, Ding S, Rando TA et al 2014 Translational strategies and chal - lenges in regenerative medicine. Nature Med 20:814–21.
230
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
STEM CEllS iN REgENERAT ivE MEdiCiNE e16SECT iON 1 Matsunari H, Nagashima H, Watanabe M et al 2013 Blastocyst complemen - tation generates exogenic pancreas in vivo in apancreatic cloned pigs. Proc Natl Acad Sci U S A 1 10:4557–62. Nature editorial. 2014 STAP retracted. Nature 51 1:5–6. Obokata H, Sasai Y, Niwa H et al 2014a Bidirectional developmental poten - tial in reprogrammed cells with acquired pluripotency. Nature 505: 676–80. Obokata H, Wakayama T, Sasai Y et al 2014b Stimulus-triggered fate conver - sion of somatic cells into pluripotency. Nature 505:641–7. Schwartz SD, Hubschman JP, Heilwell G et al 2012 Embryonic stem cell trials for macular degeneration: a preliminary report. Lancet 379: 713–20. Tachibana M, Amato P, Sparman M et al 2013 Human embryonic stem cells derived by somatic cell nuclear transfer. Cell 153:1228–38. Takahashi K, Yamanaka S 2006 Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126:663–76. Thomson JA, Itskovitz-Eldor J, Shapiro SS et al 1998 Embryonic stem cell lines derived from human blastocysts. Science 282:1 145–7. Watts G 201 1 Moorfields Eye Hospital is to host European trial of human embryonic stem cells. BMJ 343:d6124.Elliott MJ, De Coppi P, Speggiorin S et al 2012 Stem-cell-based, tissue engi - neered tracheal replacement in a child: a 2-year follow-up study. Lancet 380:994–1000. Fishman JM, De Coppi P, Elliott MJ et al 201 1 Airway tissue engineering. Expert Opin Biol Ther 1 1:1623–35. Fishman JM, Lowdell MW, Urbani L et al 2013 Immunomodulatory effect of a decellularized skeletal muscle scaffold in a discordant xenotrans - plantation model. Proc Natl Acad Sci U S A 1 10:14360–5. Hou P, Li Y, Zhang X et al 2013 Pluripotent stem cells induced from mouse somatic cells by small-molecule compounds. Science 341:651–4. Jungebluth P, Alici E, Baiguera S et al 201 1 Tracheobronchial transplantation with a stem-cell-seeded bioartificial nanocomposite: a proof-of-concept study. Lancet 378:1997–2004. Kobayashi T, Yamaguchi T, Hamanaka S et al 2010 Generation of rat pan - creas in mouse by interspecific blastocyst injection of pluripotent stem cells. Cell 142:787–99. Krupalnik V, Hanna JH 2014 Stem cells: the quest for the perfect repro - grammed cell. Nature 51 1:160–2. Langer R, Vacanti JP 1993 Tissue engineering. Science 260:920–6. Ma H, Morey R, O’Neil RC et al 2014 Abnormalities in human pluripotent cells due to reprogramming mechanisms. Nature 51 1:177–83. Macchiarini P, Jungebluth P, Go T et al 2008 Clinical transplantation of a tissue-engineered airway. Lancet 372:2023–30.
231
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
e17 COMMENTAR Y 1.3 COMMENTARY Merkel cells 1.3 The epidermis, our outermost layer of skin, serves as a protective barrier and sensory interface with the environment. In vertebrates, these func - tions are accomplished through the work of only four conserved cell types. Keratinocytes form a water-tight barrier, melanocytes protect against ultraviolet damage and Langerhans cells perform immune sur - veillance. Merkel cells, which constitute less than 1% of the epidermis, are an enigmatic cell type whose function and developmental origin have long been debated; however, recent advances show that mamma - lian Merkel cells are epithelial derivatives that collaborate with sensory neurones to initiate touch sensation. As early as the nineteenth century, Merkel cells were proposed to contribute to skin’s sensory function. These cells were first described by Friedrich Sigmund Merkel, who named them touch cells ( Tastzellen), based on contacts with nerve terminals (Merkel 1875). Merkel cells are found in highly touch-sensitive skin areas, including glabrous (hairless) skin of fingertips and lips, some hair follicles, and touch domes on the trunk (Fig. 1.3.1). Most Merkel cells are innervated by a particular class of myelinated (Aβ) sensory neurones called slowly adapting type I (SAI) afferents (Iggo and Muir 1969). These touch receptors are unusual because they produce biphasic discharges of action potentials that send two types of information to the brain (see Fig. 1.3.1): dynamic responses to moving stimuli represent an object’s spatial features (e.g. edges and curvature), whereas sustained action-potential trains signify steady pres - sure (e.g. contact with clothing; Johnson 2001). Merkel cells were initially proposed to serve as sensory receptor cells based on their synapse-like contacts with nerve terminals and dense- core vesicles filled with neurotransmitters (Halata et al 2003). Sensory receptor cells, such as taste-bud cells or hair cells of the inner ear, trans - duce the energy of a sensory stimulus into electrical signals, which then trigger neurotransmitter release to excite sensory afferents. The hypo - thesis that Merkel cells are mechanosensory receptors was challenged by neurophysiological studies that reported touch sensitivity after Merkel-cell loss, which suggested that Merkel cells act in a non-sensory capacity. Recent studies in rodent models have reconciled these views by demonstrating that both Merkel cells and their sensory afferents are mechanosensory cells. Three lines of evidence support this two-receptor- site model, which was proposed more than two decades ago (Yamashita and Ogawa 1991). First, Merkel cells are touch-sensitive. Like inner-ear hair cells, Merkel cells display fast mechanotransduction currents that convert cellular displacements into electrical signals (Ikeda et al 2014, Maksimovic et al 2014, Woo et al 2014). Piezo2, a large, mechanically activated mem - brane protein, is required for mechanotransduction currents in Merkel cells. Thus, Piezo2 is a leading candidate gene to encode sensory trans - duction channels for touch (see Fig. 1.3.1). Second, Merkel cells are excitatory cells. This was established with a state-of-the-art technique called optogenetics, which uses light-gated molecules derived from Archea to activate or inhibit electrical signalling in specific cells. Transgenic mice were engineered to selectively express light-gated membrane proteins in Merkel cells (Maksimovic et al 2014). When the skin is illuminated to activate Merkel cells in these mice, their sensory afferents produce sustained volleys of action potentials that mimic the response to static pressure (see Fig. 1.3.1). Conversely, neu- ronal firing is inhibited when Merkel cells are optogenetically silenced. These results confirm that Merkel cells are sensory cells capable of excit - ing afferents. Third, Merkel cells are necessary for sustained, but not for dynamic, firing in SAI afferents. Prototypical SAI responses are abolished in trans - genic mouse strains that lack either Merkel cells or epidermal Piezo2 expression (Maricich et al 2009, Maksimovic et al 2014, Woo et al 2014). Instead, sustained touch produces only transient bursts of action potentials in these mice (see Fig. 1.3.1). This finding indicates that Merkel cells are essential for conveying information about static pres - sure to the nervous system. Importantly, they also show that sensory afferents are independently capable of transducing dynamic touch, Fig. 1.3.1 The Merkel cell–neurite complex is a compound touch receptor. The schematic shows the innervation of a touch dome. Merkel cells (blue) cluster in these highly touch-sensitive spots in humans and other mammals. Merkel cells are mechanosensory receptor cells derived from epidermal progenitors. Adjacent Merkel cells are innervated by a myelinated sensory afferent (yellow) that produces slowly adapting type I (SAI) responses. This response has two phases: (1) Dynamic firing. During a moving stimulus (blue bar), mechanotransduction channels located in the SAI afferent open to excite a transient neuronal volley (blue action potentials). (2) Sustained firing. Static pressure (green bar) activates the Merkel cells’ Piezo2-dependent mechanotransduction channels to excite electrical signalling. Merkel cells then signal to SAI afferent terminals to produce sustained action-potential firing (green). Although numerous neuroactive molecules localize to dense-core vesicles, the neurotransmitters and receptors (orange) that mediate excitatory signalling have not been identified. (Receptor and channel symbols © motifolio.com.) Dermis Epidermis Myelin sheathSAI afferent SAI response Dynamic firing Sustained firingMerkel cell Merkel-cell progenitor Keratinocytes Mechanotransduction channel Dense-core vesicles Neurotransmitter receptor Touch Action potential trainsStatic pressureMoving stimulusEllen A Lumpkin
232
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
MERkEl CElls e18sECTION 1 serve additional, non-sensory functions in skin? This is particularly important in human skin, where a substantial fraction of Merkel cells appear to lack innervation. Are human Merkel cells derived from epi - dermis and renewed throughout life? If so, how are these processes disrupted during ageing, which is accompanied by loss of touch recep - tors in extremities? This condition is proposed to contribute to the decline of tactile acuity, grip strength and postural stability in the elderly. Mechanisms of Merkel-cell specification may also be important in a very different human pathology – Merkel cell carcinoma (MCC). This highly aggressive, non-melanoma skin cancer is linked to Merkel cells based on ultrastructure, neurosecretory markers and keratin-20 immu- noreactivity (Bhatia et al 201 1). Recent genomic profiling studies confirm that human MCC cells express many of the same genes as mouse Merkel cells (Haeberle et al 2004, Harms et al 2013). An impor - tant unresolved question is whether Merkel cells or their epidermal progenitors act as cells of origin for MCC. Immune suppression is a probable risk factor for MCC because it is prevalent on sun-exposed skin and in immunocompromised patients. Intriguingly, MCC has been linked to a newly described virus, Merkel cell polyomavirus (Feng et al 2008), raising the possibility that viral infection plays a role in MCC pathogenesis. Active areas of investigation include identifying a causal link between MCC and polyoma viral infection, and defining mecha - nisms that underlie the aggressiveness of MCC. Acknowledgements Thanks to members of the Lumpkin laboratory for helpful comments and Ms Blair Jenkins for assistance with the figure. The author is sup- ported by the National Institutes of Health (R01 AR051219 and R01 NS0731 19).albeit with reduced activity. Finally, touch-driven behaviours are com - promised in rodents lacking functional Merkel cells, which indicates that Merkel cells are important for the perception of touch (Maricich et al 2012, Ikeda et al 2014, Woo et al 2014). Together, these studies demonstrate that the Merkel cell–neurite complex is a compound touch receptor with two mechanosensory cell types arranged in series (see Fig. 1.3.1). A key open question is: which neurotransmitters convey information between Merkel cells and sensory afferents? Transgenic mouse studies have also clarified the Merkel cell’s onto - geny. Although embedded in a stratified epithelium, Merkel cells express simple epithelial keratins (e.g. keratin-8, keratin-20) and produce dozens of neurochemical markers (Moll et al 1984, Halata et al 2003, Haeberle et al 2004). Due to this unusual assortment of epithelial and neuronal markers, it was unclear whether Merkel cells derive from epithelial precursors or neural crest. These competing models were tested in transgenic mice by genetically marking specific cell lineages based on selective expression of Cre recombinase. Wnt1 Cre, which marks all neural crest cells, failed to label Merkel cells; however, Krt14Cre, a marker of keratinocyte-derived cells, labelled Merkel cells in all skin areas (Morrison et al 2009, Van Keymeulen et al 2009). Moreover, Merkel cells are replenished from epidermal progenitors in mature skin (Van Keymeulen et al 2009, Woo et al 2010, Doucet et al 2013). Finally, Merkel cells were abolished by conditional deletion of a developmental transcription factor, Atoh1, driven by Krt14 Cre but not Wnt1Cre (Morrison et al 2009, Van Keymeulen et al 2009). Together, these studies provide a starting point to define transcriptional networks that bestow a neuro - sensory fate in epidermal cells (Bardot et al 2013, Lesko et al 2013). Despite remarkable progress in Merkel-cell biology, important ques- tions about human Merkel cells remain unanswered. Do Merkel cells REFERENCES Bardot ES, Valdes VJ, Zhang J et al 2013 Polycomb subunits Ezh1 and Ezh2 regulate the Merkel cell differentiation program in skin stem cells. EMBO J 32:1990–2000. Bhatia S, Afanasiev O, Nghiem P 201 1 Immunobiology of Merkel cell carci - noma: implications for immunotherapy of a polyomavirus-associated cancer. Curr Oncol Rep 13:488–97. Doucet YS, Woo SH, Ruiz ME et al 2013 The touch dome defines an epider - mal niche specialized for mechanosensory signaling. Cell Rep 3: 1759–65. Feng H, Shuda M, Chang Y et al 2008 Clonal integration of a polyomavirus in human Merkel cell carcinoma. Science 319:1096–100. Haeberle H, Fujiwara M, Chuang J et al 2004 Molecular profiling reveals synaptic release machinery in Merkel cells. Proc Natl Acad Sci U S A 101:14503–8. Halata Z, Grim M, Bauman KI 2003 Friedrich Sigmund Merkel and his ‘Merkel cell’, morphology, development, and physiology: review and new results. Anat Rec 271A:225–39. Harms PW, Patel RM, Verhaegen ME et al 2013 Distinct gene expression profiles of viral- and nonviral-associated Merkel cell carcinoma revealed by transcriptome analysis. J Invest Dermatol 133:936–45. Iggo A, Muir AR 1969 The structure and function of a slowly adapting touch corpuscle in hairy skin. J Physiol 200:763–96. Ikeda R, Cha M, Ling J et al 2014 Merkel cells transduce and encode tactile stimuli to drive A β-afferent impulses. Cell 157:664–75. Johnson KO 2001 The roles and functions of cutaneous mechanoreceptors. Curr Opin Neurobiol 1 1:455–61.Lesko MH, Driskell RR, Kretzschmar K et al 2013 Sox2 modulates the func - tion of two distinct cell lineages in mouse skin. Dev Biol 382:15–26. Maksimovic S, Nakatani M, Baba Y et al 2014 Epidermal Merkel cells are mechanosensory cells that tune mammalian touch receptors. Nature 509:617–21. Maricich SM, Morrison KM, Mathes EL et al 2012 Rodents rely on Merkel cells for texture discrimination tasks. J Neurosci 32:3296–300. Maricich SM, Wellnitz SA, Nelson AM et al 2009 Merkel cells are essential for light-touch responses. Science 324:1580–2. Merkel F 1875 Tastzellen und Tastkörperchen bei den Hausthieren und beim Menschen. Archiv f mikrosk Anat 1 1:636–52. Moll R, Moll I, Franke WW 1984 Identification of Merkel cells in human skin by specific cytokeratin antibodies: changes of cell density and distribu - tion in fetal and adult plantar epidermis. Differentiation 28:136–54. Morrison KM, Miesegaes GR, Lumpkin EA et al 2009 Mammalian Merkel cells are descended from the epidermal lineage. Dev Biol 336:76–83. Van Keymeulen A, Mascre G, Youseff KK et al 2009 Epidermal progenitors give rise to Merkel cells during embryonic development and adult homeostasis. J Cell Biol 187:91–100. Woo SH, Ranade S, Weyer AD et al 2014 Piezo2 is required for Merkel-cell mechanotransduction. Nature 509:622–6. Woo SH, Stumpfova M, Jensen UB et al 2010 Identification of epidermal progenitors for the Merkel cell lineage. Development 137:3965–71. Yamashita Y, Ogawa H 1991 Slowly adapting cutaneous mechanoreceptor afferent units associated with Merkel cells in frogs and effects of direct currents. Somatosens Motor Res 8:87–95.
233
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
e19 COMMENTAR Y 1.4 COMMENTARY Metaplasia 1.4 Metaplasia is the appearance, in adult life, of a patch of tissue that normally belongs elsewhere in the body (Slack 1985, Slack 2007). Generally, it is understood that the metaplasia originated in situ and has not migrated from elsewhere. Therefore, metastatic tumour depos - its, for example, are not considered to be metaplasias, and neither are tissues misplaced during embryonic development, such as thyroid resi - dues in the thyroglossal duct. Metaplasias are very diverse (Willis 1962). They include ectopic bone and cartilage arising from connective tissue, squamous metaplasia of glandular epithelia, and substitution of one glandular tissue for another. The latter group is of particular interest from a biological point of view. These metaplastic transformations are normally between tissue types that originated in embryonic development as neighbours in a common cell sheet. They may remain neighbours into adult life or may become separated. For example, intestinal metaplasia of the stomach is a well- studied type of metaplasia in which patches of small intestine-like epithelium are found in the stomach (Stemmermann and Hayashi 1968, Gutierrez-Gonzalez and Wright 2008). Intestinal metaplasia found in the urinary bladder contains cell types and markers normally found in the colon (Sung et al 2006). The bladder is not part of a common cell sheet with the intestine in the adult, but in the embryo it is, originating from the allantoic diverticulum of the hindgut. Glandular metaplasias are particularly common in the alimentary canal and in the female reproductive tract (Kurita 201 1), perhaps because these are both examples of structures in which a number of different tissue types arise from a simple epithelial tube in the embryo. The clinical importance of metaplasias stems from the fact that they often predispose to cancer. For example, carcinomas of the bronchus usually arise within areas of squamous metaplasia (Auerbach et al 1961) and carcinomas of the stomach usually arise in regions of intes - tinal metaplasia (Busuttil and Boussioutas 2009). Metaplasias arise because the combination of gene activities that defines the tissue type in embryonic development becomes altered in adult life. The genes in question encode a limited number of key tran - scription factors. For example, the transcription factor p63 is essential for the formation of squamous epithelia (Koster et al 2004) and is always found to be expressed in patches of squamous metaplasia. In experiments with mice, ectopic expression of the Cdx2 gene, normally necessary for intestinal development, leads to patches of intestinal tissue in the stomach (Silberg et al 2002, Mutoh et al 2002). Conversely, dele - tion of Cdx2 from the intestine leads to oesophagus-like or stomach-like patches in the intestine (Gao et al 2009, Stringer et al 2012). In tissues that are maintained by continuous renewal from a population of stem cells, it is the stem cells whose character becomes switched in an initiat - ing event. Once the change has occurred, it is irreversible. A well-studied example of metaplasia is Barrett’s metaplasia, commonly known as Barrett’s oesophagus (Falk 2002, Gilbert et al 201 1) (Fig. 1.4.1). This is characterized by the presence of columnar SSQE BMSSQE BMBM SSQEZ-lineZ-line BM SSQE D EA B C Fig. 1.4.1 Barrett’s metaplasia. A, Normal endoscopic appearance showing the junction between gastric columnar and oesophageal squamous epithelium (Z-line). B–C, Endoscopies showing two cases of Barrett’s oesophagus. The Z-line has moved proximally and is very irregular. There are islands of squamous tissue within the columnar region. D, Histological view, haematoxylin and eosin. E, Histological view, immunostained for CDX2 (brown). Scale bar 200 microns. Abbreviations: BM, metaplastic epithelium (Barrett’s metaplasia); SSQE, stratified squamous epithelium. Jonathan MW Slack, Leonard P Griffiths, David Tosh
234
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
METAplAsiA e20sECTiON 1 epithelium, the tissue also contains mucous glands connected by ducts to the lumen, which are located in the submucosa but originate from the original embryonic oesophageal epithelium. Barrett’s metaplasia probably arises by transformation of the basal cells of the squamous epithelium, or from the mucous glands, or both (Nicholson et al 2012, Leedham et al 2008). The predisposing cause of Barrett’s metaplasia is gastro-oesophageal reflux, in which stomach acid, and bile from the intestine, enter the oesophagus (Souza 2010). It is known that this environment can induce in oesophageal cells expression of some of the key genes involved in intestinal development, such as those encoding CDX factors and hepatocyte nuclear factors (Eda et al 2003, Debruyne et al 2006, Piessen et al 2007). Obesity is a well-known risk factor for reflux, partly because of effects on intra-abdominal pressure and probably also through the effects of cytokines released from fat (Corley et al 2007). Barrett’s meta - plasia may progress to dysplasia and to adenocarcinoma. Progression is associated with loss of the tumour suppressor gene CDKN2A, encod- ing the protein p16, which is an inhibitor of cell division; and TP53, encoding the protein p53, required for death of abnormal cells (Spechler et al 2010, Chen et al 201 1). Studies of the clonal structure of Barrett’s metaplasia have shown that is it complex, containing cell populations derived from several original progenitors, with various combinations of abnormalities arising from somatic mutation (Leedham et al 2008). Gastro-oesophageal reflux is normally treated with proton pump inhibitors that reduce the levels of stomach acid. This improves associ - ated oesophagitis but does not usually result in regression of any Bar - rett’s metaplasia. Barrett’s metaplasia as such is not normally treated but, if it progresses to dysplasia, it may be extirpated by a variety of endoscopic techniques (Lim and Fitzgerald 2013).epithelium in the distal oesophagus, of which some resembles intesti - nal epithelium possessing goblet cells. Barrett’s metaplasia is important because it is a risk factor for the development of oesophageal adeno - carcinoma, a cancer whose incidence rose considerably in the late twen - tieth century (Hvid-Jensen et al 201 1). In less developed countries, most oesophageal cancers are squamous cell carcinomas, arising from the normal squamous lining of the oesophagus but, in the Western world, adenocarcinomas are now more common. At least some metaplasias are monoclonal, i.e. a particular focus consists of tissue all derived from a single cell. This is the case for intes- tinal metaplasia of the stomach, where neighbouring metaplastic crypts arise by fission (Gutierrez-Gonzalez et al 201 1). In Barrett’s metaplasia, patches of crypts can also be monoclonal (Nicholson 2012), although the lesion as a whole is polyclonal. A key property of metaplastic foci is that the new tissue is in a situation of competitive growth with the old tissue that surrounds it. Unless the metaplastic tissue can outgrow the surrounding tissue, then the transformation of a single stem cell would not give rise to a macroscopic lesion. However, this aspect of epithelial biology remains poorly understood. The exact cell of origin of Barrett’s metaplasia is still not known (Barbera and Fitzgerald 2010). The columnar lined epithelium of Bar - rett’s metaplasia is in continuation with the columnar lined gastric mucosa and appears as a proximal displacement of the junction with the stratified squamous epithelium (the Z-line). However, it is generally thought not to be a phenomenon of cell migration alone, as animal experiments indicate that it is possible to elicit patches of metaplasia separated from the gastric epithelium (Gillen et al 1988, Goldstein et al 1997). The normal oesophagus originates from an endodermal tube lined with columnar epithelium, which becomes squamous between about 6–7 weeks of gestation. Although lined with squamous REFERENCES Auerbach O, Stout AP, Hammond EC et al 1961 Changes in bronchial epi - thelium in relation to cigarette smoking and cancer of the lung. N Engl J Med 265:253–67. Barbera M, Fitzgerald RC 2010 Cellular origin of Barrett’s metaplasia and oesophageal stem cells. Biochem Soc Trans 38:370–3. Busuttil RA, Boussioutas A 2009 Intestinal metaplasia: a premalignant lesion involved in gastric carcinogenesis. J Gastroenterol Hepatol 24: 193–201. Chen H, Fang Y, Tevebaugh W et al 201 1 Molecular mechanisms of Barrett’s esophagus. Dig Dis Sci 56:3405–20. Corley DA, Kubo A, Levin TR et al 2007 Abdominal obesity and body mass index as risk factors for Barrett’s esophagus. Gastroenterology 133: 34–41. Debruyne PR, Witek M, Gong L et al 2006 Bile acids induce ectopic expres - sion of intestinal guanylyl cyclase C through nuclear factor-kappa B and Cdx2 in human esophageal cells. Gastroenterology 130:1 191–206. Eda A, Osawa H, Satoh K et al 2003 Aberrant expression of CDX2 in Barrett’s epithelium and inflammatory esophageal mucosa. J Gastroenterol 38: 14–22. Falk GW 2002 Barrett’s esophagus. Gastroenterology 122:1569–91. Gao N, White P, Kaestner KH 2009 Establishment of intestinal identity and epithelial-mesenchymal signaling by Cdx2. Dev Cell 16:588–99. Gilbert EW, Luna RA, Harrison VL et al 201 1 Barrett’s esophagus: a review of the literature. J Gastrointest Surg 15:708–18. Gillen P, Keeling P, Byrne PJ et al 1988 Experimental columnar metaplasia in the canine oesophagus. Br J Surg 75:1 13–15. Goldstein SR, Yang GY, Curtis SK et al 1997 Development of esophageal metaplasia and adenocarcinoma in a rat surgical model without the use of a carcinogen. Carcinogenesis 18:2265–70. Gutierrez-Gonzalez L, Graham TA, Rodriguez-Justo M et al 201 1 The clonal origins of dysplasia from intestinal metaplasia in the human stomach. Gastroenterology 140:1251–60.e6. Gutierrez-Gonzalez L, Wright NA 2008 Biology of intestinal metaplasia in 2008: more than a simple phenotypic alteration. Dig Liver Dis 40: 510–22. Hvid-Jensen F, Pedersen L, Drewes AM et al 201 1 Incidence of adenocarci - noma among patients with Barrett’s esophagus. N Engl J Med 365: 1375–83. Koster MI, Kim S, Mills AA et al 2004 p63 is the molecular switch for initia - tion of an epithelial stratification program. Genes Dev 18:126–31.Kurita T 201 1 Normal and abnormal epithelial differentiation in the female reproductive tract. Differentiation 82:1 17–26. Leedham SJ, Preston SL, McDonald SA et al 2008 Individual crypt genetic heterogeneity and the origin of metaplastic glandular epithelium in human Barrett’s oesophagus. Gut 57:1041–8. Lim YC, Fitzgerald RC 2013 Diagnosis and treatment of Barrett’s oesopha - gus. Br Med Bull 107:1 17–32. Mutoh H, Hakamata Y, Sato K et al 2002 Conversion of gastric mucosa to intestinal metaplasia in Cdx2-expressing transgenic mice. Biochem Biophys Res Commun 294:470–9. Nicholson AM, Graham TA, Simpson A et al 2012 Barrett’s metaplasia glands are clonal, contain multiple stem cells and share a common squamous progenitor. Gut 61:1380–9. Piessen G, Jonckheere N, Vincent A et al 2007 Regulation of the human mucin MUC4 by taurodeoxycholic and taurochenodeoxycholic bile acids in oesophageal cancer cells is mediated by hepatocyte nuclear factor 1 alpha. Biochem J 402:81–91. Silberg DG, Sullivan J, Kang E et al 2002 Cdx2 ectopic expression induced gastric intestinal metaplasia in transgenic mice. Gastroenterology 122: 689–96. Slack JMW 1985 Homeotic transformations in man: implications for the mechanism of embryonic development and for the organization of epithelia. J Theor Biol 1 14:463–90. Slack JMW 2007 Metaplasia and transdifferentiation: from pure biology to the clinic. Nat Rev Molec Cell Biol 8:369–78. Souza RF 2010 The role of acid and bile reflux in oesophagitis and Barrett’s metaplasia. Biochem Soc Trans 38:348–52. Spechler SJ, Fitzgerald RC, Prasad GA et al 2010 History, molecular mecha - nisms, and endoscopic treatment of Barrett’s esophagus. Gastroenterol - ogy 138:854–69. Stemmermann GN, Hayashi T 1968 Intestinal metaplasia of the gastric mucosa. A gross and microscopic study of its distribution in various disease states. J Natl Cancer Inst 41:627–34. Stringer EJ, Duluc I, Saandi T et al 2012 Cdx2 determines the fate of post - natal intestinal endoderm. Development 139:465–74. Sung MT, Lopez-Beltran A, Eble JN et al 2006 Divergent pathway of intesti - nal metaplasia and cystitis glandularis of the urinary bladder. Mod Pathol 19:1395–401. Willis RA 1962 The Borderland of Embryology and Pathology. London: Butterworths.
235
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
e21 COMMENTAR Y 1.5 Electron microscopy in the twenty-first centuryCOMMENTARY 1.5 Introduction Biological electron microscopy (EM) is pivotal in life science research. A measure of its ubiquity is the diversity of its capacity, e.g. three- dimensional structural information about proteins and viruses at Å resolution; three-dimensional reconstructions of cellular organelles and tissues (from a few tens of nanometres to hundreds of micrometres); and two-dimensional and topographic information and elemental microanalysis at subcellular resolution. The capacity of EM is now further enhanced by correlative light and electron microscopy (CLEM) workflows, where light microscopy (LM) events are co-localized with underlying ultrastructure. In its simplest form, LM generates a ‘Google Earth’-like map to identify areas of interest, which are then studied at the ultrastructural level by EM, so overcoming both the physical size limitations imposed on samples by EM and the diffraction limits of resolution imposed by LM (Williams and Carter 2009). CLEM can also be combined with advanced EM sample preparation and imaging tech - niques to provide previously unheralded levels of cytoarchitectural insight into biological systems (McDonnald 2009, Plitzko et al 2009, Sartori et al 2007). However, biological EM is not without its challenges. The electron beam is generated under vacuum at pressures and temperatures that are nominally incompatible with liquid water, yet water is the most abun - dant cellular constituent. Moreover, carbon-based life forms also have poor contrast in the electron microscope because they are composed mainly of light elements. In fact, carbon is so ‘transparent’ in the elec - tron microscope that it is often employed as a film to support biological samples. To overcome these caveats, hydrated ‘live’ tissue is converted to a ‘fixed’ state; conventional protocols employ a series of steps, includ - ing chemical fixation, alcohol dehydration and resin infiltration. Heavy metal salts are added for positive staining of fixed and embedded speci - mens or negative staining of whole structures that have been deposited on a support film (Glauert and Lewis 1998). More recently, cryo-fixation has been adopted to allow the ‘solidification of a biological specimen by cooling with the aim of minimal displacement of its components’ (Steinbrecht and Zierold 1987). By using low temperature as a physical fixation strategy, the morphology and dimensions of the living material are retained and soluble cellular components are not displaced, which means that processing artefacts commonly encountered in more con - ventional room-temperature EM techniques are either reduced or removed. Cryo-EM often allows direct observation of specimens that have not been stained or chemically fixed. EM in pathology Despite the decline in diagnostic histopathology by EM from the height of its popularity in the 1980s, EM remains an important diagnostic tool in several well-defined areas. Indeed, guidance issued by the UK’s Royal College of Pathologists cites and supports the Association of Clinical Pathologists Best Practice No. 160, which states that ‘Many respondents [to a review of laboratory practice in renal pathology] expressed the opinion that to carry out evaluation of renal biopsy specimens without at least having the availability of electron microscopy is negligent’ (de Haro and Furness 2012). Other specialized areas include the diagnosis of primary ciliary dyskinesia; certain skin/connective tissue disorders, e.g. inherited bullous lesions and Ehlers–Danlos syndrome; and spe-cialized areas of ophthalmic pathology (de Haro and Furness 2012). Diagnostic transmission electron microscopy (TEM) in the clinical setting has been largely stagnant and there has been little change in procedures or application for decades. In marked contrast, TEM in the research setting has experienced an explosion in capacity and capability, and there is considerable opportunity for twenty-first century EM to translate to, and to advance, clinical research and diagnostic pathology (Fig. 1.5.1, Video 1.5.1). Limitations of traditional processing techniques and current protocols Traditional EM processing strategies introduce artefacts (perturbations to structure) ( Fig. 1.5.2) that affect tissue structure, e.g. by shrinkage or swelling of the tissue under examination, the shrinkage of cellular organelles, and/or the extraction or redistribution of cellular constitu - ents such as lipids, proteins and DNA (Boyde and Maconnachie 1979). Most chemical fixatives react with proteins and cross-link peptide chains as part of their action (Glauert and Lewis 1998). These reactions can be highly deleterious to epitopes, significantly compromising immunohistochemical studies. Cryo-fixation has two distinct advantages over chemical fixation (Nermut and Frank 1971, Nermut 1991). It is rapid (measured in milli - seconds), which means that the sample is preserved, hydrated and in a ‘close-to-life’ state at the point of initiating fixation, and ensures simul - taneous immobilization of all macromolecular components (Robards and Sleytr 1995). Many protein networks are labile and prone to disrup - tion with even a slight osmotic or temperature change; cryo-fixation minimizes these deleterious effects. Cryo-techniques allow the study of biological samples with improved ultrastructural preservation and may facilitate the study of dynamic processes. Rapid freezing prevents arte - factual aggregations of proteins because tissue fixation and processing can be performed with no or minimal use of cross-linking fixatives, which means that samples retain higher antigenicity (Kellenberger 1991, Hayat 2000, Claeys et al 2004). This level of preservation relies on vitrification (the transformation of water from a liquid to an amorphous state without inducing the Fig. 1.5.1 A Masson trichrome stain of a renal biopsy taken for diagnostic histopathology. The arrow highlights a glomerular capillary, fringed with foot processes, which would be selected for investigation at higher resolution by transmission electron microscopy (see Video 1.5.1). For diagnostic pathology, the whole glomerulus would be examined. Key: red, keratin; blue/green, collagen; light red/pink, cytoplasm. (Image courtesy of M Balys.) Roland A Fleck
236
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
ElECTRON MiCRO sCOpY iN ThE TwENTY -fiRsT CENT uRY e22sECTiON 1 Fig. 1.5.2 An overview of electron microscopy specimen preparation workflow. Areas to the left of the dashed line are at room temperature, indicating that samples can be handled without the requirement for specialist cryo-operating procedures. Samples to the right of the dashed line must be handled in a manner suitable for vitreous biological samples, taking specific care to pre-cool tools before approaching or manipulating samples. Arrows denote the direction of workflow and key stages required to process a sample from ‘living/close to life’ state to final image. Vitreous thin film (VTF) is a technique in which a liquid sample is pipetted (4–6 µl) on to a transmission electron microscopy (TEM) grid and blotted to create a thin, even liquid film across the surface of the grid. Ideally, the TEM grid should have a holey carbon film and, following blotting, the liquid sample will be held as a mono- layer (approximately 12 nm thick) of particles supported within the holes. This film is then plunged rapidly into a liquid cryogen (ethane) to vitrify the sample, generating a vitreous ice film that is amorphous (no long-range structure) and thus a desirable support phase for cryo-electron microscopy of the encased particles (Cheng et al 2012). Other abbreviations: CEMOVIS, cryo-electron microscopy of vitreous sections; SEM, scanning electron microscopy. (Adapted from Fleck RA 2015 Low-temperature electron microscopy. In: Wolkers W, Oldenhof H (eds), Methods in Cryopreservation and Freeze-Drying Protocols. Methods in Molecular Biology. Berlin: Springer.)Room temperature specimen handlingRoom temperature specimen handling and processing Warm/ ‘live sample’Plunge freezing High-pressure freezingPhysically fixed vitrified specimen Freeze fractureFixation Processing VisualizationFreeze fracture Freeze-substitutionCryo-ultramicrotomyVTF Chemical fixation Conventional room temperature sample preparation techniquesTokuyasu/ immunolabelling techniqueTokuyasu CEMOVIS Ultramicrotomy Replica Imaging by either SEM or TEM at conventional imaging conditions of temperature and pressureImaging by either cryo-SEM or cryo-TEM under cryogenic conditions nucleation of ice crystals). The nucleation of ice crystals is temperature- and pressure-dependent. Crystallization depends on the cooling rate, itself dependent on the thermal properties of water, the sample thick - ness, and the heat extraction flow at the surface of the specimen. Freez - ing is a time-dependent process. Vitrification sits at the beginning of a workflow where the sample is either subsequently processed to a stable room-temperature state for imaging or maintained in its vitreous state throughout (see Fig. 1.5.2) (Fleck 2015). There are a variety of methods for rapid freeze fixation of tissues. However, the depth of vitrification is often limited (e.g. a few microns for samples plunged into liquid cryogen). Of the available technolo - gies, high-pressure freezing (HPF) is by far the most effective. First introduced by Moor and Riehle in 1968, HPF exploits the physical benefit of high pressure (210 MPa) to reduce the cooling rate required for the vitrification of water from several 100,000 K.s−1 to a few 1,000 K.s−1, making vitrification of relatively thick samples practicable (Moor and Riehle 1968, Studer et al 2008). HPF machines synchro - nize pressurization and cooling of the sample (to below the glass transition temperature (Tg)) within 20 ms, in a highly reproducible manner (Müller and Moor 1984, Riehle 1968, Studer et al 2001), extending the depth of vitrification to as much as 200 µm (Studer et al 1995, Studer et al 2008). Tissue samples prepared this way require further modification to make them compatible with the transmission electron microscope. They must either be converted to a room-temperature stable state (e.g. by freeze substitution into a resin), or be sectioned whilst remaining below the glass transition temperature (T g); when this is done, images close to the native structure of biological specimens can be achieved. However, despite its benefits, cryo-electron microscopy of vitreous sec - tions (CEMOVIS) has not been routinely adopted, principally because of the technical difficulty of producing a ribbon of frozen sections and transferring it on to an EM grid. A double micromanipulator tool that guides the ribbon using an electrically conductive fibre, and then posi - tions and attaches the ribbon to an EM grid has recently been developed (Studer et al 2014) ( Fig. 1.5.3). Advances in instrumentation and detector technology EM has undergone a revolution in recent years, largely coincident with the advent of digital charge-couple device (CCD) cameras and increasing personal computer (PC) processing power. What was princi - pally a two-dimensional technique is now one that readily embraces three dimensions by electron tomography (Gan and Jensen 201 1). In this technique, a specimen is tilted through the incident beam, sequential images are acquired, and, after alignment, a weighted back- projection is used to generate a three-dimensional tomographic volume (see Video 1.5.1). These techniques have stimulated instrument changes. In conventional TEM, low accelerating voltages (80–120 keV) provide a strong beam-to-specimen interaction in order to increase contrast from electron scattering. Cryo-TEM and tomography require higher accelerating voltages (200–300 keV), which reduce the beam-to- specimen interaction; the contribution of chromatic aberration to the image is also reduced, theoretically improving resolution (Williams and Carter 2009). Changing the emission source has been another major technical advance. A field emission gun (FEG) is becoming the preferred source of the electron beam because it provides a beam of smaller diameter, higher coherence and a current density (brightness) of up to three orders of magnitude greater than was achieved with conven - tional thermionic emitters (e.g. tungsten or lanthanum hexaboride
237
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Electron microscopy in the twenty-first century e23 COMMENTAR Y 1.5 and resolution with superior detective quantum efficiency (DQE) rela - tive to a conventional CCD camera (Milazzo et al 2010). Moreover, direct exposure to the incident electron beam significantly improves the signal-to-noise ratio in comparison to an SCCD (Milazzo et al 2010). Perhaps more importantly, DDD allow images to be recorded as movies, which means that movement caused by the electron beam can be cor - rected computationally (Grigorieff 2013). As an example, images of ribosomes of sufficient quality to permit alignment at an accuracy that approaches near-atomic resolution have been obtained using approxi - mately 30,000 images (Bai et al 2013). When combined, ZP, EFTEM and DDD offer a paradigm shift in the capability of the transmission electron microscope and are likely to drive further advances in EM-based research. Whether in isolation or combination, these techniques will translate from cutting edge to routine instrumentation, heralding generic benefits to the field. Phase contrast electron microscopy utilizing ZP has already been applied to histochemically stained strong-phase (high-contrast heavy-metal- stained) biological specimens (Atsuzawa et al 2009). In sections of various thicknesses specifically stained for the Golgi apparatus, a high-contrast gain was observed even for strong-phase objects, indicating that phase contrast electron microscopy has an application for traditional strong-phase-contrasted biological specimens (Atsuzawa et al 2009), as well as for low-phase cryo-fixed biological specimens (e.g. the visualiza - tion of cyanophage Syn5 inside its host, Synechococcus, using ZP in combination with cryo-tomography (cryo-ET) to enhance image con - trast over conventional cryo-ET) and will permit direct identification of subcellular components (Dai et al 2013). Advances in the scanning electron microscope Cryo-FEGSEM has become well established as a research tool. The sta - bility and emitter life of a FEG source, coupled with the stability and low keV sensitivity of a modern scanning electron microscope, is a powerful tool. Unlike TEM, in which high accelerating voltages are used to transmit the electrons through the specimen, SEM benefits from low accelerating voltages that limit beam interaction with the specimen. This is key to successful cryo-SEM but creates a challenge around gen - erating sufficient signal for high-resolution imaging. Modern FEGSEM systems (with improved detector sensitivity and signal-to-noise ratios) are now achieving sub-nm resolution at 1 keV, making them highly capable tools for life science research, particularly when combined with high-resolution coating and fracture techniques (Fleck 2015) (see Fig. 1.5.2; Fig. 1.5.4). Large three-dimensional volumes may be generated by the assembly of sequential two-dimensional images acquired in the scanning elec - tron microscope. There are essentially three strategies, each with unique strengths and weaknesses: focused ion beam (FIB), serial block face SEM (SBFSEM) and array tomography (AT). FIB involves milling of material in a dual-beam scanning electron microscope using a focused beam of gallium ions to cut (mill) away layers of material. After each cut, the surface is imaged, ultimately producing a stack of two- dimensional images that can be aligned to generate a three-dimensional volume. Although FIB can be combined with cryo-preparation tech - niques, elemental microanalysis and the range of imaging modes avail - able within the SEM, it can be challenging (problems include achieving registry between slices, gallium contamination of the sample, and damage to structural information caused by the milling process). SBFSEM takes a different approach from FIB and incorporates a microtome within the scanning electron microscope chamber (Denk and Horstmann 2004). The microtome cuts a thin section from the block, an image is taken, and the procedure is repeated until a three- dimensional volume is generated ( Video 1.5.2). SBFSEM is less prone to registration issues because the sample essentially remains locked in position below the detector. However, like FIB, this technique is destruc - tive since, once sectioned, the sample is lost. SBFSEM uses room tem-perature processing techniques with high levels of en bloc heavy metal staining that are required to generate a sufficiently robust back-scattered electron signal to form an image. The processing steps, and the limita - tions of the back-scattered electron detector, mean that SBFSEM is cur - rently incompatible with elemental microanalysis or cryo-techniques. However, this is a nascent technology that is advancing rapidly. Novel applications will emerge with advances in detectors, improved resins (e.g. conductive resins) and the exploitation of charge suppression approaches in the SEM (e.g. beam deceleration). Unlike FIB and SBFSEM, AT is not destructive and material may be recovered and re-imaged. The technique allows samples to be prepared by any TEM preparation strategy. Material is sectioned at room tempera - ture by ultramicrotomy; the sections are collected in order as a ribbon and may then be stained to enhance contrast, or are immunolabelled. (LaB 6)-tipped filaments) (Kersker 2001). The net results for both scan - ning electron microscopy (SEM) and TEM have been greatly improved signal-to-noise ratio and spatial resolution, coupled with a significant increase in emitter life and reliability. FEG has had a major impact on the practicality of cryo-SEM and resolution in cryo-TEM studies, although these higher accelerating voltages carry a penalty of decreased contrast. Overcoming the challenges of contrast in the transmission electron microscope Zero-loss energy filtering is employed to improve image contrast in the transmission electron microscope. The filter may be positioned either in-column or post-column. Increased scattering contrast and the exclu - sion of the blurred background (as a consequence of the interception of inelastically scattered electrons) allow specimens to be imaged closer to focus, and increases high-resolution contrast and image resolution (Grimm et al 1997, Grimm et al 1998). Energy filtered transmission electron microscopy (EFTEM) imaging to remove the contributions of inelastically scattered electrons is advan - tageous when imaging thick biological samples. Han et al (1995) dem - onstrated that only elastically scattered electrons contributed to the coherent image component when imaging biological specimens of more than 0.5 µm thick at 200 keV; optimization of the zero-loss signal required operation at intermediate to high primary accelerating voltages (over 200 keV). The gain due to EFTEM increases with specimen thick - ness, with a linear dependence for light-scattering elements (Grimm et al 1998). These results are important for the accurate recording of images of thick biological specimens by tomography and to demon - strate the increasing importance of high accelerating voltages in biologi- cal TEM (Grimm et al 1998). New technologies are overcoming previous limitations of low con- trast in the transmission electron microscope. Zernike phase plates (ZP), which are well accepted for imaging with the light microscope, represent a novel approach to imaging in the transmission electron microscope. Despite technical hurdles to their use, the potential reward in greater contrast has driven research into this technology (Fukuda et al 2009, Glaeser 2013). An approximate five-fold increase in contrast is possible with a π2 ZP that shifts the phase of only the scattered elec - trons (these phase-shifted electrons interfere with the unscattered elec - trons in the image plane), resulting in amplitude contrast rather than phase contrast. Indirect scintillator-coupled charge-coupled-device (SCCD) cameras replaced film cameras and revolutionized EM because they provided a near-instant digital readout and obviated the physical limitation of a fixed number of image plates. Direct detection devices (DDD) avoid the fundamental inefficiencies of the electron-to-light conversion process of a scintillator and achieve unmatched sensitivity Fig. 1.5.3 Cryo-electron microscopy of vitreous sections (CEMOVIS) performed using a cryo-ultramicrotome. The cryo-chamber (Leica EM FC6, Leica Microsystems, Austria) is mounted on a conventional room-temperature ultramicrotome (Leica EM UC6, Leica Microsystems, Austria). The double manipulators (Manip, Diatome, Switzerland) are then attached to the top of the cryo-chamber, allowing precise control of both the vitreous ribbon and the transmission electron microscopy grid. The system is located within a glove box to maintain a low-humidity environment around the sample that will reduce contamination during sectioning.
238
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
ElECTRON MiCRO sCOpY iN ThE TwENTY -fiRsT CENT uRY e24sECTiON 1 Fig. 1.5.5 ATUMtome and tape. The ATUMtome is mounted on an PT-PC PowerTome (RMC, USA), showing the tape being transferred from a delivery spool (top) to receiving spool (bottom) through the boat, which maintains a water reservoir behind the edge of the diamond knife (blue) to float the sections and allow their capture by tape. Tape tension is maintained throughout by a series of tensioning springs and guides. The insert shows individual sections on tape mounted on carbon tape and silicon wafer (to help dissipate specimen charging during imaging) ready for viewing in the scanning electron microscope. Fig. 1.5.4 Cryo-scanning electron microscopy (SEM) images. A, Fractured Plasmodium falciparum -infected erythrocytes. A cross-fractured infected erythrocyte is shown; the insert depicts a high-magnification image of the p-face of the infected erythrocyte with membrane proteins visible on the fracture plane. The scale bar represents 1 µm (main image) and 100 nm (inset). B, Cross-fractured Neisseria meningitidis showing surface membrane protein organization and polysaccharide coat. The scale bar represents 100 nm. Images were taken with a cold-field emission gun cryo-SEM (JSM- 7401F, JEOL, Japan), equipped with a Baltec/Leica VCT100 vacuum transfer system and cryo-stage. Coating was applied after fracturing (3 nm Pt/C) using multi-angle rotational coating (BAF060, Baltec/Leica Microsystems, Austria). A B Ribbons may be imaged by LM or SEM; individual sections can be identified, recovered and imaged at the higher spatial resolution of the transmission electron microscope. Successful AT requires high-precision x-y-z sample stages that can perform programmed, sequential (in focus) imaging of serial sections. Arguably the most challenging aspect of the workflow is the serial sectioning of the sample; to date, this is under - taken almost exclusively by highly skilled operators. However, the ATUMtome (RMC, USA) employs a tape system to automate the serial sectioning process and has recently been commercialized ( Fig. 1.5.5). New multibeam-multidetector SEMs (e.g. 61 concurrent beams) are being developed that will allow images of up to 1.2 billion pixels to be acquired in 1 s (Marx 2013) and will further speed data acquisition. Developments in thin-film technology and microfabrication are facilitating the design of liquid cells that will permit improved imaging of samples in the liquid phase (de Jonge and Ross 201 1). In combina - tion with recent advances in ZP, EFTEM and DDD, the new generation of liquid cells offers considerable potential for the study of complex dynamic cellular processes. REFERENCES Atsuzawa K, Usuda N, Nakazawa A et al 2009 High-contrast imaging of plastic-embedded tissues by phase contrast electron microscopy. J Elec - tron Microscopy 58:35–45. Bai XC, Fernandez IS, McMullan G et al 2013 Ribosome structures to near- atomic resolution from thirty thousand cryo-EM particles. Elife 2:e00461. Boyde A, Maconnachie E 1979 Volume changes during preparation of mouse embryonic tissue for scanning electron microscopy. Scanning 2:149–63. Cheng D, Mitchell DRG, Shieh D-B et al 2012 Practical considerations in the successful preparation of specimens for thin-film cryo-transmission electron microscopy. In: Méndez-Vilas A (ed) Current Microscopy Con - tributions to Advances in Science and Technology, Microscopy Series no. 5 (vol. 2). Badajoz: Formatex; pp. 880–90. Claeys M, Vanhecke D, Couvreur M et al 2004 High-pressure freezing and freeze substitution of gravid Caenorhabditis elegans (Nematoda: Rhab- ditida) for transmission electron microscopy. Nematology 6:319–27.Dai W, Fu C, Raytcheva D et al 2013 Visualizing virus assembly intermediates inside marine cyanobacteria. Nature 502:707–10. de Haro T, Furness P 2012 Current and future delivery of diagnostic electron microscopy in the UK: results of a national survey. J Clin Pathol 65: 357–61. de Jonge N, Ross F 201 1 Electron microscopy of specimens in liquid. Nat Nanotechnol 6:695–704. Denk W, Horstmann H 2004 Serial block-face scanning electron microscopy to reconstruct three-dimensional tissue nanostructure. PLoS Biol 2: e329. Fleck RA 2015 Low-temperature electron microscopy. In: Wolkers W, Oldenhof H (eds), Methods in Cryopreservation and Freeze-Drying, Lab Protocol Series, Methods in Molecular Biology. Berlin: Springer; pp. 243–74.
239
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Electron microscopy in the twenty-first century e25 COMMENTAR Y 1.5 Fukuda Y, Fukazawa Y, Danev R et al 2009 Tuning of the Zernike phase-plate for visualization of detailed ultrastructure in complex biological speci - mens. J Struct Biol 168:476–84. Gan L, Jensen GJ 201 1 Electron tomography of cells. Q Rev Biophys 45:27–56. Glaeser RM 2013 Methods for imaging weak-phase objects in electron microscopy. Rev Sci Instrum 84:1 1 1 101. Glauert AM, Lewis PR 1998 Biological specimen preparation for transmis- sion electron microscopy. London: Portland Press; p. 326. Grigorieff N 2013 Structural biology: direct detection pays off for electron cryo-microscopy. Elife 2:e00573. Grimm R, Bärmann M, Häckl D et al 1997 Energy-filtered electron tomog - raphy of ice-embedded actin and vesicles. Biophys J 72:482–89. Grimm R, Typke D, Baumeister W 1998 Improving image quality by zero- loss energy filtering – quantitative assessment by means of image cross- correlation. J Microsc-Oxford 190:339–49. Han KF, Sedat JW, Agard DA 1995 Mechanism of image formation for thick biological specimens: exit wavefront reconstruction and electron energy-loss spectroscopic imaging. J Microsc-Oxford 178:107–19. Hayat MA 2000 Chemical fixation. In: Principles and Techniques of Electron Microscopy: Biological Applications, 4th ed. Cambridge: Cambridge University Press, 4–80. Kellenberger E 1991 The potential of cryofixation and freeze substitution: observations and theoretical considerations. J Microsc-Oxford 161: 183–203. Kersker MM 2001 The modern microscope today. In: Zhang X-F, Zhang Z (eds) Progress in Transmission Electron Microscopy 1: Concepts and Techniques. Berlin: Springer; pp. 1–34. McDonnald KL 2009 A review of high-pressure freezing preparation tech - niques for correlative light and electron microscopy of the same cells and tissues. J Microsc-Oxford 235:273–81. Marx V 2013 Brain mapping in high resolution. Nature 503:147–52. Milazzo AC, Moldovan G, Lanman J et al 2010 Characterization of a direct detection device imaging camera for transmission electron microscopy. Ultramicroscopy 1 10:744–7. Moor H, Riehle U 1968 Snap-freezing under high pressure: A new fixation technique for freeze-etching. Proc Fourth Europ Reg Conf Elect Microsc 2:33–4.Müller M, Moor H 1984 Cryofixation of thick specimens by high pressure freezing. In: Revel JP, Barnard T, Haggis GH (eds) Science of Biological Specimen Preparation. AMF O’Hare, Chicago: SEM; pp. 131–8. Nermut MV 1991 Unorthodox methods of negative staining. Micron Microsc Acta 22:327–39. Nermut MV, Frank H 1971 Fine structure of Influenza A2 (SINGAPORE) as revealed by negative staining, freeze-drying and freeze-etching. J Gen Virol 10:37–51. Plitzko JM, Rigort A, Leis A 2009 Correlative cryo-light microscopy and cryo-electron tomography: from cellular territories to molecular land - scapes. Curr Opin Biotechnol 20:83–9. Riehle U 1968 Über die Vitrifizierung verdünnter wässriger Lösungen. Dis - sertation No. 4271: Eidgenössiche Technische Hochschule. Zurich: Federal Institute of Technology (ETH). Robards AW, Sleytr UB 1995 Low temperature methods in biological elec - tron microscopy. In: Glauert AM (ed.), Practical Methods in Electron Microscopy, vol. 10. Amsterdam: Elsevier; pp. 5–146. Sartori A, Gatz R, Beck F et al 2007 Correlative microscopy: bridging the gap between fluorescence light microscopy and cryo-electron tomography. J Struct Biol 160:135–45. Steinbrecht RA, Zierold K 1987 Cryotechniques in biological electron micro - scopy. Berlin: Springer; p. 293. Studer D, Graber W, Al-Amoudi A et al 2001 A new approach for cryofixation by high-pressure freezing. J Microsc-Oxford 203:285. Studer D, Humbel BM, Chiquet M 2008 Electron microscopy of high pres - sure frozen samples: bridging the gap between cellular ultrastructure and atomic resolution. Histochem Cell Biol 130:877–89. Studer D, Klein A, Iacovache I et al 2014 A new tool based on two micro - manipulators facilitates the handling of ultrathin cryosection ribbons. J Struct Biol 185:125–8. Studer D, Michel M, Wohlwend M et al 1995 Vitrification of articular carti - lage by high-pressure freezing. J Microsc-Oxford 179:321–32. Williams DB, Carter CB 2009 Transmission Electron Microscopy: A Text - book for Materials Science. Berlin: Springer; p. 760.
240
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
e26SECTION 1 COMMENTARY The reaction of peripheral nerves to injury 1.6 Main nerve trunks contain many thousands of axons. When a trunk is stretched, but not ruptured, by dislocation of, say, the femoral or humeral heads, or by the nearby passage of a bullet, the axons within that nerve may respond in different ways. The degree of injury may vary along the length of a particular nerve or transversely across its cross- section: some axons will remain intact, some will sustain conduction block but recover rapidly, and yet others will degenerate throughout their entire extent distal to the injury because they have been ruptured. A proportion recover spontaneously whilst the remainder never recover. Seddon (1943) introduced the terms neurapraxia for conduction block, axonotmesis for a degenerative lesion of favourable prognosis, and neurotmesis for a degenerative lesion of unfavourable prognosis. Sunderland (1951) expanded this classification to five degrees of peripheral nerve injury by subdividing neurapraxia according to the degree of connective tissue disruption, and MacKinnon and Dellon (1988) subsequently added a sixth category. However, the fundamental distinction between neurapraxia and either axonotmesis or neurotmesis remains the persistence of conduction in the distal segment of the nerve in neurapraxia, and the absence of conduction in axonotmesis or neu - rotmesis. In clinical practice, neurotmesis usually signifies transection of an entire nerve trunk, its constituent fascicles and their surrounding investments of perineurium and epineurium; the proximal and distal nerve stumps are frequently separated by an inter-stump gap of varying length ( Fig. 1.6.1). The early disappearance of conduction indicates impending or actual critical ischaemia and is probably the most impor - tant indicator that a lesion is deepening (see below). Neurapraxia/conduction block Conduction block exhibits certain characteristic features: paralysis exceeds loss of sensation; proprioceptive nerves are more deeply affected than those responsible for light touch; and sympathetic nerves to the smooth muscles of the blood vessels and sweat glands are least affected. There are distinct patterns of conduction block. Anoxia is important in all of them and there are often elements of mechanical deformation, inducing alterations of the nodal and paranodal apparatus in myeli - nated axons, and subsequent demyelination either of whole internodes or of paranodes.Transient ischaemia Lying in one position without moving for a long time (e.g. during coma, inebriation or anaesthesia), prolonged sitting with the legs crossed or prolonged leaning on the elbows are all situations in which a limb nerve may be transiently compressed against a bone or a hard surface. The ischaemia thus induced in the compressed nerves causes an anoxic, physiological, block of both axoplasmic trans - port and ion channel functions along affected axons. This is seen during operation for the exposure of limb nerves with an inflated cuff in posi - tion: stimulation of a nerve evokes a brisk muscular response by trans - mission through the neuromuscular junction, which then diminishes before disappearing after about 30 minutes, whereas conduction within the nerve itself can be detected for about another 30 minutes. On the other hand, direct stimulation of the muscle provokes a twitch that can be elicited for about another 2 hours; loss of this direct response signals impending muscle death. The selective vulnerability of different popu - lations of nerve fibres is demonstrated by the classical experiment of applying an inflated cuff about the arm (Lewis et al 1931). The observer experiences first a loss of superficial sensibility, then a graduated loss of muscle power. The first pain response is lost soon after superficial sensibility fails; the delayed pain response is still detectable after 40 minutes of ischaemia. Large myelinated axons are first affected; unmyelinated axons and autonomic fibres escape; and pilo- and vaso - motor functions are scarcely affected. All modalities recover within a few minutes of release of the cuff; the unpleasant quality of the residual delayed pain sensation, and the burst of painful ‘pins and needles’ after release of the cuff, give an insight into the feelings of patients affected by dysaesthesiae. Slowly progressing ischaemia Conduction block caused either by a haematoma or aneurysm, or by bleeding into compartments pro - duces early and deep autonomic paralysis and loss of power that extends over hours or days, whilst deep pressure sense and some joint position sense persist. These symptoms are seen in war wounds, when nerves become compressed and strangled by scar tissue or by the cicatrix deep to split skin grafts, and may progress over a period of weeks or months. The recovery of function and relief of pain are rapid – at times, dramatic – after removal of the cause in most of these cases. Prolonged conduction block associated with focal demyel­ ination Severe prolonged pressure causes local demyelination and a conduction block that is not directly attributable to anoxia and that may last for weeks or months (Fig. 1.6.2). Compression-induced focal paranodal myelin deformation may be involved; experimental studies using a cuff inflated to high pressure around a limb produced slippage of paranodal myelin at nodes of Ranvier lying under the edges of the cuff, i.e. where the pressure gradient between compressed and non- compressed portions of the nerve was likely to be greatest (Ochoa et al 1972, Dyck et al 200 5). Clinically, local demyelination and conduction Fig. 1.6.1 Axonotmesis (centre) and neurotmesis (bottom) at the moment of injury. Note the preservation of the Schwann cell basal lamina in axonotmesis. (Courtesy of Rolfe Birch, all rights reserved, published in Birch R, Surgical Disorders of the Peripheral Nerve. 2nd edition, 2011. Springer-Verlag, London.) Schwann cell nucleus Basal lamina Node of Ranvier Myelin sheath Axon Fig. 1.6.2 The effect of mechanical compression: paranodal and juxtaparanodal myelin slippage under the edges of an inflated pressure cuff (after Ochoa J, Fowler TJ, Gilliatt RW 1972). (Courtesy of Rolfe Birch, all rights reserved, published in Birch R, Surgical Disorders of the Peripheral Nerve. 2nd edition, 2011. Springer-Verlag, London.)Proximal Distal Cuff AxonRolfe Birch
241
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
The reaction of peripheral nerves to injury e27 COMMENTAR Y 1.6 1.6.5). (iii) A nerve displaced into a fracture or joint will usually recover if it is extricated within a few days, but will be destroyed if a compres - sion device has been used to stabilize the skeleton without retrieving the nerve (Fig. 1.6.6). (iv) Function may be retained for a long time in a nerve exposed to continuing traction from a malunited fracture. (v) Nerves may continue to work for months or years before failing when they are exposed to the progressing fibrosis caused by radiotherapy, by overlying split skin graft or by other examples of deep scar. (vi) Pain is the cardinal symptom of the unrelieved action of a noxious agent. Wallerian degeneration Trauma that physically separates axons from their cell bodies produces Wallerian degeneration (WD). All affected axons in the extreme tip of block will persist until the source of the external irritation, e.g. a bony projection, is removed (Birch and St Clair Strange 1990). Relief of pain and recovery follow rapidly in lesions that have persisted for many months, or even years. Mechanical deformation is the likely explanation for the conduction block seen in cases of ‘hourglass’ constriction of fascicles within a nerve trunk, where axons may be focally constricted within affected fascicles ( Fig. 1.6.3). Prolonged conduction block of war wounds The characteristic features of classical neurapraxia (see above), caused by the nearby passage of a missile, are likely to be provoked by a momentary displace - ment or stretching of the nerve trunks. However, a different pattern of conduction block has been recognized in recent conflicts, where the patient is exposed, at close range, to the shock wave of an explosion without any wound or fracture, or signs of significant injury to the soft tissues at the level of the nerve lesion. In these cases, the smallest fibres are most deeply affected and they may not recover. Pain is rare in both of these variants of conduction block (Birch et al 2012). Degenerative lesions There are essentially two types of degenerative lesion, with very different potentials for recovery. In the first, which approximates to axonotmesis, the tubes of Schwann cell basal lamina that surround each axon– Schwann cell unit remain intact, which means that axons may regener - ate in an orderly fashion across the lesion site and into the distal stump, providing that the causative agent is removed and that the length of the distal stump (i.e. the distance from the site of injury to the end organ) is not so long that it compromises axonal regrowth. In the second, which approximates to neurotmesis, not only the Schwann cell basal lamina tubes, but also the connective tissue sheaths around the fascicle are interrupted; spontaneous axonal regeneration following this type of injury will be imperfect and disorderly, or may not occur at all. Deepening of a lesion in a nerve that has not been divided Unrelieved activity of a noxious agent on an intact nerve is signified by pain and increasing neurological deficit, indicating that the lesion is deepening. Nerves within a swollen ischaemic limb or tense compart - ment become compressed and anoxic. Nerves displaced by an expand - ing haematoma become stretched and anoxic. Nerves that have become displaced into a fracture or joint are tethered and become subjected to compression, stretching and anoxia. These nerves are not severed by the initial injury; the first lesion is axonotmesis and spontaneous recovery is probable if the cause is removed. However, if this does not happen, the lesion continues to deepen into that of neurotmesis and spontan - eous recovery will not occur. The speed of deepening of a lesion is related to the cause: (i) A nerve crushed by a plate or strangled by an encircling suture may recover if the cause is removed within a few minutes (Fig. 1.6.4). (ii) A nerve that has become compressed and anoxic within a swollen ischaemic limb will recover if perfusion of the tissues is restored within 3 hours; the likelihood of full spontaneous recovery diminishes with the passage of every hour after that time ( Fig. Fig. 1.6.3 Hourglass ‘constriction’ of the lateral root of the median nerve: the result of traction injury. There was full spontaneous recovery. (Courtesy of Rolfe Birch, all rights reserved, published in Birch R, Surgical Disorders of the Peripheral Nerve. 2nd edition, 2011. Springer-Verlag, London.) Fig. 1.6.4 The appearance of a radial nerve after extrication from beneath a compression plate 48 hours after the first operation. There was relief of pain but only incomplete recovery so that subsequent flexor to extensor transfer was necessary. (Courtesy of Rolfe Birch, all rights reserved, published in Birch R, Surgical Disorders of the Peripheral Nerve. 2nd edition, 2011. Springer-Verlag, London.) Fig. 1.6.5 Volkmann’s ischaemic contracture. A, The ulnar nerve exposed during flexor muscle slide 8 weeks after supracondylar fracture. The epineurial vessels and also the ulnar recurrent collateral vessels are occluded, and the nerve is compressed by the swollen infarcted muscle. B , The appearance of the hand 14 years later. (Courtesy of Rolfe Birch, all rights reserved, published in Birch R, Surgical Disorders of the Peripheral Nerve. 2nd edition, 2011. Springer-Verlag, London.) A B
242
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
ThE REACTION Of pERIphERA l NER vES TO INjuRY e28SECTION 1 involved in retrograde signalling to initiate expression of RAGs, particu - larly the family of glycoprotein 130 (gp130) cytokines, see Zigmond (201 1).) Changes in the expression of ion channels and receptors in the neuronal cell bodies in human DRGs can be detected within a few minutes of an injury to the nerve (Lawson 2005). Tip of proximal stump Within a few hours of injury, the cut ends of a transected axon seal off, forming end bulbs. The end bulb that forms at the proximal tip of axons in the proximal stump is transformed into a growth cone from which multiple needle-like filopodia and broader sheet-like lamellipodia grow out. The filopodia are rich in actin, and may extend or retract within a matter of minutes. Each axon forms new branches or sprouts: collateral sprouts arise from nodes of Ranvier at levels where the axons are still intact; terminal sprouts arise from the tips of the surviving axons. The axon sprouts and fine cyto - plasmic processes derived from their associated Schwann cells form clusters, regenerating units, surrounded by Schwann cell basal laminae. Sprouts within one regenerating unit represent the regenerative effort of one neurone and its axon. Within a few days of injury, the calibre of axons in the proximal stumps is reduced and nerve conduction velocity in the proximal segment falls. Distal stump During the first 2 or 3 weeks after injury, the constitutive tissue response within a denervated distal stump involves numerous events: some sequential, others consecutive. These include the produc- tion of debris, as axons and their myelin sheaths are degraded; an increase in local blood flow; activation of resident macrophages, and recruitment and influx of exogenous macrophages; and proliferation of fibroblasts and Schwann cells. The gliotic response produces a popula - tion of dedifferentiated daughter Schwann cells that remain within the basal lamina tubes secreted by their parent cells, forming what are now called Schwann tubes but which were previously known as bands of Büngner. They behave acutely as the ‘presumed targets’ of the regrowing axons until the actual targets (muscle or sensory end organ) have been reinnervated: they secrete proteins that collectively facilitate axonal regrowth, whether by providing trophic support or by establishing a supportive growth matrix (for further reading on these interactions, see Hall (2005), Webber et al (201 1), Arthur-Farraj et al (2012), Gordon (2015)). There is a growing consensus that Schwann cells express dis - tinct motor and sensory phenotypes, and that this fundamental differ - ence affects the ability of Schwann cell tubes to selectively support regenerating neurones. For example, expression of osteopontin and clusterin is upregulated in Schwann cells in transected peripheral nerves: the two secreted factors appear to facilitate regeneration of motor or sensory neurones, respectively (Wright et al 2014). While the microenvironment of an acutely denervated distal stump facilitates axonal regrowth because it provides a vascularized segment of longitudinally orientated, laminin-rich basal lamina tubes filled with axon-responsive Schwann cells, a chronically denervated distal stump is associated with poor axonal regeneration and poor functional recov - ery (Sulaiman and Gordon 2009). Biopsies taken during late repairs, especially when the injury has been complicated by arterial injury or by sepsis, often show remarkably few Schwann cells lying within a densely collagenous matrix. Myelin fragments are detectable in such cases many months after the injury. The residual Schwann cells become progressively less receptive to ingrowing regenerating axons with time; experimental studies have shown that they downregulate expression of axon-responsive receptors that are normally important in Schwann cell–axon signalling (Li et al 1997, Li et al 1998). These findings under - score the clinical observation that there is a relatively narrow window of opportunity when surgical intervention is most likely to produce positive results (Lundborg 2000). The special case of preganglionic injury The preganglionic lesion is all too common in severe traction injuries to the brachial and lumbosacral plexuses. The roots of the spinal nerves are torn from the spinal cord, which means that the somatic afferent pathway is interrupted between the dorsal root ganglion and the spinal cord (see Fig. 1.6.6). The neuronal cell bodies and their axons, investing Schwann cells and associated basal laminae remain intact, healthy and conducting for a long time after the injury. The surviving axons include all those with cell bodies in the dorsal root ganglion, including many ‘recurrent’ fibres in the ventral root that derived from cells in the dorsal root ganglion ( Figs 1.6.7–1.6.9). Somatic efferent fibres, being sepa - rated from their cell bodies, degenerate; postganglionic autonomic efferent axons also degenerate, as a result of damage to their grey rami communicantes. Carlstedt (2007) reckons that about one-half of all motor neurones in the affected spinal cord segment have disappeared by 2 weeks after avulsion of the ventral root and he urges ‘a swift the proximal stump and throughout the distal stump degenerate, irre - spective of their calibre or functional modality; recovery is often slow and incomplete, and there may be considerable residual loss of func - tion. Moreover, although research has tended to address issues associ - ated with the molecular and cellular events that occur at the lesion site and target tissues, it is important to remember that injury to a periph - eral nerve that induces WD also initiates significant and often long-lasting central changes in sensory ganglia and ascending and descending pathways within the spinal cord and brain; these changes also have a significant impact on the likelihood of recovery. Studies of cell bodies of human dorsal root ganglia (DRG) have revealed changes in the expression of a wide range of receptors and a remarkable alteration in the expression of genes regulating neuronal activity (Lawson 2005, Rabert et al 2004, Anand et al 2008). There is an extensive literature on WD in both central and peripheral nervous systems, and only a brief summary of the response in the peripheral nervous system is presented here (see Gordon (2015)). Neuronal cell body and proximal stump The central and the peripheral effects of WD are profound and, in neurotmesis, ultimately irreversible. Not only is the neuronal cell body separated from its con - tinual supply of retrogradely transported neurotrophins, but also its central connections rapidly alter; many neurones thus isolated will die. Amputation provides a model of the effect of permanent axonotomy on the spinal cord: there is extensive loss of neurones in the dorsal root ganglia and in the anterior horn, and a diminution of the large myeli - nated axons in the ventral and dorsal roots (Dyck et al 1984, Suzuki et al 1993). Neuronal death is more severe in more proximal neurot - mesis, and less marked after axonotmesis than after neurotmesis. Neu - rotmesis in the neonate produces a more rapid and much greater incidence of sensory and motor neurone death than in the adult. Surviv - ing neurones rapidly upregulate a series of genes involved in repair and neuroprotection mechanisms following injury (repair-associated genes, RAGs), which direct many of the changes seen during WD and subse- quent recovery (e.g. Calenda et al 2012). (For a review of the molecules Fig. 1.6.6 Deepening of a lesion. A, A median nerve extricated from a supracondylar fracture in a 9-year-old girl 3 days after injury; there was complete recovery. B, A median nerve extricated from a supracondylar fracture in a 13-year-old girl 8 weeks after injury; there was no recovery. (Courtesy of Rolfe Birch, all rights reserved, published in Birch R, Surgical Disorders of the Peripheral Nerve. 2nd edition, 2011. Springer-Verlag, London.) BA
243
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
The reaction of peripheral nerves to injury e29 COMMENTAR Y 1.6 Regeneration of end organs Reinnervation of muscle spindles and Golgi tendon receptors is gener - ally good after crush lesions (axonotmesis) . Regeneration after repair of a divided nerve is much less orderly. A muscle spindle may become reinnervated by afferents normally destined for the Golgi tendon organ; many tendon organs remain denervated and the regenerated endings are frequently abnormal in appearance. The reorganization of motor units after repair causes significant alterations in the mechanical input to an individual tendon organ. Muscles usually exhibit weakness, impaired coordination and reduced stamina after nerve repair. There may be selective failure of regeneration of the largest-diameter fibres and of coactivation of the α and γ efferents. The normal movement of joints is brought about by smoothly coordinated and controlled activity in muscles that is pre - cisely and delicately regulated by inhibition and facilitation of the motor neurones. The conversion of an antagonist to an agonist is the basis of musculotendinous transfer; it is common to see patients actively extending the knee or the ankle and toes as soon as the post - operative splint is removed after hamstring to quadriceps transfer or anterior transfer of tibialis posterior. Reinnervated muscles usually fail to convert after muscle transfer, irrespective of their power. Perhaps the defective reinnervation of the deep afferent pathways from the muscle spindles and the tendon receptors blinds muscles, which are, after all, sensory as well as effector organs. Cutaneous sensory receptors similarly undergo a slow degenerative change after denervation and after 3 years they may disappear. Re - innervation tends to reverse these changes, although the longer the period of denervation, the less complete will be the regeneration. It is usual to find numerous myelinated axons in the tissues bridging nerve stumps in the human, even in failed sutures or in grafts coming to revi - sion (Fig. 1.6.10 ) (Terenghi et al 1998); these findings encapsulate the difference between regeneration and recovery of function. Plasticity in the central nervous system Nerve transfer involves connecting the proximal stump of a healthy nerve to the distal stump of one that has been injured in such a way that direct repair is not possible. There is considerable adaptation of the central receptor and effector mechanisms after nerve transfer ( Fig. 1.6.1 1). Independent flexion of the elbow without associated activity in the flexor muscles of the forearm is usual by 24 months after transfer from the ulnar nerve to the nerve to biceps brachii. Reinnervation of the muscle spindles and cor - responding reorganization and remapping of the somatosensory cortex have been confirmed after transfer of intercostal nerves to the musculo - cutaneous nerve in adult patients with severe lesions of the brachial plexus (Malessy et al 2003, Sai et al 1996). These central phenomena may explain the remarkably good recovery of sensation within the hand Fig. 1.6.8 Wallerian degeneration in the ventral root of the eighth cervical nerve, 6 weeks after avulsion from the spinal cord. A degenerate efferent myelinated fibre (right) compared with a healthy myelinated afferent fibre (left). Magnification ×11115. (Courtesy of Rolfe Birch, all rights reserved, published in Birch R, Surgical Disorders of the Peripheral Nerve. 2nd edition, 2011. Springer-Verlag, London.) Fig. 1.6.9 Afferent and efferent fibres in a ventral root. Healthy myelinated and unmyelinated fibres in the ventral root of the eighth cervical nerve avulsed from the spinal cord 6 weeks previously. The myelinated efferent fibres have undergone Wallerian degeneration and there is a notable increase in the amount of endoneurial collagen. (Courtesy of Rolfe Birch, all rights reserved, published in Birch R, Surgical Disorders of the Peripheral Nerve. 2nd edition, 2011. Springer-Verlag, London.) intervention to re-establish contact between the injured nerve cells and the periphery with its supply of neurotrophic substances to counteract nerve cell loss’ . After avulsion injury, human dorsal root ganglion neur - ones show dramatic changes in the expression of genes involved in neurotransmission, trophism, cytokine function, signal transduction, myelination, transcription regulation and apoptosis (Rabert et al 2004). Regeneration and recovery of function ‘But the journey of the axon tip to an end organ is only the most dramatic of the phases in the process of regeneration, and its arrival is alone no guarantee of the return of useful function. ’ Young (1942) The cellular microenvironment of a normal peripheral nerve does not ordinarily support axonal regrowth; indeed, there appear to be ‘molecu- lar brakes’ that prevent axonal sprouting. Somewhat counterintuitively, it appears that their expression may be upregulated after injury. WD transforms the environment throughout the distal stump of a transected nerve to one that facilitates axonal regrowth, albeit for a relatively short period (Christie and Zochodne 2013). Manipulating the many cellular and molecular responses that occur during the injury response in the peripheral nervous system remains an as yet unachieved goal of recon - structive surgery.Fig. 1.6.7 Intact (afferent) myelinated and unmyelinated fibres in a suprascapular nerve, 6 weeks after an avulsion lesion of the brachial plexus; efferent fibres have degenerated. Magnification ×6600. (Courtesy of Rolfe Birch, all rights reserved, published in Birch R, Surgical Disorders of the Peripheral Nerve. 2nd edition, 2011. Springer- Verlag, London.)
244
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
ThE REACTION Of pERIphERA l NER vES TO INjuRY e30SECTION 1 whilst afferent impulses from the little and, probably, the ring fingers entered the spinal cord through the fifth cervical nerve as shown in Figure 1.6.12. Similar findings have been observed in four other cases after transfer of intercostal nerves to the lateral cord ( Fig. 1.6.13). The situation in the adult is very different. After transfer of the intercostal nerves to the lateral or medial cords, it is usual to find that stimulation of the reinnervated skin is referred to the chest wall and only rarely to the skin of the hand (Htut et al 2006). Concepts of the mechanisms underlying cortical plasticity following nerve injury and the importance of maintaining an active sensory map of the affected part in the somato - sensory cortex during the deafferentation period are considered further by Lundborg (2003), Rosén and Lundborg (2007), Björkman et al (2009), Rosén et al (201 1) and Knox et al (2015) . Laboratory and clinical studies There are major differences between a laboratory investigation, in which a controlled, precise and limited lesion is inflicted on a periph - eral nerve in an anaesthetized experimental animal, and the situation faced by the clinician presented with a patient with a massive wound involving soft tissues (including blood vessels and muscles) and skeletal Fig. 1.6.10 Useless regeneration in a median nerve sutured 3 years previously. Proximal to the suture line, the endoneurium contains thinly remyelinated axons, some in regenerating clusters, and extensive collagenization; there was no recovery of function. Magnification ×4300. (Courtesy of Rolfe Birch, all rights reserved, published in Birch R, Surgical Disorders of the Peripheral Nerve. 2nd edition, 2011. Springer-Verlag, London.) Fig. 1.6.11 A 13-year-old boy with a right-sided lesion: avulsion of C5, 6 and 7. Repair was undertaken at 2 months as follows: accessory to suprascapular transfer; one bundle of the ulnar nerve to the nerve to biceps and the nerve to the medial head of triceps; medial cutaneous nerve of the forearm to the lateral root of the median nerve. Results at 18 months show a full range of lateral rotation; abduction to 60°; power of elbow flexion, Medical Research Council (MRC) grade 4; power of elbow extension, MRC grade 3 +. There is a full range of active flexion and extension without obvious co-contraction. (Courtesy of Rolfe Birch, all rights reserved, published in Birch R, Surgical Disorders of the Peripheral Nerve. 2nd edition, 2011. Springer-Verlag, London.) Fig. 1.6.12 The method of repair and the likely pathways for afferent function. Abbreviations: CS, cool sensation; CW, cotton wool; JPS, joint position sense; PP, pinprick; Vib, vibration threshold; WS, warm sensation. (Courtesy of Rolfe Birch, all rights reserved, published in Birch R, Surgical Disorders of the Peripheral Nerve. 2nd edition, 2011. Springer-Verlag, London.)c5 c6c7 C6, C7via T3–T5C8 via C5 RH Age 6y Normal CW , PP , VIBJPS, CS, WSlocalizationLat.SS C5 C6 C7 C8 T1 T3 T4 T5XI C8 T1 Fig. 1.6.13 Birth lesions of the brachial plexus. This boy had a group 4 lesion with rupture of C5 and avulsion of C6, 7, 8 and T1 at birth. At 10 weeks of age, C5 was grafted to the upper trunk and to the ventral root of C8 and T1; accessory nerve was transferred to the ventral root of C7; the sensory divisions of intercostal nerves T3 and T4 were transferred to the lateral root of the median nerve. He regained a useful grasp between thumb and index finger, and is an extremely well-adjusted sportsman and musician. The photograph was taken at 10 years of age. (Courtesy of Rolfe Birch, all rights reserved, published in Birch R, Surgical Disorders of the Peripheral Nerve. 2nd edition, 2011. Springer-Verlag, London.) after repair of severe birth lesions of the brachial plexus. Indeed, the sensory recovery is far better than that of skeletal muscle and sympa - thetic function. Thus, there is accurate localization to the dermatomes of avulsed spinal nerves that had been reinnervated by intercostal nerves transferred from remote spinal segments (Anand and Birch 2002). Case report Operation in a 6-month-old boy confirmed rupture of C5 with avulsion of C6, 7, 8 and T1. The lesion was repaired by transfers of the accessory nerve to the suprascapular nerve, of C5 to C8 and T1, and of intercostal nerves T3, 4 and 5 to the lateral cord of the brachial plexus. At the age of 5 years, he was able to localize cotton wool and pinprick sensation accurately in both hands and was able to feel a pin in the affected side. Monofilament, vibration thresholds and joint position sense were iden - tical in both hands. The thresholds for warming were 3.1°C (C5), 0.8°C (C6), 4.1°C (C7), 3.6°C (C8) and 3.3°C (T1) on the affected side. The thresholds for cooling were 2°C (C5), 3.9°C (C6), 3.8°C (C7), 2.3°C (C8) and normal in T1. In the unaffected hand, the thresholds were less than 3.4°C for warm sensation and less than 2.2°C for cool sensa - tion. Sweating was 70% of that of the intact palm. It seems likely that afferent impulses from the thumb, index and, probably, the middle fingers entered the spinal cord through the three intercostal nerves,
245
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
The reaction of peripheral nerves to injury e31 COMMENTAR Y 1.6 elements, and sometimes with other injuries that threaten life and limb. That said, the fundamental cellular processes that underlie subsequent regeneration are similar in the laboratory, in the injuries of civilian practice and in the wounds of war, although in clinical practice they are modified by the violence of the injury, by the effects of injury on associ - ated tissues and, in particular, by ischaemia and by delay before repair. Some general conclusions may be drawn from extensive clinical and laboratory studies. The concept of retrograde degeneration of the axon extending to the first internode applies only to the most benign lesion: that of experi - mentally crushing the nerve between the tips of a jeweller’s forceps. The extent of longitudinal damage to a nerve is greater in a rupture caused by traction than it is in ‘tidy’ transections caused by knife or glass; this effect worsens with increasing delay before repair, is worse still when healing has been complicated by sepsis, and worst of all in neglected ischaemia. Failure to repair main vessels and to ensure perfusion of tissues is profoundly deleterious to regeneration and rules out worthwhile recov- ery of function. REFERENCES Anand P, Birch R 2002 Restoration of sensory function and lack of long-term chronic pain syndromes after brachial plexus injury in human neonates. Brain 125:1 13–22. Anand U, Otto WR, Facer P et al 2008 TRPA1 receptor localisation in the human peripheral nervous system and functional studies in cultured human and rat sensory neurons. Neurosci Lett 438:221–7. Arthur-Farraj PJ, Latouche M, Wilton DK et al 2012 c-Jun reprograms Schwann cells of injured nerves to generate a repair cell essential for regeneration. Neuron 75:633–47. Birch R, Misra P, Stewart MP et al 2012 Nerve injuries sustained during warfare: part II: Outcomes. J Bone Joint Surg Br 94:529–35. Birch R, St Clair Strange FG 1990 A new type of peripheral nerve lesion. J Bone Joint Surg 72B:312–13. Björkman A, Weibull A, Rosén B et al 2009 Rapid cortical reorganisation and improved sensitivity of the hand following cutaneous anaesthesia of the forearm. Eur J Neurosci 29:837–44. Calenda G, Strong TD, Pavlovich CP et al 2012 Whole genome microarray of the major pelvic ganglion after cavernous nerve injury: new insights into molecular profile changes after nerve injury. BJU Int 109: 1552–64. Carlstedt T 2007 Central Nerve Plexus Injury. London: Imperial College Press. Christie KJ, Zochodne D 2013 Neuroscience Forefront Review. Peripheral axon regrowth: new molecular approaches. Neuroscience 240:310–24. Dyck PJ, Nukada H, Lais CA et al 1984 Permanent axotomy: a model of chronic neuronal degeneration produced by axonal atrophy, myelin remodelling and regeneration. In: Dyck PJ, Thomas PK, Lambert EH et al (eds) Peripheral Neuropathy, 2nd ed. Philadelphia: Elsevier, WB Saunders, pp. 660–90. Dyck PJ, Dyck PJB, Engelstad J 2005 Pathological alterations of nerves. In: Dyck PJ, Thomas PK (eds) Peripheral Neuropathy, 4th ed. Philadelphia: Elsevier, Saunders; Ch. 32, pp. 733–829. Gordon T 2015 The biology, limits, and promotion of peripheral nerve regeneration in rats and humans. In: Tubbs RS et al (eds) Nerves and Nerve Injuries. Elsevier, Academic Press; Ch. 61, pp. 993–1012. Hall S 2005 Mechanisms of repair after traumatic injury In: Dyck PJ, Thomas PK (eds) Peripheral Neuropathy, 4th ed. Philadelphia: Elsevier, Saun - ders; Ch. 58, pp. 1403–34. Htut M, Misra P, Anand P et al 2006 Pain phenomena and sensory recovery following brachial plexus avulsion injury and surgical repair. J Hand Surg 31B:596–605. Kawabuchi M, Tan H, Wang S 201 1 Age affects reciprocal cellular interactions in neuromuscular synapses following peripheral nerve injury. Ageing Res Rev 10:43–53. Knox ADC, Goswami R, Anastakis DJ et al 2015 Cortical plasticity after peripheral nerve injury. In: Tubbs RS et al (eds) Nerves and Nerve Inju - ries. Elsevier, Academic Press; Ch. 64, pp. 1055–71. Lawson SN 2005 The peripheral sensory nervous system: dorsal root gan- glion neurones. In: Dyck PJ, Thomas PK (eds) Peripheral Neuropathy (2 vols), 4th ed. Philadelphia: Elsevier, Saunders; Ch. 8, pp. 163–202. Lewis T, Pickering GW, Rothschild P 1931 Centripetal paralysis arising out of arrested bloodflow to the limb. Heart 16:1–32.Li H, Terenghi G, Hall SM 1997 Effects of delayed re-innervation on the expression of c-erbB receptors by chronically denervated rat Schwann cells in vivo. Glia 20:333–47. Li H, Wigley C, Hall SM 1998 Chronically denervated rat Schwann cells respond to GGF in vitro. Glia 24:290–303. Lundborg G 2000 Brain plasticity and hand surgery: an overview. J Hand Surg Br 25:242–52. Lundborg G 2003 Nerve injury and repair. A challenge to the plastic brain. The Bunge Memorial Lecture. J Periph Nerv Syst 8:209–26. Mackinnon S, Dellon AL 1988 Surgery of the Peripheral Nerve. New York: Thieme. Malessy MJ, Bakker D, Dekker AJ et al 2003 Functional magnetic resonance imaging and control over the biceps muscle after intercostal-musculocutaneous nerve transfer. J Neurosurg 98:261–8. Ochoa J, Fowler TJ, Gilliatt RW 1972 Anatomical changes in peripheral nerves compressed by pneumatic tourniquet. J Anat 1 13:433–55. Rabert D, Xiao Y, Yiangou Y et al 2004 Plasticity of gene expression in injured human dorsal root ganglia revealed by gene chip oligonucleo - tide microarrays. J Clin Neurosci 1 1:289–99. Rosén B, Lundborg G 2007 Enhanced sensory recovery after median nerve repair using cortical audio-tactile interaction. A randomised multicentre study. J Hand Surg Eur 32:31–7. Rosén B, Björkman A, Lundborg G 201 1 Improving hand sensibility in vibra - tion induced neuropathy: a case-series. J Occup Med Toxicol 6:13. Sai K, Kanamura A, Sibuya M et al 1996 Reconstruction of tonic vibration reflex in the biceps brachii reinnervated by transferred intercostal nerves in patients with brachial plexus injury. Neurosci Lett 206:1–4. Seddon HJ 1943 Three types of nerve injury. Brain 66:237–88. Sulaiman OA, Gordon T 2009 Role of chronic Schwann cell denervation in poor functional recovery after nerve injuries and experimental strategies to combat it. Neurosurgery 65:A105–14. Sunderland S 1951 A classification of peripheral nerve injuries producing loss of function. Brain 74:491–516. Suzuki H, Oyanagi K, Takahishi H et al 1993 A quantitative pathological investigation of the cervical cord, roots and ganglia after long term ampu - tation of the unilateral upper arm. Acta Neuropathol (Berl) 85:666–73. Terenghi G, Calder JS, Birch R et al 1998 A morphological study of Schwann cells and axonal regeneration in chronically transected human periph-eral nerve. J Hand Surg 23B:583–7. Webber CA, Christie KJ, Cheng C et al 201 1 Schwann cells direct peripheral nerve regeneration through the Netrin-1 receptors, DCC and Unc5H2. Glia 59:1503–17. Wright MC, Mi R, Connor E et al 2014 Novel roles for osteopontin and clusterin in peripheral motor and sensory axon regeneration. J Neurosci 34:1689–700. Young JZ 1942 Functional repair of nervous tissue. Physiol Rev 22:318–74. Zigmond RE 201 1 Gp130 cytokines are positive signals triggering changes in gene expression and axon outgrowth in peripheral neurons following injury. Front Mol Neurosci 4:62.As time goes by, the cellular response in both stumps changes from one that facilitates regeneration to one that is less favourable. Dense collagenization, a profusion of fibroblasts and a loss of Schwann cells are characteristics of the chronically denervated distal stump and are typical of late cases. The normal architecture of a nerve is most closely restored to normal in a well-executed, tension-free primary suture. Regeneration through a graft falls away along its length and not solely at the suture lines. Delay before repair leads to increasing fibrosis and to shrinking of the distal segment so that it becomes impossible to ensure an accurate topographical match. Destruction of the target tissues, of muscle and skin, limits function even when there is strong regeneration. Chronological age plays an important role: for example, signalling mechanisms involved in the reinnervation of skeletal muscle may become impaired with age (Kawabuchi et al 201 1).
246
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
163 CHAPTER 8 Preimplantation development involve modifications of membrane sterols or surface proteins. They traverse the cumulus oophorus and corona radiata, then bind to specific glycoprotein receptors on the zona pellucida, ZP3 and ZP2. Interaction of ZP3 with the sperm head induces the acrosome reaction, in which fusion of membranes on the sperm head releases enzymes, such as acrosin, which help to digest the zona around the sperm head, allowing the sperm to reach the perivitelline space. In the perivitelline space, the spermatozoon fuses with the oocyte microvilli, possibly via two disintegrin peptides in the sperm head and integrin in the oolemma (Figs 8.4, 8.5A). Fusion of the sperm with the oolemma causes a weak membrane depolarization and leads to a calcium wave, which is triggered by the sperm at the site of fusion and crosses the egg within 5–20 seconds. The calcium wave amplifies the local signal at the site of sperm–oocyte interaction and distributes it throughout the oocyte cytoplasm. The increase in calcium concentration is the signal that causes the oocyte to resume cell division, initiating the completion of meiosis II and setting off the developmental programme that leads to embryogenesis. The pulses of intracellular calcium that occur every few minutes for the first few hours of development also trigger the fusion of cortical granules with the oolemma. The cortical secretory granules release an enzyme that hydrolyses the ZP3 receptor on the zona pellucida and so prevents other sperm from binding and undergoing the acrosome reaction, thus establishing the block to polyspermy. The same cortical granule secre - tion may also modify the vitelline layer and oolemma, making them less susceptible to sperm–oocyte fusion and providing a further level of polyspermy block. The sperm head undergoes its protamine → histone transition as the second polar body is extruded. The two pronuclei grow, move together and condense in preparation for syngamy and cleavage after approxi - mately 24 hours ( Fig. 8.5B). Nucleolar ribosomal ribonucleic acid (rRNA), and perhaps some messenger RNA (mRNA), are synthesized in pronuclei. A succeeding series of cleavage divisions produces eight even-sized blastomeres at approximately 55 hours, when embryonic mRNA is transcribed. Several examples of cells that contain male and female pronuclei, termed ootids, have been described. Pronuclear fusion as such does not occur; the two pronuclear envelopes disappear and the two chromo - some groups move together to assume positions on the first cleavage spindle. No true zygote containing a membrane-bound nucleus is formed. The presence of the pronuclei from both parental origins is crucial for spatial organization and the controlled growth of cells, tissues and organs. In the mouse, embryos in which the paternal pronucleus has been removed and replaced with a second maternal pronucleus develop to a relatively advanced state (25 somites), but with limited develop - ment of the trophoblast and extraembryonic tissues. In contrast, embryos in which the maternal pronucleus has been replaced by a second paternal pronucleus develop very poorly, forming embryos of only six to eight somites, but with extensive trophoblast. Thus it seems that the maternal genome is relatively more important for the develop - ment of the embryo, whereas the paternal genome is essential for the development of the extraembryonic tissues that would lead to placental formation. This functional inequivalence of homologous parental chromo - somes is called parental imprinting. The process causes the expression of particular genes to be dependent on their parental origin; some genes are expressed only from the maternally inherited chromosome and others from the paternally inherited chromosome. The genes involved are called imprinted genes. The requirement for both parental genomes is limited to a subset of the chromosomes. Uniparental disomy can arise through meiotic and mitotic non-disjunction events, and results in individuals who are completely disomic or who exhibit mosaicism of disomic and non-disomic cells. If imprinted genes reside on the affected chromosomes, then the uniparental disomic cells will either express a Understanding the spatial and temporal developmental processes that take place within an embryo as it develops from a single cell into a recognizable human is the challenge of embryology. The control of these processes resides within the genome; fundamental questions remain concerning the genes and interactions involved in development. STAGING OF EMBRYOS For the purposes of embryological study, prenatal life is divided into an embryonic period and a fetal period. The embryonic period covers the first 8 weeks of development (weeks following ovulation and fertil - ization resulting in pregnancy). The ages of early human embryos have previously been estimated by comparing their development with that of monkey embryos of known postovulatory ages. Because embryos develop at different rates and attain different final weights and sizes, a classification of human embryos into 23 stages occurring during the first 8 weeks after ovulation was developed most successfully by Streeter (1942), and the task was continued by O’Rahilly and Müller (1987). An embryo was initially staged by comparing its development with that of other embryos. On the basis of correlating particular maternal men - strual histories and the known developmental ages of monkey embryos, growth tables were constructed so that the size of an embryo (specifi - cally, the greatest length) could be used to predict its presumed age in postovulatory days (synonymous to postfertilizational days). O’Rahilly and Müller (2000, 2010) emphasize that the stages are based on external and internal morphological criteria and are not founded on length or age. Ultrasonic examination of embryos in vivo has necessitated the revision of some of the ages related to stages, and embryos of stages 6–16 are now thought to be up to 3 to 5 days older than the previously used embryological estimates (O’Rahilly and Müller 1999, 2010). Within this staging system, embryonic life commences with fertilization at stage 1; stage 2 encompasses embryos from two cells, through com - paction and early segregation, to the appearance of the blastocele. The developmental processes occurring during the first 10 stages of embry- onic life are shown in Figure 8.1. Much of our knowledge of the early developmental processes is derived from experimental studies on amniote embryos, particularly the chick, mouse and rat. Figure 8.2 shows the comparative timescales of development of these species and human development up to stage 12. The size and age, in postovulatory days, of human development from stage 10 to stage 23 is given in Figure 8.3. Information on developmental age after stage 23 (8 weeks post ovulation) is shown in Figure 14.3, where the developmental staging used throughout this text is juxtaposed with the obstetric estimation of gestation that is used clinically. A critique of staging terminology and the hazards of the concurrent use of gestational age and embryonic age is given in Chapter 14; sizes and ages of fetuses towards the end of gestation are illustrated in Fig. 14.9. FERTILIZATION The central feature of reproduction is the fusion of the two gamete pronuclei at fertilization. In humans, the male gametes are spermato - zoa, which are produced from puberty onwards. Female gametes are released as secondary oocytes in the second meiotic metaphase, usually singly, in a cyclical fashion. The signal for the completion of the second meiotic division is fertilization, which stimulates the cell division cycle to resume, completing meiosis and extruding the second polar body (the second set of redundant meiotic chromosomes). Fertilization normally occurs in the ampullary region of the uterine tube, probably within 24 hours of ovulation. Very few spermatozoa reach the ampulla to achieve fertilization. They must undergo capacita - tion, a process that is still incompletely understood, and which may
248
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Preim Plantation develo Pment 164SeCtion 2 tation genetic diagnosis (PGD). A sample (biopsy) is removed from either the oocyte (polar bodies), the embryo itself (a blastomere) or the blastocyst (small piece of trophectoderm), and subjected to specific genetic testing. Oocytes and embryos can also be biopsied and screened for aneuploidy (errors in chromosome copy numbers) in preimplanta - tion genetic screening (PGS). Unaffected embryos can then be identified for transfer to the patient. Embryos that are surplus to immediate thera - peutic requirements can also be cryopreserved in liquid nitrogen for later use. Propanediol or dimethylsulphoxide are commonly used as cryoprotectants for early embryos, and, like glycerol, may be used for blastocysts. Conception rates per cycle using ovarian stimulation, IVF and successive transfers of fresh and cryopreserved embryos exceed those obtained during non-assisted, or natural, conception. PREIMPLANTATION DEVELOPMENT Cleavage The first divisions of the zygote are termed cleavages. They distribute the cytoplasm approximately equally among daughter blastomeres, so, although the cell number of the preimplantation embryo increases, its total mass actually decreases slightly ( Fig. 8.6). Cell division can be asynchronous and daughter cells may retain a cytoplasmic link through much of the immediately subsequent cell cycle via a midbody, as a result of the delayed completion of cytokinesis. No centrioles are present until 16–32 cells are seen, but amorphous pericentriolar mat - erial is present and serves to organize the mitotic spindles, which are characteristically more barrel- than spindle-shaped at this time.double dose of the gene or have both copies repressed. For example, the gene encoding the embryonal mitogen insulin-like growth factor II is expressed from the paternally inherited chromosome, and repressed when maternally inherited. In vitro fertilization In vitro fertilization (IVF) of human gametes is a successful way of overcoming most forms of infertility ( Video 8.1). Controlled stimula - tion of the ovaries (e.g. pituitary downregulation using gonadotrophin-releasing hormone superactive analogues, followed by stimulation with purified follicle stimulating hormone or urinary menopausal gonadotrophins) enables many preovulatory oocytes (often 10 or more) to be recruited and matured, and then aspirated transvaginally using ultrasound guidance, 34–38 hours after injection of human cho - rionic gonadotrophin (which is given to mimic the luteinizing hormone surge). These oocytes are then incubated overnight with motile spermatozoa in a specially formulated culture medium, in an attempt to achieve successful in vitro fertilization. In cases of severe male-factor infertility, in which there are insufficient normal spermato - zoa to achieve in vitro fertilization, individual spermatozoa can be directly injected into the oocyte in a process known as intracytoplas - mic sperm injection (ICSI), which is as successful as routine IVF. In cases in which there are no spermatozoa in the ejaculate, suitable material can sometimes be directly aspirated from the epididymis or surgically retrieved from the testes, and the extracted sperm are then used for intracytoplasmic injection of sperm. It is also now possible, in some cases, to test embryos for the pres - ence of a particular genetic or chromosomal abnormality by preimplan - Fig . 8 .1 Developmental processes occurring during the first 10 stages of development . In the early stages, a series of binary choices determine the cell lineages . Generally, the earliest stages are concerned with formation of the extraembryonic tissues, whereas the later stages are concerned with the formation of embryonic tissues . Oocyte Ootid ZygoteTropho- blastSyncytio-trophoblast Cyto-trophoblast Hypoblast Innercellmass EpiblastLacunae Chorion VilliIntervillousspaces AllantoisSecondary yolk sacExtraemb. mesenchyme Haemopoiesis Primary yolk sac Connecting stalkVisceralhypoblast Parietalhypoblast Extraemb.mesoblast Amnion Primordial germ cells Primitive streak Notochordal process/plate Embryonic endoderm Mesoblast Caudal eminenceMesenchyme Somatopleuric and splanchnopleuric coelomic epithelium Epithelial somites Neuralectoderm Neuralcrest Ectodermal placodes SurfaceectodermDescription of stage FertilizationFirst cleavage Preimplantation Compaction Free blastocyst hatched from zona Implantation Implanted previllus Secondary yolk sac Primitive streak Notochord Neurenteric canal Before folding Neural groove Somites Intraemb. coelom Neural tube Neural crestAfter folding1 2 3 4 5a 5b 5c 6a 6b 7 9 10 118 Approx. age in days2–3 4–5 18–21 24–27 28–30 26–29 21–256 7–1216–18 1
249
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Preimplantation development 165 CHaPter 8 All cleavage divisions after fertilization are dependent on continuing protein synthesis. In contrast, passage through the earliest cycles, up to eight cells, is independent of mRNA synthesis. Thereafter, experimental inhibition of transcription blocks further division and development, indicating that activation of the embryonic genome is required. There is also direct evidence for the synthesis of embryonically encoded pro - teins at this time. As the genes of the embryo first become both active and essential, the previously functional maternally derived mRNA is destroyed. However, protein made on these maternal templates does persist at least during blastocyst growth. Spontaneous developmental arrest of embryos cultured in vitro seems to occur during the cell cycle of gene activation, but it is not caused by total failure of that activation process. Early cleavage, up to the formation of eight cells, requires pyruvate or lactate as metabolic substrates, but thereafter more glucose is metabolized and may be required. The earliest time at which different types of cells can be identified within the cleaving embryo is when 8–16 cells are present. Up to the formation of eight cells, cells are essentially spherical, touch each other loosely, and have no specialized intercellular junctions or significant extracellular matrix; the cytoplasm in each cell is organized in a radially symmetric manner around a centrally located nucleus. Once eight cells have formed, a process of compaction occurs. Cells flatten on each other to maximize intercellular contact, initiate the formation of gap and focal tight junctions, and radically reorganize their cytoplasmic confor - mation from a radially symmetric to a highly asymmetric phenotype. This latter process includes the migration of nuclei towards the centre of the embryo, the redistribution of surface microvilli and an underly- ing mesh of microfilaments and microtubules to the exposed surface, and the localization of endosomes beneath the apical cytoskeletal mesh. As a result of the process of compaction, the embryo forms a primitive protoepithelial cyst, which consists of eight polarized cells, in which the apices face outwards and basolateral surfaces face internally. The focal tight junctions, which align to become increasingly linear, are localized to the boundary between the apical and basolateral surfaces. Fig . 8 .2 Within developmental biology, evidence concerning the nature of developmental processes has come mainly from studies in vertebrate embryos, most commonly amniote embryos of the chick, mouse and rat . This chart illustrates the comparative timescale of development in these animals and in humans . Fertilization Incubation Implantation Primitive streak Neural plate Pharyngula arches 15–20 somites, rostral neuroporeclosed Upper limb bud Lowerlimb bud Days 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30Chick Mouse Rat Human(days from onset of incubation) (days from time of mating) (days from time of mating) (days post ovulation) Fig . 8 .3 Human developmental stages 10–23 . The greatest embryonic length in mm (ordinate) is plotted against age in postfertilizational weeks (abscissa), with the stages superimposed according to current information . (Data provided by courtesy of Professor R O’Rahilly . See also O’Rahilly and Müller (2010) .)Days post fertilization20 1530 Stage 23 22 21 20 19 18 17 16 15 14 13 1112 10 9 87625 10 5 14 21 28 35 42 49 560Size (mm)
250
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Preim Plantation develo Pment 166SeCtion 2 elements of compaction. The entire process can function in the absence of both mRNA and protein synthesis. Post-translational controls are sufficient and seem to involve regulation through protein phosphoryl - ation. Significantly, although E-cadherin is not synthesized and present on the surface of cleaving blastomeres, it first becomes phosphorylated when eight cells are visible, at the initiation of compaction. The process of compaction is important for the generation of cell diversity in the early embryo. As each polarized cell divides, it retains significant elements of its polar organization, so that its daughter cells inherit cytocortical domains, the nature of which reflects their origin and organization in the original parent cell in the eight-celled embryo. Thus, if the axis of division is aligned approximately at right angles to the axis of cell polarity, the more superficially placed daughter cell inherits all the apical cytocortex and some of the basolateral cytocortex and is polar, whereas the more centrally placed cell inherits only baso - lateral cytocortex and is apolar. In contrast, if the axis of division is aligned approximately along the axis of the cell polarity, two polar daughter cells are formed. In this way, two-cell populations are formed in the 16-cell embryo that differ in phenotype (polar, apolar) and posi - tion (superficial, deep). The number of cells in each population in any one embryo will be determined by the ratio of divisions along, and at right angles to, the axis of eight-cell polarity. The theoretical and observed limits of the polar to apolar ratio are 16 : 0 and 8 : 8. The outer polar cells contribute largely to the trophectoderm, whereas the inner Gap junctions form between apposed basolateral surfaces and become functional. The process of compaction involves the cell surface and the calcium dependent cell–cell adhesion glycoprotein, E-cadherin (also called L-CAM or uvomorulin). Neutralization of its function disturbs all three Fig . 8 .4 The fertilization pathway: a succession of steps . After a sperm binds to the zona pellucida, the acrosome reaction takes place (see detail at top) . The outer acrosomal membrane (blue), an enzyme-rich organelle in the anterior of the sperm head, fuses at many points with the plasma membrane surrounding the sperm head . Then those fused membranes form vesicles, which are eventually sloughed off from the head, exposing the acrosomal enzymes (red) . The enzymes digest a path through the zona pellucida, enabling the sperm to advance . Eventually, the sperm meets and fuses with the secondary oocyte plasma membrane and this triggers cortical and zona reactions . First, enzyme-rich cortical granules in the oocyte cytoplasm release their contents (yellow) into the zona pellucida, starting at the point of fusion and progressing right and left . Next, in the zona reaction, the enzymes modify the zona pellucida, transforming it into an impenetrable barrier to sperm as a guard against polyspermy (multiple fertilization) . Neck MidpieceHead Perivitelline space Zona pellucidaContinuation of reaction Acrosome-reacted sperm Penetration Calcium wave starting at sperm entry point Cortical reactionFusionZona reaction Excluded spermInitiation of acrosome reaction Binding TailNucleusPlasma membrane Acrosomal contentsFusion of plasma membraneand outer acrosomal membrane Vesiculation Acrosome-reacted sperm Inner acrosomal membraneOuter acrosomal membrane Cortical granule Plasma membrane Fig . 8 .5 A, An unfertilized human secondary oocyte surrounded by the zona pellucida; the first polar body can be seen . Spermatozoa are visible outside the zona pellucida . B, A fertilized human ootid before fusion of the pronuclei . Two polar bodies can be seen beneath the zona pellucida . A B
251
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Preimplantation development 167 CHaPter 8 apolar cells contribute almost exclusively to the inner cell mass in most embryos. In cleavage, the generation of cell diversity, to either trophectoderm or inner cell mass, occurs in the 16-cell morula and precedes the forma - tion of the blastocyst. During the 16-cell cycle, the outer polar cells continue to differentiate an epithelial phenotype, and display further aspects of polarity and intercellular adhesion typical of epithelial cells, while the inner apolar cells remain symmetrically organized. During the next cell division (16–32 cells), a proportion of polar cells again divide differentiatively, as in the previous cycle, each yielding one polar and one apolar progeny, which enter the trophectoderm and inner cell mass lineages, respectively. Although differentiative division at this time is less common than at the 8- to 16-cell transition, it has the important function of regulating an appropriate number of cells in the two tissues of the blastocyst. Thus, if differentiative divisions were relatively infre - quent at the 8- to 16-cell transition, they will be more frequent at the 16- to 32-cell transition, and vice versa. After division to 32 cells, the outer polar cells complete their dif - ferentiation into a functional epithelium, display structurally complete zonular tight junctions and begin to form desmosomes. The nascent trophectoderm engages in vectorial fluid transport in an apical to basal direction to generate a cavity that expands in size during the 32- to 64-cell cycles and converts the ball of cells, the morula, to a sphere, the blastocyst (Fig. 8.7). Once the blastocyst forms, the diversification of the trophectoderm and inner cell mass lineages is complete, and troph - ectoderm differentiative divisions no longer occur. In the late blastocyst, the trophectoderm is referred to as the trophoblast, which can be divided into polar trophoblast, lying in direct contact with the inner cell mass, and mural trophoblast, surrounding the blastocyst cavity (Fig. 8.8). Blastocyst The blastocyst ‘hatches’ from its zona pellucida at 6–7 days, possibly assisted by an enzyme similar to trypsin (see Figs 8.7C, 8.8). Tropho - blast oozes out of a small slit; many embryos form a figure-of-eight shape bisected by the zona pellucida, especially if it has been hardened during oocyte maturation and cleavage. Such half-hatching could result in the formation of identical twins. Hatched blastocysts expand and differentiation of the inner cell mass proceeds (see Fig. 8.8). The outer cells of the blastocyst – the trophoblast or trophectoderm – are flattened polyhedral cells, which possess ultrastructural features typical of a transporting epithelium. The trophoblast covering the inner cell mass is the polar trophoblast and that surrounding the blastocyst cavity is the mural trophoblast. The free, unattached blastocyst is assigned to stage 3 of development at approximately 4 days post ovula - tion, whereas implantation (before villus development) occurs within a period of 7–12 days post ovulation and over the next two stages of Fig . 8 .6 Successive stages of cleavage of a human ootid . A, The two-cell stage . B, The three-cell stage . C, The five-cell stage . D, The eight-cell stage . A B C D Fig . 8 .7 Human embryos . Formation of a morula and blastocyst within the zona pellucida and blastocyst hatching from the zona pellucida . A, A ball of cells, the morula, with the cells undergoing compaction . B, The blastocyst cavity is developing and the inner cell mass can be seen on one side of the cavity . C, The blastocyst is beginning to hatch from the zona pellucida . B A C Fig . 8 .8 A human blastocyst nearly completely hatched from the zona pellucida . The blastocyst can now expand to its full size . development. Even at this early stage, cells of the inner cell mass are already arranged into an upper layer (i.e. closest to the polar tropho - blast), the epiblast, which will give rise to the embryonic cells, and a lower layer, the hypoblast, which has an extraembryonic fate. Thus, the dorsoventral axis of the developing embryo and a bilaminar arrange - ment of the inner cell mass are both established at or before implanta - tion. (The earliest primordial germ cells may also be defined at this stage.) Attachment to the uterine wall On the sixth postovulatory day, the blastocyst adheres to the uterine mucosa and the events leading to the specialized, intimate contact of trophoblast and endometrium begin. Implantation, which is the term used for this complicated process, includes the following stages: dissolu - tion of the zona pellucida; orientation and adhesion of the blastocyst on to the endometrium; trophoblastic penetration into the endometrium; migration of the blastocyst into the endometrium; and spread and pro - liferation of the trophoblast, which envelops and specifically disrupts
252
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Preim Plantation develo Pment 168SeCtion 2 single ovum, or both. It is most likely to be seen in women treated with drugs to stimulate ovulation. The range of separation of twin embryos is reflected in the separation of the extraembryonic membranes. The types of placentation that can occur are shown in Figure 8.9. Monoamniotic, monochorionic placen- tae are associated with the greatest perinatal mortality (50%), caused both by entanglement of the umbilical cords impeding the blood supply and by various vascular shunts between the placentae, which may divert blood from one fetus to the other. Artery–artery anastomoses are the most common, followed by artery–vein anastomoses. If the shunting of blood across the placentae from one twin to the other is balanced by more than one vascular connection, development may proceed unimpaired. However, if this is not the case, one twin may receive blood from the other, leading to cardiac enlargement, increased urination and polyhydramnios in the recipient, and anaemia, oligo - hydramnios and atrophy in the donor. Dizygotic twins have either completely separate chorionic sacs or sacs that have fused. Such placentae are separated by four membranes, two amnia and two choria; in addition, these placentae have a ridge of firmer tissue at the base of the dividing membranes, caused by the abut - ting of two expanding placental tissues against each other. FORMATION OF EXTRAEMBRYONIC TISSUES The earliest developmental processes in mammalian embryos involve the production of those extraembryonic structures that will support and nourish the embryo during development. Production of these layers and invades the maternal tissues (Ch. 9). The interactions between the hatched blastocyst and the maternal endometrium are not well under - stood. Developmentally competent embryos appear to enhance the uterine environment and promote their own implantation (Brosens et al 2014, Macklon and Brosens 2014). The site of implantation is normally in the endometrium of the posterior wall of the uterus, nearer to the fundus than to the cervix; it may be in the median plane or to one or other side, but may occur elsewhere in the uterus, or in an extrauterine or ectopic site. Ectopic implantation The conceptus may be arrested at any point during its migration through the uterine tube and implant in its wall. Previous pelvic inflammation damages the tubal epithelium and may predispose to such delay in tubal transport. The presence of an intrauterine contraceptive device or the use of progesterone-based oral contraceptives may also predispose to ectopic pregnancy, probably because of alteration in the normal tubal transport mechanisms. Nidation of the embryo as an ectopic pregnancy most frequently occurs in the wider ampullary portion of the uterine tube, but may also occur in the narrow intramural part or even in the ovary itself. Most ectopic pregnancies are anembryonic, although the continuing growth of the trophoblast will produce a positive pregnancy test, and may cause rupture of the uterine tube and significant intraperitoneal haemorrhage. Ectopic pregnancies with a live embryo are the most dangerous because they grow rapidly and may be detected only when they have eroded the uterine tube wall and surrounding blood vessels, as early as 8 weeks of pregnancy. Similarly, cornual ectopics (in the intramural part of the tube) may present with catastrophic haemorrhage because there is a substantial blood supply in the surrounding muscularis. Ovarian or abdominal pregnancies are exceptionally rare. Although some are presumed to have been caused by fertilization occurring in the vicinity of the ovary (primary), most are probably caused secondar - ily and result from an extrusion of the conceptus through the abdomi - nal ostium of the tube. Apart from their important clinical implications, these conditions emphasize the fact that the conceptus can implant successfully into tissues other than a normal progestational endometrium. Prolonged development can occur in such sites and is usually terminated by a mechanical or vascular accident and not by a fundamental nutritive or endocrine insufficiency or by an immune maternal response. Abdomi - nal implantation may occur on any organ, e.g. bowel, liver or omentum. If such a pregnancy continues, this makes removal of the placenta at delivery or abortion hazardous as a result of haemorrhage; conse - quently, the placenta is usually left in situ to degenerate spontaneously. Twinning Spontaneous twinning occurs once in about every 80 births. Monozy - gotic twins arise from a single ovum fertilized by a single sperm. At some stage up to the establishment of the axis of the embryonic area and the development of the primitive streak, the embryonic cells sepa- rate into two parts, each of which gives rise to a complete embryo. The process of hatching of the blastocyst from the zona pellucida may result in constriction of the emerging cells and separation into two discrete entities. There is a gradual decrease in the average thickness of the zona pellucida with increasing maternal age, which may be causally related to the increase in frequency of monozygotic twinning with increased maternal age. The resultant twins have the same genotype but the description ‘identical twins’ is best avoided, since most monozygotic twins have differences in phenotypes. Late separation of twins from a single conceptus may result in conjoined twins; these may be equal or unequal, as in acardia. After twinning, monozygotic embryos enter a period of intense catch-up growth. Despite starting out at half the size, each twin embryo or fetus is of a size comparable to a singleton fetus in the second trimester of pregnancy, but declines in relative size in the last 10 weeks of pregnancy. The sex of monozygotic twins will be the same. Monoamniotic, monochorionic, monozygotic twins are most likely to be female, as are acardiac twins. The male to female ratio for all monozygotic twins is 0.487, and for monoamniotic, monochorionic twins it is 0.231. Dizygotic twins represent the most frequent form of twinning. They result from multiple ovulations, which can be induced by gonado - trophins or drugs commonly used in patients with infertility. Dizygotic twins may be different sexes; like-sex pairs are more common. The male to female ratio is 0.518. Multiple births greater than twinning, such as triplets or quadruplets, can arise from multiple ovulations, or from a Fig . 8 .9 Relationships of the extraembryonic membranes in different types of twinning . A, Diamnionic, dichorionic separated; i .e . separation of the first two blastomeres results in separate implantation sites . B, Diamnionic, dichorionic fused; here the chorionic membranes are fused but the fetuses occupy separate choria . C, Diamnionic, monochorionic; reduplication of the inner cell mass can result in a single placenta and chorionic sacs but separate amniotic cavities . D, Monoamnionic, monochorionic; duplication of the embryonic axis results in two embryos sharing a single placenta, chorion and amnion . E, Incomplete separation of the embryonic axis results in conjoined twins . F, Unequal division of the embryonic axis or unequal division of the blood supply may result in an acardiac monster . A EB C D F
253
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Formation of extraembryonic tissues 169 CHaPter 8mural trophoblast and, rarely, may also share gap junctions. The visceral hypoblast cells are cuboidal; they have a uniform apical surface towards the blastocyst cavity but irregular basal and lateral regions, with flanges and projections underlying one another and extending into intercel - lular spaces. There is no basal lamina subjacent to the visceral hypo - blast, and the distance between the hypoblast cells and the epiblast basal lamina is variable. A series of modifications of the original blastocystic cavity develops beneath the hypoblast later than those developing above the epiblast. While the amniotic cavity is enlarging within the sphere of epiblast cells, the parietal hypoblast cells are proliferating and spreading along the mural trophoblast until they extend most of the way around the circumference of the blastocyst, converging towards the abembryonic pole. At the same time, a space appears between the parietal hypoblast and the mural trophoblast that limits the circumference of the hypo - blastic cavity. A variety of terms have been applied to the parietal hypoblast layer: extraembryonic hypoblast, extraembryonic endoderm and exocoelomic (Heuser’s) membrane. The cavity that the layer ini - tially surrounds is the primary yolk sac (primary umbilical vesicle), and the resultant smaller cavity lined by hypoblast is the secondary yolk sac. It has been suggested that the secondary yolk sac forms in a variety of ways, including cavitation of visceral hypoblast (a method similar to formation of the amnion), rearrangement of proliferating visceral hypoblast and folding of the parietal layer of the primary yolk sac into the secondary yolk sac. Further development of the yolk sac is described on page 177. The visceral hypoblast cells are thought to be important in many aspects of the early specification of cell lines and control of the timing of early development. They induce the formation and position of the primitive streak in the midline of the embryo, thus establishing the first axis of the embryonic disc, and they influence the timing of epiblast epithelial-to-mesenchyme transition at the primitive streak. Hypoblast cells remain beneath the primitive streak; their experimental removal in avian embryos causes multiple embryonic axes to form. Hypoblast cells are also believed to be necessary for successful induction of the head region and for the successful specification of the primordial germ cells. With the later formation of the embryonic cell layers from the epiblast, the visceral hypoblast appears to be sequestered into the sec - ondary yolk sac wall by the expansion of the newly formed embryonic endoderm beneath the epiblast, although there is now evidence that some hypoblast cells may remain in the final endoderm layer and con - tribute to the gut, and that they may also play a role in the induction of cardiogenesis. For a review of the roles of hypoblast cells, see Stern and Downs (2012). After the formation of the secondary yolk sac, a diverticulum of the visceral hypoblast, the allantois, forms towards one end of the embryonic region and extends into the local extraembryonic mesoblast. It passes from the roof of the secondary yolk sac to the same plane as the amnion. Further development of the allantois is described on page 178. Extraembryonic mesoblast By definition, extraembryonic tissues encompass all tissues that do not contribute directly to the future body of the definitive embryo and, later, the fetus. At stage 5, blastocysts are implanted but do not yet display trophoblastic villi (see Fig. 8.10); they range from 7 to 12 days in age. A feature of this stage is the first formation of extraembryonic meso - blast, which will come to cover the amnion, secondary yolk sac and the internal wall of the mural trophoblast, and will form the connecting stalk of the embryo with its contained allantoenteric diverticulum. The origin of this first mesoblastic extraembryonic layer is by no means clear; it may arise from several sources, including the caudal region of the epiblast, the parietal hypoblast and subhypoblastic cells. The tro - phoblastic origin of extraembryonic mesoblast is questioned because there is always a complete basal lamina underlying the trophoblast; the migration of cells out of an epithelium is usually associated with previ - ous disruption of the basal lamina. Certainly, the origin of extraembry - onic cells will change over time as new germinal populations are established. The first mesoblastic extraembryonic layer gives rise to the layer known as extraembryonic mesoblast, arranged as a mesothelium with underlying extraembryonic mesenchymal cells; this also appears to form an extracellular structure corresponding to the magma reticulare, between the mural trophoblast and the primary yolk sac in the stage 5 embryo. Later extraembryonic mesoblast populations mushroom beneath the cytotrophoblastic cells at the embryonic pole, forming the cores of the developing villus stems, and villi and the angioblastic begins before implantation is complete. At present, it is unclear where the extraembryonic cell lines arise. The trophoblast was considered to be a source but evidence now points to the inner cell mass as the site of origin. Figure 8.1 shows the sequence of development of various tissues in the early embryo. Epiblast and amniotic cavity Epiblast cells are closest to the implanting face of the trophoblast and have a definite polarity; they are arranged in a radial manner with extensive junctions near the centre of the mass of cells, supported by supranuclear organelles. A few epiblast cells are contiguous with cytotrophoblast cells; apart from this contact, a basal lamina surrounds what is initially a spherical cluster of epiblast cells, and isolates them from all other cells. Those epiblast cells adjacent to the hypoblast become taller and more columnar than those adjacent to the tropho - blast, and this causes the epiblast sphere to become flattened and the centre of the sphere to be shifted towards the polar trophoblast. Amni- otic fluid accumulates at the eccentric centre of the now lenticular epiblast mass, which is bordered by apical junctional complexes and microvilli. As further fluid accumulates, an amniotic cavity forms, roofed by low cuboidal cells that possess irregular microvilli. The cells share short apical junctional complexes and associated desmosomes, and rest on an underlying basal lamina. The demarcation between true amnion cells and those of the remaining definitive epiblast is clear. The columnar epiblast cells are arranged as a pseudostratified layer with microvilli, frequently a single cilium, clefted nuclei and large nucleoli; the cells have a distinct, continuous basal lamina. Cell division in the epiblast tends to occur near the apical surface, causing this region to become more crowded than the basal region. At the margins of the embryonic disc, the amnion cells are contiguous with the epiblast; there is a gradation in cell size from columnar to low cuboidal within a two- to three-cell span ( Fig. 8.10; see Fig. 9.1). Further development of the amnion and amniotic fluid is described on page 178. Hypoblast and yolk sac Hypoblast is the term used to delineate the lower layer of cells of the early bilaminar disc, most commonly in avian embryos. This layer is also termed anterior, or distal, visceral endoderm in the mouse embryo. Just before implantation, the hypoblast consists of a layer of squamous cells that is only slightly larger in extent than the epiblast. The cells exhibit polarity, in that apical microvilli face the cavity of the blastocyst and apical junctional complexes, but they lack a basal lamina. During early implantation, the hypoblast extends beyond the edges of the epiblast and can now be subdivided into those cells in contact with the epiblast basal lamina, the visceral hypoblast, and those cells in contact with the mural trophoblast, the parietal hypoblast. The parietal hypo - blast cells are squamous; they may share adhesion junctions with the Fig . 8 .10 The implanting conceptus at stage 5b . The embryo at this stage is composed only of the abutting epiblast and hypoblast layers . It is suspended within the blastocyst cavity (chorionic cavity) and surrounded by a layer of cytotrophoblast . A large mass of syncytiotrophoblast, with interconnecting lacunae, is penetrating the maternal endometrium . Cytotrophoblast SyncytiotrophoblastHypoblast
254
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Preim Plantation develo Pment 170SeCtion 2cells that will give rise to the capillaries within them and the earliest blood cells. Initially, the extraembryonic mesoblast connects the amnion to the chorion over a wide area. Continued development and expansion of the extraembryonic coelom means that this attachment becomes increasingly circumvented to a connecting stalk, which is a permanent connection between the future caudal end of the embryonic disc and the chorion. The connecting stalk forms a pathway along which vascular anastomoses around the allantois establish communication with those of the chorion. KEY REFERENCES Brosens JJ, Walker MS, Teklenburg G 2014 Uterine selection of human embryos at implantation. Sci Reports 4:3894. Macklon NS Brosens JJ 2014 The human endometrium as a sensor of embryo quality. Biol Reprod 91:98. O’Rahilly R, Müller F 1987 Developmental Stages in Human Embryos. Washington: Carnegie Institution. O’Rahilly R, Müller F 1999 The Embryonic Human Brain. An Atlas of Developmental Stages, 2nd ed. New York: Wiley-Liss. O’Rahilly R, Müller F 2000 Minireview: Prenatal ages and stages – measures and errors. Teratology 61:382–4.O’Rahilly R, Müller F 2010 Developmental stages in human embryos: revised and new measurements. Cells Tissues Organs 192:73–84. Stern CD, Downs KM 2012 The hypoblast (visceral endoderm): an evo-devo perspective. Development 139:1059–69. Streeter GL 1942 Developmental horizons in human embryos. Descriptions of age group XI, 13 to 20 somites, and age group XII, 21 to 29 somites. Contrib Embryol Carnegie Inst Washington 30:21 1–45.Video 8 .1 Human in vitro fertilization and early development .  Bonus  e-book  video
255
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Preimplantation development 170.e1 CHaPter 8REFERENCES Brosens JJ, Walker MS, Teklenburg G 2014 Uterine selection of human embryos at implantation. Sci Reports 4:3894. Macklon NS Brosens JJ 2014 The human endometrium as a sensor of embryo quality. Biol Reprod 91:98. O’Rahilly R, Müller F 1987 Developmental Stages in Human Embryos. Washington: Carnegie Institution. O’Rahilly R, Müller F 1999 The Embryonic Human Brain. An Atlas of Developmental Stages, 2nd ed. New York: Wiley-Liss. O’Rahilly R, Müller F 2000 Minireview: Prenatal ages and stages – measures and errors. Teratology 61:382–4.O’Rahilly R, Müller F 2010 Developmental stages in human embryos: revised and new measurements. Cells Tissues Organs 192:73–84. Stern CD, Downs KM 2012 The hypoblast (visceral endoderm): an evo-devo perspective. Development 139:1059–69. Streeter GL 1942 Developmental horizons in human embryos. Descriptions of age group XI, 13 to 20 somites, and age group XII, 21 to 29 somites. Contrib Embryol Carnegie Inst Washington 30:21 1–45.
256
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
171 CHAPTER 9 Implantation and placentation and during leukocyte emigration from the blood into the tissues. Flanges of syncytial trophoblast grow between the cells of the uterine luminal epithelium towards the underlying basal lamina without apparent damage to the maternal cell membranes or disruption of the intercellular junctions. Instead, shared junctions, including tight junc ­ tions, are formed with many of the maternal uterine epithelial cells. Implantation continues with erosion of maternal vascular endothe ­ lium and glandular epithelium, and phagocytosis of secretory products, until the blastocyst occupies an uneven implantation cavity in the stroma (interstitial implantation) (Fig. 9.1). In the early postimplant­ ation phase, the maternal surface is resealed by re ­epithelialization and the formation of a plug, which may contain fibrin. As the blastocyst burrows more deeply into the endometrium, syncytial trophoblast forms over the mural cytotrophoblast, but never achieves the thickness of the syncytial trophoblast over the embryonic pole. The syncytiotrophoblast, which expresses neither class I nor class II major histocompatibility complex (MHC) antigens, secretes numerous hormones; human chorionic gonadotrophin (hCG) can be detected in maternal urine from as early as 10 days after fertilization and forms the basis for tests for early pregnancy. This hCG prolongs the life of the corpus luteum, which continues to secrete progesterone and oestrogens during approximately the first 2 months of pregnancy, until these essen ­ tial hormones are produced by the placenta. Menstruation ceases on successful implantation. The endometrium, now known as the decidua in pregnancy, thickens to form a suitable nidus for the conceptus. Decidualization of the endometrial stroma may occur without an intrauterine pregnancy, e.g. in the presence of an ectopic pregnancy, after prolonged treatment with progesterone, and in the late secretory phase of a non ­conception cycle. Decidual differentiation is not evident in the stroma at the earliest stages of implantation, and it may not be until a week later that fully IMPLANTATION Implantation involves the initial attachment of the trophoblastic wall of the blastocyst to the endometrial luminal epithelium and its decidual response. It is now apparent that, during this process, there is an inter­ active dialogue between the implanting embryo and endometrial decidu al stromal cells. Competent preimplantation human embryos actively enhance the uterine environment for their implantation, whereas developmentally impaired embryos induce endoplasmic retic ­ ulum stress responses in decidual cells that inhibit implantation (Brosens et al 2014). The blastocyst/trophoblast lineage gives rise to three main cell types in the human placenta: syncytiotrophoblast cells form the epithelial covering of the villous tree and are the main endocrine component of the placenta; villous cytotrophoblast cells represent a germinative popu ­ lation that proliferates throughout pregnancy, fusing to generate syn­ cytiotrophoblast; and extravillous trophoblast cells are non ­proliferative and invade the maternal endometrium. The first two cell lines can be seen from stages 4 and 5 onwards. The cytotrophoblast cells that form the mural and polar trophoblast are cuboidal and covered externally with syncytial trophoblast (syncytiotrophoblast), a multinucleated mass of cytoplasm that forms initially in areas near the inner cell mass after apposition of the blastocyst to the uterine mucosa (see Fig. 8.10). Preimplantation embryos produce matrix metalloproteinases (MMPs) that mediate penetration of the maternal subepithelial basal lamina by the syncytiotrophoblast. Trophoblast cells express L ­selectin (usually seen as a mediator of neutrophil rolling and tethering in inflamed endothelium), and the maternal epithelium upregulates selectin ­oligosaccharide ­based ligands. Thus, differentiating cytotro ­ phoblast cells appear to use processes that also occur in vasculogenesis Fig. 9.1 The implanting conceptus at stage 6. The embryo is composed of epiblast, with the amniotic cavity above it, and hypoblast, with the secondary yolk sac below it (see also Figs 8.10, 10.1). Both cavities are covered externally with extraembryonic mesoblast, which also lines the larger chorionic cavity. Primary villi cover the outer aspect of the conceptus and extend into the maternal endometrium, and in places, maternal blood fills the lacunae. Cytotrophoblast SyncytiotrophoblastDecidua Extraembryonic mesoblastUterine glands Amniotic cavity Secondary yolk sac Lacuna Extraembryonic coelom (chorionic cavity)Uterine blood vessel Amnion Yolk sac wall Chorion Epithelium of endometrial surface Coagulum over implanting blastocystVillous stems
257
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Implantat Ion and placentat Ion 172SectIon 2differentiated cells are present. During decidualization, the interglan ­ dular tissue increases in quantity (Burton et al 2010). It contains a substantial population of leukocytes (large granular lymphocytes, macrophages and T cells) distributed amongst large decidual cells. The most numerous are uterine natural killer (NK) cells, which accumulate in the endometrium during the secretory phase of the cycle and persist until mid ­pregnancy; they interact with invading extravillous tropho­ blast cells expressing human leukocyte antigen G and C (HLA ­G, HLA­C). Decidual cells are stromal cells that contain varying amounts of glycogen, lipid and vimentin ­type intermediate filaments in their cytoplasm. They are generally rounded but their shape may vary, depending on the local packing density. They may contain one, two or sometimes three nuclei, frequently display rows of club ­like cytoplas ­ mic protrusions enclosing granules, and are associated with a charac ­ teristic capsular basal lamina. Decidual cells produce a range of secretory products, including insulin ­like growth factor binding protein 1 (IGF ­ BP1) and prolactin, which may be taken up by the trophoblast. These secretions probably play a role in the maintenance and growth of the conceptus in the early part of postimplantational development, and can be detected in amniotic fluid in the first trimester of pregnancy. Extracellular matrix, growth factors and protease inhibitors pro ­ duced by the decidua all probably modulate the degradative activity of the trophoblast and support placental morphogenesis and placental accession to the maternal blood supply. Once implantation is complete, distinctive names are applied to different regions of the decidua ( Fig. 9.2). The part covering the conceptus is the decidua capsularis; that between the conceptus and the uterine muscular wall is the decidua basalis (where the placenta subsequently develops); and that which lines the remainder of the body of the uterus is the decidua parietalis. There is no evidence that their respective resident maternal cell popula ­ tions exhibit site ­specific properties. DEVELOPMENT OF THE PLACENTA Formation of the human placenta requires a developmental progression that proceeds in a specific chronological order: the development of primary, secondary and anchoring villi, and the local differentiation of haemangioblast cells within them, occur concurrently with the modifi ­ cation of maternal blood vessels to ensure their patency. With the onset of the embryonic heart beat, a primitive circulation exists between the embryo and the secondary yolk sac. An embryonic–placental circula ­ tion starts around week 8 of gestation; over the course of 40 weeks, the placenta develops into a highly vascular organ.Fig. 9.2 The gravid uterus in the second month. A placental site precisely in the uterine fundus, as indicated on the figure, is rather unusual; the dorsal, ventral or lateral wall of the corpus uteri is more usual. The maternal endometrium is now termed decidua; different regions are distinguished. Uterine tube Umbilical cordConnecting stalk Decidua parietalis Chorion Chorionic cavity Myometrium Plug of mucusEmbryowithinamnionDeciduacapsularisAmnioticcavityYolk sac Vitelline ductVillousstemsof chorionfrondosumIntervillous space Decidua basalis Fig. 9.3 A, The growing syncytiotrophoblast erodes decidual glands and spiral arteries, and their contents form lacunae within the syncytiotrophoblast. Cytotrophoblastic cells extend as villi into the syncytiotrophoblast. B, Cells originating from the cytotrophoblast move into the decidua, forming interstitial extravillous trophoblast cells, the third line of cytotrophoblastic cells. These surround the opened glands and spiral arteries in the decidua and inner myometrium. Cells that move through the tunicae adventitia, media and interna to enter the lumen of the vessels become endovascular extravillous trophoblast. These cells remodel the walls and plug the lumen of the spiral arteries, permitting only plasma to enter the forming intervillous space. Glandular secretions in the intervillous space provide histiotrophic nutrition to the embryo. Syncytiotrophoblast Cytotrophoblast Mesenchyme AmnionLacunae Endometrial gland Spiral artery DeciduaSpiral arteryEndometrial gland Decidua Glandular secretions in intervillous spaceEndovascular extravillous trophoblast Plasma in intervillous spaceCytotrophoblast Interstitial extravillous trophoblastSyncytiotrophoblast A BAs the blastocyst implants, the syncytiotrophoblast invades the uterine tissues, including the glands and walls of maternal blood vessels (see Fig. 9.1; Fig. 9.3), and increases rapidly in thickness over the embryonic pole ( Fig. 9.4). A progressively thinner layer covers the rest of the wall towards the abembryonic pole. Microvillus ­lined clefts and lacunar spaces develop in the syncytiotrophoblastic envelope (days 9–1 1 of pregnancy) and establish communications with one another. Initially, many of these spaces contain maternal blood derived from dilated uterine capillaries and veins, as the walls of the vessels are par ­ tially destroyed. As the conceptus grows, the lacunar spaces enlarge, and become confluent to form intervillous spaces. The projections of syncytiotrophoblast into the maternal decidua are called primary villi. They are invaded initially with cytotrophoblast and
258
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
development of the placenta 173 cHapte R 9 Fig. 9.5 Placental development. CytotrophoblastLacunae in syncytiumDEVELOPMENT Fetal vessels Stem villus True villus Secondary syncytial fusion Terminal villus Connective tissue AmnionMaternal blood Ingrowth of cytotrophoblast and mesoblast bearing fetal vesselsLacunarcirculationLacunaeenlarging asintervillousspacesCytotrophoblasticcell columnSyncytiotrophoblastCotyledonary septum Maternal vesselsin deciduaBasal plate Intervillous space Chorionic platethen with extraembryonic mesenchyme (days 13–15) to form second ­ ary placental villi. Capillaries develop in the mesenchymal core of the villi during the third week post conception. The cytotrophoblast within the villi continues to grow through the invading syncytiotrophoblast and makes direct contact with the decidua basalis, forming anchoring villi. Further cytotrophoblast proliferation occurs laterally so that neigh ­ bouring outgrowths meet to form a spherical cytotrophoblastic shell around the conceptus (Fig. 9.5; see Fig. 9.3B). Lateral projections from the main stem villus form true villi. Elongation of growing capillaries outstrips that of the containing villi, leading to looping of vessels. This obtrusion of both capillary loops and new sprouts results in the forma ­ tion of terminal villi. As secondary villi form, single mononuclear cells become detached from the distal (anchoring) cytotrophoblast and infiltrate the maternal decidua (Fig. 9.6; see Fig. 9.3B). These extravillous cells are the third line of the original trophoblastic cells. Interstitial extravillous trophob ­ last cells invade the maternal spiral arteries from their adventitia. The smooth muscle and internal elastic lamina are replaced with extracel ­ lular fibrinoid deposits (Harris 2010). Those cells that pass through the endothelial basement membrane and gain access to the vessel lumen are termed endovascular extravillous trophoblast (see Figs 9.3B, 9.6). These cells plug the maternal spiral arteries until the end of the first trimester; the cells also migrate antidromically, against the flow of maternal blood, along the spiral arteries as far as the inner myometrial region (Huppertz et al 2014). The definitive placenta is composed of a chorionic plate on its fetal aspect and a basal plate on its maternal aspect, separated by an interven ­ ing intervillous space containing villous stems with branches in contact with maternal blood (see Fig. 9.5). During the first trimester, develop ­ ment takes place in a low ­oxygen environment supported by histio ­ trophic nutrition from the endometrial glands, which discharge into the intervillous space until at least 10 weeks (see Fig. 9.3A) (Burton et al 2002, Burton et al 2007, Burton and Fowden 2012). When maternal blood bathes the surfaces of the chorion that bound the intervillous space, the human placenta is defined as haemochorial. Different grades of fusion exist between the maternal and fetal tissues in many other mammals (e.g. epitheliochorial, syndesmochorial, endotheliochorial). The chorion is vascularized by the allantoic blood vessels of the body stalk, and so the human placenta is also termed chorio ­allantoic (whereas, in some mammals, a choriovitelline placenta either exists alone or supplements the chorio ­allantoic variety). The human placenta is also defined as discoidal, in contrast to other shapes in other mammals, and deciduate because maternal tissue is shed with the pla­ centa and membranes at parturition as part of the afterbirth. Growth of the placenta Expansion of the entire conceptus is accompanied by radial growth of the villi and, simultaneously, an integrated tangential growth and expansion of the trophoblastic shell. Eventually, each villous stem forms a complex that consists of a single trunk attached by its base to the chorion, from which second ­ and third ­order branches (intermedi ­ ate and terminal villi) arise distally. Terminal villi are specialized for exchange between fetal and maternal circulations; each one starts as a syncytial outgrowth and is invaded by cytotrophoblastic cells, which then develop a core of fetal mesenchyme as the villus continues to grow. The core is vascularized by fetal capillaries (i.e. each villus passes through primary, secondary and tertiary grades of histological differen ­ tiation). The germinal cytotrophoblast continues to add cells that fuse with the overlying syncytium and so contribute to the expansion of the haemochorial interface. Terminal villi continue to form and branch within the confines of the definitive placenta throughout gestation, projecting in all directions into the intervillous space (see Fig. 9.5). From the third week until about the second month of pregnancy, the entire chorion is covered with villous stems. They are thus continu ­ ous peripherally with the trophoblastic shell, which is in close appos­ition with both the decidua capsularis and the decidua basalis. The villi adjacent to the decidua basalis are stouter and longer, and show a greater profusion of terminal villi. As the conceptus continues to expand, the decidua capsularis is progressively compressed and thinned, the circulation through it is gradually reduced, and adjacent villi slowly atrophy and disappear. This process starts at the abembryonic pole; by the end of the third month, the abembryonic hemisphere of the Fig. 9.4 The conceptus at about stage 14. The embryonic pole shows extensive villous formation at the chorion frondosum, whereas the abembryonic pole is smooth and villous-free at the chorion laeve. (Photograph courtesy of P Collins.)
259
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Implantat Ion and placentat Ion 174SectIon 2 Chorionic plate The chorionic plate is covered on its fetal aspect by amniotic epithe ­ lium; on the stromal side of the epithelium, a connective tissue layer carries the main branches of the umbilical vessels (see Fig. 9.5; Fig. 9.7). Subjacent to this are a diminishing layer of cytotrophoblast and then the inner syncytial wall of the intervillous space. The connective tissue layer is formed by fusion between the mesenchyme ­covered surfaces of amnion and chorion, and is more fibrous and less cellular than Whar ­ ton’s jelly (of the umbilical cord), except near the larger vessels. The umbilical vessels radiate and branch from the cord attachment, with variations in the branching pattern, until they reach the bases of the trunks of the villous stems; they then arborize within the intermediate and terminal villi. There are no anastomoses between vascular trees of adjacent stems. The two umbilical arteries are normally joined at, or just before they enter, the chorionic plate, by some form of substantial transverse anastomosis (Hyrtl’s anastomosis). Basal plate The basal plate, from fetal to maternal aspect, forms the outer wall of the intervillous space. The trophoblast and adjacent decidua are enmeshed in layers of fibrinoid and basement membrane ­like extracel ­ lular matrix to form a complex junctional zone. In different places, the basal plate may contain syncytium, cytotrophoblast or fibrinoid matrix, remnants of the cytotrophoblastic shell, and, at the site of implantation, areas of necrotic maternal decidua (the so ­called Nitabuch’s stria) (see Figs 9.5 and 9.7). Nitabuch’s stria and the decidua basalis contain cytotrophoblast and multinucleate trophoblast giant cells derived from the mononuclear extravillous interstitial cytotrophoblast population that infiltrate the decidua basalis during the first 18 weeks of pregnancy. These cells penetrate as far as the inner third of the myometrium but can often be observed at or near the decidual–myometrial junction. They are not found in the decidua parietalis or the adjacent myo ­ metrium, suggesting that the placental bed giant cell represents a dif­ ferentiative end stage in the extravillous trophoblast lineage. The striae of fibrinoid are irregularly interconnected and variable in prominence. Strands pass from Nitabuch’s stria into the adjacent decidua, which contains basal remnants of the endometrial glands and large and small decidual cells scattered in a connective tissue framework that supports an extensive venous plexus. Initially, only the central area of a placenta contains extravillous trophoblast cells, both within and around the spiral arteries. These cells conceptus is largely denuded. Eventually, the whole chorion apposed to the decidua capsularis is smooth and is now termed the chorion laeve; this will later become apposed to the amnion, forming chorio ­ amnion (see Figs 9.2, 9.9). In contrast, the villous stems of the disc ­ shaped region of chorion apposed to the decidua basalis increase greatly in size and complexity, and the region is now termed the chorion frondosum (see Fig. 9.4). The chorion frondosum and the decidua basalis constitute the definitive placental site (see Fig. 9.2). Abnormali ­ ties in this process may account for the persistence of villi at abnormal sites in the chorion laeve of the gestational sac and hence the presence of accessory or succenturiate lobes within the membranes of the defini ­ tive placenta. The presence of the entire villous ring of the primitive placenta beyond the second month post ­menstruation leads to the development of placenta membranacea, a giant definitive placental structure surrounding the whole gestational sac during the remainder of pregnancy. Coincidentally with the growth of the embryo and the expansion of the amnion, the decidua capsularis is thinned and distended, and the space between it and the decidua parietalis is gradually obliterated. By the second month of pregnancy, the three endometrial strata recogniz ­ able in the premenstrual phase, i.e. compactum, spongiosum and basale, are better differentiated and easily distinguished. The glands in the spongiosum are compressed and appear as oblique slit ­like fissures lined by low cuboidal cells. By the beginning of the third month of pregnancy, the decidua capsularis and decidua parietalis are in contact. By the fifth month, the decidua capsularis is greatly thinned, and it virtually disappears during the succeeding months. Focal bleeding often occurs in the periphery of the developing pla ­ centa at the time of the formation of the membranes (8–12 weeks). This complication, which is termed threatened miscarriage, is a common clinical complication of pregnancy and can lead to a complete miscar ­ riage if the haematoma extends to the definitive placenta (Jauniaux et al 2006). At term, the placental diameter varies from 200 to 220 mm, the mean placental weight is 470 g, its mean thickness is 25 mm and the total villous surface area is 12–14 m2, providing an extensive and inti ­ mate interface for materno–fetal exchange; the uteroplacental circula ­ tion carries approximately 600 ml of maternal blood per minute. There are no lymph vessels in the placenta and, equally, no aggregates of lymphoid cells. Systematic evaluation and careful description of the placenta after delivery have been recommended for correlation with later neonatal neurodevelopmental outcomes (Roescher et al 2014).Fig. 9.6 As the placenta grows, the interstitial and endovascular extravillous trophoblast cells continue to remodel the spiral arteries into large-bore, low-resistance uteroplacental vessels. These arteries remain occluded by endovascular extravillous trophoblast until the end of the first trimester, promoting a low oxygen environment. Myometrium Spiral artery Uterine gland Decidua Cotyledonary septum Glandular secretions in intervillous spaceEndovascular extravillous trophoblast Plasma in intervillous spaceCytotrophoblast Interstitial extravillous trophoblastSyncytiotrophoblast
260
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
development of the placenta 175 cHapte R 9 feature occupying the peripheral margin of the placenta and commu ­ nicating freely with the intervillous space, has not been confirmed. Recent anatomic and in vivo studies have shown that human placen ­ tation is, in fact, not truly haemochorial in early pregnancy ( Video 9.1) (Jauniaux et al 2003). From the time of implantation, the extravillous trophoblast not only invades the uterine tissues but also forms a con ­ tinuous shell at the level of the decidua. The cells of this shell anchor the placenta to the maternal tissue and also form plugs in the tips of the uteroplacental arteries (see Figs 9.3B, 9.6) (Burton et al 1999). The shell and the plugs act like a labyrinthine interface that filters maternal blood, permitting a slow seepage of plasma but no true blood flow into the intervillous space (Burton et al 2003). This creates a physiological placental hypoxia, which may protect the developing embryo against the deleterious and teratogenic effects of oxygen free radicals. As a result, there is a uteroplacental O 2 gradient that exerts a regulatory effect on placental tissue development and function (Jauniaux et al 2003). In particular, it influences cytotrophoblast proliferation and differenti­ation along the invasive pathway, villous vasculogenesis and the form­ ation of the chorion laeve. An O 2 gradient persists within lobules of the placenta throughout pregnancy, with the central region well oxygenated compared to the periphery, owing to the direction of maternal blood flow (Burton and Jauniaux 201 1). At the end of the first trimester, the endovascular extravillous tro ­ phoblastic plugs are progressively dislocated, allowing maternal blood to flow progressively more freely and continuously within the intervil ­ lous space (see Figs 9.6, 9.7, Video 9.1 ). During the transitional phase of 10–14 weeks’ gestation, two ­thirds of the primitive placenta disap ­ pears, the chorionic cavity is obliterated by the growth of the amniotic sac, and maternal blood flows progressively throughout the entire pla ­ centa (Jauniaux et al 2003). These events bring the maternal blood closer to the fetal tissues, facilitating nutrient and gaseous exchange between the maternal and fetal circulations (Gutierrez ­Marcos et al 2012). Structure of a placental villus Chorionic villi are the essential structures involved in exchanges between mother and fetus. The villous tissues separating fetal and gradually extend laterally, reaching the periphery of the placenta around mid­gestation; the extent of invasion is progressively shallower towards the periphery. Invaded maternal vessels show a 5–10 ­fold dilation of the vessel mouth and altered responsiveness to circulating vasoactive compounds (Hung et al 2001, Burton et al 2009). The conversion of maternal musculoelastic, spiral arteries to large ­ bore, low ­resistance uteroplacental vessels is considered key for a suc ­ cessful human pregnancy. Failure to achieve these changes is a feature of common complications of pregnancy, such as early ­onset pre ­ eclampsia, miscarriage and fetal growth restriction. From the third month onwards, the basal plate develops placental or cotyledonary septa, which are ingrowths of the cytotrophoblast covered with syncytium that grow towards, but do not fuse with, the chorionic plate (see Fig. 9.5). The septa circumscribe the maternal surface of the placenta into 15–30 lobes, often termed cotyledons. Each cotyledon surrounds a limited portion of the intervillous space associ ­ ated with a villous trunk from the chorionic plate. From the fourth month, these septa are supported by tissue from the decidua basalis. Throughout the second half of pregnancy, the basal plate becomes thinned and progressively modified: there is a relative diminution of the decidual elements, increasing deposition of fibrinoid, and admix ­ ture of fetal and maternal derivatives. Intervillous space The intervillous space contains the main trunks of the villous stems and their arborizations into intermediate and terminal villi (see Figs 9.5, 9.7). A villous trunk and its branches may be regarded as the essential structural, functional and growth unit of the developing placenta. At term, from the inner myometrium to the intervillous space, the walls of most spiral arteries consist of fibrinoid matrix, within which cytotrophoblast is embedded. This arrangement allows expansion of the arterial diameter (and so slows the rate of arterial inflow and reduces the perfusion pressure), independent of the local action of vasoconstric ­ tive agents. Endothelial cells, where present, are often hypertrophic. The veins that drain the blood away from the intervillous space pierce the basal plate and join tributaries of the uterine veins. The presence of a marginal venous sinus, which hitherto has been described as a constant Fig. 9.7 The arrangement of the placental tissues from the chorionic plate (fetal side) to the basal plate or decidua basalis (maternal side). Fibrinoid deposit CytotrophoblastCytotrophoblastic cell column Syncytiotrophoblast Syncytiotrophoblast Cytotrophoblast Fetal mesenchyme Amniotic epitheliumAnchoring villus Syncytial sprout Hofbauer cellMaternal blood vessels Fetal blood vesselsOrifices of maternal vessels Terminal villi Syncytial fusion Intermediate villus Stem villusMyometrium Basal plate Intervillous space Chorionic plate
261
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Implantat Ion and placentat Ion 176SectIon 2 plasm is more strongly basophilic than that of the cytotrophoblastic cells and is packed with organelles consistent with its secretory pheno ­ type. Where the plasma membrane adjoins basal lamina, it is often infolded into the cytoplasm, whereas the surface bordering the intervil ­ lous space is set with numerous long microvilli, which constitute the brush border seen by light microscopy. Glycogen is thought to be present in both layers of the trophoblast at all stages, although it is not always possible to demonstrate its pres ­ ence histochemically. Lipid droplets occur in both layers and are free in the core of the villus. In the trophoblast, they are found principally within the cytoplasm, but they also occur in the extracellular space between cytotrophoblast and syncytium, and between the individual cells of the cytotrophoblast. The droplets diminish in number with advancing age and may represent fat in transit from mother to fetus, and/or a pool of precursors for steroid synthesis. Membrane ­bound granular bodies of moderate electron density occur in the cytoplasm, particularly in the syncytium, some of which are probably secretion granules. Other membrane ­bound bodies, lysosomes and phagosomes, are involved in the degradation of materials engulfed from the intervil ­ lous space. Although many nuclei are dispersed within the syncytioplasm, other are aggregated into specializations referred to as syncytial knots and syncytial sprouts (Burton et al 2003). Knots must be distinguished from syncytial sprouts, which are markers of trophoblast proliferation. In the immature placenta, syncytial sprouts represent the first stages in the development of new terminal villi, which later become invaded by cytotrophoblast and villous mesenchyme (Fogarty et al 2013). Occa ­ sionally, adjacent syncytial sprouts make contact and fuse to form slender syncytial bridges. The sprouts may become detached, forming maternal syncytial emboli, which pass to the lungs. It has been com ­ puted that some 100,000 sprouts pass daily into the maternal circula ­ tion; deported trophoblast may be a mechanism by which the maternal immune system is maintained in a state of tolerance towards paternal antigens (Chamley et al 201 1). In the lungs, syncytial sprouts provoke little local reaction and apparently disappear by lysis. However, they may occasionally form foci for neoplastic growth. Syncytial sprouts are present in the term placenta. Syncytial knots are aggregates of nuclei with particularly condensed heterochromatin, and may represent a sequestration phenomenon by which senescent nuclear material is removed from adjacent metabolically active areas of syncytium. Fibrinoid deposits are frequently found on the villous surface in areas lacking syncytiotrophoblast. They may constitute a repair mecha ­ nism in which the fibrinoid forms a wound surface that is subsequently re­epithelialized by trophoblast. The extracellular matrix glycoprotein tenascin has been localized in the stroma adjacent to these sites.maternal blood are therefore of crucial functional importance. From the chorionic plate, progressive branching occurs into the villous tree, as stem villi give way to intermediate and terminal villi. Each villus has a core of connective tissue containing collagen types I, III, V and VI, as well as fibronectin. Cross ­banded fibres (30–35 nm) of type I collagen often occur in bundles, whereas type III collagen is present as thinner (10–15 nm) beaded fibres, which form a meshwork that often encases the larger fibres. Collagens V and VI are present as 6–10 nm fibres closely associated with collagens I and III. Laminin and collagen type IV are present in the stroma associated with the basal laminae that surround fetal vessels and also in the trophoblast basal lamina. Overly ­ ing this matrix are ensheathing cyto ­ and syncytiotrophoblast cells bathed by the maternal blood in the intervillous space (see Figs 9.5, 9.7; Fig. 9.8). Cohesion between the cells of the cytotrophoblast, and also between the cytotrophoblast and the syncytium, is provided by numerous desmosomes between the apposed plasma membranes. In earlier stages, the cytotrophoblast forms an almost continuous layer on the basal lamina. After the fourth month, it gradually expends itself producing syncytium, which comes to lie on the basal lamina over an increasingly large area (56% at term), and becomes progressively thinner. Cytotrophoblastic cells persist until term but, because the increase in villous surface area outstrips their proliferation, they are usually disposed singly. In the first and second trimesters, cytotropho­ blastic sprouts, covered in syncytium, are present and represent a stage in the development of new villi. Cytotrophoblast columns at the tips of anchoring villi extend from the villous basal lamina to the maternal decidual stroma. The cells of the villous cytotrophoblast (Langhans cells) are pale ­ staining with a slight basophilia. Ultrastructurally, they have a rather electron ­translucent cytoplasm and relatively few organelles. They contain intermediate filaments, particularly in association with desmo ­ somes. Between the desmosomes, the membranes of adjacent cells are separated by approximately 20 nm. Sometimes, the intercellular gap widens to accommodate microvillous projections from the cell surfaces; occasionally, it contains patches of fibrinoid. A smaller population of intermediate cytotrophoblast may also be found in the chorionic villi. This postmitotic population represents a state of partial differentiation between the cytotrophoblast stem cell and the overlying syncytium. The syncytiotrophoblast is the multinucleated epithelium of the placenta and is an intensely active tissue layer, across which most trans ­ placental transport must occur (Ellery et al 2009). It is a selectively permeable barrier that allows water, oxygen and other nutritive sub ­ stances and hormones to pass from mother to fetus, and some of the products of excretion to pass from fetus to mother. It secretes a range of placental hormones into the maternal circulation. Syncytial cyto ­Fig. 9.8 A, A chorionic villus and its arterio–capillary–venous system carrying fetal blood. The artery carries deoxygenated blood and waste products from the fetus, and the vein carries oxygenated blood and nutrients to the fetus. B–C, Transverse sections through individual chorionic villi at 10 weeks (B) and at full term ( C). The villi would be within the intervillous space, bathed externally in maternal blood. The placental membrane, composed of fetal tissues, separates the maternal blood from the fetal blood. A B C Artery Stem villusVeinArterio–capillary– venous network Fetal capillariesEndothelium of fetal capillary SyncytiotrophoblastSyncytiotrophoblast Fibrinoid material Nuclear aggregation or syncytial knotCytotrophoblastAnchoring villus Hofbauer cellConnective tissue corePlacental membraneFetal blood Fetal capillary Persisting cytotrophoblast cells
262
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Fetal membranes 177 cHapte R 9acute phase of a newly acquired maternal infection. As the maternal circulation inside the placenta only starts to function at the end of the first trimester, transmission of Toxoplasma to the embryos is rare during that period, but highly damaging to the developing fetus. Plasmodium falciparum is transferred by villous sequestration of maternal infected erythrocytes or a breakdown of the villous barrier during labour. The presence of maternal rubella in the early months of pregnancy is of especial importance in relation to the production of congenital anoma ­ lies. (For further reading about TORCH perinatal infections, caused by Toxoplasma gondii, rubella virus, CMV and herpes simplex virus (HSV), see Stegmann and Carey (2002).) FETAL MEMBRANES The implanting conceptus consists initially of three cavities and their surrounding epithelia. The original blastocyst cavity, surrounded by trophoblast, is now termed the chorionic cavity (synonymous with extraembryonic coelom). It is a large cavity containing the much smaller amniotic cavity and secondary yolk sac (see Fig. 9.1). The apposition of the latter two cavities delineates the extent of the early embryo. The chorionic cavity becomes lined with extraembryonic mesoblast, which is also reflected over the outer surface of the amnion and yolk sac. A fourth cavity, the allantois, develops later as a caudal hypoblastic diver ­ ticulum that becomes embedded within the extraembryonic mesen­chyme, forming the connecting stalk of the embryo. It does not have a direct mesothelial covering. Chorion The chorion consists of developing trophoblast and extraembryonic mesothelium. It varies in thickness during development, both tempor­ ally and spatially. It is thickest at the implantation site throughout gesta ­ tion as the chorion frondosum and then the placenta, and thinner as gestation progresses over the remainder of its surface as the chorion laeve (see Fig. 9.4). At term, the chorion consists of an inner cellular layer containing fibroblasts and a reticular layer of fibroblasts and Hof ­ bauer cells, which resembles the mesenchyme of an intermediate villus. The outer layer consists of cytotrophoblast 3–10 cells deep, resting on a pseudo ­basement membrane, which extends beneath and between the cells. Occasional obliterated villi within the trophoblast layer are the remnants of villi present in the chorion frondosum of the first tri ­ mester. Although the interface between the trophoblast and decidua parietalis is uneven, no trophoblast infiltration of the decidua parietalis occurs. Yolk sac As the secondary yolk sac forms, it delineates a cavity lined with par­ ietal, and perhaps visceral, hypoblast, continuous with the developing endoderm from the primitive streak (Ch. 10; see Fig. 10.10). The sec ­ ondary yolk sac is the first structure that can be detected ultrasono ­ graphically within the chorionic cavity (Jauniaux et al 1991). Its diameter increases slightly between 6 and 10 weeks of gestation, reach ­ ing a maximum of 6–7 mm, after which its size decreases. The inner cells of the yolk sac (denoted endoderm in many studies, although this layer is restricted to the embryo itself) display a few short microvilli and are linked by juxtaluminal tight junctions (Jones and Jauniaux 1995). Their cytoplasm contains numerous mitochondria, whorls of rough endoplasmic reticulum, Golgi bodies and secretory droplets, giving them the appearance of being highly active synthetic cells. With further development, the epithelium becomes folded to form a series of cyst ­like structures or tubules, only some of which com ­ municate with the central cavity. The cells synthesize several serum proteins in common with the fetal liver, such as α ­fetoprotein (AFP), α­1­antitrypsin, albumin, pre ­albumin and transferrin (Gulbis et al 1998). With rare exceptions, the secretion of most of these proteins is confined to the embryonic compartments (Jauniaux and Gulbis 2000b). The yolk sac becomes coated with extraembryonic mesenchyme, which forms mesenchymal and mesothelial layers. A diffuse capillary plexus develops between the mesothelial layer and the underlying sec­ ondary yolk sac wall, and subsequently drains through vitelline veins to the developing liver. The mesothelial layer bears a dense covering of microvilli; the presence of numerous coated pits and pinocytotic vesi ­ cles gives it the appearance of an absorptive epithelium (Jones and Jauniaux 1995, Burke et al 2013). The secondary yolk sac plays a major role in the early embryonic development of all mammals. In laboratory rodents, it has been The core of a villus contains small and large reticulum cells, fibro­ blasts and macrophages (Hofbauer cells). Early mesenchymal cells probably differentiate into small reticulum cells, which, in turn, produce fibroblasts or large reticulum cells. The small reticulum cells appear to delimit a collagen ­free stromal channel system through which Hofbauer cells migrate. Mesenchymal collagen increases from a network of fine fibres in early mesenchymal villi to a densely fibrous stroma within stem villi in the second and third trimesters. After approximately 14 weeks, the stromal channels found in immature intermediate villi are infilled by collagen to give the fibrous stroma characteristic of the stem villus. Fetal placental vessels include arterioles and capillaries; lymphatic vessels are not present. Pericytes may be found in close association with the capillary endothelium, and from late first trimester, the vessels are surrounded externally by a basal lamina. From the second trimester (and a little later in terminal villi), dilated thin ­walled capillaries are found immediately adjacent to the villous trophoblast; their respective basal laminae apparently fuse to produce a vasculosyncytial interface; the distance separating the maternal and fetal circulations may be reduced to as little as 2–3 µm. The endothelial cells are non ­fenestrated, display numerous caveolae, and are linked by conspicuous junctional complexes incorporating tight and adherens junctions. transport across placental villi The mechanism of transfer of substances across the placental barrier (membrane) is complex. The volume of maternal blood circulating through the intervillous space has been assessed at 500 ml per minute. Simple diffusion suffices to explain gaseous exchange. Transfer of ions and other water ­soluble solutes is by paracellular and transcellular dif ­ fusion and transport; the relative importance of each of these for most individual solutes is unknown, and the paracellular pathway is mor ­ phologically undefined. Glucose transfer involves facilitated diffusion, while active transport mechanisms carry calcium and at least some amino acids. The fat ­soluble and water ­soluble vitamins are likely to pass the placental barrier with different degrees of facility (Jauniaux and Gulbis 2000a, Jauniaux et al 2004, Jauniaux et al 2005). The water ­ soluble vitamins B and C pass readily. Water is interchanged between fetus and mother (in both directions) at approximately 3.5 litres per hour. The transfer of substances of high molecular weight, such as complex sugars, some lipids and hormonal and non ­hormonal pro ­ teins, varies greatly in rate and degree and is not so well understood; energy ­dependent selective transport mechanisms, including receptor ­ mediated transcytosis, are likely to be involved. Lipids may be transported unchanged through and between the cells of the trophoblast to the core of the villus. The passage of maternal antibodies (immunoglobulins) across the placental barrier confers some degree of passive immunity on the fetus; it is widely accepted that transfer is by micropinocytosis. Investigation of transplacental mechan­ isms is complicated by the fact that the trophoblast itself is the site of synthesis and storage of certain substances, e.g. glycogen. The placenta is an important endocrine organ. Some steroid hor ­ mones, various oestrogens, β­endorphins, progesterone, hCG and human chorionic somatomammotropin (hCS), also known as placental lactogen (hPL), are synthesized and secreted by the syncytium. The tro ­ phoblast contains enzyme systems that are associated with the synthesis of steroid hormones, as well as enzymes that inactivate maternal hor ­ mones, allowing the fetus to develop and mature in a protected environ ­ ment; for example, the fetus is protected from 10 times higher maternal glucocorticoid levels by the placental enzyme 1 1beta ­hydroxysteroid dehydrogenase 2 (1 1beta ­HSD2), which converts biologically active maternal cortisol to inactive cortisone (Murphy et al 1974). It has been suggested that leukocytes may migrate from the maternal blood through the placental barrier into the fetal capillaries. It has also been shown that some fetal and maternal red blood cells may cross the barrier. The former may have important consequences, e.g. in Rhesus incompatibility. The majority of drugs are small molecules and are sufficiently lipophilic to pass the placental barrier from 7 weeks of gestation in quantities that yield measurable concentrations in both coelomic and amniotic fluid. Many are tolerated by the fetus but some may exert grave teratogenic effects on the developing embryo/fetus, depending on the gestational age and the doses used by the mother (e.g. thalidomide, cocaine) (Ross et al 2015). A well ­documented association exists between maternal alcohol ingestion and fetal abnormalities. Addiction of the fetus can occur to substances of maternal abuse such as heroin. A wide variety of bacteria (Listeria), protozoa ( Toxoplasma gondii, Plasmodium falciparum) and viruses, including cytomegalovirus (CMV), human immunodeficiency virus (HIV) and rubella, are known to pass the placental barrier from mother to fetus. Toxoplasma and CMV are transferred by cellular invasion of the villous trophoblast during the
263
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Implantat Ion and placentat Ion 178SectIon 2intra ­abdominal urachus; this, in turn, continues into the apex of the bladder. Amnion (chorio-amnion) The original amniotic cells develop from the edges of the epiblast of the embryonic disc, which ultimately form the interface with the skin at the umbilical region. Between the tenth and twelfth weeks of preg­ nancy, the amniotic cavity expands until it makes contact with the chorion to form the chorio ­amnion, an avascular membrane that per ­ sists to term. The amniotic membrane extends along the connecting stalk and forms the outer covering of the umbilical cord. After birth, the site of this embryonic/extraembryonic junction is important because the extraembryonic cell lines will die, causing the umbilical cord to degenerate and detach from the body wall. In cases of anomalous development of the ventral body wall, e.g. gastroschisis and exompha ­ los, the reflections of the amnion along the forming umbilical cord may be incomplete (see below). The inner surface of the amnion consists of a simple cuboidal epi ­ thelium. It has a microvillous apical surface, beneath which is a cortical web of intermediate filaments and microfilaments. There are no tight junctional complexes between adjacent cells and cationic dyes penetrate between the cells as far as the basal lamina. The intercellular clefts present scattered desmosomes, but elsewhere the clefts widen and contain interlacing microvilli. These features are consistent with selec ­ tive permeability properties. The epithelium synthesizes and deposits extracellular matrix into the compact layer of acellular stroma located beneath the basal lamina, as well as the basal lamina itself. Towards the end of gestation, increasing numbers of amniotic cells undergo apoptosis. Apoptotic cells become detached from the amnion and are found in the amniotic cavity at term. The highest inci ­ dence is in weeks 40–41, independent of the onset of labour. Apoptosis may play a role in the fragility and rupture of the fetal membranes at term. Human amniotic epithelial cells are thought to be pluripotent because they arise so early from the conceptus. They can be distin ­ guished from the epiblast cells from day 8. Amniotic cells lack the MHC antigen and so the amnion can be exposed to the maternal immune system without eliciting a maternal immune response. Cultured human amniotic epithelial cells express a range of neural and glial markers, including glial fibrillary acidic protein, myelin basic protein, vimentin and neurofilament proteins, suggesting that these cells may supply neurotrophic factors to the amniotic fluid. They also appear to have a hepatocyte gene expression profile, showing albumin production, gly ­ cogen storage and albumin secretion in culture. In organ culture, they have been shown to secrete 30 ­fold larger amounts of albumin than in monolayer culture, and to secrete α­1­antitrypsin (Takashina et al 2004). Amnion is used in the repair of corneas after trauma and as a graft material for reconstructing vaginas in women with cloacal abnormalities. Prenatal changes to the chorio-amnion At term, the surface area of the chorio ­amnion is 1000–1200 cm2, with 30% overlying the placenta and the remaining 70% in contact with the decidua (Myatt and Sun 2010). There is a diminution of the decidual component of the chorion and an increase in chorionic apoptosis towards term at the supracervical site, overlying the cervical os, which encompasses the site of rupture (Chai et al 2013). The amnion and choriodecidua show varying degrees of separation over the putative rupture site; the mean rupture strength of the zone is 60% of the remaining membranes, associated with changes in collagen organiza ­ tion and increase in MMP ­9 (Strauss 2013). Term vaginal delivery is associated with a three ­fold increase in apoptosis in the chorion close to the rupture site, compared to fetal membranes from non ­labour, elective caesarean delivery (Harirah et al 2012). In preterm, premature rupture of membranes (PPROM), the chorion is reported to be thinner at all sites (Fortner et al 2014). The presence of bacteria in the chorion at term, ascending via the vagina, has been reported without concomitant chorio ­amnionitis. However, chorio ­ amnionitis rates are higher in preterm labour and this may be a con ­ tributing cause (Harirah et al 2012, Fortner et al 2014). Loss of chorio ­amnion intracellular cytokeratins, which become downregu ­ lated in infection, is thought to make the amnion vulnerable to apop ­ tosis, shear stress and rupture (Vanderhoeven et al 2014). There is increasing evidence that infection and inflammation of the chorio ­amnion is associated with prenatal changes to the developing brain, especially the cortex, periventricular white matter and the devel ­ oping cerebral blood vessels (Harteman et al 2013, Roescher et al demonstrated as one of the initial sites of haemopoiesis. Human data indicate that it has an absorptive role for molecules of maternal and placental origin found in the chorionic cavity (Gulbis et al 1998), and mediates the main movement of molecules passing from the chorionic cavity to the yolk sac and, subsequently, to the embryonic gut and circulation. After week 9, the cellular components of the wall of the secondary yolk sac start to degenerate, and their function is subsumed into exchanges at the placental chorionic plate. With embryonic develop ­ ment of the midgut, the connection of the yolk sac to the embryo becomes attenuated to a slender and elongated vitelline intestinal duct. Both the yolk sac and its duct remain within the extraembryonic coelom (chorionic cavity) throughout gestation, located between the amnion and chorion as they fuse, near the placental attachment of the umbilical cord. Allantois The allantoenteric diverticulum (see Fig. 10.10) arises early in the third week as a solid, endodermal outgrowth from the dorsocaudal part of the yolk sac into the mesenchyme of the connecting stalk. It soon becomes canalized. When the hindgut is developed, the proximal (enteric) part of the diverticulum is incorporated in its ventral wall. The distal (allantoic) part remains as the allantoic duct and is carried ven ­ trally to open into the ventral aspect of the cloaca or terminal part of the hindgut ( Fig. 9.9A). The allantois is a site of angiogenesis, giving rise to the umbilical vessels that connect to the placental circulation. The extraembryonic mesenchyme around the allantois forms the con ­ necting stalk, which is later incorporated into the umbilical cord. In the fetus, the allantoic duct, which is confined to the proximal end of the umbilical cord, elongates and thins. However, it may persist as an interrupted series of epithelial strands at term, in which case the proximal strand is often continuous at the umbilicus with the median Fig. 9.9 A, A Longitudinal section of a conceptus showing the cavities associated with development. The amniotic cavity and yolk sac are both covered with extraembryonic mesoblast and are contained within the larger chorionic cavity, which is lined with extraembryonic mesoblast. The embryo is attached to the chorion frondosum by the connecting stalk into which the allantois projects. B, A longitudinal section of a conceptus at a later stage, showing the diminution of the chorionic cavity, expansion of the amniotic cavity, relative attenuation of the yolk sac and the structures that give rise to the umbilical cord. A Connecting stalk attachmentbecomes moreventrally placed Chorionic cavity/ extraembryoniccoelom Pericardial coelomYolk sacPlacental area (chorion frondosum) Allantoic canal Amniotic cavity Chorion laeve Yolk sac Line of fusion of amnion withchorion, i.e. lineof obliteratedextraembryoniccoelom Amniotic cavityUmbilical cord Obliterated yolk duct Umbilical coelom Allantoic ductB
264
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Umbilical cord 179 cHapte R 9of cytokines, have also been detected in amniotic fluid; the concentra ­ tions of most cytokines differ from those in maternal blood and reflect the physiological status of the fetus (Tong 2013). Proteomic analysis of human amniotic fluid at 16–18 weeks has identified more than 800 proteins and peptides (Cho et al 2007). It appears that amniotic fluid may be acting as a pathway for the transport of signalling molecules to target cells. In animal studies, substances injected into the amniotic cavity were detected within the embryos within 2 minutes, demonstrat ­ ing a rapid exchange with all fetal tissues (Tong 2013). A number of clinical markers in amniotic fluid are used as predic ­ tors of fetal health, including interleukin (IL) ­10, IL ­6, cell free fetal DNA and cell free fetal mRNA. Genes and proteins specific to the fetal brain, lung, kidney, blood and heart have been identified in amniotic fluid, differential gene expression investigated in a range of trisomic fetuses and lung maturity predicted by measurement of lecithin, sphin ­ gomyelin and phosphatidylglycerol (Edlow and Bianchi 2012, Ross and Beall 2014). Bacterial species have been isolated from amniotic fluid, and microbial invasion of the amniotic cavity is considered a leading cause of preterm premature delivery (DiGiulio 2012). It is thought such bacteria access the amniotic cavity via the vagina and cervix, although many oral bacteria have also been identified, suggest ­ ing that a haematogenous route is also likely. The establishment of pathogenesis is influenced by microbial adhesion, biofilm formation, and immune evasion and the maternal and fetal immune systems (DiGiulio 2012). UMBILICAL CORD The formation of the connecting stalk and the early formation of the umbilical cord are described in Chapter 10. The umbilical cord ulti ­ mately consists of an outer covering of flattened amniotic epithelial cells and an interior mass of mesenchyme of diverse origins (see Fig. 9.9). It contains two tubes of hypoblastic origin, the vitelline ­intestinal and allantoic ducts, and their associated vitelline and allantoic (umbili ­ cal) blood vessels. The yolk stalk and continuing duct extend the length of the cord, whereas the allantoic duct extends only into its proximal part. The mesenchymal core is derived from the somatopleuric extra­ embryonic mesenchyme covering the amniotic folds, splanchnopleuric extraembryonic mesenchyme of the yolk stalk (which carries the vitel ­ line vessels and clothes the yolk duct), and similar allantoic mesen ­ chyme of the connecting stalk (which clothes the allantoic duct and initially carries two umbilical arteries and two umbilical veins). These various mesenchymal compartments fuse and are gradually trans ­ formed into the loose connective tissue (Wharton’s jelly) that character ­ izes the more mature cord. The tissue consists of widely spaced elongated fibroblasts separated by a delicate three ­dimensional meshwork of fine collagen fibres, which contains a variety of hydrated glycosaminogly ­ cans and is particularly rich in hyaluronic acid. The vitelline and allantoic (umbilical) vessels, which are initially symmetrical, become modified as a result of changes in the circulation. The vitelline vessels involute, whereas most of the allantoic (umbilical) vessels persist. The right umbilical vein disappears but the two umbilical arteries normally remain. Occasionally, one umbilical artery may disap ­ pear; there is some correlation within structural anomalies, most often cardiac, in such cases. The vessels of the umbilical cord are rarely straight, and are usually twisted into either a right ­ or left ­handed cylin ­ drical helix. The number of turns involved ranges from a few to over 300. This conformation may be produced by unequal growth of the vessels, or by torsional forces imposed by fetal movements. Its func ­ tional significance is obscure; perhaps the pulsations and contractions of the helical vessels assist the venous return to the fetus in the umbili ­ cal vein. Anomalies of the fetal anterior abdominal wall, such as exomphalos and gastroschisis, may affect the arrangement of the outer covering of amnion cells along the proximal end of the umbilical cord. Exomphalos arises from a failure of the lateral folds along the ventral surface of the embryo, resulting in failure of the normal embryonic regression of the midgut from the umbilical stalk into the abdominal cavity. The abdomi ­ nal contents, including intestines and liver or spleen, covered by a sac of parietal peritoneum and amnion, protrude into the base of the umbilical cord. In gastroschisis, the insertion of the umbilical cord is intact and there is evisceration of the intestine through a small abdomi­nal wall defect that is usually located to the right of the umbilical cord; this results in free loops of bowel in the amniotic cavity. Theories con ­ cerning the aetiology of this defect include abnormal involution of the right umbilical vein or disruption of the omphalomesenteric artery by ischaemia.2014). This association also extends to the neonatal period and post ­ natal gut maturation (Wynn and Neu 2012). Histologically confirmed chorio ­amnionitis is associated with changes in cerebral blood flow soon after birth; male infants appear to be more affected than female (Koch et al 2014). AMNIOTIC FLUID The amniotic fluid, or liquor amnii, provides a buoyant medium that supports the delicate tissues of the young embryo and allows free move ­ ment of the fetus up to the later stages of pregnancy. It also diminishes the risk to the fetus of injury from outside. Amniotic fluid is derived from multiple sources throughout gestation. These include secretions from amniotic epithelium, filtration of fluid from maternal vessels via the parietal decidua and amniochorion, filtration from the fetal vessels via the chorionic plate or the umbilical cord, and fetal urine and fetal lung secretions. In early pregnancy, diffusion from intracorporeal vessels via fetal skin provides another source. Once the gut is formed, fetal swallowing of amniotic fluid is a normal occurrence; the fluid is absorbed into the fetal circulation. Amniotic fluid volume increases logarithmically during the first half of pregnancy, from less than 10 ml at 8 weeks’ gestation to 30 ml at 22 weeks and 770 ml at 28 weeks. After 30 weeks, the volume may remain unchanged to 36 weeks. It decreases towards the expected delivery date, and volumes decrease sharply in post ­term gestation, averaging 515 ml at 41 weeks (Beall et al 2007, Ross and Beall 2014). Human fetal urine output from the metanephric kidney increases from 1 10 ml/kg/day at 25 weeks to almost 200 ml/kg/day at term; this corresponds to almost 1000 ml/day. The rate of human fetal lung secre ­ tion has not been measured but is presumed to be in the range of 60–100 ml/kg/day near term (Callen 2008). Animal studies indicate that lung fluid production is about one ­third that of urine production, and that half of the fluid leaving the lungs enters the amniotic fluid and half is swallowed (Beall et al 2007, Ross and Beall 2014). It is estimated that human fetuses swallow up to 760 ml/day of amniotic fluid near term, although this decreases in the days before delivery. The remainder of the amniotic fluid is thought to be absorbed via an intramembranous pathway directly across the amniotic cavity and fetal surface of the placenta into fetal blood vessels (Beall et al 2007, Ross and Beall 2014). Intramembranous flow is thought to reach 400 ml/day at term (Callen 2008). Amniotic fluid volume is estimated at routine antenatal ultrasound scans, despite some concerns about its objectivity (Beattie and Rich 2007, Callen 2008, Gilbert 2012); it may be expressed as an amniotic fluid index. A deficiency of amniotic fluid is termed oligohydramnios and absent amniotic fluid is anhydramnios. Oligohydramnios in the second or third trimester is usually the result of urinary tract malforma ­ tions, e.g. bilateral renal agenesis or obstruction of the lower urinary tract, uteroplacental insufficiency or premature rupture of the mem ­ branes. The urachus may play a critical role in the resolution of oligo ­ hydramnios in lower urinary tract obstruction by acting as a fistula between the bladder and the amniotic space. The major concern with oligohydramnios at less than 20 weeks is the significant risk of pulmo ­ nary hypoplasia and neonatal death. A volume of amniotic fluid in excess of 2 litres is generally considered to be abnormal and constitutes polyhydramnios (see Fig. 14.5A). Maternal causes include cardiac and renal problems and diabetes mellitus, which causes fetal hyperglycae ­ mia and polyuria. Fetal causes include reduced fetal swallowing due to congenital malformations, e.g. anencephaly; upper intestinal tract obstruction (oesophageal and duodenal atresia); compressive pulmo ­ nary disorders (congenital diaphragmatic hernia); and neuro muscular impairment of swallowing. Amniotic fluid has, for many years, been regarded as a medium that physically supports the developing embryo and fetus. More recently, it has also been conceptualized as part of the embryonic and early fetal extracellular matrix, and its composition throughout pregnancy has been investigated (Tong et al 2009). Techniques for biochemical analy ­ sis have revealed similarities and differences between amniotic fluid and fetal umbilical cord blood and maternal serum. Although, early in development, amniotic fluid resembles blood plasma, it also demon ­ strates extremely low oxygen tension, reflecting the physiological pla ­ cental hypoxia noted in the first trimester (Ross and Beall 2014, Jauniaux et al 2003). In the second and third trimesters, primary electrolytes are similar in amniotic fluid, umbilical cord blood and maternal serum, whereas glucose, cholesterol and some enzymes are markedly lower in amniotic fluid. Umbilical cord blood and maternal serum contain 8 and 12.5 times more protein than amniotic fluid, respectively (Tong et al 2009). More than 100 metabolites, including cortisol and a range
265
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Implantat Ion and placentat Ion 180SectIon 2Mature umbilical vessels, particularly the arteries, have a strong mus ­ cular coat that contracts readily in response to mechanical stimuli. The outermost bundles pursue an interlacing spiral course, and when they contract, they produce shortening of the vessel and thickening of the media, with folding of the interna and considerable narrowing of the lumen. This action may account for the periodic sharp constrictions of contour, the so ­called valves of Hoboken, which often characterize these vessels. The fully developed umbilical cord is, on average, some 50 cm long and 1–2 cm in diameter. Its length varies from 20 to 120 cm; exception ­ ally short or long cords are associated with fetal problems and compli­ cations during labour. A long umbilical cord may prolapse through the cervix into the vagina once the fetal membranes rupture and this may be exacerbated by conditions that prevent the fetal head from fully occupying the maternal pelvis, e.g. pelvic tumours (fibroids), ovarian cysts, placenta praevia and prematurity. Compression of the cord by the presenting part of the fetus, or an umbilical artery spasm, will lead to fetal hypoxia and death, if untreated. The risk of perinatal death rises as the interval from diagnosis to delivery increases. The treatment is either funic replacement (pushing the cord back above the fetal head) or, more commonly, immediate caesarean section, depending on factors such as fetal viability. The distal end of the umbilical cord usually attaches in the central portion of the placenta, but in 0.2% of pregnancies, velamentous inser ­ tion is observed (i.e. into the membranes) and this may be associated Video 9.1 Ultrasound features of the maternal placental blood flow.  Bonus  e-book  video with vulnerability to injury and fetal haemorrhage. This is especially important if the placenta is low ­lying, and may be associated with vasa praevia, in which case fetal blood vessels run across the internal os. Inadvertent rupture of the fetal vessels in spontaneous labour or at the time of amniotomy (artificial rupture of membranes to induce labour) will cause fetal haemorrhage and may prove fatal. Infection of the umbilical cord, funisitis, is associated with chorio ­amnionitis and later neurodevelopmental outcomes (Roescher et al 2014, Buhimschi and Buhimschi 2012). Recent studies have examined the optimal time for cutting the umbilical cord after delivery and found that early clamping is associated with neonatal anaemia and iron deficiency. Delayed clamping, defined as ligation of the umbilical cord 2–3 minutes after birth or when cord pulsation stops, reduces the prevalence of iron deficiency at 4 months of age and results in improved ferritin levels (Andersson et al 201 1, McDonald et al 2013). KEY REFERENCES Brosens JJ, Salker MS, Teklenburg G et al 2014 Uterine selection of human embryos at implantation. Sci Rep 4:3894. Study showing that developmentally impaired human embryos elicit an endoplasmic stress response in human decidual cells and suggesting that distinct positive and negative mechanisms contribute to active selection of human embryos at implantation. Burton GJ, Jauniaux E 201 1 Oxidative stress. Best Pract Res Clin Obstet Gynaecol 25:287–99.A review of the evidence implicates oxidative stress in the pathophysiology of placental related complications of human pregnancy ranging from miscarriage to pre-eclampsia and premature rupture of the membranes. Burton GJ, Jauniaux E, Watson AL 1999 Maternal arterial connections to the placental intervillous space during the first trimester of human preg­nancy: the Boyd collection revisited. Am J Obstet Gynecol 181: 718–24.Using hysterectomy specimens from the Boyd anatomy collection, this study shows that the maternal circulation to the placenta must be extremely sluggish before the eighth week of pregnancy, supporting the concept that development of the human fetoplacental unit during most of the first trimester takes place in a low-oxygen environment. Burton GJ, Watson AL, Hempstock J et al 2002 Uterine glands provide his ­ tiotrophic nutrition for the human fetus during the first trimester of pregnancy. J Clin Endocrinol Metab 87:2954–9.The first study to demonstrate that the uterine glands are an important source of nutrients during organogenesis, when metabolism is essentially anaerobic. Uterine glands remain active until at least week 10 of pregnancy, and their secretions are delivered freely into the placental intervillous space, supporting the concept of histiotrophic nutrition as opposed to hemochorial nutrition. Cho CK, Shan SJ, Winsor EJ, Diamandis EP 2007 Proteomics analysis of human amniotic fluid. Mol Cell Proteomics 6:1406–15.This paper presents an extensive profile of the normal human amniotic fluid proteome and considers the molecular functions of amniotic fluid including the development of biomarkers for pregnancy-associated abnormalities.Gulbis B, Jauniaux E, Cotton F et al 1998 Protein and enzyme patterns in the fluid cavities of the first trimester gestational sac: relevance to the absorptive role of secondary yolk sac. Mol Hum Reprod 4:857–62.Study demonstrating that the secondary yolk sac membrane is an important zone of transfer between the extra-embryonic and embryonic compartments and suggesting that therapeutic protocols making use of fetal somatic gene therapy could be performed by injecting transduced cells into the exocoelomic cavity. Jauniaux E, Gulbis B 2000a Fluid compartments of the embryonic environ ­ ment. Hum Reprod Update 6:268–78.This paper reviews the composition, biology and role of the extraembryonic coelomic or chorionic fluid cavity and its connections with the secondary yolk sac and developing placenta during the first trimester of human pregnancy. Ross MG, Beall MH 2014 Amniotic fluid dynamics. In: Creasy R, Resnik R, Iams JD et al (eds) Saunders’ Maternal ­Fetal Medicine: Principles and Practice, 7th ed. Philadelphia: Elsevier, Saunders; Ch. 3, pp. 47–52. This chapter provides an overview of the production, circulation and absorption of amniotic fluid during gestation. Strauss JF 2013 Extracellular matrix dynamics and fetal membrane rupture. Repro Sci 20:140–53.This paper presents the composition and biomechanical properties of the fetal membrane extracellular matrix. It considers the changes in extracellular matrix composition during normal pregnancy and term delivery and those seen in preterm premature rupture of membranes. Tong X 2013 Amniotic fluid may act as a transporting pathway for signalling molecules and stem cells during the embryonic development of amni ­ otes. J Chinese Med Ass 76:606–10.This paper considers the bi-directional substance exchange between amniotic fluid and fetal tissues, its role in transporting signaling molecules and the dynamic nature of amniotic fluid throughout pregnancy.
266
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Implantation and placentation 180.e1 cHapte R 9REFERENCES Andersson O, Hellström ­Westas L, Andersson D et al 201 1 Effect of delayed versus early umbilical cord clamping on neonatal outcomes and iron status at 4 months: a randomised controlled trial. BMJ 343:d7157. Beall MH, van den Wijngaard JPHM, van Gemert MJC et al 2007 Amniotic fluid water dynamics. Placenta 28:816–23. Beattie RB, Rich DA 2007 Disorders of amniotic fluid, placenta and mem ­ branes. In: Twining P, McHugo JM, Pilling DW (eds) Textbook of Fetal Abnormalities. London: Elsevier, Churchill Livingstone; Ch 5, pp. 75–94. Brosens JJ, Salker MS, Teklenburg G et al 2014 Uterine selection of human embryos at implantation. Sci Rep 4:3894. Study showing that developmentally impaired human embryos elicit an endoplasmic stress response in human decidual cells and suggesting that distinct positive and negative mechanisms contribute to active selection of human embryos at implantation. Buhimschi IA, Buhimschi CS 2012 Proteomics/diagnosis of chorioamnion­ itis and of relationships with the fetal exposome. Sem Fetal Neonatal Med 17:36–45. Burke KA, Jauniaux E, Burton GJ et al 2013 Expression and immunolocalisa ­ tion of the endocytic receptors megalin and cubilin in the human yolk sac and placenta across gestation. Placenta 34:1 105–9. Burton GJ, Fowden AL 2012 Review: The placenta and developmental pro ­ gramming: balancing fetal nutrient demands with maternal resource allocation. Placenta 33 Suppl:S23–7. Burton GJ, Jauniaux E 201 1 Oxidative stress. Best Pract Res Clin Obstet Gynaecol 25:287–99.A review of the evidence implicates oxidative stress in the pathophysiology of placental related complications of human pregnancy ranging from miscarriage to pre-eclampsia and premature rupture of the membranes. Burton GJ, Jauniaux E, Watson AL 1999 Maternal arterial connections to the placental intervillous space during the first trimester of human preg­nancy; the Boyd Collection revisited. Am J Obstet Gynecol 181:718–24.Using hysterectomy specimens from the Boyd anatomy collection, this study shows that the maternal circulation to the placenta must be extremely sluggish before the eighth week of pregnancy, supporting the concept that development of the human fetoplacental unit during most of the first trimester takes place in a low-oxygen environment. Burton GJ, Watson AL, Hempstock J et al 2002 Uterine glands provide his ­ tiotrophic nutrition for the human fetus during the first trimester of pregnancy. J Clin Endocrinol Metab 87:2954–9.The first study to demonstrate that the uterine glands are an important source of nutrients during organogenesis, when metabolism is essentially anaerobic. Uterine glands remain active until at least week 10 of pregnancy, and their secretions are delivered freely into the placental intervillous space, supporting the concept of histiotrophic nutrition as opposed to hemochorial nutrition. Burton GJ, Skepper JN, Hempstock J et al 2003 A reappraisal of the contrast ­ ing morphological appearances of villous cytotrophoblast cells during early human pregnancy; evidence for both apoptosis and primary necro ­ sis. Placenta 24:297–305. Burton GJ, Jauniaux E, Charnock ­Jones DS 2007 Human early placental development: potential roles of the endometrial glands. Placenta 28 Suppl A:S64–9. Burton GJ, Woods AW, Jauniaux E et al 2009 Rheological and physiological consequences of conversion of the maternal spiral arteries for uteropla ­ cental blood flow during human pregnancy. Placenta 30:473–82. Burton GJ, Jauniaux E, Chernock ­Jones DS 2010 The influence of the intra­ uterine environment on the human placental development. Int J Dev Biol 54:303–12. Callen PW 2008 Amniotic fluid volume: its role in fetal health and disease. In: Callen PW (ed) Ultrasonography in Obstetrics and Gynecology, 5th ed; Ch. 20, pp. 758–79. Chai M, Walker SP, Riley C et al 2013 Effect of supracervical apposition and spontaneous labour on apoptosis and matrix metalloproteinases in human fetal membranes. Biomed Res Int 2013:316146. Chamley LW, Chen Q, Ding J et al 201 1 Trophoblast deportation: just a waste disposal system or antigen sharing? J Reprod Immunol 88:99–105. Cho CK, Shan SJ, Winsor EJ et al 2007 Proteomics analysis of human amni ­ otic fluid. Mol Cell Proteomics 6:1406–15.This paper presents an extensive profile of the normal human amniotic fluid proteome and considers the molecular functions of amniotic fluid including the development of biomarkers for pregnancy-associated abnormalities.DiGiulio DB 2012 Diversity of microbes in amniotic fluid. Sem Fetal Neo ­ natal Med 17:2–1 1. Edlow AG, Bianchi DW 2012 Tracking fetal development through molecular analysis of maternal biofluids. Biochim Biophys Acta 1822:1970–80. Ellery PM, Cindrova ­Davies T, Jauniaux E et al 2009 Evidence for transcrip ­ tional activity in the syncytiotrophoblast of the human placenta. Pla ­ centa 30:329–34. Fogarty NM, Ferguson ­Smith AC, Burton GJ 2013 Syncytial knots (Tenney ­ Parker changes) in the human placenta: evidence of loss of transcrip ­ tional activity and oxidative damage. Am J Pathol 183:144–52. Fortner KB, Grotegut CA, Ransom CE et al 2014 Bacteria localization and chorion thinning among preterm premature rupture of membranes. PLOS One 9:e83338. Gilbert WM 2012 Amniotic fluid disorders. In: Gabbe SG, Niebyl JR, Simpson JL et al (eds). Obstetrics: Normal and Problem Pregnancies, 6th ed. Philadelphia: Elsevier, Saunders; Ch. 33, pp. 759–68. Gulbis B, Jauniaux E, Cotton F et al 1998 Protein and enzyme pattern in the fluid cavities of the first trimester human gestational sac: relevance to the absorptive role of the secondary yolk sac. Mol Hum Reprod 4:857–62.Study demonstrating that the secondary yolk sac membrane is an important zone of transfer between the extra-embryonic and embryonic compartments and suggesting that therapeutic protocols making use of fetal somatic gene therapy could be performed by injecting transduced cells into the exocoelomic cavity. Gutierrez ­Marcos JF, Constância M, Burton GJ 2012 Maternal to offspring resource allocation in plants and mammals. Placenta 33 Suppl 2: e310. Harirah HM, Borahay MA, Zaman W et al 2012 Increased apoptosis in chorionic trophoblasts of human fetal membranes with labor at term. Int J Clin Med 3:136–42. Harris LK 2010 Review: Trophoblast ­vascular cell interactions in early pregnancy: how to remodel a vessel. Placenta 31 Suppl:S93–8. Harteman JC, Nikkels PGJ, Benders MJNL et al 2013 Placental pathology in full­term infants with hypoxic ­ischemic neonatal encephalopathy and association with magnetic resonance imaging pattern of brain injury. J Pediatr 163:968–95. Hung TH, Skepper JN, Burton GJ 2001 In vitro ischemia ­reperfusion injury in term human placenta as a model for oxidative stress in pathological pregnancies. Am J Pathol 159:1031–43. Huppertz B, Weiss G, Moser G 2014 Trophoblast invasion and oxygenation of the placenta: measurements versus presumptions. J Reprod Immunol 101–2:74–9. Jauniaux E, Gulbis B 2000a Fluid compartments of the embryonic environ ­ ment. Hum Reprod Update 6:268–78.This paper reviews the composition, biology and role of the extraembryonic coelomic or chorionic fluid cavity and its connections with the secondary yolk sac and developing placenta during the first trimester of human pregnancy. Jauniaux E, Gulbis B 2000b In vivo investigation of placental transfer early in human pregnancy. Eur J Obstet Gynecol Reprod Biol 92:45–9. Jauniaux E, Jurkovic D, Henriet Y et al 1991 Development of the secondary human yolk sac: correlation of sonographic and anatomic features. Hum Reprod 6:1 160–6. Jauniaux E, Gulbis B, Burton GJ 2003 The human first trimester gestational sac limits rather than facilitates oxygen transfer to the foetus: a review. Placenta Suppl A 24:S86–93. Jauniaux E, Cindrova ­Davies T, Johns J et al 2004 Distribution and transfer pathways of antioxidant molecules inside the first trimester human gestational sac. J Clin Endocrinol Metab 89:1452–8. Jauniaux E, Hempstock J, Teng C et al 2005 Polyol concentrations in the fluid compartments of the human conceptus during the first trimester of pregnancy: maintenance of redox potential in a low oxygen environ ­ ment. J Clin Endocrinol Metab 90:1 171–5. Jauniaux E, Poston L, Burton GJ 2006 Placental ­related diseases of preg ­ nancy: involvement of oxidative stress and implications in human evo ­ lution. Hum Reprod Update 12:747–55. Jones CPJ, Jauniaux E 1995 Ultrastructure of the materno ­embryonic inter ­ face in the first trimester of pregnancy. Micron 2:145–73. Koch FR, Wagner CL, Jenkins DD et al 2014 Sex differences in cerebral blood flow following chorioamnionitis in healthy term infants. J Perinatol 34: 197–202.
267
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Implantat Ion and placentat Ion 180.e2 SectIon 2McDonald SJ, Middleton P, Dowswell T et al 2013 Effect of timing of umbili ­ cal cord clamping of term infants on maternal and neonatal outcomes. Cochrane Database Syst Rev 7:CD004074. Murphy BE, Clark SJ, Donald IR et al 1974 Conversion of maternal cortisol to cortisone during placental transfer to the human fetus. Am J Obstet Gynecol 1 18:538–41. Myatt L, Sun K 2010 Role of fetal membranes in signaling of fetal maturation and parturition. Int J Dev Biol 54:545–53. Roescher AM, Timmer Erwich JJ, Bos AF 2014 Placental pathology, perinatal death, neonatal outcome, and neurological development: a systematic review. Plos One 9:e89419. Ross MG, Beall MH 2014 Amniotic fluid dynamics. In: Creasy R, Resnik R, Iams JD et al (eds) Saunders’ Maternal ­Fetal Medicine: Prin ­ ciples and Practice, 7th ed. Philadelphia: Elsevier, Saunders; Ch. 3, pp. 47–52.This chapter provides an overview of the production, circulation and absorption of amniotic fluid during gestation. Ross EJ, Graham DL, Money KM et al 2015 Developmental consequences of fetal exposure to drugs: what we know and what we still must learn. Neuropsychopharmacology 40:61–87. Stegmann BJ, Carey JC 2002 TORCH Infections. Toxoplasmosis, Other (syphilis, varicella ­zoster, parvovirus B19), Rubella, Cytomegalovirus (CMV), and Herpes infections. Curr Womens Health Rep. 2:253–8. Strauss JF 2013 Extracellular matrix dynamics and fetal membrane rupture. Reprod Sci 20:140–53.This paper presents the composition and biomechanical properties of the fetal membrane extracellular matrix. It considers the changes in extracellular matrix composition during normal pregnancy and term delivery and those seen in preterm premature rupture of membranes. Takashina S, Ise H, Zhao P et al 2004 Human amniotic epithelial cells possess hepatocyte ­like characteristics and functions. Cell Struct Funct 3:13–84. Tong X 2013 Amniotic fluid may act as a transporting pathway for signalling molecules and stem cells during the embryonic development of amni ­ otes. J Chin Med Assoc 76:606–10. This paper considers the bi-directional substance exchange between amniotic fluid and fetal tissues, its role in transporting signaling molecules and the dynamic nature of amniotic fluid throughout pregnancy. Tong X, Wang L, Gao T et al 2009 Potential function of amniotic fluid in fetal development – novel insights by comparing the composition of human amniotic fluid with umbilical cord and maternal serum at mid and late gestation. J Chin Med Assoc 72:368–73. Vanderhoeven JP, Bierle CJ, Kapur RP et al 2014 Group B streptococcal infection of the choriodecidua induces dysfunction of the cytokeratin network in amniotic epithelium: a pathway to membrane weakening. PLOS Pathog 10:e1003920. Wynn JL, Neu J 2012 The neonatal gastrointestinal tract as a conduit to systemic inflammation and developmental delays. Neu J: Gastroenterol ­ ogy and Nutrition: Neonatology Questions and Controversies, 2nd ed. Philadelphia: Elsevier, Saunders; Ch.19, pp. 293–304.
268
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
181 CHAPTER 10 Cell populations at gastrulation the region of the disc closest to the streak ‘caudal’, and the region of the disc furthest from the streak ‘cranial’ or ‘rostral’ . With the develop ­ ment of the streak, the terms medial and lateral can be used. The relative dimensions of the primitive streak and the fates of the cells that pass through it change with the developmental stage. Thus, the streak extends half way along the disc in the stage 6 embryo, reaching its greatest relative length in stage 7 and its maximum length in stage 8. Formation of the primitive streak is induced by the underlying vis ­ ceral hypoblast, which remains beneath the streak even at later stages. CONCEPTUS WITH A BILAMINAR EMBRYONIC DISC At stage 6, the conceptus is composed of the walls of three cavities: the large chorionic cavity is surrounded by a meshwork of trophoblast and developing villi, and lined with extraembryonic mesoblast. The chorion, trophoblast and extraembryonic mesoblast enclose the extraembryonic coelom and contain the much smaller amniotic cavity and yolk sac (see Fig. 9.1). These latter cavities abut at the embryonic bilaminar disc where the epithelial epiblast and visceral hypoblast are approximated. A fourth cavity, the allantois, will form as a hypoblastic diverticulum in stage 7. The ‘bilaminar disc’ commonly referred to in embryology texts does not yet possess the definitive layers of embryonic ectoderm and endoderm that will give rise to embryonic structures. Only the epiblast will give rise to the embryo; all other layers produced so far are extra­ embryonic. The amnion and chorion (and surrounding mesoblast) are part of the extraembryonic somatopleure, whereas the yolk sac, allan ­ tois and surrounding extraembryonic mesoblast constitute extraembry ­ onic splanchnopleure. At the junctional zone surrounding the margins of the embryonic area, where the walls of the amnion and yolk sac converge, the somatopleuric and splanchnopleuric layers of extraem ­ bryonic mesoblast are continuous. The terms epiblast and hypoblast are used to make the distinction between the earliest bilaminar disc layers and the later embryonic layers. Epiblast and hypoblast contain mixed populations of cells with little restriction, which establish the placental structures and extraem ­ bryonic tissues before the production of embryonic cell lines at gastrula ­ tion. The older terminology depicting three germ layers that give rise to the skin, gut lining and intervening tissues is thus incorrect for the bilaminar and trilaminar embryonic disc. The application and retention of this aged terminology for the early stages of embryology continue to cause confusion and inhibit the development of more pertinent descrip­ tive language to describe these early events. The earliest three cell line ­ ages, not layers, that give rise to the extraembryonic membranes and the embryo are trophoblast, hypoblast and epiblast; each expresses dif ­ ferent lineage ­specific genes for its establishment, mainten ance and differentiation (Artus and Hadjantonakis 2012). At early stage 6, the epiblast is producing extraembryonic mesen­ chyme from its caudal margin. With the appearance of the primitive streak, a process is begun whereby cells of the epiblast either pass deep to the epiblast layer to form the populations of cells within the embryo, or remain on the dorsal aspect of the embryo to become the embryonic ectoderm. Although human embryos do not form a ‘gastrula’ as such, the term gastrulation is used here to denote an early period of develop ­ ment during which significant rearrangements, migrations and folding of the early embryo occur. The primitive streak is the site of organizer cells analogous to those found in embryos that do undergo gastrula ­ tion. The appearance of the primitive streak therefore marks the begin ­ ning of a period when gross alterations in morphology and complex rearrangements of cell populations occur. During this time, the epiblast will give rise to a complex multilaminar structure with a defined cranio ­ caudal axis. By the end of gastrulation, cell populations from different, often widely separated, regions of the embryonic disc will become spatially related and the embryonic shape will have been produced. Primitive streak and node Seen from the dorsal (epiblastic) aspect, at stage 6, the embryonic disc appears elongated. The primitive streak is first seen in the caudal region of the embryonic disc at this stage as a collection of pluripotent cells, orientated along its long axis in the median position, conferring the future craniocaudal axis of the embryo ( Figs 10.1–10.2). Although the future cranial and caudal regions of the embryo are well within the boundaries of the embryonic disc, it has become the practice to term Fig. 10.1 A longitudinal section through an early conceptus. Ingression of mesoblast is occurring at the primitive streak and the notochord is ingressing via the primitive (Hensen’s) node. Epiblast Primitive node NotochordHypoblast MesenchymeYolk sacAllantoisExtraembryonic mesoblast Primitive streak Amniotic cavity Fig. 10.2 A transverse section through the embryonic plate at the level of the primitive streak to show the early movement of mesoblast between the epiblast and underlying hypoblast. Primitive streak Extraembryonic mesoblastEpiblast Hypoblast MesenchymeEndodermal and notochord cells
269
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Cell populations at gastrulation 182seCtion 2 along the long axis of the embryonic disc. By stages 9 and 10, the cells at the lateral edge of the plate have begun to migrate laterally as free mesenchymal cells and the plate reduces in height to two cells deep. At stage 1 1, the migrating prechordal mesenchyme forms bilateral pre­ mandibular mesenchymal condensations and is no longer a median structure. The extent of prechordal cells remaining within the endoderm is not clear. Notochord The notochord, also called chordamesoderm, the head process or chorda, arises from epiblast cells of the medial part of the primitive node. It passes through several stages during development. The cells of the early notochordal process express myogenic markers transitorily as they migrate beneath the epiblast, but later they become epithelial, forming junctions and a basal lamina. The notochordal cells are inti ­ mately mixed with endodermal cells, as both cell lines ingress at the same time (see Figs 10.1–10.2; Fig. 10.4). In the stage 8 embryo, the ingressing notochordal cells remain in the midline along the cephalo ­ caudal axis. They form a rostral part, which is composed of a cell mass continuous with the prechordal mesenchyme; a mid portion, in which cells are arranged in a tube with a central notochordal canal; and a caudal epithelial layer of cells, the notochordal plate, which is contigu ­ ous with the embryonic endoderm and forms a roof to the secondary yolk sac. There is a transitory opening between the primitive node (and amniotic cavity) and the secondary yolk sac called the neurenteric canal (so named because its upper opening is in the future caudal floor of the neural groove, and its lower opening is into the archenteron, which is the primitive gut); it may still be found at stage 9, and the site of the neurenteric canal can be recognized in stage 10 embryos. The ingression of notochordal cells at the primitive node is matched by specification of the overlying neural ectodermal cells, and the notochordal plate is thus matched in length by the future neural floor plate. Both the noto ­ chord and the region of the floor plate of the neural tube may arise from a common progenitor cell. The early notochord is important for the maintenance and subsequent development of the neural floor plate and the induction of motor neurones. Removal of the notochord results in elimination of the neural floor plate and motor neurones, and expression of sensory cell types. Caudal eminence From stages 9 and 10, the region between the neurenteric canal and the cloacal membrane (see below), including the primitive streak, is termed the caudal eminence. It consists of the caudal region of the trunk, com ­ posed of mesenchyme derived from the primitive streak and epiblast, and covered with surface ectoderm. Whereas ingression of cells through the primitive streak gives rise to the prechordal and notochordal plates, and cells rostral to the neurenteric canal (see below), the cells of the caudal eminence arise from local division of a mesenchymal popula ­ tion positioned caudal to the neurenteric canal. The caudal portions of the notochord, which form later in development when secondary neur­ulation processes begin, arise from these cell populations, sometimes termed the caudoneural hinge or junction. This tissue is thicker and Fig. 10.3 The predictive fates of the epiblast cell population at the time the primitive streak is present. Neural ectoderm Surface ectoderm Mesoblast Extraembryonic mesoblastMedial halves of the somites Lateral halves of the somitesNotochord Endoderm Primordial germ cellsThe primitive streak may be considered to be generally homologous with the blastopore of lower vertebrates (e.g. amphibia), with the nodal region corresponding to the dorsal lip. Experiments clearly show the lip of the blastopore to be a dynamic wave front on which cells are carried into the interior to form the roof of the archenteron, a situation analogous to ingression through the node of the prechordal plate and endoderm. The primitive streak similarly may be considered analogous to the coapted, or fused, lateral lips of the blastopore, and the cloacal membrane and its immediate environs are considered analogous to the ventral lip of the blastopore. At the primitive streak, epiblast cells undergo a period of intense proliferation, the rate of division being much faster than that of blasto­ meres during cleavage. Streak formation is associated with the local production of several cell layers, extensive disruption of the basal lamina, increase in adhesive plaques and gap junctions, synthesis of vimentin, and loss of cytokeratins by the emerging cells. As the epiblast cells proliferate, two ridges are formed on each side of the primitive streak, which appears to sink between them. The lower midline portion of the streak is termed the primitive groove. The process by which cells become part of the streak and then migrate away from it beneath the epiblast is termed ingression. The primitive node, or Hensen’s node, is the most rostral region of the primitive streak. It appears as a curved ridge of cells similar in shape to the top of an old ­fashioned keyhole. Cells ingressing from the ridge pass into the primitive pit (the most rostral part of the primitive groove), and then migrate rostrally beneath the epiblast. The primitive node has been recorded in all stage 7 human embryos; it produces axial cell populations, the prechordal plate, notochord, embryonic endoderm and the medial halves of the somites. Experimental removal of the node results in complete absence of the notochord and a failure of neurulation. Recent studies on the dynamics of ingression at the primitive streak have shown movement of the underlying extracellular tissues as well as of the overlying cells. Re ­examination of the processes of cell movement relative to the surrounding extracellular molecules in embryos has found that vertical and convergent extension motion patterns, previ ­ ously thought to be limited to the epiblast, also occur in sub ­epiblastic extracellular matrix fibrils of fibronectin (Zamir et al 2008, Szabó et al 201 1). Position and time of ingression through the primitive streak Studies of cell fate have shown that epiblast cells that will pass through the streak are randomly located within the epiblast layer before their ingression, and that epiblast fate is determined at or before the time of ingression through the streak, indicating that passage through the prim ­ itive streak is the most important factor for future differentiation. The position and time of ingression through either streak or node directly affect the developmental fate of cells. Passage through the streak is specified according to position, e.g. via the node, or rostral, middle or caudal regions of the streak. Cells that ingress through the primitive node give rise to the axial cell lines, the prechordal mesenchyme and notochord, and to the endoderm and the medial halves of the somites. The rostral portion of the primitive streak produces cells for the lateral halves of the somites, whereas the middle streak produces the lateral plate mesoblast. The adjacent caudal portion of the streak gives rise to the primordial germ cells, which can be distinguished histologically and histochemically, and the most caudal portion of the streak contributes cells to the extraembryonic mesoblast until the somites are visible. A composite of the information on the position of ingression through the streak and node is shown in Figure 10.3. The epiblast cells that do not pass through the streak but instead remain within the epiblast popula ­ tion give rise to the neural and surface ectoderm of the embryo. Prechordal plate The earliest cells migrating through the primitive node and streak give rise to both the embryonic endoderm and the notochord. The pre ­ chordal plate is first seen at stage 7. It has been defined as a localized thickening of the endoderm rostral to the notochordal process, although it is seen as a highly developed mesenchymal mass in contact with the floor of the neural groove, rostral to the notochordal process, rather than as an epithelial layer. The prechordal plate is a temporary collec ­ tion of cells that underlies the neural plate during stage 9. It is com­posed of cells that are similar to, or larger and more spherical than, the ingressing endodermal cells (Müller and O’Rahilly 2003). In stage 8 embryos, the prechordal plate is up to eight cells deep and extends
270
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Conceptus with a bilaminar embryonic disc 183 CHapter 10 Intraembryonic mesoblast (mesenchyme) Epiblast cells ingress through the cranial and middle parts of the streak individually, maintaining their apical epithelial contacts while elong­ ating ventrally. The cells become flask ­shaped, with thin, attenuated apical necks and broad basal regions. Their basal and lateral surfaces form lamellipodia and filopodia, and the apical contact is released. The cells are now free mesoblast cells, their fibroblastic, stellate morphology reflecting the release from the epithelial layer. Once through the streak, the cells migrate away from it, using the basal lamina of the overlying epiblast and extracellular matrix as a substratum. The cells contact one another by filopodia and lamellipodia, with which they also contact the basal lamina. Gap junctions have been observed between filopodia and cell bodies. With the appearance of the mesoblast, spaces form between the epiblast and visceral hypoblast that are filled with extracel ­ lular matrix rich in glycosaminoglycans. The migrating mesoblast has a leading edge of cells that open up the migration routes, and the follow ­ ing cells seem to be pulled along behind in a coordinated mass move ­ ment. Mesoblast formed by cells migrating through the primitive node and rostral primitive streak will form the paraxial mesenchyme, whereas cells migrating through the middle to caudal streak will form the lateral plate mesenchyme (see Figs 10.1–10.2). Embryonic ectoderm When the ingression of cells through the primitive streak is completed, the epithelial cells remaining in the epiblast layer are termed embryonic more advanced in differentiation than the tissues derived from the early primitive streak. Embryonic endoderm Before ingression, definitive embryonic endoderm cells are found in the epiblast, located at the primitive node and rostral primitive streak. In the mouse, the endodermal cells lie beneath the epiblast mainly in the midline, interspersed with presumptive notochordal cells, forming the roof of the secondary yolk sac. The ingressing endoderm displaces the visceral hypoblast into the secondary yolk sac wall by a dramatic territorial expansion that is brought about by a change in the morph­ ology of the cells (see Figs 10.1–10.2, 10.4). The putative endoderm cells are cuboidal epithelial cells within the node but they become squamous in the endoderm layer; this could result in a four ­fold increase in the surface area covered by the cells. A complete replacement of the visceral hypoblast has not yet been confirmed and there may be a mixed popu ­ lation of cells in the endodermal layer in the early stages. Ingression of cells through the streak and node in the human is apparent at stage 6, and, by stage 7, a population of endoderm and notochord cells is present beneath the epiblast (see Figs 10.1–10.2, 10.4). During stages 6–1 1, the midline roof of the secondary yolk sac becomes populated mainly by the notochordal plate, which remains in direct lateral conti ­ nuity with the endodermal cells. It is not until stage 1 1, after the defini ­ tive notochord is formed, that the endoderm cells can join across the midline. For the developmental fate of the embryonic endoderm, see Figure 12.3.Fig. 10.4 A, The unfolded embryo, showing the disposition of the intraembryonic coelom within the embryonic disc. The lines across the embryo show the level of transverse sections through the disc shown in B–D. B–D, Transverse sections through the disc at the points indicated in A. E, A longitudinal section through the disc. Primitive streak AllantoisNeural tissue Primitive streak Connecting stalk Cloacal membrane Entrance to intraembryoniccoelomA Ectoderm Bucco-pharyngealmembrane Endoderm Yolk sac MesenchymeIntraembryonic coelom Buccopharyngeal membraneIntraembryonic coelom Neural tissue Primitive streak Cloacal membrane Notochordal plateNotochordal plate Intraembryonic coelom Intraembryonic coelomB C DAmniotic cavity Amniotic cavity Connecting stalk Allantois Yolk sacE
271
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Cell populations at gastrulation 184seCtion 2Mesoblast that passes in a cranial direction flanks the notochordal plate and passes around the prechordal plate region, converging medi ­ ally to fuse in the midline beyond its cephalic border. This transmedian mass, in which the heart and pericardium will develop, is initially termed the cardiogenic mesoblast. It fuses with the junctional zone of extraembryonic mesoblast around the extreme cephalic margin of the embryonic area. This region will eventually form the septum transver ­ sum and primitive ventral mesentery of the foregut. Mesoblast passing laterally from the streak soon approaches and becomes confluent with the extraembryonic mesoblast around the margins of the disc, i.e. at the junctional zone where the splanchnic and somatic strata of extraembry­ onic mesoblast merge. Mesoblast that streams caudally from the primi­ tive streak skirts the margins of the cloacal membrane and then converges towards the caudal midline extremity of the embryonic disc to become continuous with the extraembryonic mesoblast of the con ­ necting stalk. It is unclear if the lower layer of the cloacal membrane consists of visceral hypoblast, like the more cranial primitive streak (the hypoblast is necessary for maintaining the streak), or if it is replaced by migrating embryonic endoderm, or if there is a region for ingression of endoderm at the caudal end of the streak, similar to the node cranially. Still further caudally, the embryonic disc develops a midline diver ­ ticulum adjacent to the cloacal membrane. This diverticulum, the allantois, projects into the extraembryonic connecting stalk (see Figs 9.9, 10.4; Fig. 10.6). There is little information about which cells form the allantois, i.e. whether it is composed of visceral hypoblast, parietal hypoblast or embryonic endoderm. The allantois later devel­ ops a rich anastomotic blood supply around it, in the manner of the yolk sac. The generation of cells at the primitive node produces midline endo ­ derm, notochord and the floor plate of the future neural tube. As the notochord grows and elongates, there is a matched growth of neural floor plate cells until both cell lines extend to the buccopharyngeal membrane. The epiblast lateral to the midline contains both future surface and neural ectoderm. The latter becomes arranged between the primitive node and buccopharyngeal membrane; cells destined to be in the neural plate lie medially, and those destined for the neural crest lie along the junction between the neural plate and surface ectoderm (see Fig. 10.5). A smaller subpopulation of neuronal cells, the ectodermal placodes, are arranged either close to the neural crest or within the rostral limit of the neural plate itself. FOLDING OF THE EMBRYO In a diagrammatic representation of the trilaminar disc prior to folding and viewed from the ectodermal aspect, all of the future external surface of the body is delimited (see Fig. 10.5). The ends of the gut tube are specified on the ectodermal surface at the buccopharyngeal and cloacal membranes, which are regions where the ectoderm and underlying endoderm are apposed without intervening mesoblast. In the midline between these membranes, proliferation of the neural ectoderm matches the underlying migration of mesoblast from the primitive streak, so that the neural plate covers the paraxial mesenchyme on each side of the notochord (see Fig. 10.6E). As the paraxial mesenchyme segments, the formation of the epithelial somites elevates the edges of the neural plate and initiates primary neurulation (see Fig. 10.6F; Figs 10.7–10.8). The neural plate itself undergoes concurrent morphological changes. The most medial cells become wedge ­shaped, forming the neural groove. Further elevation of the edges of the neural groove permits fusion of the neuronal populations in the dorsal midline to form the neural tube. The surface ectoderm forms the putative dorsal epidermis (see Figs 10.6G, 10.7–10.8). Cells at the lateral edge of the neural plate, termed neural crest cells, remain as a linearly arranged mesenchymal population between these two epithelia. Fusion of the neural tube begins in the future rhombencephalic region of the embryo and pro ­ ceeds rostrally and caudally to about the level of somite 29. Neurulation is described further in Chapter 17. A population of neural epithelial cells remain within the surface ectoderm; at this stage they are termed ectodermal placodes. The representation of a person on the trilaminar disc ( Fig. 10.9) shows, to some extent, the way in which the positions of the main body structures are already specified in the unfolded embryo. Ectoderm lateral to the neural plate and the paraxial mesenchyme will form structures within the back. The portion of the disc between the bucco­pharyngeal membrane and the edge of the disc will become the ventral thoracic wall and the ventral abdominal wall cranial to the umbilicus. Further caudally, midway along the neural axis, the lateral portions of the disc will become the lateral and ventral abdominal walls of the ectoderm cells. This layer still contains a mixed population because both surface ectoderm cells and neural ectoderm cells are present. It is believed that these cells were originally in the cranial half of the disc when the primitive streak first appeared, at which time the neural ­fated cells were closest to the streak, and the surface ectoderm cells were most cranial (see Fig. 10.3). The process of primary neurulation relocates most of the neuroepithelial cells (see below). Primordial germ cells Although early studies on human embryos have reported primordial germ cells, and described their development from the early endoderm of the yolk sac and allantois, it is now clear from animal experimenta ­ tion that the primordial germ cells arise from epiblast ingressing at the caudal end of the primitive streak (see Fig. 10.3). It is not known whether these cells originate from rostral regions that migrate to the streak or from local caudal regions. Extremely early segregation of the germ cells, when the epiblast layer consists of only 10–13 cells, has been demonstrated. It has been suggested that the primordial germ cells remain sequestered in the extraembryonic mesenchyme at the caudal end of the embryo until the embryonic endoderm has been produced and gastrulation completed, and that they start to migrate along the allantoic and hindgut endoderm as the folding of the embryo begins. The formation of the tail fold brings the proximal portion of the allan ­ tois within the body, so reducing the final distance over which the cells migrate to the genital ridges. Further development of the germ cells is described in Chapter 72. TRILAMINAR DISC Although the stage 8 embryo is termed a trilaminar disc, the concept of three epithelial layers forming a trilaminar disc is incorrect; the middle, mesoblast, layer is several cells thick with intervening extracel ­ lular matrix. The embryo at this stage, approximately 23 days after ovu ­ lation, is pear ­shaped, and broader cranially than caudally ( Fig. 10.5). The upper epiblast cells are tall and form a pseudostratified columnar epithelial layer with a basal lamina, except at the primitive streak, where the cells are ingressing to form the other layers. The more centrally placed epiblast will give rise to neural ectoderm (neurectoderm) and the more laterally placed epiblast will give rise to surface ectoderm. The future neural ectoderm is seen as a neural plate that matches the length of the notochordal plate directly beneath, being slightly wider near the prechordal plate. The lower embryonic endoderm, a simple squamous layer with a developing basal lamina, is not always complete at this stage, particularly in the midline caudal to the prechordal plate, which is still occupied by the notochordal process or plate. The middle, mesoblast, layer is composed of free cells migrating cranially, laterally and caudally from the primitive streak (see Fig. 10.4). They produce extracellular matrix, which separates the epiblast and endoderm of the embryonic area and permits their passage. The streams of mesoblast extend between the epiblast and endoderm over all of the disc area except cranially at the buccopharyngeal membrane (where the endoderm and ectoderm become apposed once the prechordal mesen ­ chyme has migrated laterally), and caudally at the cloacal membrane (a patch of thickened endoderm, similar to the buccopharyngeal mem ­ brane, caudal to the primitive streak). The mesoblast on each side of the notochord is termed paraxial mesenchyme. Fig. 10.5 The extent and shape of the neural plate in an unfolded embryo. Buccopharyngeal membrane Neural plate Cloacal membrane
272
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Formation of the intraembryonic coelom 185 CHapter 10 trunk. The portion of the disc beyond the cloacal membrane will form the ventral abdominal wall caudal to the umbilicus. The circumference of the disc, where the embryonic tissue meets the extraembryonic mem ­ branes, will become restricted to the connection between the ventral abdominal wall and the umbilical cord, i.e. the umbilicus. Head folding begins at stage 9, when the fusing cranial neural plate rises above the surface ectoderm and the portion of the disc rostral to the buccopharyngeal membrane (which contains the cardiogenic mes ­ enchyme) moves to lie ventral to the developing brain (see Fig. 10.6). The prosencephalon and buccopharyngeal membrane are now the most rostral structures of the embryo. The previously flat region of endoderm, which may contain cells from the prechordal plate, is now modified into a deep tube, the primitive foregut. Tail folding can be seen in stage 10 embryos, when the entire embryo comes to rise above the level of the yolk sac. Similar movement of the part of the disc caudal to the cloacal membrane results in its repositioning ventral to the neural plate. Generally, as the embryo rises above the edges of the disc, the lateral regions of the disc are drawn ventrally and medially, contributing to the lateral folding of the embryo. FORMATION OF THE INTRAEMBRYONIC COELOM At and just before stage 9 (before formation of the head fold), vesicles appear between the mesenchymal cells cranial to the buccopharyngeal membrane and within the cranial lateral plate mesenchyme. At the periphery of the vesicles, the mesenchymal cells develop junctional complexes and apical polarity, and form an epithelium. The vesicles become confluent to form a horseshoe ­shaped tube, the intraembryonic coelom, which extends caudally to the level of the first somite and lat­ erally into the lateral plate mesenchyme towards the extraembryonic mesenchyme. The lateral plate mesenchyme thus develops somatopleu ­ ric coelomic epithelium subjacent to the ectoderm, and a splanchno­ pleuric coelomic epithelium next to the embryonic endoderm (see Fig. 12.2C(iv)). At this stage, the intra ­ and extraembryonic coeloms do not communicate. During development of the head fold, the morphological move ­ ments that organize the foregut and buccopharyngeal membrane have a similarly profound effect on the shape of the intraembryonic coelom. The midline portion of the originally flat, horseshoe ­shaped coelom moves ventrally, leaving the caudal arms of the horseshoe in their origi ­ nal position. In this way, the midline part of the coelom, which was originally just rostral to the buccopharyngeal membrane, comes to lie ventral to the foregut (caudal to the buccopharyngeal membrane), and the two lateral extensions of the coelom pass close to the lateral walls of the foregut on each side. The caudal portions of the coelom (the two arms of the horseshoe), which, in the unfolded disc, communicated laterally with the extraembryonic coelom, turn 90° to lie lateral to the gut, and communicate with the extraembryonic coelom ventrally. Compartments of the coelom that will give rise to the body cavities later in development can already be seen. The midline ventral portion, caudal to the buccopharyngeal membrane, becomes the pericardial cavity. The canals lateral to the foregut (pericardioperitoneal canals) become the pleural cavities and the uppermost part of the peritoneal cavity. The remaining portion of the coelom becomes the peritoneal cavity. By stage 1 1, the intraembryonic coelom within the lateral plate mesenchyme extends caudally to the level of the caudal wall of the yolk sac. The intra ­ and extraembryonic coeloms communicate widely on each side of the midgut along the length of the embryo from the level of the fourth somite ( Fig. 10.10). In the early embryo, the intraembryonic coelom provides a route for the circulation of coelomic fluid and, with the beating of the heart tube, functions as a primitive circulation that takes nutritive fluid deep into the embryo until it is superseded by the blood vascular system. The coelomic channel, and the primitive circulation that passes through it, is of paramount importance up to stage 13. Whereas the superficial tissues of the embryo can receive nutrients via the amniotic sac and yolk sac fluids, the deeper tissues are, until the formation of the coelom, under conditions similar to those found in tissue culture. However, from stage 10, exocoelomic fluid, propelled by the first contractions of the developing heart, is brought into contact with the deeply placed mesenchyme. This early ‘circulation’ ensures that an adequate supply of nutrients reaches the rapidly increasing amount of embryonic tissue, and meets most of the requirements of the deeper mesenchymal deriva ­ tives. From stage 12, the endothelial system expands and fills rapidly with plasma, which passes across the locally thinned coelomic epithe ­ lium into the large hepatocardiac channels that project into the peri ­ cardioperitoneal canals at the level of the seventh somite.Neural groove Somite Split lateral plateNotochordF Neural crest Neural tube Somite Notochord Intraembryonic coelomMidgutGB C Foregut Hindgut Midgut D Neural plate Amniotic cavity Paraxial mesenchyme Intraembryonic coelomE Notochordal plateYolk sacIntraembryonic coelom Bucco- pharyngealmembraneYolk sacCloacal membraneAllantoisNeural tissue Amniotic cavity Connecting stalkA Notochord Fig. 10.6 Head and tail (A–D) and lateral (E–G) folding of the embryo. A–D, Median sagittal (longitudinal or axial) sections through the embryonic disc at successive stages; the relative positions of the buccopharyngeal and cloacal membranes have been maintained; thus, the movement of the most rostral and caudal portions of the disc can be followed. As these portions of the disc move ventrally, the initially wide-open yolk sac becomes constricted and fore- and hindgut divisions can be seen; the midgut is that region remaining in wide connection to the yolk sac. E–G, Transverse sections through the midpoint of the embryonic disc at successive stages to illustrate lateral folding that occurs as neurulation proceeds.
273
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Cell populations at gastrulation 186seCtion 2 Fig. 10.7 Scanning electron micrographs of rat embryos at the time of neurulation. A, Ventral view, showing the neural fold (NF), and the heart (H), with the somatopleuric pericardial membrane and surface ectoderm removed; the arrow indicates the entrance to the foregut via the cranial intestinal portal. B, Dorsolateral view; the arrows indicate the extent of rostral (to the right) and caudal (to the left) neural tube formation. (Photographs by P Collins; printed by S Cox, Electron Microscopy Unit, Southampton General Hospital.) NF H A B Fig. 10.8 A human embryo at stage 10, 2.1 mm long, with nine somites: right lateral and dorsal aspects. Nearly all the yolk sac and the caudal amnion have been excised. First pharyngeal arch, dorsal extremityBrain(early expansionof neural fold) Yolk sac wall coveredin extraembryonicmesoderm Unsegmented paraxial mesenchymeSomites (5 and 6)Rostral neuroporeNeural crest Caudal neuroporeAmnion, cut edgesFig. 10.9 A representation of a person on the flat embryonic disc. The position of the central nervous system has been matched to the dimensions of the neural plate, and the position of the heart in the thorax to the position of the pericardial coelom. The limbs, although represented in this diagram, are not present on the disc at this stage. The usefulness of this diagram lies in its illustration of the extent of the anterior body wall both rostral to the buccopharyngeal membrane and caudal to the cloacal membrane. The future dorsal regions of the body are found medially on the disc, while the ventral regions of the body are situated laterally and peripherally on the disc. After head and tail folding and lateral folding, the peripheral edge of the disc becomes constricted as the edge of the umbilicus. Position of the pericardial coelom In spite of the importance of the coelom in defining the body cavi ­ ties, and of the coelomic epithelium in the production of the major mesenchymal populations of the trunk (Streeter 1942, Langemeijer 1976, O’Rahilly and Müller 1987) (see Fig. 12.2), until recently the overall contribution of the coelom and its epithelium to the embryo has received relatively little attention (Carmona et al 2013). The coelom is a single, tubular organ that may be compared to the neural tube, in that it possesses a specialized wall that encloses a cavity. Like neural ectoderm, proliferating coelomic epithelium is a pseudostratified columnar epithelium with an inner germinal layer, from which cellular progeny migrate. After the germinal phase, both epithelia ultimately form the lining of a cavity, ependyma for the neural epithelium, and mesothelium for the coelomic epithelium. Coelomic epithelium, like neural epithelium, produces cells destined for different fates from dif ­ ferent sites and at different developmental times. Coelomic cells are like the neural epithelium, in that they have apical epithelial specializations and tapering basal processes that are in direct contact with the underly ­ ing mesenchyme, without an intervening basal lamina. The possibility of the tapering processes forming directional signals for migrating progeny, similar to the radial glia of the neural tube, has not been examined. EMBRYONIC CELL POPULATIONS AT GASTRULATION After gastrulation, the cells of the embryo contribute to two fundamen ­ tal types of tissue, namely: epithelial and mesenchymal. Differentiation of specialized circulating blood cells and other cell types occurs in sequence. Embryonic and fetal cell types are replaced later in develop ­ ment or after birth. Epithelia Epithelial populations in the embryo have many of the morphological characteristics of differentiated epithelia, i.e. they are composed of sheets of closely packed, polarized cells, with narrow intercellular clefts containing minimal extracellular material, and a developed basal
274
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
embryonic cell populations at gastrulation 187 CHapter 10 their early patterning are termed organizers, e.g. the primitive streak. Later germinal epithelia give rise to system ­specific progenitor popula ­ tions, e.g. the ventricular zone of the neural tube. All epithelia other than special germinal epithelia divide to produce embryonic growth throughout development and may retain stem cells that will divide throughout life. Characteristically, epithelia clothe internal and external surfaces as simple or compound cellular sheets that separate phases of differing composition (e.g. the external environment and the subepithelial tissue fluids; intravascular and extravascular fluids, and so on). Traffic of mat­ erials in the intercellular clefts is limited. Traffic occurs across the cells because their limiting membranes, which function as energy ­dependent selective barriers, enhance the passage of some materials and impede the passage of others. Mesenchyme The terms mesoblast and mesenchyme are used in this text in a specific manner and are not interchangeable. Previously, cells forming a popula ­ tion between the epiblast and hypoblast were termed mesoderm and, more recently, mesenchyme. The terms primary and secondary mesen ­ chyme have been used to distinguish between those cells that arise from ingression through the primitive streak and those that arise from lamina containing specific proteins synthesized by the epithelium itself. The cells usually show juxtaluminal lateral surface specializations, such as desmosomes, tight junctions, gap junctions, and so on, and special­ izations of the apical surface, such as microvilli or cilia. Epithelial polarity factors establish the extent of the apical, lateral and basal domains of the individual cells forming an epithelium. Three main control pathways have been identified in maintaining apicobasal polarity (St Jonston and Sanson 201 1). The apical domain is specified by the Crumbs transmembrane protein; Baz/PAR ­3 specifies the posi ­ tion and extent of the intercellular junction; and Scribble restricts the size of the junctional domains. Once established with an underlying basal lamina, the basal surface displays integrins. Changes in the mor­ phology of an epithelial sheet will involve relative changes in the apical, lateral and basal domains of all cells. Planar polarity of epithelia is driven by intercellular cortical actin and myosin. Epithelial cells are contiguous via actomyosin contractile networks, which contact the lateral cell membrane through E ­cadherin adhesion (Gorfinkiel and Blanshard 201 1). Wnt/planar cell polarity signalling pathways are involved in these epithelial morphogenetic movements and also in orientated cell division which occurs in epithelial sheets (Tao et al 2009). Embryonic epithelia differ from those in the fetus and adult. Two distinct types can be identified. Early germinal epithelia, which give rise to epithelial or mesenchymal populations of the embryo and confer Fig. 10.10 A, An early stage in development of a human blastocyst. B, A blastocyst sectioned through the longitudinal axis of an embryo, showing the early formation of the allantois and the connecting stalk. C, A longitudinal section of an embryo at a later stage of development; the pericardial cavity can be seen at the most rostral part of the embryonic area. D, A longitudinal section of an embryo at a later stage, showing formation of the head and tail folds, the expansion of the amnion and the delimitation of the umbilicus. E, A transverse section along the line a–b in D; observe that the intraembryonic coelom communicates freely with the extraembryonic coelom. F, A longitudinal section of an embryo at a later stage, showing full expansion of the amniotic cavity and the umbilical cord. A a bD EFEmbryonic pole Amniotic duct ChorionTrophoblastConnecting stalk Secondary yolk sacChorionic mesoblastAllantoentericdiverticulum Early villous stemsAmniotic cavity Extraembryonic mesoblast Yolk sac Chorionic cavity/extraembryonic coelom Connecting stalk attachment becomes more ventrally placed Extraembryonic coelomTail fold Pericardial coelom Head foldBuccopharyngeal membranePlacental area (chorion frondosum)C Amniotic cavity Pericardial coelomEmbryonic area Chorionic mesoblastAllantoenteric diverticulum Villous stemConnecting stalk Extraembryonic coelom Splanchnopleuric mesoblast ofyolk sac Secondary yolk sacAllantoic canalHindgut Amniotic cavityMidgut Chorion laeveSeptum transversum Intercoelomic communicationChorionAmnion Intraembryonic coelomSomite Yolk sac Septum transversumBuccopharyngeal membrane Line of fusion ofamnion with chorion,i.e. line of obliteratedextraembryonic coelomAmnioticcavityPericardialcavity ForegutLateral body fold Extraembryonic coelomNeural groove Notochord Midgut Yolk duct Yolk sac wall ExtraembryonicmesoblastYolk sacAmniotic cavityForegut Umbilical cord Obliterated yolk duct Umbilical coelom Cloacal membrane Allantoic ductCloacal membraneB
275
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Cell populations at gastrulation 188seCtion 2routes, removes the support for overlying epithelia, and disrupts the normal branching of glandular systems. Fibronectin deposited extracellularly along a migration pathway will affect cells that touch it later, causing realignment of their intracellular actin filaments and thus of their orientation; it also induces cell migra ­ tion. The receptors for extracellular matrix molecules such as fibronectin and laminin were originally termed integrins because they integrate extracellular proteins and intracellular cytoskeletal elements (via α and β subunits that span the cell membrane), allowing them to act together; the binding preference of integrins depends on their combination of subunits and environmental conditions. It is known that extracellular matrix is structured rather than random. Epithelial and mesenchymal cell populations can structure the space around them by secretion of particular matrix molecules or growth factors, which, in turn, can organize the cells that contact them. Cell– matrix interactions and matrix–cell interactions control the position of migration routes and cellular ‘decisions’ to migrate or to begin to dif ­ ferentiate. Matrix molecules propagate developmental instructions from cell to cell and form a far ­reaching four ­dimensional (spatial and tem ­ poral) mechanism of communication. Mapping the molecular path ­ ways and the biomechanical processes that underlie morphogenesis, the dynamic epigenetics of early mesenchyme cells and their secreted three ­ dimensional extracellular matrices, and the mechanisms by which the stiffness of the surrounding matrix affects the actions of the cells within it and the specific deposition of fibrillary proteins creates tissue bounda ­ ries and zones of tension, are all the subject of ongoing research (e.g. Davidson et al 2010, Loganathan et al 2012). Transition between epithelial and mesenchyme states Transformations of cell morphology from epithelium to mesenchyme, and vice versa, occur in specific places and times during development, and can be seen as ways of dispersing germinal centres with increasing restriction. The first epithelial ­to­mesenchyme transition occurs at the primitive streak, a germinal epithelium that confers embryonic specifi ­ cation on the resultant mesoblast population. The mesoblast so formed migrates and the cells undergo mesenchyme ­to­epithelial transitions when they reach their final destinations. Series of small epithelial ger ­ minal centres, the somites, are formed, as are larger, more extensive, germinal epithelial sheets that line the walls of the intraembryonic coelom. The coelomic walls, especially those derived from somato­ pleure and splanchnopleure, form germinal epithelia that give rise to the major mesenchymal populations that form the viscera. The early epithelial somites undergo further local epithelial ­to­mesenchyme tran ­ sitions to form the sclerotomes, and subsequently form several germinal epithelia in the epithelial plate of each somite. Later mesenchyme ­to­ epithelium transitions are not associated with the formation of germi ­ nal epithelia; the most common involve the transition of mesenchyme into the endothelium of the vascular system. The nephrons of the mes ­ onephric and metanephric systems also form from mesenchyme ­to­ epithelial transition.neural crest ingression, respectively. Primary mesenchymal cells revert to epithelia at their destinations. However, whereas some primary mesen chymal cells may become epithelial within a short timeframe, e.g. somites and lateral plate, other cells may transform later, e.g. the epithe lium lining blood vessels. To address these terminological con ­ flicts, the mixed population of epiblast cells that ingress through the primitive streak and come to lie between the epiblast and embryonic endoderm is termed mesoblast until the cells have migrated to their final position, at which time the populations of mesenchyme can be identified and their fates inferred. Mesoblastic and mesenchymal cells have no polarity. They form junctional complexes, which are not exclusively juxtaluminal, and they produce extracellular matrix molecules and fibres from the entire cell surface. Mesenchymal populations are formed from a range of germinal epithelia and by proliferation of mesenchymal cells directly, and occupy all the regions between the various epithelial layers described above. The term mesoderm is reserved for the coelomic epithelia that later form mesothelia. Mesenchymal cells support epithelia throughout the developing body, both locally where they contribute to the basement membrane and form the lamina propria and smooth muscle of tubes, and generally where they differentiate into connective tissue. Specific mesenchymal populations control the patterning of local regions of epithelium (e.g. the zone of polarizing activity on the postaxial limb border posterior to the apical ectodermal ridge). Extracellular matrix The space beneath epithelia and between mesenchyme cells is filled with extracellular matrix; both epithelial and mesenchymal cells syn ­ thesize extracellular matrix molecules and their receptors. Epithelial cells produce a two ­dimensional basal lamina, which contains a variety of matrix molecules, including laminin, fibronectin, type IV collagen and various proteoglycans. The particular molecules can vary during development according to spatial and temporal patterns, resulting in changes in the behaviour of the underlying mesenchymal cells (e.g. in patterning of the basal regions of the skull). Mesenchymal cells produce extracellular matrix molecules in three dimensions. Those adjacent to an epithelial layer will connect with its basal lamina, forming a base ­ ment membrane that secures the epithelial layer to the underlying tissue. Cells deep within a mesenchymal population may synthesize matrix molecules (fibrillar or granular) to separate cells locally, open up migration routes or leave information within the matrix to act on cell populations passing at a later time. Molecules of the extracellular matrix are complex; they include more than 19 individual types of collagen (some of which are capable of being individually spliced to give more than 100 variants), proteogly ­ cans and glycoproteins (which come in a wide variety of forms, with and without binding proteins), and elastic fibres (see Ch. 2). Hyaluronic acid, a glycosaminoglycan, has a vast capacity to bind water molecules and so create and structure the space between the mesenchymal cells, thereby producing much of the overall shape of an embryo. Experimen ­ tal removal of hyaluronic acid prevents the formation of cell migration KEY REFERENCES Müller F, O’Rahilly R 2003 The prechordal plate, the rostral end of the notochord and nearby median features in staged human embryos. Cells Tissues Organs 173:1–20. O’Rahilly R, Müller F 1987 Developmental Stages in Human Embryos. Washington: Carnegie Institution.St Jonston D, Sanson B 201 1 Epithelial polarity and morphogenesis. Curr Opin Cell Biol 23:540–6. Streeter GL 1942 Developmental horizons in human embryos. Descriptions of age group XI, 13 to 20 somites, and age group XII, 21 to 29 somites. Contrib Embryol Carnegie Inst Washington 30:21 1–45.
276
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Cell populations at gastrulation 188.e1 CHapter 10REFERENCES Artus J, Hadjantonakis A 2012 Troika of the mouse blastocyst: lineage seg ­ regation and stem cells. Curr Stem Cell Res Ther 7:78–91. Carmona R, Cano E, Mattiotti A et al 2013 Cells derived from the coelomic epithelium contribute to multiple gastrointestinal tissues in mouse embryos. PLoS One 8:e55890. Davidson LA, Joshi SD, Kim HY et al 2010. Emergent morphogenesis: elastic mechanics of a self ­deforming tissue. J Biomech 43:63. Gorfinkiel N, Blanshard GB 201 1 Dynamics of actomyosin contractile activ ­ ity during epithelial morphogenesis. Curr Opin Cell Biol 23:531–9. Langemeijer RA 1976 Le Coelome et son revêtement comme organoblast­ ème. Bull Assoc Anat 60:547–58. Loganathan R, Potetz BR, Rongish BJ et al 2012 Spatial anisotropies and temporal fluctuations in extracellular matrix network texture during early embryogenesis. PLoS One 7:e38266. Müller F, O’Rahilly R 2003 The prechordal plate, the rostral end of the notochord and nearby median features in staged human embryos. Cells Tissues Organs 173:1–20.O’Rahilly R, Müller F 1987 Developmental Stages in Human Embryos. Washington: Carnegie Institution. St Jonston D, Sanson B 201 1 Epithelial polarity and morphogenesis. Curr Opin Cell Biol 23:540–6. Streeter GL 1942 Developmental horizons in human embryos. Descriptions of age group XI, 13 to 20 somites, and age group XII, 21 to 29 somites. Contrib Embryol Carnegie Inst Washington 30:21 1–45. Szabó A, Rupp PA, Rongish BJ et al 201 1 Extracellular matrix fluctuations during early embryogenesis. Phys Biol 8:045006. Tao H, Suzuki M, Klyonarl H et al 2009 Mouse prickle 1, the homolog of a PCP gene is essential for epiblast apical ­basal polarity. PNAS 106: 14426–31. Zamir EA, Rongish BJ, Little CD 2008 The ECM moves during primitive streak formation – computation of ECM versus cellular motion. PLoS Biol 6:e247.
277
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
189 CHAPTER 11 Embryonic induction and cell division (further restrictions), they are said to be determined. Determined cells are programmed to follow a process of development that will lead to differentiation. The determined state is a heritable characteristic of cells and is the final step in restriction. Once a cell has become determined, it will progress to a differentiated phenotype if the environmental factors are suitable. The process of determination and differentiation within embryonic cell populations is reflected by the ability of these populations to produce specific proteins. Primary proteins (colloquially termed house - keeping proteins) are considered essential for cellular metabolism, whereas proteins synthesized as cells become determined, those specific to the state of determination, are termed secondary proteins; for example, liver and kidney cells, but not muscle cells, produce arginase. Fully differentiated cells produce tertiary proteins, which no other cell line can synthesize, e.g. haemoglobin in erythrocytes. The range of housekeeping, regulatory, and tissue-specific proteomes in adult cells is presented in the Human Protein Atlas (www.proteinatlas.org). As populations of cells become progressively determined, they can be described within a hierarchy of cellular development as transiently amplifying cells, progenitor cells, stem cells and terminally differenti - ated cells. Transiently amplifying cells Transiently amplifying cells undergo proliferative cell mitosis and produce cells that are equally determined. At some stage, and as a result of an inductive stimulus, these cells will enter a quantal cycle that culminates in a quantal mitosis. This will result in an increase in the restriction of their progeny, which continue to undergo proliferative mitoses at a progressive level of determination. The quantal mitosis corresponds to the time of binary choice when the commitment of the progeny is different from that of the parent. Progenitor cells Progenitor cells are already determined along a particular pathway. They may individually follow that differentiation pathway, or may proliferate and produce large numbers of similarly determined progenitor cells that subsequently differentiate; neuroblasts or myoblasts are examples of progenitor cells. Stem cells Either individually or as a population, stem cells can both produce determined progeny and reproduce themselves (see Commen - tary 1.2). It is generally believed that stem cells undergo asymmetric divisions, in which one daughter cell remains as a stem cell, while the other proceeds along a differentiation pathway; in marked contrast, proliferative cell division may be symmetrical, producing derived cells with an identical determination. Human embryonic stem cells (hESCs) are pluripotent cells that can be derived from the inner cell mass of human blastocysts in vitro; or obtained surplus to in vitro fertilization fertility programmes; or created from oocytes donated and fertilized for that purpose. Although not yet achieved, it is hoped that hESCs can be coaxed down particular pathways under appropriate pharmaceutical conditions to produce differentiated cells that will be effective in revers - ing some degenerative diseases (e.g. dopamine-producing neurones for Parkinson’s disease, insulin-producing islet cells for diabetes), or to replace acutely damaged tissues (e.g. motor neurones for acute spinal cord injury, cardiomyocytes in acute myocardial infarction). Proof of principle has been demonstrated in some animal models, and multipo - tent haemopoietic progenitor stem cells from human umbilical cord blood are now used as an alternative treatment to bone marrow trans - plantation for the treatment of some inherited genetic disorders (thalas - saemia) and blood malignancies (leukaemias). Terminally differentiated cells By virtue of their extreme special - ization, terminally differentiated cells can no longer divide, e.g. eryth- rocytes and neurones. Apoptosis is a particular variety of terminal differentiation in which the final outcome is the death of the individual cells or cell populations. It occurs in the developing limb, where cells EMBRYONIC INDUCTION AND CELL DIVISION The stages of genomic activity that unfold in the fertilized zygote from cleavage onwards are not yet fully elucidated. An admittedly simplistic but useful approach likens developmental processes to a series of binary choices (see Fig. 8.1), the first being the choice between trophoblast and inner cell mass, which specifies if the future tissue will be placenta and fetal membranes or embryo. Bioinformatic databases use embryological ontologies wherein the lineage of organs and tissues to be retraced to the earliest time (see Commentary 2.1). They provide a temporal and spatial framework against which the sequential activation of genes preceding changes in cell and tissue morphology can be matched. They also provide a means of viewing all of the complex partonomy, lineage data and temporal information occurring within embryos within a specific timeframe and of linking these to adult anatomical tissue types. As our knowledge of the complexity of early developmental events increases, ontological databases will assist the correlation of embryological information: current limitations are discussed in Bard (2012). Our knowledge of the early events that occur in a range of mam - malian species and our interpretation of the significance of particular sequences of gene expression is limited. The upregulation of early genes appears to reflect the time period of implantation. In the mouse, implantation occurs one day after blastocyst formation, thus genes relevant to implantation are expressed early, whereas in cattle, implan - tation occurs after 14 days, by which time gastrulation has occurred and neurulation has been initiated (Oron 2012). Within the developing blastocyst, the position of cells relative to each other leads to epigenetic changes. The acquisition of cell polarity is fundamental, because this specifies apical and basolateral domains, cell surface specializations, the position and nature of cell:cell junctions and the concomitant positioning of intracellular organelles and cytoskeletal elements, which in turn leads to the location of future deposition of extracellular proteins. The two cell phenotypes, epithelial and non-epithelial cells, are specified during cleavage and early compac - tion within the zona pellucida. All relevant genes required to maintain these cell shapes and behaviours are upregulated at these very early stages. During the formation of trophoblast, hypoblast (primitive endo - derm) and epiblast, the cells respond locally to their environment, specifically to other cells that are positioned lateral and basal to them. The genes operating at these early times in the mouse embryo are given in Takaoka and Hamada (2012). Cell populations within the embryo interact to provide the develop - mental integration and fine control necessary to achieve tissue-specific morphogenesis. In the early embryo, such interactions may occur only if particular regions of the embryo are present, e.g. signalling centres or organizers. As the embryo matures, some interactions tend to occur between adjacent cell populations, e.g. epithelium and mesenchyme, and later between adjacent differentiating tissues, e.g. between nerves and muscle, or between muscle and skeletal elements. The interactions between adjacent epithelia and underlying connective tissue continue throughout embryonic and fetal life and extend into postnatal life. In the adult, these interactions also permit the metaplastic changes that tissues can undergo in response to local environmental conditions. Tissue interactions result in changes or reorganization of one or both tissues, which would not have occurred in the absence of the tissue interactions. The process of tissue interaction is also called induc - tion, i.e. one tissue is said to induce another. The ability of a tissue to respond to inductive signals is called competence, and denotes the ability of a cell population to develop in response to the environments present in the embryo at that particular stage. After a cell population has been induced to develop along a certain pathway, it will lose com - petence and become restricted. Once restricted, cells are set on a par - ticular pathway of development; after a number of binary choices
278
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Embryonic induction and c Ell division 190sEction 2mesenchyme, which results in the morphogenesis of mammary gland-like structures; chickens do not normally develop mammary glands. Tissue interactions continue into adult life and are probably respon - sible for maintaining the functional heterogeneity of adult tissues and organs. This is exemplified by the complex tissue heterogeneity, with sharply compartmentalized boundaries, that occurs in the oral cavity. The junctions between the mucosa of the vestibule and the lip, and between the vermilion border and the facial skin, are distinct bound - aries of specific epithelial and mesenchymal differentiation, and are almost certainly maintained by continuing epithelial–mesenchymal interactions in adult life. Perturbation of these interactions throughout the body may underlie a wide variety of adult diseases, including sus - ceptibility to cancer and proliferative disorders. Signalling between embryonic cells and tissues Cellular interactions may be signalled by four principal mechanisms: direct cell–cell contact; cell adhesion molecules and their receptors; extracellular matrix molecules and their receptors; and growth factors and their receptors. Many of these mechanisms interact, and it is likely that combinations of them are involved in development. Figure 1 1.1 illustrates some ways by which mesenchymal cells could signal to epi - thelial cells. An additional set of identical mechanisms could operate for epithelial–mesenchymal cell signalling. Clearly, the complexity of these mechanisms will increase in reciprocal interactions; moreover, a single molecule may have different effects on epithelial and mesen - chymal cells. Direct cell–cell contact permits the construction of gap junctions, which are important for communication and the transfer of informa - tion between cells. The transient production of gap junctions is seen as epithelial somites are formed, between neuroepithelial cells within rhombomeres, and in the tunica media of the outflow tract of the heart. die along the pre- and postaxial limits of the apical ectodermal ridge, limiting its extent, and also between the digits, permitting their separation. Tissue interactions There are two types of cell and tissue interaction, namely: permissive and instructive. In a permissive interaction, a signal from an apposing tissue is neces - sary for the successful self-differentiation of the responding tissue. This means that a particular cell population (or the matrix molecules secreted by the cells that it contains) will maintain mitotic activity in an adjacent cell population. Since a variety of different cell populations may permit a specific cell population to undergo mitosis and cell dif - ferentiation, no specific instruction or signal that may limit the devel - opmental options of the responding tissue is involved; this signal therefore does not influence the developmental pathway selected and there is no restriction. The responding tissue has the intrinsic capacity to develop, and only needs appropriate environmental conditions in order to express this capacity. Permissive interactions often occur later in development, when a tissue whose fate has already been determined is maintained and stabilized by another. An instructive (directive) interaction (induction) changes the cell type of the responding tissue, so that the cell population becomes restricted. Wessells (1977) proposed four general principles in most instructive interactions: in the presence of tissue A, responding tissue B develops in a certain way; in the absence of tissue A, responding tissue B does not develop in that way; in the absence of tissue A, but in the presence of tissue C, tissue B does not develop in that way; and in the presence of tissue A, a tissue D, which would normally develop differ- ently, is changed to develop like tissue B. These principles are exemplified during induction of the lens vesicle by the optic cup. An example of last-mentioned principle is the experi - mental association of chicken flank ectoderm with mouse mammary Epithelium Basal lamina Mesenchyme cell 1 1 Direct cell–cell contact by gap junctions. 3 A soluble factor (growth factor) reacting with a receptor for that factor on the epithelial cells. 4 Extracellular matrix molecule secreted by the mesenchyme cells interacting with a receptor on the epithelial cell. 5 A soluble factor (growth factor) secreted by a mesenchymal cell having a biphasic action interacting (i) with a receptor on an epithelial cell, causing it to express a specific extracellular matrix molecule receptor; (ii) with a receptor on a mesenchyme cell, causing it to secrete a specific extracellular matrix molecule which then interacts with the induced epithelial receptor. 6 A soluble factor (growth factor) secreted by a mesenchyme cell interacting with a receptor on an epithelial cell, causing it to express a receptor or secrete a factor, which interacts with another factor synthesized, or receptor expressed, by another mesenchyme cell. 7 A soluble factor secreted by a mesenchyme cell interacting with a receptor on an epithelial cell, causing it to synthesize an extracellular matrix molecule (or a receptor for such a molecule) which then interacts with a specific receptor for that molecule on another mesenchyme cell. 8 A soluble factor secreted from a mesenchyme cell interacting with a receptor on an epithelial cell, causing it to synthesize a molecule which stabilizes or enhances the interaction between a mesenchymal-derived factor and its epithelial receptor. 9 A soluble factor secreted by a mesenchyme cell interacting with a receptor on an epithelial cell, causing the inhibition of synthesis/assembly of a factor or receptor. 10 A soluble factor secreted by a mesenchyme cell binding to the extracellular matrix of the basal lamina, where it remains active and subsequently interacts with a receptor on an epithelial cell which appears at a later developmental time. 2 Cell–cell contact by cell adhesion molecules.2 3 4 5 6 7 8 9 10 Fig. 11.1 The many ways by which mesenchyme cells could signal to epithelial cells. Precisely the same mechanisms can operate in reverse, i.e. epithelium to mesenchyme.
279
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
morphogenesis and pattern formation 191 cHaPtEr 11at the tip of the acinus is therefore split into two, and two branches develop from this point. Pattern formation concerns the processes whereby the individual members of a mass of cells, initially apparently homogeneous, follow a number of different avenues of differentiation that are precisely related to each other in an orderly manner in space and time. The pat - terns embraced by the term apply not only to regions of regular geo - metric order, e.g. the crystalline lens, but also to asymmetric structures such as the limb. For such a process to occur, individual cells must be informed of their position within the embryo, and utilize that informa - tion for appropriate differentiation. Patterning of regions is seen in: the Fig. 11.2 In addition to the mechanisms described in Figure 11.1, cells may also communicate by the reception, production and secretion of growth factors. A typical embryonic mesenchyme cell may receive and produce growth factors in this way. 1 Endocrine Delivery of growth factor by the blood stream from a distant biosynthetic site to the target mesenchyme cell. 2 Autocrine Synthesis of growth factor by the cell, its secretion, binding and activation of a surface receptor elsewhere on its own surface. 3 Paracrine Synthesis of growth factor by the cell, its secretion and diffusion to an adjacent cell (or group of cells) where it binds to and activates a cell surface receptor. 4 Juxtacrine Synthesis of growth factor by the cell. The growth factor remains on the cell surface and binds to and activates a receptor on an immediately adjacent cell. 5 Intracrine Synthesis of growth factor within the cytoplasm of the cell. The growth factor moves to the nucleus and binds and activates its own nuclear receptors. 6 Matricrine Synthesis and export of growth factor from the cell. The growth factor binds to the extracellular matrix where it remains active and subsequently binds to and activates a receptor for that growth factor on the same or a different cell.Extracellular matrix Blood vessel Mesenchyme cell 43 5 61 2 Fig. 11.3 Branching of a tubular duct may occur as a result of an interaction between the proliferating epithelium of the duct and its surrounding mesenchyme and extracellular matrix. A, Mesenchymal cells initiate cleft formation by producing collagen III fibrils locally within the development clefts and hyaluronidase over other parts of the epithelium. Collagen III prevents local degradation of the epithelial basal lamina by hyaluronidase and slows the rate of mitosis of the overlying epithelial cells. B, In regions where no collagen III is produced, hyaluronidase breaks down the epithelial basal lamina and locally increases epithelial mitoses, forming an expanded acinus (see arrows) . (Adapted from Gilbert SF 1991 Developmental Biology. Sunderland, MA: Sinauer Associates.)Narrow cleft Mesenchyme cellsCollagen fibrils HyaluronidaseCollagen Epithelial cells Collagen fibrilsHyaluronidase produced by mesenchyme cellsA BEndogenous electrical fields are also believed to have a role in cell–cell communication. Such fields have been demonstrated in a range of amphibian embryos, and in vertebrate embryos during primitive streak ingression. Neuroepithelial cells are electrically coupled, regardless of their position relative to inter-rhombomeric boundaries. The spatial and temporal distribution of a variety of cell adhesion molecules has been localized in the early embryo. The appearance of these molecules correlates with a variety of morphogenetic events that involve cell aggregation or disaggregation; e.g. an early response of groups of cells to embryonic inductive influences is the expression of cadherins, calcium-dependent adhesion molecules typically found in epithelial populations. Other molecules found in the extracellular matrix, e.g. fibronectin and laminin, inter alia, can modulate cell adhe - sion by their degree of glycosylation. Self-assembly or cross-linking by matrix molecules may affect cell adhesiveness by increasing the avail - ability of binding sites or by obscuring them. Extracellular matrix molecules include localized molecules of the basal lamina, e.g. laminin, fibronectin, and much larger complex asso - ciations of collagen, glycosaminoglycans, proteoglycans and glycopro - teins between the mesenchyme cells. Mutations of the genes that code for extracellular matrix molecules give rise to a number of congenital disorders, e.g. mutations in type I collagen produce osteogenesis imper - fecta; mutations in type II collagen produce disorders of cartilage; and mutations in fibrillin are associated with Marfan’s syndrome. Growth factors are distinguished from extracellular matrix mol - ecules. They can be delivered to, and act on, cells in a variety of ways, namely: endocrine, autocrine, paracrine, intracrine, juxtacrine and matricrine (Fig. 1 1.2). Many growth factors are secreted in a latent form, e.g. associated with a propeptide (latency-associated peptide) in the case of transforming growth factor beta, or attached to a binding protein, in the case of insulin-like growth factors. MORPHOGENESIS AND PATTERN FORMATION Morphogenesis may be described as the assumption of form by the whole, or part, of a developing embryo. As a term, it is used to denote the movement of cell populations and the changing shape of an embryo, particularly during early development. The most obvious examples of morphogenesis are the large migra - tions that occur during gastrulation; local examples include branching morphogenesis, which occurs in the developing lungs and kidneys, for example, and in most glandular organs. The development of branches from a tubular duct occurs over a period of time. An interaction between the proliferating epithelium of the duct and its surrounding mesen - chyme and extracellular matrix results in a series of clefts that produce a characteristic branching pattern (Fig. 1 1.3). During tubular and acinar development, hyaluronidase secreted by the underlying mesenchymal cells breaks down the basal lamina produced by the epithelial cells; this increases epithelial mitoses locally and results in an expanding acinus. Cleft formation is initiated by the mesenchyme, which produces col- lagen III fibrils within putative clefts. (If the collagen is removed, no clefts develop, whereas if excess collagen is not removed, supernumer - ary clefts appear.) The collagen acts to protect the basal lamina from the effects of the hyaluronidase, which means that the overlying epithe - lia have a locally reduced rate of mitosis. The region of rapid mitoses
280
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Embryonic induction and c Ell division 192sEction 2progress zone and zone of polarizing activity within the limbs; the fates of the medial and lateral, and later the cranial and caudal, halves of the somites; and the neural crest mesenchyme within the pharyngeal arches. For details of patterning in vertebrate development, see Tickle (2003). Genes in development Two related themes have emerged from experimental studies of devel - opment: first, that the control of embryonic morphology has been highly conserved in evolution between vertebrates and invertebrates; and second, that this control involves families of genes coding for proteins that act as transcriptional regulators. Homeobox genes are believed to be responsible, at least in part, for the evolutionary origin of the embryonic body plan (Robert 2001). Experimental study of transgenic animals in which the homeobox genes have been knocked out provide some evidence of their function; however, because developmental processes permit significant recovery from insult, some of the outcomes cannot be directly interpreted as demonstrating the effect of such gene loss. Other gene families required for normal development include the T-box family, Helix-Loop-Helix transcription factors and Sox genes; the signalling factors, transforming growth factor beta (TGF- β), bone mor- phogenetic proteins (BMPs), fibroblast growth factor family (FGF), the Wnt family and hedgehog signalling molecules. A range of cell receptor molecules are also required. For details on individual members within this range of factors, see Carlson (2014). Experimental approaches to embryology One of the most exciting techniques to provide information on cell movements and fates during development is the use of chimeric embryos. Small portions of an embryo are excised and replaced with similar portions of an embryo from a different species at the same stage, and the resulting development is then studied. This technique has been particularly effective using chick and quail embryos because the nucle - olus is especially prominent in all quail cells, whereas it is not promi - nent in chick cells, which means that quail cells may be easily identified within a chick embryo after chimeric transplantation (Le Douarin 1969). The technique has also confirmed the reciprocity of tissue inter - action between the embryonic species, a phenomenon that had previ - ously been illustrated, for a limited period, in co-cultures of embryonic avian and mammalian tissues. Somite development and vertebral for - mation have been studied in mouse–chick chimeras (Fontaine-Pérus 2000). The production, in vitro, of human–animal chimeric cell lines is providing new ways of studying cellular pathways, as is the introduction of human artificial chromosome vectors into animal cells to study their interaction. KEY REFERENCES Bard J 2012 A new ontology (structured hierarchy) of human developmental anatomy for the first 7 weeks (Carnegie stages 1–20). J Anat 221: 406–16. Carlson BM 2014 Human Embryology and Developmental Biology, 5th ed. Philadelphia: Elsevier, Saunders; Ch. 4. Fontaine-Pérus J 2000 Mouse–chick chimera: an experimental system for study of somite development. Curr Top Dev Biol 48:269–300. Le Douarin NM 1969 Particularités du noyau interphasique chez la caille japonaise (Coturnix coturnix japonica). Utilisation de ces particularités comme ‘marquage biologique’ dans les recherches sur les interactions tissulaires et les migrations cellulaires au cours de l’ontogenèse. Bull Biol Fr Belg 103:435–52.Oron E, Ivanova N 2012 Cell fate regulation in early mammalian develop - ment. Phys Biol 9:045002. Robert JS 2001 Interpreting the homeobox: metaphors of gene action and activation in development and evolution. Evol Dev 3:287–95.Takaoka K, Hamada H 2012 Cell fate decisions and axis determination in the early mouse embryo. Development 139:3–14. Tickle C 2003 Patterning in Vertebrate Development. Oxford: Oxford Uni - versity Press. Wessells NK 1977 Tissue Interaction and Development. Menlo Park, CA: Benjamin.
281
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Embryonic induction and cell division 192.e1 cHaPtEr 11REFERENCES Bard J 2012 A new ontology (structured hierarchy) of human developmen - tal anatomy for the first 7 weeks (Carnegie stages 1–20). J Anat 221: 406–16. Carlson BM 2014 Human Embryology and Developmental Biology, 5th ed. Philadelphia: Elsevier, Saunders; Ch. 4. Fontaine-Pérus J 2000 Mouse–chick chimera: an experimental system for study of somite development. Curr Top Dev Biol 48:269–300. Le Douarin NM 1969 Particularités du noyau interphasique chez la caille japonaise (Coturnix coturnix japonica). Utilisation de ces particularités comme ‘marquage biologique’ dans les recherches sur les interactions tissulaires et les migrations cellulaires au cours de l’ontogenèse. Bull Biol Fr Belg 103:435–52.Oron E, Ivanova N 2012 Cell fate regulation in early mammalian develop - ment. Phys Biol 9:045002. Robert JS 2001 Interpreting the homeobox: metaphors of gene action and activation in development and evolution. Evol Dev 3:287–95. Takaoka K, Hamada H 2012 Cell fate decisions and axis determination in the early mouse embryo. Development 139:3–14. Tickle C 2003 Patterning in Vertebrate Development. Oxford: Oxford Uni - versity Press. Wessells NK 1977 Tissue Interaction and Development. Menlo Park, CA: Benjamin.
282
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
193 Cell populations at the start of organogenesisCHAPTER 12 contains the transmedian pericardial cavity, which communicates dorsocaudally with right and left pericardioperitoneal canals. These pass dorsally to the transverse septum mesenchyme and open caudally into the extraembryonic coelom on each side of the midgut. The intra­ embryonic mesenchyme has begun to differentiate and the paraxial mesenchyme is being segmented into somites. Neural groove closure is progressing caudally, so that a neural tube is forming between the newly segmenting somites. Rostrally, the early brain regions, which have not yet fused, can be discerned. The neuroepithelium is separated from the dorsal aspect of the gut by the notochord. The earliest blood vessels have appeared and a primitive tubular heart occupies the pericardium. The chorionic circulation is soon to be established, after which the embryo rapidly becomes completely dependent on the maternal blood stream for its requirements. The embryo is connected to the developing placenta by a mesenchymal connecting stalk in which the umbilical vessels develop, and which also contains the allantois, a hindgut diver ­ ticulum. The lateral body walls are still widely separated. The embryo has contact with three different vesicles: the amnion, which is in contact with the surface ectoderm; the yolk sac, which is in contact with the endoderm; and the chorionic cavity, containing the extraembryonic coelom, which is in contact with the intraembryonic coelomic lining (see Fig. 10.10). The early body plan of the embryo is segmented. The boundaries between the segments are maintained by the differential expression of genes and proteins that restrict cell migration in these regions. Organo ­ genetic processes either retain the segmental plan, e.g. spinal nerves, or replace it locally, e.g. the modifications of somatic intersegmental vessels by the development of longitudinal anastomoses. Abnormalities may result from improper specification of segments along the cephalo ­ caudal axis and may fail to produce the appropriately modified segmen ­ tal plan. The degree to which vertebrate embryos are developmentally con ­ strained at this period of development is controversial. Comparative studies on the timing at which specific embryonic structures appear, heterochrony, have shown that other embryonic species do not follow the same developmental sequence as humans (Richardson and Keuck 2002). Although some developmental mechanisms are highly con ­ served, e.g. the homeobox gene codes, others may have been dissociated and modified in different vertebrate species during evolution. Organogenesis, the further development of body regions and organs that is described elsewhere in this book, starts from about stage 10 (approximately 28 days). Although it is both conventional and conven ­ ient to consider the further development of each body system on an individual basis, not only do all systems develop simultaneously, but also they interact and modify each other as they develop. This necessary interdependence is supported by the evidence of experimental embry­ ology and reinforced by the phenomena of growth anomalies, which cut across the artificial boundaries of systems in most instances. For these reasons, it is recommended that the development of an individual system or body region should be studied in relation to others, especially those most closely associated with it, whether spatiotemporally or causally. EMBRYONIC CELL POPULATIONS AT THE START OF ORGANOGENESIS The developmental processes operating in the embryo between stages 5 and 9 enabled the construction of the bi ­ and trilaminar embryonic disc, the intraembryonic coelom and new proliferative epithelia. From the end of stage 10, a range of local epithelial and mesenchymal popula ­ tions now interact to produce viscera and appendages. The inductive influences on these tissues and their repertoire of responses are very different from those seen at the onset of gastrulation. The range of tissues present at the start of organogenesis, when the body plan is clear, SPECIFICATION OF THE BODY AXES AND THE BODY PLAN Embryos may be thought of as being constructed with three orthogonal spatial axes (cephalocaudal, dorsoventral and laterolateral), plus a tem ­ poral axis. In mammalian embryos, axes cannot be specified at very early stages; embryonic axes can be defined only after the early extra­ embryonic structures have been formed and the inner cell mass can be seen. The position of the future epiblast can be predicted in human embryos when the hollow blastocyst has formed. The inner cell mass becomes (seemingly) randomly located on the inside of the trophecto ­ derm and forms a population of epiblast cells subjacent to the tropho­blast. This region implants first. It is not known whether the trophectoderm in contact with the inner cell mass initiates implanta ­ tion, so that the future dorsal surface of the embryo is closest to the disrupted maternal vessels at the implantation site, or whether the inner cell mass can travel around the inside of the trophoblast to gain a posi ­ tion subjacent to the implantation site once implantation has started. Axes may be conferred on the whole embryonic disc, which is ini ­ tially flat and mainly two ­dimensional. However, their subsequent ori ­ entation in the folded three ­dimensional embryo will be completely different. The dorsal structures of the folded embryo form from a cir ­ cumscribed central ellipse of the early flat embryonic disc (see Fig. 10.5). Lateral and ventral structures form from the remainder of the disc, and the peripheral edge of the disc eventually becomes constricted at the umbilicus (see Figs 10.9–10.10). Although the appearance of part of the epiblast is taken to specify the dorsal surface of the embryo, the inner layer, i.e. the hypoblast, is not, by default, a ventral embryonic structure. The primary, cephalocaudal, axis is conferred by the appearance of the primitive streak in the bilaminar disc. The primitive streak patterns cells during ingression, and so also specifies the dorsoventral axis, which becomes apparent after embryonic folding. The position of ingression through the streak confers axial, medial or lateral character­ istics on the forming mesenchyme cells. The axial and medial popula ­ tions remain as dorsal structures in the folded embryo, and the surface ectoderm above them will exhibit dorsal characteristics. The lateral plate mesenchyme will assume lateral and ventral positions after embryonic folding, and the surface ectoderm above this population will gain ventral characteristics. The third and last spatial axis is the bilateral, or laterolateral, axis, which appears as a consequence of the development of the former two axes. Initially, the right and left halves of the embryonic body are bilat ­ erally symmetric. Lateral projections, the upper and lower limbs, develop in two places on each side of the body wall (somatopleure). With the last axis established, the temporal modification of the original embryonic axes can be seen. The segmental arrangement of the cephalocaudal axis is very obvious in the early embryo and is retained in many structures in adult life. Similarly, dorsal embryonic structures remain dorsal and undergo relatively little change. However, structures that were originally midline and ventral, especially those derived from splanchnopleuric mesenchyme, e.g. the cardiovascular system and the gut, are subject to extensive shifts, and change from a bilaterally sym ­ metric arrangement to an entire body that is now chiral, i.e. has distinct left and right sides. The development of all the body organs and systems, organogenesis, begins after the dramatic events of gastrulation, when the embryo has attained a recognizable body plan. In human embryos, this corresponds to the end of stage 10 ( Fig. 12.1). The head and tail folds are well formed, with enclosure of the foregut and hindgut (proenteron and metenteron), although the midgut (mesenteron) is only partly con ­ stricted from the yolk sac. The forebrain projection dominates the cranial end of the embryo, and the buccopharyngeal membrane and cardiac prominence are caudal and ventral to it. The cardiac prominence
283
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Cell populations at the start of organogenesis 194seCtion 2 then passes over the occipital and cervicothoracic parts of the embryo, superficial to the four occipital somites and later to the occipitocervical junction. Further caudally, it is associated with the upper limb field, where it will give rise to the apical ectodermal ridge. O’Rahilly and Müller (1985) have called the portion of the ring between the upper and lower limbs the intermembral part. It overlies the underlying intraembryonic coelom, and later (between stages 12 and 13), the mesonephric duct and ridge. In stages 14 and 15, this portion of the ectodermal ring gives rise to the mammary line. Caudal to the lower limb field, in the unfolded embryo, the ring passes distal to the cloacal membrane. In the folded embryo, this region becomes superior to the cloacal membrane and corresponds to the ectoderm associated with the external genitalia, particularly the genital tubercle and urogenital swell­ ings, where it influences the internal and external descent of the testis and the formation of the inguinal canal. Neural ectoderm The neuroepithelium at the time of primary neurulation is pseudos­ tratified. It has a midline hinge region, which, with concomitant wedging of the cells in the lateral wall of the neural groove, promotes neural tube formation. The processes of the neuroepithelium abut on to internal and external limiting membranes. This epithelium prolifer ­ ates to form all the cell lines of the central nervous system and, via the production of neural crest, all the cell lines of the peripheral nervous system. Notochord During stage 10, the notochordal plate undergoes a process that is similar to, but a mirror image of, neurulation, and forms an epithelial tube from caudal to rostral, ending with the pharynx. The notochordal plate forms a deep groove, the vertical edges of the groove move medi ­ ally and touch, and then the endodermal epithelium from each side fuses ventral to the notochord. The cells swell and develop an internal pressure (turgor) that confers rigidity on the notochord. The notochord is surrounded by a basal lamina, which is initially referred to as a peri ­ notochordal sheath; this term is subsequently applied to mesenchymal populations that surround the notochord. After stage 1 1, the tubular notochord is in contact with the neural tube dorsally and the endoderm ventrally. It is not a proliferative epithelium, but it has inductive effects on the overlying neural tube and the adjacent somites, and later pro ­ vides a focus for sclerotomal migration.is given below and shown in Figures 12.1 and 12.2. For a summary of the fates of the embryonic cell populations, see Figure 12.3 . Epithelial populations in the embryo Surface ectoderm During embryogenesis, the surface ectoderm shows regional differences in thickness. Ectoderm over the dorsal region of the head and trunk is thin, as is the pericardial covering; this has been interpreted as a con ­ sequence of the expansion of this epithelium over structures that are enlarging rapidly as development proceeds. After the surface ectoderm has completed a number of early interactions it forms the periderm, which remains throughout fetal life and differentiates into epidermis. Ectodermal ring and ectodermal placodes The ectoderm on the head and lateral borders of the embryo shows a zone of epithelial thickening, the ectodermal ring, which can be dis ­ cerned from stage 10 and is completed by stage 12. Rostrally, it contains populations of neuroectoderm that remain in the ectoderm after primary neurulation and are termed ectodermal placodes; these pla ­ codes may be considered to be neuroepithelial cells that remain within the surface ectoderm until central nervous system development has progressed sufficiently for their inclusion into sensory epithelia and cranial nerve ganglia. The neuronal placodes may invaginate in toto to form a vesicle, or remain as a neuronal layer, or contribute individually to neuronal structures with cells of other origins. The midline ectoder ­ mal thickening, the adenohypophysial placode, invaginates as Rathke’s pouch and forms a vesicle immediately rostral to the buccopharyngeal membrane. The ectodermal ring then passes bilaterally to encompass the olfactory and optic placodes, which give rise to the olfactory sensory epithelium and the lens of the eye, respectively. It then overlays the pharyngeal arches, where it gives rise to epibranchial placodes; these remove themselves individually from the ectoderm at stage 10–1 1 and become associated with the neural crest cells within the cranial sensory ganglia supplying the arches. It also forms specializations of the ecto ­ derm on the frontonasal, maxillary and mandibular processes, which give rise to the tooth buds and the outer coating of the teeth. The paired otic placodes overlying the rhombencephalon at the lateral portion of the second pharyngeal arch invaginate to form the otic vesicles, which give rise to the membranous labyrinth of the ear. The ectodermal ring Fig. 12.1 A, The embryo at stage 11, showing the position of the intraembryonic coelom (contained by the walls coloured blue). B, The three major epithelial populations within a stage 11 embryo, viewed from a ventrolateral position. The neural tube lies dorsal to the gut. Ventrally, the intraembryonic coelom crosses the midline at the level of the foregut and hindgut, but is lateral to the midgut and a portion of the foregut. Pericardial cavityPericardial cavity Amnion Yolk sac wall Entrance to pericardioperitoneal canal Yolk sac wall Peritoneal cavityPharynx Neural tube Midgut HindgutPericardioperitoneal canal ForegutNeural tube Somite Left umbilical vein Umbilical cordB A
284
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
embryonic cell populations at the start of organogenesis 195 Chapter 12 Somites Somites are discrete epithelial spheres formed by the transformation of paraxial mesoblast cells to epithelium once their migration is complete. Specific regions of the somitic epithelium form local proliferative centres and, like neural and coelomic epithelium, it produces cells destined for different fates, at different developmental times. Somites can be seen on each side of the fusing neural tube in the human embryo from stage 9. Development proceeds in a craniocaudal direction. The original epithelial somites form the base of the skull, and the vertebral column and ribs; the dorsolateral portion, the dermomyotome, gives rise to the majority of the skeletal muscle of the body, including that in the limbs. The relative dispositions of the neural, endodermal and coelomic epithelia are shown in Figure 12.1. Mesenchymal populations in the embryo In the stage 10 embryo, the major mesenchymal populations are in place. Mesoblast is still being generated at the primitive streak and moving into the presomitic mesenchymal population adjacent to the notochord. Some mesoblast is also contributing to the lateral regions Endoderm The craniocaudal progression of development means that the endo ­ derm of the early stomodeum develops ahead of other portions of the endodermal epithelium. The development of the pharyngeal arches and pouches (Ch. 36) is closely associated with the development of the neural ectoderm and proliferation of the neural crest. The respiratory diverticulum arises slightly later, when the postpharyngeal gut may also be distinguished. The endoderm gives rise to the epithelial lining of the respiratory and gastrointestinal tracts, the biliary system, and the bladder and urethra. Coelomic epithelium The coelomic epithelium lines the intraembryonic coelom, which is subdivided into a midline pericardial cavity, two bilateral pericardio­ peritoneal canals, and the initially bilateral peritoneal cavities; the latter are continuous with the extraembryonic coelom. The coelomic epithelium is a germinal epithelium. It produces the myocardium and connective tissue populations for the viscera, and also gives rise to the supporting cells for the germ cells, the epithelial lining of the uro ­ genital tracts, and the mesothelial lining of the pericardial, pleural and peritoneal cavities.Fig. 12.2 A, Mesoblast populations within the early embryonic disc. B, A stage 11 embryo, showing the position of the intraembryonic coelom (contained by the walls coloured blue) and the positions of the sections (i)–(iv) shown in C. C, Transverse sections, arranged cranial to caudal, from a stage 11 embryo; the populations of mesenchyme and the sites of mesenchymal proliferation are indicated. Unsegmented paraxial mesenchymeA C (i)(i) (ii)(ii) (iii) (iii) (iv) (iv)Lateral plate mesenchymeBuccopharyngeal membrane Cloacal membrane Pericardial cavityNeural tube Amnion Yolk sac wall Entrance to pericardio- peritoneal canalPrechordal mesenchyme Notochord Neural crest mesenchyme Aorta Endocardium Developing myocardium Notochordal plate Aorta Gut Pericardioperitoneal canal Umbilical cordSomite MidgutLeft umbilical veinPharynx Pericardio- peritoneal canalUnsegmented paraxial mesenchyme Proliferative coelomic epithelium Pericardial cavity Neural crest mesenchyme Epithelial somite Aorta Proliferative splanchnopleuric epitheliumNeural crest mesenchyme Notochordal plate Pericardioperitoneal canal Gut Intraembryonic coelom (peritoneal cavity)GutProliferative somatopleuriccoelomic epitheliumNotochordal plate AortaParaxial mesenchyme Proliferative somatopleuric coelomic epithelium Endothelium of sinus venosus Septum transversum mesenchymeGut Neural crest mesenchyme Epithelial somite Region of intermediate mesenchyme Proliferative splanchnopleuric coelomic epitheliumB
285
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Cell populations at the start of organogenesis 196seCtion 2 Fig. 12.3 Structures that will be derived from specific epithelial and mesenchymal populations in the early embryo. Abbreviation: CNS, central nervous system. Endoderm epithelium Primitive gut Foregut – recesses, diverticula and glands of the pharynx. General mucous glandular and duct-lining cells and the main follicular cells of the thyroid. Epithelium of pharyngeal pouches (tonsil, middle ear cavity, thymus, parathyroids 3 and 4, C cells of thyroid), adenoids, epithelial lining of the auditorytube, tympanic cavity, tympanic antrum, internallamina of the tympanic membrane. Respiratory tract – epithelial lining, secretory and duct-lining cells of the trachea, bronchi, bronchiolesand alveolar sacs. Epithelial lining, secretory and duct-lining cells of the oesophagus, stomach and duodenum. Hepatocytes of liver, biliary tract, exocrine and endocrine cells of the pancreas. Midgut – epithelial lining, glandular and duct-lining cells of the duodenum, jejunum, appendix, caecum,part of transverse colon. Hindgut – epithelial lining, glandular and duct-lining cells of part of the transverse,descending and sigmoid colon, rectum, upper partof anal canal. Allantois – urinary bladder, vagina, urethra, secretory cells of the prostate and urethral glands.Coelomic wall epithelium Mesenchyme Surface ectoderm epithelium Neural plate epithelium Neural crestWalls of intraembryonic coelom Primitive pericardium – myocardium, parietal pericardium. Pericardioperitoneal canals – visceral, parietal and mediastinal pleura, pleuroperitonealmembranes contributing to diaphragm. Splanchnopleuric epithelium – visceral peritoneum of stomach, peritoneum of lesser and greateromenta, falciform ligament, lienorenal andgastrosplenic ligaments. Somatopleuric epithelium – parietal peritoneum.Primitive peritoneal cavity Splanchnopleuric epithelium – visceral peritonealcovering of mid- and hindgut, the mesentery,transverse and sigmoid mesocolon. Pronephros, epithelial lining of mesonephric ducts, vas deferens, epididymis, seminal vesicles,ejaculatory duct, ureters, vesical trigone. Müllerian ducts, epithelial lining of uterine tubes, body and cervix of uterus, vagina, broad ligament ofuterus. Germinal epithelium of gonad (note the germ cells are not included on this chart because of their earlysequestration into the extraembryonic tissues). Germinal epithelium forming cortex of suprarenal gland. Somatopleuric epithelium – parietal peritoneum, tunica vaginalis of testis.Paraxial mesenchyme(somites and somitomeres)Sclerotome – vertebrae and portions of theneurocranium, axial skeleton.Myotome – all voluntary muscles of the head, trunkand limbs.Dermatome – dermis of skin over dorsal regions. Intermediate mesenchyme – connective tissue of gonads, mesonephric and metanephric nephrons,smooth muscle and connective tissues of thereproductive tracts.Septum transversum – epicardium, fibrouspericardium, portion of diaphragm, oesophagealmesentery, sinusoids of liver, tissue within lesseromentum and falciform ligament. Lateral plate mesenchyme Splanchnopleuric layer – smooth muscle andconnective tissues of respiratory tract andassociated glands. Smooth muscle and connective tissues of intestinal tract, associated glands and abdominal mesenteries. Smooth muscle and connective tissue of blood vessels (also see below). Somatopleuric layer – appendicular skeleton, connective tissue of limbs and trunk, includingcartilage, ligaments and tendons. Dermis of ventral body wall and limbs. Mesenchyme of external genitalia. Angiogenic mesenchyme Endocardium of heart, endothelium of blood andlymphatic vessels, vessels of choroid plexus,sinusoids of liver and spleen, circulating blood cells,microglia, tissue macrophages. Ectodermal placodesAdenohypophysis. Sensory neurones of the cranial ganglia V, VII, VIII, IX, X. Olfactory receptor cells and olfactory epithelium.Epithelial walls of the membranous labyrinth, the cochlear organ of Corti. Lens of the eye. Enamel organs of the teeth.Cranial structures Secretory and duct-lining cells of the lacrimal,nasal, labial, palatine, oral and salivary glands. Epithelia of the cornea and conjunctiva.Epithelial lining of the external acoustic meatus and external epithelium of the tympanic membrane. Epithelial lining of the lacrimal canaliculi and nasolacrimal duct. Epithelial lining of the paranasal sinuses, lips, cheeks, gums and palate. Epidermal structures Most of the cutaneous epidermal cells, thesecretory, duct-lining and myoepithelial cells of thesweat, sebaceous and mammary glands. Hair and nails. Proctodeal epithelium and epithelium of the terminal male urethra.CNS – Brain and spinal cordNeurohypophysis. Prosencephalon (telencephalon and diencephalon) – cerebral hemispheres, basal nuclei. Mesencephalon – cerebral peduncles, tectum, tegmentum. Rhombencephalon (metencephalon and myelencephalon) – cerebellum, pons, medullaoblongata. Spinal cord.All cranial and spinal motor nerves.All CNS neurones, including preganglionic efferent neurones, with somata within the CNS. Astrocytes and oligodendrocytes.Ependyma lining the cerebral ventricles, aqueduct and central canal of brain and spinal cord,tanycytes, cells covering the choroid plexuses,circumventricular cells. Retina and optic nerve (II), epithelium of the iris, ciliary body and processes.Neural derivativesSensory neurones of the cranial ganglia V, VII, VIII, IX, X. Sensory neurones of the spinal dorsal root ganglia and their peripheral sensory receptors. Satellite cells in all sensory ganglia.Sympathetic ganglia and plexuses: neurones and satellite cells. Parasympathetic ganglia and plexuses: neurones and satellite cells. Enteric plexuses: neurones and glial cells.Schwann cells of all the peripheral nerves.Medulla of the suprarenal gland. Chromaffin cells. Carotid body type I cells (and type II, satellite type cells).Calcitonin-producing cells (C cells). Melanocytes. Mesenchymal derivatives in the head Frontal, parietal, squamous temporal, nasal, vomer,palatine bones, maxillae and mandible. Meninges. Choroid and sclera of eye. Connective tissue of lacrimal, nasal, labial, palatine, oral and salivary glands. Dentine of teeth. Connective tissues of head, including cartilage, ligaments and tendons. Connective tissues of thyroid gland and of the pharyngeal pouches, i.e. parathyroid glands, thymus. Tunica media of the outflow tract of the heart and the great vessels.
286
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
embryonic cell populations at the start of organogenesis 197 Chapter 12the primitive streak, none of the cells that arise from the neural crest become arranged as epithelia. The development and fate of head and trunk neural crest cells are very different and therefore they will be considered separately. head neural crest Head neural crest migrates before the neural tube closes. Two popula ­ tions of crest cells develop. Some retain a neuronal lineage and con ­ tribute to the somatic sensory and parasympathetic ganglia in the head and neck. Others produce extensive mesenchymal populations; the crest cell population arising from the head is larger than that found at any trunk level. Each brain region has its own crest population that migrates dorsolaterally around the sides of the neural tube to reach the ventral side of the head. Crest cells surround the prosencephalic and optic vesicles, and occupy each of the pharyngeal arches (Ch. 36). They provide mesenchyme cells that will produce the connective tissue in parts of the neuro ­ and viscerocrania. All cartilage, bone, ligament, tendon, dermal components and glandular stroma in the head are derived from the head neural crest. Head neural crest also contributes to the tunica media of the aortic arch arteries. trunk neural crest Unlike its counterpart in the head, trunk neural crest is formed as the neural tube closes craniocaudally, which means that various stages of crest development can be found in the more caudal regions of an embryo. As the neural tube begins to fuse dorsally in the midline, the neural crest cells lose their epithelial characteristics and junctional con ­ nections, and form a band of loosely arranged mesenchyme cells imme ­ diately dorsal to the neural tube and beneath the ectoderm. Initially, most of the crest cells lie with their long axes perpendicular to the long axis of the neural tube. Later, the cell population expands laterally and around the neural tube as a sheet. Trunk neural crest cells move from their position dorsal to the neural tube via three routes (see Fig. 17.1 1): dorsolaterally, to form dorsal root ganglia throughout the trunk; ven ­ trally, to form sympathetic ganglia, enteric nerves and the suprarenal medulla; and rostrocaudally, along the aorta, to form the pre ­aortic ganglia. In a second migration route, crest cells pass dorsolaterally between the ectoderm and the epithelial plate of the somite into the somatopleure, where they eventually form melanocytes in the skin. Lateral plate Lateral plate is the term for the early unsegmented mesoblast popula­ tion lateral to the paraxial mesenchyme. Mesoblastic cells, which arise from the middle of the primitive streak (primary mesenchyme), migrate cranially, laterally and caudally to reach their destinations, where they revert to epithelium and form a continuous layer that adheres to the ectoderm dorsally and the endoderm ventrally. The epithelium faces a new intraembryonic cavity, the intraembryonic coelom, which becomes confluent with the extraembryonic coelom and provides a route for the circulation of coelomic fluid through the embryo. Once formed, the intraembryonic coelomic wall becomes a proliferative epithelium that produces new populations of mesenchymal cells. The mesenchymal population subjacent to the ectoderm is termed somatopleuric mesen ­ chyme and is produced by the somatopleuric coelomic epithelium. The mesenchymal population surrounding the endoderm is termed splanch ­ nopleuric mesenchyme and is produced by the splanchnopleuric coel­omic epithelium (see Fig. 12.2). It is important to note that these terms are relevant only caudal to the third pharyngeal arch. Rostral to this, there is a sparse mesenchymal population between the pharynx and the surface ectoderm prior to migration of the head neural crest, and there are no landmarks with which to demarcate lateral from paraxial mesenchyme. This unsplit lateral plate is believed to contribute to the cricoid and arytenoid car ­ tilages, the tracheal rings and the associated connective tissue. somatopleuric mesenchyme Somatopleuric mesenchyme produces a mixed population of connec ­ tive tissues and has a significant organizing effect at the level of the developing limbs. The pattern of limb development is controlled by information contained in the somatopleuric mesenchyme. Regions of the limb are specified by interaction between the surface ectoderm (apical ectodermal ridge) and underlying somatopleuric mesenchyme; together, these tissues form the progress zone of the limb. The somato­pleuric mesenchyme in the limb bud also specifies the postaxial border of the developing limb. Somatopleuric mesenchyme gives rise to the connective tissue elements of the appendicular skeleton, including the pectoral and pelvic girdles and the bones and cartilage of the limbs, and their associated ligaments and tendons. It also gives rise to the dermis of the skin of the ventral and lateral body walls and of the limbs.of the embryo. The different mesenchymal populations within the embryo from stage 10 onwards are described below. The relative dispositions of the early mesenchymal populations are shown in Figure 12.2. Axial mesenchyme The first epiblast cells to ingress through the primitive streak form the endoderm and notochord, and initially occupy a midline position. The earliest population of endodermal cells rostral to the notochordal plate is termed the prechordal plate. The notochordal cells remain medially and the endodermal cells subsequently flatten and spread laterally. The population of cells that remain mesenchymal in contact with the floor of the neural groove, just rostral to the notochordal plate, is termed prechordal mesenchyme (Fig. 12.4). These axial mesenchyme cells are tightly packed, unlike the more lateral paraxial cells but, unlike the notochord, they are not contained in an extracellular sheath. They are displaced laterally at the time of head flexion and form bilateral pre ­ mandibular mesenchymal condensations. They become associated with the local paraxial mesenchyme. Orthotopic grafting has demonstrated that these cells leave the edges of the prechordal mesenchyme and migrate laterally into the periocular mesenchyme, where they give rise to all of the extrinsic ocular muscles. Paraxial mesenchyme Paraxial mesoblast is a transient structure that forms somites at its cranial end, whilst unsegmented mesoblast is added caudally by the primitive streak. Epiblast cells that migrate through the primitive node and rostral primitive streak during gastrulation form mesoblast cells that migrate to a position lateral to the notochord and beneath the developing neural plate. Cells that ingress through the primitive node form the medial part of this paraxial mesoblast, and cells that ingress through the rostral streak form the lateral part (see Fig. 10.3). The paraxial mesoblast extends cranially from the primitive streak to the prechordal plate, which is immediately rostral to the notochord. Before somite formation, this mesenchymal tissue is also termed presomitic or unsegmented mesenchyme in mammals (analogous to the segmen ­ tal plate in birds). Paraxial mesoblast rostral to the otic vesicle was previously believed not to segment. Caudal to the otic vesicle, the paraxial mesoblast on each side of the rhombencephalon segments into somites as the neural folds elevate and neurulation begins; somites are therefore post ­otic. During somito­ genesis, mesoblastic cells show changes in shape and in cell–cell adhe ­ sion, and become organized into epithelial somites; this process begins at the eighth somitomere, just caudal to the midpoint of the noto ­ chordal plate. Somite one is also termed the first occipital somite. The post otic paraxial cell population is termed paraxial mesoderm in con ­ temporary literature. Skeletal muscle throughout the body is derived from paraxial mesoblast ­derived somatic epithelium, which proliferates to form myoblastic populations that migrate to the head, body and limbs. Septum transversum Early mesenchyme that invaginates through the middle part of the primitive streak comes to lie rostral to the buccopharyngeal membrane, where the cells form the epithelial wall of the pericardial coelom. As this epithelium proliferates, the visceral pericardial wall gives rise to myocardium. The parietal pericardial wall forms mesenchyme, initially termed precardiac, or cardiac, mesenchyme, which is able to induce proliferation of hepatic endodermal epithelium. With further prolifera ­ tion, the precardiac mesenchyme forms a ventral mass caudal to the heart, the septum transversum, separating the foregut endoderm from the pericardial coelom (see Fig. 12.2). By stage 1 1, the septum transver ­ sum extends dorsally on each side of the developing gut and becomes continuous with the mesenchymal populations that proliferate from the walls of the pericardioperitoneal canals. Cells of the septum trans ­ versum give rise to the sinusoids of the liver, the central portions of the diaphragm and the epicardium. Neural crest The neural crest is unique: it gives rise to neural populations in the head and trunk, and also provides an extensive mesenchymal popula ­ tion in the head with attributes similar, in terms of patterning, to somatopleuric mesenchyme. Neural crest cells arise from cells that lie initially at the outermost edges of the neural plate, between the pre ­ sumptive epidermis and the neural tube, and are committed to a neural crest lineage before the neural plate begins to fold. After neurulation, neural crest cells form a transient axial population and then disperse, in some cases migrating over considerable distances, to a variety of dif ­ ferent developmental fates. Unlike mesoblast, which is produced from
287
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Cell populations at the start of organogenesis 198seCtion 2 Fig. 12.4 The organization of the head and pharynx in an embryo at about stage 14. The individual tissue components have been separated but are aligned in register through the numbered zones. PATTERNS OF GENE EXPRESSIONHOX-b5 HOX-b4HOX-b3HOX-b2 HOX-a1 CRABP RAR-β Wnt-1Engrailed-2 HOX-b1 Krox 20 Wnt-2 AXIAL STRUCTURES MIGRATION OF NEURAL CREST MESENCHYME PARAXIAL MESENCHYME BRAIN AND CRANIAL MOTOR NERVES CRANIAL SENSORY AND PARASYMPATHETIC GANGLIA SUGGESTED ORIGIN OF STRIATED MUSCLES FROM PARAXIAL MESENCHYME PHARYNGEAL ARCH ARTERIES ENDODERMAL PHARYNGEAL POUCHES4th 3rd 2nd 1st 4th 3rd 2nd 1st1 2 34 5 12 3 4 5 6 7 12 3 4 5 6 7 1234 5 1 234 1 2 3 41 2 3 456 7XII X IX VII VVI IIIIVPharyngeal arches:Rhombomeres 765432 1 SomitesSomitomeres Proximal VII (root) OticNeural crest shown in green and ectodermal placodes in grey Neural crest shown in green around great vessels Neural crest shown in green around pharyngeal pouchesRhombencephalon Otic vesicle Otic vesicle Proximal X (jugular) Proximal IX (superior) VIII (vestibulocochlear) Distal X (nodose) Distal IX (petrosal) 3rd arch Laryngeal Glossal2nd arch (hyoid) 1st arch (mandibular) Descending aorta Ductus arteriosus Pulmonary artery Ascending aorta Inferior parathyroid Superior parathyroid Oesophagus Lung budUltimobranchial bodySpinal cord Notochord Optic cup MandibularMaxillary Frontonasal Ciliary Prechordal mesenchyme migrates to form extraocular musclesV (trigeminal) PterygopalatineDistal VII (geniculate)Submandibular Lateral rectus Inferior rectus EyeSuperior rectus Superior oblique Medial rectus Internal carotid artery Tubotympanic recessPharyngeal arch arteries Pharyngeal tonsilThyroid ThymusInferior obliqueProsencephalonPrechordal mesenchymeMesencephalon
288
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
199 Chapter 12Key references endodermal tissues are necessary for endothelial differentiation, par ­ ticularly the early foregut. Angioblastic mesenchyme forms early in the third week of development from extraembryonic mesenchyme in the splanchno pleure of the yolk sac, in the body stalk (containing the allantois), and in the somatopleure of the chorion. The peripheral cells flatten as a vascular endothelium, whereas the central cells transform into primitive red blood corpuscles. Later, contiguous islands merge, forming a continuous network of fine vessels. Intraembryonic blood vessels are first seen at the endoderm–mesenchyme interface within the lateral splanchnic mesenchyme at the caudolateral margins of the cranial intestinal portal. Angioblastic competence has been demon­ strated within the ventral (splanchnopleuric) mesenchymes with which the endoderm interacts. However, the notochord and prechordal plate do not contain angiogenic cells. Similarly, ectodermal tissues do not appear to give rise to angiogenic cells. Somites, derived from paraxial mesenchyme, have been shown to be a source of angioblasts that either differentiate with the somite derivatives, or migrate to the neural tube, ventrolateral body wall, limb buds, mesonephros and the dorsal part of the aorta. The earliest angiogenic mesenchymal cells form blood vessels by vasculogenesis, a process in which new vessels (e.g. endothelial heart tubes, dorsal aortae, umbilical and early vitelline vessels) develop in situ (Ch. 13). Later vessels develop by angiogenesis, sprouting and branch ­ ing from the endothelium of pre ­existing vessels; this process is the means by which most other vessels develop. The ultimate pattern of vessels is controlled by the surrounding, non ­angiogenic mesenchyme; vessels become morphologically specific for the organ in which they develop, and also immunologically specific, expressing organ ­specific proteins.splanchnopleuric mesenchyme Splanchnopleuric mesenchyme surrounds the developing gut and res ­ piratory tubes, contributing connective tissue cells to the lamina propria and submucosa, and smooth muscle cells to the muscularis mucosae and muscularis externa. It plays a patterning role in endodermal devel ­ opment, specifying the region and villus type in the gut, and the branch ­ ing pattern in the respiratory tract. Intermediate mesenchyme Intermediate mesenchyme is a loose collection of cells found between the somites and the lateral plate (see Fig. 12.2). Its development is closely related to the progress of differentiation of the somites and the proliferating coelomic epithelium from which it is derived. Intermedi ­ ate mesenchyme is not present before somitogenesis or the formation of the eighth somite. In embryos with 8–10 somites, it is present lateral to the sixth somite but does not extend cranially. The mesenchyme cells are arranged as layers, one continuous with the dorsal side of the par­ axial mesenchyme and the somatopleure, the other with the ventral side of the paraxial mesenchyme and the splanchnopleure. As development proceeds, the intermediate mesenchyme forms a loosely packed dorsolateral cord of cells, which lengthens at the caudal end and ultimately joins the cloaca. It gives rise to the nephric system, gonads and reproductive ducts. Angioblastic mesenchyme Mesenchymal cells, which give rise to the cellular elements of the blood, the endothelium and the mesenchymal layers of the tunica externa and adventitia of blood and lymphatic vessels, arise from extraembryonic and intraembryonic sources. Evidence suggests that KEY REFERENCES O’Rahilly R, Müller F 1985 The origin of the ectodermal ring in staged human embryos of the first 5 weeks. Acta Anat 122:145–57.Richardson MK, Keuck G 2002 Haeckel’s ABC of evolution and develop ­ ment. Biol Rev Camb Philos Soc 77:495–528.
289
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Cell populations at the start of organogenesis 199.e1 Chapter 12REFERENCES O’Rahilly R, Müller F 1985 The origin of the ectodermal ring in staged human embryos of the first 5 weeks. Acta Anat 122:145–57.Richardson MK, Keuck G 2002 Haeckel’s ABC of evolution and develop ­ ment. Biol Rev Camb Philos Soc 77:495–528.
290
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
200CHAPTER 13 Early embryonic circulation guidance, where the distal tip cell extends supported by proximal stalk cells that may divide; sprout elongation in response to cues from the extracellular matrix; lumen formation, which is initially by the conflu - ence of intracellular vacuoles, followed by establishment of apical basal polarity of the cells of the endothelial sprout; and fusion of the tip cell to another sprout tip with confluence of the two lumina established (Chappell et al 2012, Kuijper et al 2007). The developmental processes of angiogenesis are similar to those seen in both neoplasia and acute inflammation in adult tissue. The scientific literature on angiogenesis is extensive and the process will not be considered in detail here; the interested reader is directed to Senger and Davis (201 1), Beets et al (2013), Herbert and Stainier (201 1) and Eichmann and Pardanaud (2014). Studies have considered the development of capillary beds within the surrounding mesenchymal tissues destined to become laminae pro - priae, but few have looked at how capillary beds form within tissues undergoing morphogenetic movements. Czirok et al (201 1) discussed the temporal imperative for free-living embryos to develop rapidly in contrast to amniote embryos, which do not have to fend for themselves. They concluded that amniote vascular patterns of development, prior to the establishment of a circulation, are not predetermined by a network of patterning genes and cell signalling pathways, but rather that the tissue movements during organogenesis and the realignment and tension that develop within extracellular matrix fibres may guide cell migration and the position of early vessels. Extracellular matrix plays a critical role in angiogenesis and vessel stabilization in the embryo and in the adult; differences in the composi - tion and abundance of specific extracellular matrix components between The early embryonic circulation is symmetrical ( Fig. 13.1). It is modi - fied throughout development to produce a functioning fetal circulation that is connected to the placenta, and changes rapidly at birth to accom - modate disconnection from the placenta and the start of gaseous exchange in the lungs. Major restructuring of early vessels occurs as the embryo grows; anastomoses form and then disappear, capillaries fuse and give rise to arteries or veins, and the direction of blood flow may reverse several times before the final arrangement of vessels is completed. ANGIOGENESIS The earliest circulatory components develop by vasculogenesis in the extraembryonic tissues. The endothelial heart tubes, dorsal aortae, umbilical and early vitelline vessels arise by vasculogenesis within the embryo. Further vessel development occurs by a process of angiogenesis in which angioblasts, arising in splanchnic and somitic tissues, add endothelial sprouts and branches to earlier vessels. None of the main vessels of the adult arises as a single trunk in the embryo. A capillary network is first laid down along the course of each vessel; the larger arteries and veins are defined by selection and enlargement of definite paths in this network. Lymphatic vessels develop after the main arteries and veins are formed; they arise initially by angiogenesis from the car - dinal veins and subsequently by proliferation of lymphangioblasts to form lymphatic capillaries. Five stages have been described during angiogenesis: endothelial tip cell specification and sprout initiation; sprout elongation and local Fig. 13.1 The early, symmetrical blood vascular system. A, Ventrolateral view of the endothelial profile of the heart, the first aortic arch arteries and the dorsal aorta shown in relation to the major epithelial populations. B, Ventrolateral view of the main venous channels shown in relation to the major epithelial populations. C, Left lateral view of the blood vascular system of a stage 11 human embryo; only the endothelial lining of the heart tube is shown. At this stage, the arteries and veins are in the process of development and no true circulation is yet established. Vitelline plexus Yolk stalk (cut) Umbilical arteriesOtic placodeRudiment of second aortic archFirst aortic arch Primitive head vein Right atrium Rudiment of common cardinal vein Rudiments of postcardinal vein Dorsal aortaVentricleAortic sac A B C Left umbilical arteryDorsal aortaVentricleAtriumLeft first aortic arch Left precardinal vein Left postcardinal veinLeft common cardinal vein Sinus venosus Septum transversumPericardial cavity Left umbilical vein
291
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Angiogenesis 201 CHAPTER 13 As the heart muscle thickens, compacts and strengthens, the cardiac orifices become both relatively and absolutely reduced in size, the valves increase their efficiency, and the large arteries acquire their muscular walls and undergo a relative reduction in size. From this time onwards, the embryo is dependent for its nourishment on the expanding capillary beds, and the function of the larger arteries becomes restricted to that of controllable distribution channels to keep the embryonic tissues constantly and appropriately supplied. The heart starts to beat early, before the development of the conduc - tion system, and a circulation is established before a competent valvular mechanism has formed. Cardiac output increases in proportion with the weight of the embryo and cardiac rate increases with development. However, most of the increase in cardiac output results from a geomet - ric increase in stroke volume. When dorsal aortic blood flow is matched to embryonic weight, blood flow remains constant over a more than 150-fold change in mass of the embryo. After head folding, the embryo has bilateral primitive aortae, each consisting of ventral and dorsal parts that are continuous through the first embryonic aortic arches (Ch. 36). The ventral aortae are fused and form a dilated aortic sac. The dorsal aortae run caudally, one on each side of the notochord. In the fourth week, they fuse from about the level of the fourth thoracic to that of the fourth lumbar segment to form a single definitive descending aorta (see Fig. 13.1A,C; Fig. 13.2B). In general, more mature endothelial channels are seen in the rostral, more advanced regions of the embryo whereas, more caudally, a changing capillary plexus constantly remodels until it becomes conflu - ent with the vascular channels of the connecting stalk. The dorsal continuation of the primitive dorsal aortae directs blood into an anas - tomosing network around the allantois, which will form the umbilical arteries. Blood is channelled back to the developing heart from the allantois via umbilical veins, from anastomoses in the primitive yolk sac via the vitelline veins, and from the body via pre- and postcardinal veins that join to form the common cardinal veins (see Figs 13.1B,C, 13.2A). Vascular anomalies Localized defects of vascular development, characterized by focal increases in the number of vessels that are abnor - mally tortuous and enlarged, have been shown to have genetic causes and mapped to the predisposing gene and chromosomal locus (Boon et al 201 1). It has been suggested that a range of capillary malforma - tions, venous malformations, telangiectasia and lymphatic malforma - tions, including lymphoedema, is associated with factors identified in the early formation of blood and lymphatic vessels (Brouillard and Vikkula 2007, Boon et al 201 1).the embryo and adult mean that there are likely to be significant dif - ferences in the process at different ages. Fibronectin is required for normal vascular development; its deposition is related to interactions between endothelial cells and pericytes, and local tensional forces that occur between blood vessel basement membrane and local extracellular matrix assembly (Davis and Senger 2005; Senger and Davis 201 1). Early blood vessels are initially surrounded by a fibronectin-rich matrix that is later incorporated into the endothelial basal lamina along with laminin, a particularly early constituent. Several layers of fibronectin- expressing cells are seen around larger vessels, such as the dorsal aortae. The endothelium does not synthesize a basal lamina in those regions where remodelling and angiogenesis are active, and the mesenchyme around such endothelium does not express α-actin or laminin until branching has stopped and differentiation of the tunica media begins (after a stable vascular pattern has formed). It is not known how dif - ferentiation of pericytes and smooth muscle cells is induced; the major - ity of arteries accumulate medial smooth muscle from the surrounding mesenchyme. Although studies have elucidated the genes involved in vasculogen - esis and angiogenesis, the drivers that lead to development of asym - metric vessels from an early symmetrical pattern and the specification of the position of particular arteries, veins and lymphatic vessels are still not clear. In early development, the arteries of the embryo are dispro - portionately large and their walls consist of little more than a single layer of endothelium. The cardiac orifices are also relatively large and the force of the cardiac contraction is weak; consequently, the circula - tion is sluggish, despite the rapid rate of contraction of the developing heart. However, the tissues are able to draw nourishment not only from the capillaries but also from the large arteries and the intraembryonic coelomic fluid. It has been suggested that the rapidly expanding cardiovascular system is filled with plasma by the movement of fluid from the intra - embryonic coelom to the veins. In general, the wall of the intraembry - onic coelom is composed of proliferating cells that produce the splanchnopleuric and somatopleuric mesenchymal populations. How - ever, the walls of a portion of the pericardioperitoneal canals are thin - ner, and possibly more permeable to coelomic fluid, at the time when the canals surround the hepatocardiac channels (veins). The latter lie between the hepatic plexus and the sinus venosus; the hepatocar - diac channel on the right side is more developed and on the left it is more plexiform, with only a transitory connection to the sinus veno - sus. The differentiation of this specific coelomic region occurs just in advance of the expansion and filling of the right and left atria, at about stage 12.Fig. 13.2 Profile reconstructions of the blood vascular system of a stage 13 human embryo. The early circulation is now asymmetrical; the venous vessels are enlarging on the right and diminishing on the left. A, Seen from the right side, showing the main venous channels to the heart. B, Seen from the left side, showing the aortic arch arteries, the main vessels arising from the dorsal aorta and umbilical arteries. Note that only the endothelial lining of the heart chambers is shown and, because the muscular wall has been omitted, the pericardial cavity appears much larger than the contained heart. Otocyst Second aortic arch Left umbilical arteryThird aortic arch Internal carotid artery Truncus arteriosus Right atrium Left atrium Left ventricle Right umbilical artery Caudal plexusDeveloping liver sinusoids in septum transversumAorta Lung bud Primary plexus to intestineLeft common cardinal vein Left umbilical vein Left vitelline vein Vitelline arteryPrimary head vein Internal carotid artery Optic vesicle Truncus arteriosus Right ventricle Left ventricle Right sinus hornRight vitelline veinRight atrium Precardinal veinThird aortic archSecond aortic arch Somite IV Right common cardinal vein Aorta Hepatocardiac channel Upper limb bud Postcardinal vein Right umbilical vein 0.2 cm 0.2 cmA B Hindgut
292
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
EARly EmbRyoni C CiRCulATion 202SECT ion 2renal artery arises from the most caudal. Additional renal arteries are frequently present and may be regarded as branches of persistent lateral splanchnic arteries. Ventral splanchnic arteries The ventral splanchnic arteries are originally paired vessels distributed to the capillary plexus in the wall of the yolk sac. After fusion of the dorsal aortae, they merge as unpaired trunks that are distributed to the increasingly defined and lengthening primitive digestive tube. Longitu - dinal anastomotic channels connect these branches along the dorsal and ventral aspects of the tube, forming dorsal and ventral splanchnic anastomoses (see Fig. 13.3). These vessels obviate the need for so many ‘subdiaphragmatic’ ventral splanchnic arteries, and these are reduced to three: the coeliac trunk and the superior and inferior mesenteric arter - ies. As the viscera supplied descend into the abdomen, their origins migrate caudally by differential growth: the origin of the coeliac artery is transferred from the level of the seventh cervical segment to the level of the twelfth thoracic; the superior mesenteric from the second thoracic to the first lumbar segment; and the inferior mesenteric from the twelfth thoracic to the third lumbar segment. However, above the diaphragm, a variable number of ventral splanchnic arteries persist, usually four or five, and supply the thoracic oesophagus. The dorsal splanchnic anas - tomosis persists in the gastroepiploic, pancreaticoduodenal and primary branches of the colic arteries, whereas the ventral splanchnic anasto - mosis forms the right and left gastric and the hepatic arteries. EMBRYONIC VEINS The early embryonic veins are often segregated into two groups, visceral and somatic, for convenience and apparent simplicity. The visceral group contains the derivatives of the vitelline and umbilical veins, and the somatic group includes all remaining veins. It should be noted that, with time, embryonic veins change the principal tissues they drain. They may receive radicles from obviously parietal tissues, which become confluent with drainage channels that are clearly visceral, and so form a compound vessel. The arrangement of the early embryonic veins is initially symmetrical. The primitive tubular symmetric heart receives its venous return through the right and left sinual horns, which are initially embedded in the mesenchyme of the septum transversum. Each horn receives, most medially, the termination of the principal vitelline vein; more laterally, the umbilical vein; and, most laterally, having encircled the pleuroperitoneal canal, the common cardinal vein. This symmetric pattern changes as the heart and gut develop and the cardiac return is diverted to the right side of the heart. Vitelline veins The vitelline veins drain capillary plexuses that develop in the splanch - nopleuric mesenchyme of the secondary yolk sac. With head, tail and lateral fold formation, the upper recesses of the yolk sac are enclosed within the embryo as the splanchnopleuric gut tube, which extends from the stomodeal buccopharyngeal membrane to the proctodeal cloacal membrane. Derivatives from all these levels possess a venous drainage that is originally vitelline. Early umbilical veins The umbilical veins form by the convergence of venules that drain the splanchnopleure of the extraembryonic allantois. The human endoder- mal allantois is very small; it projects into the embryonic end of the connecting stalk, which is therefore regarded as precociously formed allantoic mesenchyme, whereas the umbilical vessels are considered to be allantoic. The peripheral venules drain the mesenchymal cores of the chorionic villous stems and terminal villi (extraembryonic somatopleu - ric structures). These are the radicles of the vena umbilicalis impar (usually single), which traverses the compacting mixed mesenchyme of the umbilical cord to reach the caudal rim of the umbilicus. Here, the single cordal vein divides into primitive right and left umbilical veins. Each curves rostrally in the somatopleuric lateral border of the umbili - cus, i.e. where intraembryonic and extraembryonic or amniotic somato - pleure are continuous, lying lateral to the communication between both the intraembryonic and the extraembryonic coeloms. Rostrolateral to the umbilicus, the two umbilical veins reach, enter and traverse the junctional mesenchyme of the septum transversum and connect with septal capillary plexuses. They then continue, entering their correspond - ing cardiac sinual horns lateral to the terminations of the vitelline veins. EMBRYONIC ARTERIES Initially, the dorsal aortae are the only longitudinal vessels present. Their branches all run at right angles to the long axis of the embryo. Later, these transverse arteries become connected by longitudinal anas- tomosing channels that persist in part, forming arteries such as the vertebral, internal thoracic, superior and inferior epigastric, and gastro - epiploic. Each primitive dorsal aorta gives off somatic arteries (interseg - mental branches to the body wall), a caudal continuation that passes into the body stalk (the umbilical arteries), lateral splanchnic arteries (paired segmental branches to the mesonephric ridge), and ventral splanchnic arteries (paired segmental branches to the digestive tube). Somatic arteries The somatic arteries are intersegmental in position. They persist, almost unchanged, in the thoracic and lumbar regions, as the posterior inter - costal, subcostal and lumbar arteries. Each gives off a dorsal ramus, which passes backwards in the intersegmental interval and divides into medial and lateral branches to supply the muscles and superficial tissues of the back ( Fig. 13.3). The dorsal ramus also gives off a spinal branch, which enters the vertebral canal and divides into a series of branches that supply the walls and joints of the osteoligamentous canal, and neural branches to the spinal cord and spinal nerve roots. After giving off its dorsal ramus, each intersegmental artery runs ventrally in the body wall, gives off a lateral branch and terminates in muscular and cutaneous rami. Early umbilical arteries Initially, the umbilical arteries are the direct caudal continuation of the primitive dorsal aortae. They are present in the body stalk before any vitelline or visceral branches emerge, indicating the dominance of the allantoic over the vitelline circulation in the human embryo. (On a comparative basis, the umbilical vessels are chorio-allantoic and there - fore ‘somatovisceral’ .) After the dorsal aortae fuse, the umbilical arteries arise from their ventrolateral aspects and pass medial to the primary excretory duct to the umbilicus. Later, the proximal part of each umbili - cal artery is joined by a new vessel that leaves the aorta at its termination and passes lateral to the primary excretory duct. This, possibly the fifth lumbar intersegmental artery, constitutes the dorsal root of the umbili - cal artery (the original stem, the ventral root). The dorsal root gives off the axial artery of the lower limb, branches to the pelvic viscera and, more proximally, the external iliac artery. The ventral root disappears entirely, and the umbilical artery now arises from that part of its dorsal root distal to the external iliac artery, i.e. the internal iliac artery. Lateral splanchnic arteries The lateral splanchnic arteries supply, on each side, the mesonephros, metanephros, testis or ovary, and the suprarenal gland. All these struc - tures develop, in whole or in part, from the intermediate mesenchyme, later termed the aorta-gonad-mesonephros region. One testicular or ovarian artery and three suprarenal arteries persist on each side. The phrenic artery branches from the most cranial suprarenal artery, and the Fig. 13.3 The segmental and intersegmental arteries. The small red dilations indicate the positions of the longitudinal anastomoses. Anterior branchLateral branchVentral branch of somatic artery Lateral splanchnic arteryIntersegmental somatic artery Dorsal branch of somatic arterySpinal branch Dorsal aorta Kidney Ventral splanchnicartery Gut
293
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Embryonic veins 203 CHAPTER 13 mented by a range of bilateral longitudinal channels that anastomose with the posterior cardinal system and with each other. These channels are the subcardinal, supracardinal, azygos line, subcentral and precostal veins (Fig. 13.4). Subcardinal veins Subcardinal veins form in the ventromedial parts of the mesonephric ridges and become connected to the postcardinal veins by a number of vessels traversing the medial part of the ridges. The subcardinal veins assume the drainage of the mesonephros and intercommunicate by a pre-aortic anastomotic plexus, which later constitutes the part of the left renal vein that crosses anterior to the abdominal aorta. Supracardinal veins Supracardinal veins form dorsolateral to the aorta and lateral to the sympathetic trunk and take over the intersegmental venous drainage from the posterior cardinal vein. The supracardinal veins are also referred to as the thoracolumbar line or lateral sympathetic line veins. Azygos line veins Azygos line veins form dorsolateral to the aorta and medial to the sympathetic trunk. These channels, also referred to as the medial This early symmetric disposition of the vitelline veins and anastomoses, umbilical and common cardinal veins, and the locus of the hepatic primordial complex are summarized in Figures 13.1B and 13.2A. For further development of the vitelline and umbilical veins, see Chapter 60 (Figs 60.8–60.10). Cardinal veins and somatic venous complexes The initial venous channels in the early embryo have traditionally been termed cardinal because of their importance at this stage. The cardinal venous complexes are first represented by two large vessels on each side, the precardinal portion being rostral and the postcardinal being caudal to the heart. The two veins on each side unite to form a short common cardinal vein, which passes ventrally, lateral to the pleuropericardial canal, to open into the corresponding horn of the sinus venosus (see Figs 13.1B and 13.2A; Fig. 13.4B). The precardinal veins undergo remodelling as the head develops. The postcardinal veins, which, in the early embryo, drain the body wall, are insufficient channels for venous return from the developing mesonephros and gonads and for the growing body wall. As the embryo increases in size, they are supple -Fig. 13.4 Somatic venous development. A, A schematic section through the embryonic trunk. Principal longitudinal veins are colour-coded. Interconnections and intersegmental veins are uncoloured. B, The development of the principal somatic veins, from the early symmetric state, through states of increasing asymmetry, to the definitive arrangement. Abbreviations: IVC, inferior vena cava. Precardinal veins Internal jugular veins Subcardinal veins PostcardinalveinPostcardinal veinHepatocardiac veinsSinus venosus Superior vena cava Azygos veinCommon cardinal vein Superior intercostal vein Oblique vein and ligamentof left atriumBrachiocephalic veins Obliqueinterprecardinalanastomosis Azygos line vein Supracardinal veinIntersubcardinal anastomosis (renal collar) Interpostcardinal anastomosisHemiazygos vein Common iliac veinsIVC hepatic segment IVC subhepatic segment IVC subcardinal segment Left suprarenal vein Left renal veinLeft gonadal vein IVC supracardinal segmentA (i) B (ii)(iii) (iv)Postcardinal vein Supracardinal vein (thoracolumbar line vein) Subcardinal vein Early symmetrical disposition of veinsProgressive asymmetry; right-sided dominance; some channels enlarge; others retrogressMaturation and tributaries of superior vena cava; segments of definitive inferior vena cavaAzygos line vein (medial sympathetic line vein)Subcentral vein Hepatic segment of inferior vena cava (and right vitelline vein) Subhepatic segment of IVC
294
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
EARly EmbRyoni C CiRCulATion 204SECT ion 2 mechanotransduction with lymphatic endothelial cells elongating as a consequence of high interstitial fluid volume (Planas-Paz et al 2012). For more details on the morphogenesis of lymphatic vessels, see Tatin and Makinen (2014). The formation of jugular lymph sacs and their normal remodelling as the lymphatic system develops are of particular interest because delay or disruption is noted in aneuploidic fetuses. Increased accumulation of tissue fluid between the skin and soft tissues, posterior to the forming cervical spine – termed nuchal translucency – can be identified on ultrasound examination between 10 and 14 weeks’ gestation. Approxi - mately 75% of trisomy 21 fetuses show an increased distance of nuchal translucency at this time (see Fig. 14.4). In the majority of cases, early nuchal fluid accumulation resolves for normal and chromosomally abnormal fetuses between 14 and 20 weeks, and is thought to reflect increased lymphatic system development, the onset of urine production by the metanephric kidneys and a decrease in placental resistance (Nafziger and Vilensky 2014). In the fetus, lymphatic flow rates appear to be higher than in the neonate, which, in turn, are higher than in the adult; overall fetal lymph flow is thought to be five times greater than in adults (Bellini et al 2006). The production of pulmonary fluid secreted into the amniotic cavity is related to the lymphatic drainage of the lungs. In the neonate, pulmonary ventilation is important in regulating lymphatic flow in the lungs. The process of parturition causes significant changes in the dis- tribution of body water in the neonate, including movement of blood from the placenta to the fetus and a temporary shift of fluid from the intravascular compartment to the interstitial compartment; restoration of fluid balance between intravascular, lymphatic and interstitial com - partments occurs concomitant with an increase in postnatal oxygena - tion (Bellini et al 2006). LYMPH NODES AND LYMPHOID TISSUES Lymph vessels can be seen in the embryo in the cervical region from stage 16. Lymph nodes, which provide regional proliferative foci for lymphocytes, have been identified from week 9. Early lymph sacs become infiltrated by lymphoid cells, and the outer portion of each sac becomes the subcapsular sinus of the lymph node. Morphological dif - ferentiation of medullary and cortical compartments has not been observed until the end of week 10 (Tonar et al 2001). At the same time as these early lymph nodes are developing, the nasopharyngeal wall is infiltrated by lymphoid cells that are believed to herald the early devel - opment of the tubal and pharyngeal tonsils. In the neonate, a considerable proportion of the total amount of lymphoid tissue is localized in lymph nodes; the subsequent increase in the amount of lymphoid tissues that occurs during childhood reflects the growth of these nodes. Definitive follicles with germinal centres are formed during the first postnatal year. The pharyngeal tonsil reaches its maximal development at 6 years and its subsequent involution is com-pleted by puberty. Details of the development of gut-associated lym - phoid tissue are given in Chapter 60.sympathetic line veins, gradually take over the intersegmental venous drainage from the supracardinal veins. The intersegmental veins now reach their longitudinal channel by passing medial to the autonomic trunk, a relationship that the lumbar and intercostal veins subsequently maintain. Cranially, the azygos lines join the persistent cranial ends of the posterior cardinal veins. Subcentral veins Subcentral veins form directly dorsal to the aorta in the interval between the origins of the paired intersegmental arteries. These veins communi - cate freely with each other and with the azygos line veins; these con - nections ultimately form the retro-aortic parts of the left lumbar veins and of the hemiazygos veins. Precostal or lumbocostal venous line A precostal or lumbocostal venous line, anterior to the vertebrocostal element and posterior to the supracardinal, is recognized by some authorities. A possible derivative is the ascending lumbar vein. Further development of the somatic veins The supracardinal veins lie lateral to the aorta and the sympathetic trunks, which therefore intervene between them and the azygos lines (see Fig. 13.4). They communicate caudally with the iliac veins and cranially with the subcardinal veins in the neighbourhood of the pre-aortic intersubcardinal anastomosis. The supracardinal veins also com - municate freely with each other through the medium of the azygos lines and the subcentral veins. The most cranial of these connections, together with the supracardinal–subcardinal and the intersubcardinal anasto- moses, complete a venous ring around the aorta below the origin of the superior mesenteric artery, termed the ‘renal collar’ . The ultimate arrangement of these embryonic abdominal and thor - acic longitudinal cardinal veins may be summarized as follows. The terminal part of the left postcardinal vein forms the distal part of the left superior intercostal vein. On the right side, its cranial end persists as the terminal part of the azygos vein. The caudal part of the subcar - dinal vein is, in part, incorporated in the testicular or ovarian vein and partly disappears. The cranial end of the right subcardinal vein is incor - porated into the inferior vena cava and also forms the right suprarenal vein. The left subcardinal vein, cranial to the intersubcardinal anasto - mosis, is incorporated into the left suprarenal vein. The renal and tes - ticular or ovarian veins on both sides join the supracardinal–subcardinal anastomosis. On the left side, this is connected directly to the part of the inferior vena cava that is of subcardinal status via an intersubcardi - nal anastomosis. The right supracardinal vein forms much of the post - renal (caudal) segment of the inferior vena cava. The left supracardinal vein disappears entirely. The right azygos line persists in its thoracic part to form all but the terminal part of the azygos vein. Its lumbar part can usually be identified as a small vessel that leaves the vena azygos on the body of the twelfth thoracic vertebra and descends on the vertebral column, deep to the right crus of the diaphragm, to join the posterior aspect of the inferior vena cava at the upper end of its postrenal segment. The left azygos line forms the hemiazygos veins. The subcentral veins give rise to the retro-aortic parts of the left lumbar veins and of the hemiazygos veins (see Fig. 13.4). LYMPHATIC VESSELS The earliest lymphatic vessels arise from budding of lymphatic endothe - lial cells from the cardinal veins to form lymph sacs (Eichmann et al 2005). Six early lymph sacs can be identified; two are paired (the jugular and the posterior lymph sacs) and two are unpaired (the retroperitoneal sac and the cisterna chyli). In lower mammals, an additional pair (sub - clavian) is present, but in the human embryo, these are merely exten - sions of the jugular sacs. The jugular lymph sac is the first to appear, at the junction of the subclavian vein with the precardinal vein, with later prolongations along the internal and external jugular veins. The posterior lymph sac encircles the left common iliac vein. The retroperitoneal sac appears in the root of the mesentery near the suprarenal glands. The cisterna chyli appears opposite the third and fourth lumbar vertebrae ( Fig. 13.5). The lymph vessels bud out from the lymph sacs along lines that correspond more or less closely with the course of embryonic blood vessels (most commonly, veins); many also arise de novo in the mesenchyme and establish connections with existing vessels. In the body wall and the wall of the intestine, the deeper plexuses are the first to be developed; the vessels in the superficial layers are gradually formed by continued growth. Lymph vessel expansion appears to be influenced by local Fig. 13.5 Relative positions of the primary lymph sacs (as originally determined by Sabin (1912) ). Internal jugular vein External jugular vein Left common cardinal vein Left postcardinal vein Left suprarenal vein Left renal veinRetroperitoneal lymph sacLeft brachiocephalic vein Left suprarenal vein Left renal veinRetroperitoneal lymph sac Posterior lymph sacCisterna chyliPostrenal part of inferior vena cavaPrerenal part of inferior vena cavaSuperior vena cavaRight brachiocephalic veinJugular lymph sac
295
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Early embryonic circulation 204.e1 CHAPTER 13Lymphatic endothelial cells in lymphatic capillaries display a dis - continuous basal lamina and basement membrane. They are held by anchoring filaments to the surrounding extracellular matrix. Their dif - ferentiation in the mouse starts with expression of Sox-18 and Prox-1, a Prospero homeobox transcription factor, in a subset of cardinal veins (Park et al 201 1, Zhou et al 2010). Prox-1 is the most specific and func - tional lymphatic marker: its disruption leads to failure of lymphatic vessel development. Abnormal or absent lymphatic development is also seen with lack of vascular endothelial growth factor 3 (VEGFR3) and lack of angiopoietin 1, respectively (Yamashita 2007). Abnormalities of development and of remodelling of primary lymphatic vessels also occur if Akt, a serine/theonine protein kinase, is absent, possibly due to insufficient recruitment of smooth muscle cells to larger lymphatic vessels (Zhou et al 2010).
296
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
EARly EmbRyoni C CiRCulATion 204.e2 SECT ion 2REFERENCES Beets K, Huylenbroeck D, Moya IM et al 2013 Robustness in angiogenesis: notch and BMP shaping waves. Trends Genet 29:140–9. Bellini C, Boccardo F, Bonioli E et al 2006 Lymphodynamics in the fetus and newborn. Lymphology 39:1 10–17. Boon LM, Ballieux F, Vikkula M 201 1 Pathogenesis of vascular anomalies. Clin Plast Surg 38:7–19. Brouillard P, Vikkula M 2007 Genetic causes of vascular malformations. Hum Mol Genet 16:R140–9. Chappell JC, Wiley DM, Bautch VL 2012 How blood vessel networks are made and measured. Cells Tissues Organs 195:94–107. Czirok A, Ronglish BJ, Little CD 201 1 Vascular network formation in expand - ing versus state tissues: embryos and tumours. Genes Cancer 2: 1072–80. Davis GE, Senger DR 2005 Endothelial extracellular matrix. Circ Res 97: 1093–107. Eichmann A, Pardanaud L 2014 Emergence of endothelial cells during vas - cular development. In: Feige JJ, Pagès G, Soncin F (eds) Molecular Mechanisms of Angiogenesis. From Ontogenesis to Oncogenesis. Paris: Springer-Verlag France; Ch. 1, pp. 3–23. Eichmann A, Yuan Li, Moyon D et al 2005 Vascular development: from precursor cells to branched arterial and venous networks. Int J Dev Biol 49:259–67. Herbert SP, Stainier D 201 1 Molecular control of endothelial cell behaviour during blood vessel morphogenesis. Nat Rev Mol Cell Biol 12:551–64. Kuijper S, Turner CJ, Adams RH 2007 Regulation of angiogenesis by Eph- Ephrin interactions. Trends Cardiovasc Med 17:145–51.Nafziger E, Vilensky JA 2014 The anatomy of nuchal translucency at 10–14 weeks gestation in foetuses with Trisomy 21: an incredible medical mystery. Clin Anat 27:353–9. Park C, Lee JY, Yoon Y 201 1 Role of bone marrow derived lymphatic endothe - lial progenitor cells for lymphatic neovascularisation. Trends Cardiovasc Med 21:135–40. Planas-Paz L, Strilić B, Goedecke A et al 2012 Mechanoinduction of lymph vessel expansion. EMBO J 31:788–804. Sabin FR 1912 On the origin of the abdominal lymphatics in mammals from the vena cava and the renal glands. Anat Rec 6: 335–42. Senger DR, Davis GE 201 1 Angiogenesis. Cold Spring Harb Perspect Biol 3:1–19. Tatin F, Makinen T 2014 Lymphatic vascular morphogenesis. In: Feige JJ, Pagès G, Soncin F (eds) Molecular Mechanisms of Angiogenesis. From Ontogenesis to Oncogenesis. Paris: Springer-Verlag France; Ch. 2, pp. 25–44. Tonar Z, Kocova J, Liska V et al 2001 Early development of the jugular lym - phatics. Sb Lek 102:217–25. Yamashita JK 2007 Differentiation of arterial, venous and lymphatic endothelial cells from vascular progenitors. Trends Cardiovasc Med 17:59–63. Zhou F, Chang Z, Zhang L et al 2010 Akt/protein kinase B is required for lymphatic network formation, remodelling and valve development. Am J Pathol 177:2124–33.
297
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
205 CHAPTER 14 Pre- and postnatal development related to early embryonic stages, and once the number of somites is too great to count with accuracy, the degree of development of the pharyngeal arches is often used. External staging becomes more obvious when the limb buds appear. The upper limb bud is clearly visible at stage 13 and, by stage 16, the acquisition of a distal paddle on the upper limb bud is characteristic. At stage 18, the lower limb bud now has a distal paddle, whereas the upper limb bud has digit rays that are begin - ning to separate. By stage 23, the embryo has a head that is almost erect and rounded, and eyelids are beginning to form. The limbs look far more in proportion and fingers and toes are separate. At this stage, the external genitalia are well developed, although they may not be suffi - ciently developed for the accurate determination of the sex. Historically, the onset of bone marrow formation in the humerus was used by Streeter to indicate the end of the embryonic and the begin - ning of the fetal period of prenatal life. The fetal period occupies the remainder of intrauterine life; growth is accentuated, although differ - entiative processes continue up to and beyond birth. Overall, the fetus increases in length from 30 mm to 500 mm, and increases in weight from 2–3 g to more than 3000 g. Fetal staging Currently, there is no satisfactory system of morphological staging of the fetal period of development, and the terminology used to describe this time period reflects this confusion. The terms ‘gestation’, ‘gesta - tional age’ and ‘gestational weeks’ are considered ambiguous by O’Rahilly and Müller (2000), who recommend that they should be avoided. However, they are widely used colloquially within obstetric practice. Staging of fetal development and growth is based on an esti - mate of the duration of a pregnancy. Whereas development of a human from fertilization to full term averages 266 days, or 9.5 lunar months (28 day units), the start of pregnancy is traditionally determined clini - cally by counting days from the last menstrual period; estimated in this manner, pregnancy averages 280 days, or 10 lunar months (40 weeks). Figure 14.3 shows the embryonic timescale used in all descriptions of embryonic development and the obstetric timescale used to gauge the stage of pregnancy. The predicted date of full term and delivery is revised after routine ultrasound examination of the fetus. Early ultrasound estimation of gestation increases the rate of reported preterm delivery (delivery at <37 weeks) compared with estimation based on the date of the last menstrual period (Yang et al 2002), possibly because delayed ovulation is more frequent than early ovulation; the predicted age of a fetus esti- mated from the date of the last menstrual period may differ by more than 2 weeks from estimates of postfertilization days. Obstetric staging In obstetric practice, the duration of the period of gestation is regarded as 9 calendar months, which is approximately 270 days. The period of pregnancy is divided into thirds, termed trimesters. The first and second trimesters each cover a period of 12 weeks, and the third trimester covers the period from 24 weeks to delivery. Although the expected date of delivery is computed at 40 weeks of pregnancy, the term of the preg - nancy, i.e. its completion resulting in delivery, is considered normal between 37 and 42 weeks. Neonates delivered before 37 weeks are called preterm (or premature); those delivered after 42 weeks are post-term. The period from the end of week 24 and up to 7 days after birth is termed the perinatal period. Fetuses that are delivered and die before 24 weeks are considered to be miscarriages of pregnancy, although tech - nological advances in neonatal care can now assist the delivery and support of infants younger than 24 weeks. Infants born after 24 weeks of pregnancy who subsequently die are classed as stillborn and contrib- ute to the statistics of perinatal mortality. Studies that discuss fetal PRENATAL STAGES The absolute size of an embryo or fetus does not afford a reliable indi - cation of either its chronological age or the stage of structural organiza - tion, even though graphs based on large numbers of observations have been constructed to provide averages. All such data suffer from the dif - ficulty of timing the moment of conception in humans. It has long been customary to compute the age, whether in a normal birth or an abortion, from the first day of the last menstrual period of the mother but, as ovulation usually occurs near the fourteenth day of a menstrual cycle, this ‘menstrual age’ is an overestimate of about 2 weeks. Where a single coitus can be held to be responsible for conception, a ‘coital age’ can be established, and the ‘fertilization age’ cannot be much less than this because of the limited viability of both gametes. It is usually held that the difference may be several days, which is a highly significant interval in the earlier stages of embryonic development. Even if the time of ovulation and coitus were known in instances of spontaneous abortion, not only would some uncertainty still persist with regard to the time of fertilization, but also there would remain an indefinable period between the cessation of development and the actual recovery of the conceptus. To overcome these difficulties, early embryos have been graded or classified into developmental stages or ‘horizons’, on the basis of both internal and external features. The study of the Carnegie collection of embryos by Streeter (1942, 1945, 1948), and the continuation of this work by O’Rahilly and Müller (1987), provided, and continues to provide, a sound foundation for embryonic study and a means of com - paring stages of human development with those of the animals rou - tinely used for experimental study, namely: the chick, mouse and rat. Recent use of ultrasound for the examination of human embryos and fetuses in utero has confirmed much of the staging data. The development of a human from fertilization to birth is divided into two periods: embryonic and fetal. The embryonic period has been defined by Streeter as 8 weeks post fertilization, or 56 days. This time - scale is divided into 23 Carnegie stages, a term introduced by O’Rahilly and Müller (1987) to replace developmental ‘horizons’ . The designation of stage is based on external and internal morphological criteria and not on length or age. Embryonic staging Embryonic stages 1–10 are shown in detail in Figure 8.1. Estimations of embryonic length may be 1–5 mm less than equivalent in vivo esti- mates, reflecting the shrinkage caused by the fixation procedures that are inevitably used in embryological studies. O’Rahilly and Müller (2000, 2010) have revised some of the ages that were previously assigned to early embryonic stages, pointing out that interembryonic variation may be greater than had been thought and that, consequently, some ages may have been underestimated. They note that, as a guide, the age of an embryo can reasonably be estimated from the embryonic length within the range 3–30 mm, by adding 29 to the length. Correlating the age of any stage of development to an approximate day may be unreli - able, and a generalization using the number of weeks of development might now be more appropriate. The stages of development encompass all aspects of internal and external morphogenetic change that occur within the embryo within the duration of the stage. They are used to convey a snapshot of the status of the development of all body systems within a particular time - frame. Figure 14.1 shows the external appearance of embryos from stage 6 to stage 23, with details of their size and age in days. The correl - ation of external appearance of the embryo with internal development is shown in Figure 14.2. Obvious external features provide some guidance to the changes occurring within embryos during successive stages. Somite number is
298
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Pre- and Postnatal develo Pment 206seCtIon 2 defects, lung immaturity, renal agenesis, polyhydramnios and defects of the anterior abdominal wall (gastroschisis and exomphalos). Estima- tions are made of the following: biparietal skull diameter, taken through a plane of section that traverses the third ventricle and thalami; head circumference through a plane that traverses the third ventricle and thalami plus the cavum septi pellucidi anteriorly and the tentorial hiatus posteriorly; abdominal circumference through a plane where the transverse diameter of the liver is greatest, the appearance of the lower ribs is symmetrical and the junction of the left and right portal veins is identified; and femoral length, which measures the ossified portions of the diaphysis and metaphysis ( Fig. 14.4). Between 15 and 28 weeks, the biparietal diameter is the most accu - rate index of fetal menstrual age and the expected date of delivery. Other measurements that may also be taken include transverse cerebellar diameter and foot length. The amount and type of fetal movement, breathing movements and visceral functions, exemplified by bladder emptying, peristaltic action and colonic echogenicity, are noted. Three- and four-dimensional ultrasound scans are now routinely constructed (Video 14.1). Gender can be identified later in development but this information is not always routinely passed on to parents. Constructions of ultrasound biometry charts for fetal aging now take into account the ethnic population under consideration; it is recom - mended that locally developed charts specific to the population should be used. The use of these charts means that factors that may influence fetal biometry, including maternal age and nutritional status, maternal height, weight, parity and smoking habits, are noted, facilitating accu - rate prediction of small-for-date and growth-retarded fetuses. The World Health Organization (WHO) has published a protocol for the collec- tion of data for a multicentre study of fetal growth (Merialdi et al 2014). INTERGROWTH-21st, an international, multicentre project, also aims to standardize protocols for collecting anthropometric measurements and constructing charts describing optimal fetal and preterm postnatal growth (Sarris et al 2013, Uauy et al 2013). An outcome of routine ultrasound examination of embryos and fetuses for anomalies may be a change to the perinatal management, development and the gestational age of neonates, particularly those born before 40 weeks’ gestation, use the obstetric estimated stages and age (menstrual age), unless they specifically correct for this. If a fetal ageing system is used, it is important to remember that the age of the fetus may be 2 weeks more than a comparable fetus that has been aged from postovulatory days. Ultrasound staging The difficulty of correlating the appearance of a chorionic sac, embryo or fetus on an ultrasound scan with age during the first trimester is related to the specificity of reporting the age. An age reported as within a week of pregnancy, e.g. week 12, will cover a period of 6 days (12 weeks 0 days up to 12 weeks 6 days). Therefore, it is recommended that sono - graphic estimation of age should be given as menstrual weeks and days (i.e. 12 weeks indicates 12 weeks 0 days) (Galan et al 2008). First trimester scan An early ultrasound scan will detect implanta- tion and viability of the embryo once a heart beat is detected, confirm multiple pregnancy and estimate the date of delivery. Cardiac activity can be identified by the sixth menstrual week (Galan et al 2008). Crown– rump length measurement is the most accurate predictor of menstrual age of the embryo during the first trimester (Galan et al 2008). Nuchal translucency is measured between 10 and 14 weeks to diagnose trisomy 21. The development of three-dimensional ultrasound scanning has enabled early detection of many anomalies previously diagnosed in the second trimester, including anencephaly, hydrocephalus and encepha - locele (Pretorius et al 2014). Normograms of fetal spine growth have been constructed for a Taiwan population, which show mean spine length increasing linearly between 1 1 and 14 weeks (Cheng et al 2010). The study also noted a tendency for spinal extension as early as 1 1 weeks. Second trimester scan Routine scanning at 18–20 weeks (obstetric staging) is used to confirm the delivery date, and assess not only the position of the placenta but also the presence of fetal anomalies that would require special intervention following delivery, such as cardiac Fig. 14.1 The external appearance and size of embryos between stages 6 and 23. Early in development, external features are used to describe the stage, e.g. somites, pharyngeal arches or limb buds. (Adapted with permission from Rodeck CH, Whittle MJ 1999 Fetal Medicine. London: Churchill Livingstone.)Size (mm) 0.4 1.5–36 10 Approximate age (days)16–18 26–294–613 30–338–1116 35–4013–1718 41–4521–2320 46–5028–3023 53–58Embryonic stageStage 18 Stage 16 Stage 13Eyelids Pinna Mandibular processMaxillaryprocess Opticvesicle Otic vesicleFirst pharyngeal arch Neurulation Somites Cerebralhemisphere Connectingstalk Connectingstalk ConnectingstalkYolksac Reflection of amnion Epiblast population Primitive node Primitive streakConnecting stalkPericardial cavityand heart Lower limbbudUpper limbbudHand Umbilicalcord Stage 10 Stage 6
299
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Prenatal stages 207 CHaPter 14 Fig. 14.3 The two timescales used to depict human development. Embryonic development, in the upper scale, is counted from fertilization (or from ovulation, i.e. in postovulatory days; see O’Rahilly and Müller (1987)). Throughout this book, times given for development are based on this scale. The clinical estimation of pregnancy is counted from the last menstrual period and is shown on the lower scale; throughout this book, fetal ages relating to neonatal anatomy and growth will have been derived from the lower scale. Note that there is a 2-week discrepancy between these scales. The perinatal period is very long because it includes all preterm deliveries. Embryonic stagesImplantation period Early neonatal period (birth–7 days)Late neonatal period (7–28 days) Fetal period Perinatal period 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 1 Implantation Last menstruation Estimated date of deliveryFertilization2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 41 1 2 3 4 5 6 7 8 937 38 39 40 10Age of embryo (weeks) Age of embryo (months) Pregnancy (weeks) Pregnancy (lunar months) First trimester Second trimester Third trimester (of pregnancy)Fig. 14.2 A timetable of development of the body systems. The development of individual systems can be seen progressing from left to right. Embryonic stages and weeks of development are shown. Embryonic stages are associated with external and internal morphological features rather than embryonic length. To identify the systems and organs at risk at any time of development, follow a vertical progression from top to bottom. Embryonic stages Weeks post ovulation External appearance Nervous Respiratory Gastrointestinal Urinary Reproductive Cardiovascular MusculoskeletalHead and tail folding Neurulation Fore-, mid-, hindgutMidgut loop returns to abdomenPharyngeal pouches dorsal and ventral PancreasThyroid Liver Midgut loop rotating Urorectal septum Rotation of stomachTrachea Lung buds Further division of bronchi Primary bronchi Mesonephros Mesonephric duct Ureteric bud Germ cells in allantois wall Primitive vascular systemSeptum primum Somite period Cartilaginous part of skullMembranous part of skull Forelimb bud Forelimb digit rays Hindlimb bud20 days .................................. 30 daysSeptation of ventricles Heart beats Spleen Septum secundum Heart tubeMüllerian ducts Uterus and uterine tubes Testis at inguinal canal Indifferent gonad Testis differentiating Vagina Prostate External genitalia indifferent External genitalia differentiatingMetanephric nephrons Kidneys ascend Major calyces Minor calycesAnterior lobe pituitaryPosterior lobe pituitaryFirst neural crest cellsOtic vesicle Optic cup Membranous labyrinthPharyngeal arches Palate External earUpper lip Digits on hand Eyelids fuse 1 2 3 4 5 6 7 8 9 10 11 127 6 8 9 10 11 12 13 14 1516 23 17 22 21 18 19 20
300
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Pre- and Postnatal develo Pment 208seCtIon 2 Fig. 14.4 Ultrasound planes examined in a routine second-trimester antenatal scan (left) and measurements taken to predict the estimated date of delivery (right). The top three images are in the sagittal plane ( A). The other planes are transverse, apart from J, which is longitudinal. Abbreviations: Ao, aorta; DA, ductus arteriosus; DV, ductus venosus; LA, left atrium; LPV, left portal vein; LV, left ventricle; PA, pulmonary artery; RA, right atrium; RPV, right portal vein; RV, right ventricle; SVC, superior vena cava. Nuchal translucency (A) Transventricular (B) Heart - three-vessel view (E) Cord insertion (H) Bladder (I)Heart - four-chamber view (F)Transcerebellar (D)A B C D E F G H I JLongitudinal spine (A) Aortic arch (A) Transthalamic – head circumference (C)Transthalamic – biparietal diameter (C) Abdominal circumference (G) Femur length (J)Ultrasound planes examined in 20-week scan 20-week scan biometry for delivery date estimation Posterior ventricle Cerebellum BladderCisternamagna MidlineMidline Thalami Midline ThalamiMidline CordinsertionCordinsertionStomachRPVLPV DV BowelRA LALVRVPA DA Spine Spine Spine SpineAo SVC Spine
301
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Growth in utero 209 CHaPter 14still disproportionately large. (Although the rates of growth of the trunk and lower limbs increase during the remainder of intrauterine life, the disproportion is present after birth and, to a diminishing degree, is retained throughout childhood and on into the years of puberty.) A covering of primary hair, lanugo, appears. Towards the end of this period, sebaceous glands become active; the sebum that is secreted blends with desquamated epidermal cells to form a cheesy covering over the skin, the vernix caseosa, which is usually considered to protect the skin from maceration by the amniotic fluid. At about this time, the mother becomes conscious of fetal movement, formerly termed ‘quickening’ . In the sixth month, the lanugo darkens, the vernix caseosa is more abundant and the skin becomes markedly wrinkled. The eyelids and eyebrows are now well developed. During the seventh month, the hair of the scalp is lengthening and the eyebrow hairs and the eyelashes are well developed. The eyelids themselves separate and the pupillary mem - brane disappears. The body becomes more plump and rounded in contour and the skin loses its wrinkled appearance as a result of the increased deposition of subcutaneous fat. Fetal length has increased to approximately 350 mm and weight to about 1.5 kg. Towards the end of this month, the fetus is viable; if born prematurely, it may be able to survive without the technological assistance found in neonatal inten - sive care units and its postnatal development can proceed normally. Throughout the remaining lunar months of normal gestation, the covering of vernix caseosa is prominent. There is a progressive loss of lanugo, except for the hairs on the eyelids, eyebrows and scalp. The bodily shape becomes more infantile but, despite some acceleration in their growth, the legs have not quite equalled the arms in length pro - portionately, even at the time of birth. The thorax broadens relative to the head, and the infra-umbilical abdominal wall shows a relative increase in area, so that the umbilicus gradually becomes more centrally situated. Average lengths and weights for the eighth, ninth and tenth months are 40, 45 and 50 cm and 2, 2.5 and 3–3.5 kg, respectively. The rate of fetal growth slows from 36 to 40 weeks in response to the physi - cal limitation imposed by the maternal uterus. Birth weight thus reflects the maternal environment more than the genotype of the child. This slowing of the growth rate enables a genetically larger child developing within a small mother to be delivered successfully. After birth, the growth rate of the neonate increases; the rate of weight gain reaches a peak some 2 months postnatally. Just before birth, the lanugo almost disappears and the umbilicus is central. The testes, which begin to descend with the processus vaginalis of peritoneum during the seventh month and are approaching the scrotum in the ninth month, are usually scrotal in position. The ovaries are not yet in their final position at birth; although they have attained their final relationship to the uterine folds, they are still above the level of the pelvic brim. Fetal surgery The use of routine ultrasound examination and MRI of fetuses has led to the development of a number of prenatal surgical interventions: e.g. to correct placental or membrane anomalies resulting in twin-to-twin transfusion and the production of amniotic bands (Deprest et al 2014); or to offer improved outcomes for meningomyelocele on the basis that the neural tissue may become secondarily damaged due to exposure to amniotic fluid and mechanical traumatic injury during gestation (Adzick 2013, Danzer and Johnson 2014, Cohen et al 2014). GROWTH IN UTERO Alterations in the availability of nutrients to the fetus at particular stages of pregnancy elicit adaptive responses by the fetus. These may ensure fetal coping but may also result in pathology in adult life; the nutri - tional status of pregnant females is therefore of fundamental impor - tance for the health of the next generation. Under- or overnutrition during fetal life may lead to metabolic changes in childhood, adoles - cence and adulthood, and pass to successive generations. Poor nutrition at critical stages of fetal life permanently alters the normal developmen - tal pattern of a range of organs and tissues, e.g. the endocrine pancreas, liver and blood vessels, resulting in their pathological responses to certain conditions in later adult life (Barker et al 1993, Catalano and Hauguel-De Mouzon 201 1). Maternal hyperglycaemia leads to changes in fetal metabolism and pathology in the child and adult (Pedersen 1952). An increase in placental size occurs in pregnancy as an adaptive response to both high altitude and mild undernutrition, particularly i.e. to the time, method and place of delivery of the fetus, or the parents may choose to terminate the pregnancy to minimize concerns about fetal and neonatal suffering and long-term disability. Termination may be chosen for severe, untreatable inherited metabolic disorders (e.g. Tay–Sachs disease), severe chromosomal anomalies (e.g. trisomy 13), lethal bone dysplasias, lethal anomalies such as anencephaly and other extreme neurological defects, and bilateral renal agenesis (Cass 201 1). Improvements in prenatal screening and diagnosis have led to an overall increase in the prevalence of reported birth defects and overall lower perinatal mortality rates, reflecting increased early terminations of pregnancy (Cass 201 1). Magnetic resonance imaging Advances in prenatal ultrafast magnetic resonance imaging (MRI) now provide further ways to detect fetal anomalies, capturing particularly clear images (Sepulveda et al 2012) ( Fig. 14.5). MRI is considered a useful adjunct to ultrasound imaging at 20–22 weeks because it enables better management planning for known or suspected anomalies (Reddy et al 2014). Fetal development Although accurate morphological stages are not available for the fetal period, the developmental progression is broadly clear. During the fourth and fifth months, the fetus has a head and upper limbs that are Fig. 14.5 MRI scans of fetal anomalies. A, Polyhydramnios may arise as a result of oesophageal compression, impaired swallowing and lack of absorption of amniotic fluid from the gut. B, Gastroschisis with loops of bowel in the extracellular coelom (chorionic cavity) and normal insertion of the umbilical cord. (With permission from Sepulveda W, Ximenes R, Wong AE, et al 2012 Fetal magnetic resonance imaging and three- dimensional ultrasound in clinical practice: Applications in prenatal diagnosis. Best Pract Res Clin Obstet Gynaecol 26:593–624.) A Bowel Umbilical cord B
302
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Pre- and postnatal development 209.e1 CHaPter 14Other conditions have also been treated prenatally with limited improvement in mortality. Thus, congenital diaphragmatic hernia may be treated by percutaneous fetoscopic endoluminal tracheal occlusion from 25 to 33 weeks’ gestation to prevent loss of lung fluid and enhance lung growth (Deprest et al 2004). Fetal airway patency has also been preserved during birth through perinatal ex utero intrapartum (EXIT) treatment, where a portion of the fetus is delivered through a hysterec - tomy incision for surgery while the fetus remains attached to the utero - placental circulation. After the airway procedure is completed, the fetus is completely delivered (Deprest et al 2014).
303
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Pre- and Postnatal develo Pment 210seCtIon 2the rapid maturation of some systems and the compensatory growth of others (in terms of responses to the effect of gravity or enteral feeding or exposure to microorganisms). Details of the relative positions of the viscera and the skeleton in a full-term neonate are shown in Figures 14.6–14.8. The newborn infant is not a miniature adult, and extremely preterm infants are not the same as full-term infants. There are immense differences in the relations of some structures between the full-term neonate, child and adult, and there are also major differences between the 20-week-gestation fetus and the 40-week fetus, just before birth. The study of fetal anatomy at 20, 25, 30 and 35 weeks is vital for the investigative and life-saving procedures carried out on preterm infants today. Neonatal measurements and period of time in utero The tenth to ninetieth centile ranges for length of full-term neonates are 48–53 cm ( Fig. 14.9A). Length of the newborn is measured from crown to heel. In utero, length has been estimated either from crown– rump length, i.e. the greatest distance between the vertex of the skull and the ischial tuberosities, with the fetus in the natural curved posi - tion, or from the greatest length exclusive of the lower limbs. Greatest length is independent of fixed points and thus much simpler to measure; it is generally taken to be the sitting height in postnatal life, and is the measurement recommended by O’Rahilly and Müller (2000) as the standard in ultrasound examination. At birth, weight reflects the maternal environment, the number of conceptuses, the sex of the baby and the parity of the mother. Generally, full-term female babies are lighter than full-term males, twins are lighter than singletons, and later children tend to be heavier than the first-born. The tenth to ninetieth centile ranges for the weight of a full-term infant at parturition are 2700–3800 g (Fig. 14.9B ), the average being 3400 g; 75–80% of this weight is body water and a further 15–28% is composed of adipose tissue. Birth weight is noted against charts appropriate for ethnicity and categorized as low, normal and high. Low birth weight has been defined as less than 2500 g, very low birth weight as less than 1500 g, and extremely low birth weight as less than 1000 g. Infants may weigh less than 2500 g but not be premature by gestational age. Measurement of the range of weights that fetuses may attain before birth has led to the production of weight charts, which allow babies to be described accord - ing to how appropriate their birth weight is for their gestational age, e.g. small for gestational age, appropriate for gestational age or large for gestational age ( Fig. 14.10). Small-for-gestational-age infants, also termed ‘small-for-dates’, are often the outcome of intrauterine growth retardation. For both premature and growth-retarded infants, an assessment of gestational age, which correlates closely with the stage of maturity, is desirable. Gestational age at birth is predicted by its proximity to the estimated date of delivery and the results of ultrasonographic examin - ations during pregnancy. It is currently assessed in the neonate by evalu - ation of a number of external physical and neuromuscular signs. Scoring of these signs results in a cumulative score of maturity that is usually within ±2 weeks of the true age of the infant. The scoring scheme has been devised and improved over many years. Estimation of large-for-dates infants is based on assessment of fetal weight through ultrasound evaluation and some biometrical indices. Assessment of anterior abdominal wall width is thought to predict large-for-gestational-age babies (Walsh and McAuliffe 2012). Fetal mac - rosomia is defined as an absolute birth weight greater than 4000 g, 4500 g or 5000 g, or as a customized birth weight centile greater than the ninetieth, ninety-fifth or ninety-seventh centile for the infant’s gesta - tion age (Walsh and McAuliffe 2012, Schwartz et al 2014). The precise definition may not necessarily be helpful, as some at-risk infants, not identified as large for dates from growth curve charts, might go unrec - ognized (Larma and Landon 201 1). There is a correlation between macrosomia and short maternal stature; macrosomic fetuses are at risk of shoulder dystocia and brachial plexus injuries during vaginal delivery. GROWTH IN INFANCY AND CHILDHOOD After birth, there is a general decrease in the total body water but a rela - tive increase in intracellular fluid. Normally, the newborn loses about 10% of its birth weight by 3–4 days postnatally because of loss of excess extracellular fluid. By 1 year, total body water makes up 60% of the body weight.during mid-pregnancy. However, although a larger placenta may be better able to deliver the full nutritional requirements of the fetus, the perfusion of a larger placenta is not without problems. It may produce changes in fetal blood flow and placental enzymes, and in the normal structure of the fetal vessel wall or of its responses to circulating trophins, e.g. catecholamines or angiotensin II, which will continue into adult life. Undernutrition in later pregnancy does not produce the same sequelae and placental enlargement does not occur. However, fetal growth slows and fetal wasting may occur as oxygen, glucose and amino acids are redistributed to the placenta to maintain its function. Transition to extrauterine life The type of birth confers developmental outcomes on the neonate that affect its subsequent growth. Caesarean section for preterm delivery appears not to be associated with improved neonatal outcomes but is associated with increased risk of respiratory distress syndrome due to slower postnatal movement of fluid out of the lungs (Werner et al 2012, Bhatta and Keriakos 201 1, Bellini et al 2006). The dramatic rise of this delivery method is not necessarily driven by improved neonatal out - comes because caesarean section has been shown to increase the odds of developing asthma and type 1 diabetes in later life by 20%. Meta- analyses indicate that the endocrine milieu during caesarean section may lead to epigenetic effects on hepatic and metabolic function and so affect immune function in adult life, leading, in some instances, to food allergy and obesity (Steer and Modi 2009, Hyde and Modi 2012, Song et al 2013). Prior to birth, the gut is exposed to hundreds of proteins, metabo - lites and cytokines that are swallowed in amniotic fluid; there is some evidence for the presence of bacterial species in amniotic fluid prior to delivery. Establishment of the gut microbiota by exposure to mater - nal vaginal and colonic bacteria during vaginal delivery leads to normal maturation of the gut wall. Gut colonization in preterm infants is delayed (Cilieborg et al 2012). Infants born via caesarean section may have primary gut flora disturbance for up to 6 months after birth and associated delay in postnatal immune development (Neu and Rushing 201 1, Neu and Mai 2012, Matamoros et al 2013). The establishment of enteral feeding has profound effects on early postnatal maturation. Human breast milk contains many bioactive sub - stances that enhance gut maturity. It improves gastric emptying, encour - ages the growth of the microbiotome and contains a range of proteins, including large amounts of secretory immunoglobulin A and cytokines (e.g. interleukin (IL)-10) (Jakaitis and Denning 2014). Feeding preterm infants human milk is associated with less feeding intolerance and is thought to provide protection against diabetes and obesity in later life (Valentine and Morrow 2012). The rate of growth of infants fed breast milk for the first 6 months of postnatal life is different to that of formula-fed infants. Delivery of low-weight preterm babies, followed by parenteral feeding with no enteral nutrition, disrupts normal gut maturation and may lead to necrotizing enterocolitis (Wynn and Neu 2012). However, a balance with the gut microbiome is necessary; although full-term neonates born in tropical countries have similar villous height to those born in temperate climates, villous length in the small intestine shortens within 2–12 months of birth as a consequence of tropical enteropathy (Ramakrishna et al 2006). The time of transition to extrauterine life is permanently recorded in the primary teeth and can be reliably demonstrated in forensic dental identification. The circadian growth rhythm of dentine and enamel deposition in tooth germs is temporarily blocked by the meta - bolic stress of delivery; the resultant change in enamel prism deposi - tion is seen as a neonatal line or ring that can be detected in all deciduous teeth and permanent first molar teeth of live births. The thickness of the line is related to birth difficulties, being thinner in caesarean section and thicker in vaginal delivery (Sabel et al 2008, Canturk et al 2014). NEONATE The neonatal period extends from birth to 28 days postnatally; it is divided into an early neonatal period from birth to 7 days, and a late neonatal period from 7 to 28 days. Technological advances have enabled successful management of preterm infants, many at ages that were considered non-viable a decade or two previously. Maturational pro - cesses involving local interactions and pattern formation still drive development at local and body-system levels in preterm infants. The sudden release of such fetuses into a gaseous environment of varying temperature, with full gravity and a range of microorganisms, promotes
304
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Pre- and postnatal development 210.e1 CHaPter 14Animal studies have verified that maternal starvation during preg - nancy decreases fetal intrauterine growth factor (IGF)-I concentrations and, together with a general hypoglycaemia, impairs the development of the fetal β cells of the pancreas. Moreover, fetal undernutrition induces insulin resistance in the tissues. The coexistence of insulin resistance and impaired β-cell development in the fetus appears to be important in the pathogenesis of type 2 diabetes. The risk of developing this type of diabetes is greatest in those individuals with low weight at birth and at 1 year, and who become obese as adults, thus challenging an already impaired glucose–insulin metabolism. Fetal IGF-I concentra - tions are also lower in infants who are short at birth as a result of a long period of maternal undernutrition; these individuals have an exag - gerated response to growth hormone-releasing factor, which, together with low IGF-I concentrations, suggests a degree of growth hormone resistance. It is now thought that the balance of hormonal environment in intrauterine and early postnatal life is necessary for future adult health. The presence of altered concentrations of hormones during critical periods of development may act as endogenous functional teratogens (Plagemann 2004). Different birth phenotypes have been correlated with different path - ological sequelae. Infants who are thin at birth, with a low ponderal index (weight/length 3), tend to develop a combination of insulin resist - ance, hypertension, type 2 diabetes and lipid disorders, whereas those who are short in relation to head size tend to develop hypertension and high plasma fibrinogen concentrations. These associations have been reported in babies born small for dates, rather than in those born pre - maturely. Some babies of average weight also develop cardiovascular pathology; they are either thin at birth and small in relation to the size of their placenta, or of average weight but short in relation to head size and have below average weight gain during the first year (Barker et al 1993). For more recent views on this concept, the reader is directed to consult Godfrey and Barker (2000), Ross and Beal (2008), and Kel- ishadi and Poursafa (2014). Obesity has become increasingly prevalent since the end of the twentieth century, leading to an increase in type 2 diabetes, hyperten-sion, hyperlipidaemia, atherosclerosis and inflammation in later life. In pregnancy, an observed decrease in insulin sensitivity as pregnancy advances leads to increased nutrient transfer to the fetus. In response, the fetus increases insulin secretion, stimulates IGF-I production, and fetal growth and fat deposition are promoted (Vrachnis et al 2012). Obese females start their pregnancy with greater insulin resistance than females of normal weight, which means that their fetuses produce more IGF-I and continue to grow. The risk of fetal macrosomia is three times higher in females with poorly controlled gestational diabetes because their increased weight reflects fat mass rather than lean body mass (Catalano and Hauguel-De Mouzon 201 1). Adiposity at birth is related to obesity and metabolic dysfunction in childhood, which may be perpetuated through adulthood in an ongoing cycle through the genera - tions (Catalano and Hauguel-De Mouzon 201 1, Poston 2012). A range of risk factors for pre-, peri- and postnatal issues in maternal obesity is given in McGuire et al (2010).
305
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Growth in infancy and childhood 211 CHaPter 14 Fig. 14.6 A–C, Topographical representations of the anatomy of a full-term neonate. The surface markings of all organs are shown, with some coloured and others only in outline. The female genital tract is shown on the right of the body, and the male tract is shown on the left. A B C From cranial to caudal, the following structures are indicated: Superior sagittal sinus Brain Thyroid gland Thymus Liver Gallbladder Spleen Pancreas Blood vessels Oxygenated blood (red) is returned to the fetal heart via the umbilical vein, which passes from the umbilicus to the liver . The right atrium contains oxygenated blood, which mainlypasses to the left atrium. The right ventricle receives some oxygenated blood from this flow and also the deoxygenated blood from the head and neck. Blood is returned to theplacenta via two umbilical arteries.Note that the apex of the bladder continues as the urachusto the umbilicus.Note that the lower border of the lung is below the central,upper border of the liver .Note that the suprarenal glands are relatively large andsuperomedial to the lobulated kidneys.From cranial to caudal, the following structures are indicated: Larynx and trachea Gastrointestinal tract Pancreas Urinary bladder Prostate glandFrom cranial to caudal, the following structures are indicated: Maxillae and mandible Costal cartilages Lungs Suprarenal glands Lobulated kidneys Ureter passing to the posterior surface of the urinary bladder Right uterine tube, uterus and vagina Left testis and vas deferens
306
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Pre- and Postnatal develo Pment 212seCtIon 2 Fig. 14.8 The extent of the ossified skeleton in a full-term neonate. Note the derivation of the parts of the skeleton: the skull is derived from paraxial mesenchyme and neural crest mesenchyme; the axial skeleton, vertebrae and ribs are derived from paraxial mesenchyme; and the skeletal elements in the limbs are derived from somatopleuric mesenchyme, which forms the limb buds. Neural crest mesenchyme Paraxial mesenchymeSomatopleuric mesenchyme From cranial to caudal, the following structures are indicated: Superior sagittal sinus Brain Larynx and trachea Maxillae and mandible Thyroid gland Thymus Sternum Heart Liver Vertebrae Pancreas Blood vessels Gastrointestinal tract Visceral peritoneum Urinary bladder and urethra Uterus and vagina Fig. 14.7 A topographical representation of the anatomy of a full-term female neonate. The cut surfaces of the organs are the same colours as in Figure 14.6.
307
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Growth in infancy and childhood 213 CHaPter 14 Growth has always been regarded as a regular process. The rates of prenatal and postnatal growth can be indicated by increments in body length or weight, which, when plotted, form a growth curve (see Fig. 14.9). Growth curves can be plotted for individuals if accurate measure - ments are taken, preferably by the same person, for the entire period of growth, i.e. a longitudinal study. An alternative method is to collect a series of averages for each year of age obtained from different indi - viduals, i.e. a cross-sectional study. Cross-sectional studies are valuable for the construction of standards for height and weight attained by healthy children at specific ages, and can establish centile limits of normal growth, but they cannot reveal individual differences in either the rate of growth or the timing of particular phases of growth. The data from longitudinal and cross-sectional studies can also be used to plot the increments in height or weight from one age to the next. This produces a velocity curve, which reflects a child’s state at any particular time much better than the growth curve, in which each point is dependent on the preceding one. The oldest published longitudinal study, still of great value today, was made by Count Philibert de Mont - beillard on his son ( Fig. 14.1 1). It shows that the velocity of growth in height decreases from birth onwards, and that a marked acceleration of growth, the adolescent growth spurt, occurs from 13 to 15 years (see below; see Fig. 14.14). Cross-sectional data permit comparison of prenatal and postnatal growth. Childhood growth charts are used to predict normal childhood development. The velocity curve for the prenatal and postnatal period (Fig. 14.12) shows that the peak velocity for length is reached at about 4 months (note that these prenatal charts use the obstetric measure - ments of gestational time, in which fetal age is estimated from the last menstrual period, 2 weeks before fertilization). Growth in weight usually reaches its peak velocity after birth. Tanner (in Harrison et al (1964)) noted that the more carefully measurements are taken, the more regular is the succession of points on a growth curve. However, a longitudinal study of growth measured weekly, semi-weekly and daily recorded that growth in length and head circumference occurred by saltatory increments, with a mean amplitude of 1.01 cm for length (Lampl 2002); growth stasis, steep changes in growth and incremental growth have all been recorded in infancy, child - hood and adolescence (Caino et al 2006). Charts of height and weight correlated to age are compiled from extensive cross-sectional growth studies. Such charts show the mean height or weight attained at each age, termed the fiftieth centile, and also the centile lines for the seventy-fifth, ninetieth and ninety-seventh centiles, in addition to the twenty-fifth, ninth and second centiles. The data shown in Figure 14.13 are derived from United Kingdom cross- sectional references. Any comparison of an individual growth curve Growth rates of infants born prematurely differ from those born naturally at term. Normally, fetuses grow rapidly between 20 and 40 weeks. Optimum growth of premature infants is considered to be equivalent to intrauterine rates. However, whereas, in utero, fetuses show a slowing growth velocity before term, preterm infants show linear growth. Growth charts for preterm infants have been constructed (Fenton and Kim 2013). The INTERGROWTH-21st project is collecting data for preterm growth (see above).Fig. 14.9 Standardized graphs of ( A) fetal length and ( B) weight from 24 weeks of pregnancy, showing the tenth, fiftieth and ninetieth centiles. 4600 4400 42004000380036003400320030002800260024002200200018001600140012001000 800600400200 24 26 28 30 Preterm Term Post-term32 34 36 38 40 4290% 50% 10% 0 Week of gestationB Weight (g)53 52 515049484746454443424140393837363534333231 24 26 28 30 Preterm Term Post-term32 34 36 38 40 4290% 50% 10% Week of gestationA Length (cm) Fig. 14.10 Intrauterine growth status and its appropriateness for gestational age. Gestational age is more closely related to maturity than birth weight. The mortality for the weight ranges is indicated. 25–50% mortality4600 440042004000380036003400320030002800260024002200200018001600140012001000 800600400200 24 26 28 30 Preterm Term Post-term32 34 36 38 40 4290% 50% 10% 0 Week of gestationWeight (g) More than 50% mortalityLess than 4% mortality Appropriate for gestational age Small for gestational age4–25% mortalityLarge for gestational age
308
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Pre- and Postnatal develo Pment 214seCtIon 2 with these data must also take into account ethnicity and the nutritional and family history of that individual. Plotting the growth of children on such charts provides guidelines for the prediction of normal growth and indicates when investigation of possible growth anomalies should occur. Children who grow in an environment that does not constrain their growth exhibit a pattern of growth that is mainly parallel to a particular centile, a phenomenon that has been termed homeorhesis or canalization (following the same imaginary ‘canal’ on the growth chart). After deviation from this centile as a consequence of the adolescent growth spurt, most children return to the same centile position in adulthood, a finding that suggests that this pattern is genetically determined within individuals (Cameron 2002). For a comprehensive account of all aspects of postnatal growth, the reader is directed to consult Human Growth and Development (Cameron 2002). The WHO published growth charts for birth to 5 years and 5–19 years (2006). The charts were derived from children considered to be growing under optimal environmental conditions: they were breastfed, raised in environments that minimized poor diets and infection, and had non-smoking mothers. Similar growth charts have been published by WHO for specified regions (Mansourian et al 2012). ADOLESCENT GROWTH SPURT AND ADULT SIZE Growth charts reveal that body length increases from a neonatal range of 48–53 cm to about 75 cm during the first year after birth, and increases by 12–13 cm in the second year. Thereafter, 5–6 cm is added each year. In individual longitudinal growth curves, an increase in the velocity of growth occurs between 10.5 and 1 1 years in girls, and 12.5 and 13 years in boys. This rapid increase in growth is the adolescent growth spurt (Fig. 14.14; see Fig. 14.13). In both sexes, it lasts for 2–2.5 years. Girls gain about 16 cm in height during the spurt, with a peak velocity at 12 years of age. Boys gain about 20 cm in height (mostly by growth of the trunk), with a peak velocity at 14 years of age, during which time they may be growing at the rate of 10 cm a year. Humans seem to be the only species to have a long quiescent interval between the rapid growth that takes place immediately after birth and the adolescent growth spurt. This type of growth has been termed the Infancy/Childhood/Puberty (ICP) growth model (Stevens et al 2013). It has been suggested that the quiescent interval allows the brain to mature, and complex social learning to take place, before individuals pass through puberty and become sexually active. Apart from the expected change in growth velocity that occurs during puberty, it is now apparent that taller child height is prospectively associated with elevated risk of obesity (Johnson et al 2012, Van Dom - melen 2014). Growth in height continues at a slower rate after the adolescent growth spurt. Noticeable growth is said to stop at 18 years in females and 20 years in males; longitudinal studies have indicated an average figure of 16.25 years (females) and 17.75 years (males) with a normal variation of ±2 years (Harrison et al 1964). After this time, any increments that occur as a result of appositional growth at the cranial and caudal ends of the vertebral bodies and intervening intervertebral Fig. 14.11 A longitudinal study of growth. A, The height of de Montbeillard’s son (1759–1777) from birth to 18 years. B, A growth velocity curve, plotting increments in height from year to year. (After D’Arcy Thompson, On Growth and Form, 1942, Cambridge University Press.)Height (cm) Age (years)A B de Montbeillard’s son 1759–1777 60100 80120160 140180200 24 22 201816141210 86420 B 2 4 6 8 10 12 14 16 18 Height gain (cm per year) Age (years)de Montbeillard’s son 1759–1777 B 2 4 6 8 10 12 14 16 18 Fig. 14.12 Cross-sectional data showing growth. A, Growth in length in the prenatal and early postnatal periods. B, The corresponding velocity curve for this period. (After D’Arcy Thompson, On Growth and Form, 1942, Cambridge University Press.)cm Birth Months020 103050 40607080 2 4 6 8 10 12 14 16 18 20 22cm per month Birth Months04 2610 812 2 4 6 8 10 12 14 16 18 20 22A B
309
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
adolescent growth spurt and adult size 215 CHaPter 14 Fig. 14.13 Standard growth charts of British boys and girls showing the ninetieth, fiftieth and tenth centiles. (© Child Growth Foundation.)63 57514587 817569 39 33272115 93 0 0 2 4 6 8 10 12 14 16 1890% 50% 10% Age (years) BoysWeight (kg)150 140130120190 180170160 110 100 9080706050 0 2 4 6 8 10 12 14 16 1890% 50% 10% Age (years) BoysHeight (cm) 63 57514587 817569 39 33272115 93 0 0 2 4 6 8 10 12 14 16 1890% 50% 10% Age (years) GirlsWeight (kg)150 140130120190 180170160 110 100 9080706050 0 2 4 6 8 10 12 14 16 1890% 50% 10% Age (years) GirlsHeight (cm) discs are so small as to be difficult to measure. There is a loss of height after middle age. Development-related gene expression correlates with three phases of human growth (infancy/early childhood; childhood/puberty; final height); the expression of clusters of growth-related, evolutionarily con - served genes varies in a development-dependent manner in human tissues (Stevens et al 2013). The outcomes of the Human Genome Project have been linked to anthropomorphic traits in meta-analyses of genome-wide association data that have identified genetic variants at hundreds of loci relevant to growth in pre-puberty, puberty and adult - hood, particularly those associated with childhood and adult obesity (Genome-wide Investigation of ANThropometric measures – GIANT). A genome-wide association study has shown genetic loci linking pubertal timing, height and growth with childhood obesity (Cousminer et al 2013). The phenomenal growth rates of adolescence are most obvious in the increase in height. Weight gain is more variable. The birth weight is normally tripled by the end of the first year, and quadrupled by the end of the second year. Thereafter, weight increases by 2.25–2.75 kg annually until the adolescent growth spurt, when boys may add 20 kg to their weight and girls 16 kg. The peak velocity for weight gain lags behind the peak velocity for height by about 3 months. Body weight does not reach adult values until some time after adult height is attained. In recent years, the global estimates of the numbers of adults who are overweight or obese in adult life has exceeded the numbers who do not have sufficient food (Mascie-Taylor and Goto 2007).
310
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
Pre- and Postnatal develo Pment 216seCtIon 2 all organs and systems with time; it does not occur in the developing embryo, which displays differential rates of growth. Allometric growth describes the differences in the relative rates of growth between one part of the body and another, and is most clearly seen in the changes in body proportion between fetuses, neonates, children and adults. Between 6 and 7 weeks after fertilization, the head is nearly one-half of the total embryonic length. It subsequently grows proportionally more slowly and, at birth, it is one-quarter of the entire length. During child - hood, this pattern of growth continues with lengthening of the torso and limbs until, in adults, the head is one-eighth of the total length (Fig. 14.16). Growth of the liver, spleen, kidneys, skeletal and muscular tissues generally follows pre- and postnatal growth curves given for the INTEGRATION OF TYPES OF GROWTH DURING DEVELOPMENT AND LIFE In the later prenatal months and in the postnatal period, the various types of growth occur in differing patterns. The extent of tissue growth in organs depends on the specific duration of multiplicative growth for the cell types ( Fig. 14.15). Different cell populations complete their initial developmental proliferation and become differentiated at differ - ent times; the final stage of differentiation is usually the cessation of cell division. Growth of a body can be described in two ways: isometric and allo- metric. Isometric growth implies a progressive proportional increase in Fig. 14.14 Typical individual velocity curves for height: British boys and girls. (After Tanner JM, Whitehouse RH, Takaishi IM 1966 Standards from birth to maturity for height, weight, height velocity and weight velocity. Arch Dis Child 41:454–71, with permission from BMJ Publishing.) Height gain (cm per year) Girls Boys Age (years)04 2610 81216 141822 2024 2 4 6 8 10 12 14 16 18 Fig. 14.15 The duration of multiplicative growth for various human tissues. Prenatal months Postnatal years Duration of hyperplasia1 90 60 30 10 9 6 3 1 9 3 6Determinate tissues (number fixed near birth)Indeterminate tissues (number not fixed near birth) Birth MaturityNeurones Skeletal muscle fibres Seminiferous tubules Renal nephrons Heart muscle fibres Pulmonary alveoli Ovarian follicles Thyroid follicles Hepatocytes Exocrine acini Endocrine cells Blood cellsIntestinal villi Fig. 14.16 Allometric growth in humans. The head is very large in proportion to the rest of the body during the embryonic period. After this time, the head grows more slowly than the torso and limbs and, by adulthood, the head is only one-eighth of the body length. 9 16 Neonate 2 5 15 Adult Age in yearsWeeks of development
311
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader
217 CHaPter 14Key references entire body. Other tissues have very different growth rates; the brain, skull, lymphoid tissues and reproductive organs all show differing growth rates during childhood and adolescence (Kappelou et al 2006) (Fig. 14.17). Changes in growth at a tissue level are complex; differential growth rates during puberty can lead to temporal-specific associated pathology. For example, 30% of childhood fractures affect the radius, mostly the distal metaphysis, during the pubertal growth spurt, after which the incidence of these fractures decreases rapidly (Rauch 2012). Local dif - ferential rates of bone growth are seen in bone growth plates, peripheral and central trabecular bone, and in bone remodelling. During prepu - bertal growth, the distal radial growth plate adds about 9 mm to the length of the radius per year; the spaces between the peripheral trabecu - lae are filled with mineralized bone. It is suggested that incomplete trabecular coalescence during this rapid growth, rather than bone remodelling (which takes at least 6 months to complete), leads to lower bone mineral density, especially in boys, (which decreases rapidly after the pubertal growth spurt), and higher local metaphysial cortical por - osity (Wang et al 2010, Rauch 2012). Fig. 14.17 Growth curves of different tissues, regions of the body and systems. Note that the growth of lymphoid tissue, thymus, lymph nodes and intestinal lymph masses decreases after puberty. (Adapted with permission from Tanner JM 1962 Growth at Adolescence, 2nd ed. Oxford: Blackwell Publishing.)Size attained as percentage of total postnatal growthLymphoid Brain and head General Reproductive Age (years)040 2060100% 80120160 140180200% 2 4 6 8 10 12 14 16 18 20 KEY REFERENCES Barker DJP, Osmond C, Simmonds SJ et al 1993 The relation of head size and thinness at birth to death from cardiovascular disease in adult life. BMJ 306:422–6.A source of evidence for a link between birth weight and size and predictions of later cardiovascular pathology. Catalano PM, Hauguel-De Mouzon S 201 1 Is it time to revisit the Pedersen hypothesis in the face of the obesity epidemic? Am J Obstet Gynecol 204:479–87.This paper considers the changes in the demographics of obesity and its consequences for epigenetic changes to the feto-placental unit leading to generational chronic disease. Galan HL, Pandipati S, Filly RA 2008 Ultrasound evaluation of fetal bio - metry and normal and abnormal fetal growth. In: Callan PW (ed) Ultrasonography in Obstetrics and Gynecology. Philadelphia: Elsevier, Saunders; Ch. 7, pp. 225–65.This chapter gives an overview of dating a pregnancy, measuring growth and assessing fetal status. Godfrey KM, Barker DJ 2000 Fetal nutrition and adult disease. Am J Clin Nutr 71:1344S–52S.This paper considers the effect of maternal diet on the future risk factors for cardiovascular disease in her offspring. Harrison GA, Weiner JS, Tanner JM et al 1964 Human Biology. Oxford: Clarendon Press; Ch. 19.A presentation of data on the human growth curve.Lampl M 2002 Saltation and stasis. In: Cameron N (ed) Human Growth and Development. New York: Academic Press; Ch. 12.A presentation of the evidence for saltatory growth in humans. O’Rahilly R, Müller F 1987 Developmental Stages in Human Embryos. Washington: Carnegie Institution.The most comprehensive and detailed morphological account of human development during the first 8 weeks of life; it sets out the basis for the staging system used for human embryos. O’Rahilly R, Müller F 2010 Developmental stages in human embryos: revised and new measurements. Cells Tissues Organs 192:73–84.This paper presents the latest thoughts on matching stages to age and embryo length. Sepulveda W, Ximenes R, Wong AE et al 2012 Fetal magnetic resonance imaging and three-dimensional ultrasound in clinical practice: applica - tions in prenatal diagnosis. Best Pract Res Clin Obstet Gynaecol 26: 593–624.This paper gives examples of the clarity of 3D MRI. Uauy R, Casanello P, Krause B et al 2013 International Fetal and Newborn Growth Consortium for the 21st Century. Conceptual basis for prescrip - tive growth standards from conception to early childhood: present and future. BJOG 120 Suppl 2:3–8.This paper is part of the INTERGROWTH-21st study, which aims to develop prescriptive intrauterine and newborn growth standards to establish what is meant by ‘normal‘ fetal growth.Video 14.1 Ultrasound features of the fetus at 26 weeks.  Bonus  e-book  video
312
Gray's Anatomy
temp.pdf
https://ia802802.us.archive.org/30/items/GraysAnatomy41E2015PDF/Grays%20Anatomy-41%20E%20%282015%29%20%5BPDF%5D.pdf
PyPDF2TextLoader