Which of the following becomes the outer layer of skin, the nails, hair, and teeth?

Pharyngeal ectoderm and clefts

Surface ectoderm lines the roof of the embryonic pharynx from the outer margin up to and including the invagination of the adenohypophysial (Rathke’s) pouch (seeFigs 14.9–14.10). It also completely covers the first arch, including the lateral walls and floor of the stomodeum, unlike the more caudal arches, which are lined with endoderm. The external surface ectoderm of the first arch ultimately produces the keratinized stratified squamous epithelium of the epidermis, including hair follicles, sweat and sebaceous glands, and the specialized epithelium of the vermilion part of the lips. Within the oral cavity, first arch ectoderm forms the mucous membranes of the internal surface of the lips and cheeks, the palate, the presulcal part of the tongue, the epithelial components of the salivary and mucous glands, and the enamel organs of the developing teeth (seeFigs 17.15–17.16). The first pharyngeal cleft is obliterated ventrally; its dorsal end deepens to form the external acoustic meatus, including its ceruminous glands and the outer surface of the tympanic membrane (seeFig. 17.4). Developmental studies of gene expression in the pharyngeal arches in chick and mouse embryos are calling into question the traditional view of pharyngeal cleft development based on external appearance. In mouse embryos the external acoustic meatus forms more rostrally from an invagination of the ectoderm of the first arch rather than at the first pharyngeal cleft (Minoux et al 2013;Tucker 2017). Three auricular hillocks are observed on the first and second arches each side of the first pharyngeal cleft, during stages 15–16. It was thought that these contributed equally to the auricle (pinna) of the external ear, however, 3D reconstruction in human embryos up to stage 23 has shown that cells from the second arch form almost the entire auricle, whereas the first arch ectoderm and mesenchyme give rise to the tragus and anterior part of the external auditory meatus (Veugen et al 2019) (Ch. 16).

The external contours of the arches and clefts are modified as the skeletal and muscular elements develop. During stages 18–19 the second, third and fourth pharyngeal clefts form the rostral and caudal parts of a depression, the cervical sinus. As shown inFig. 17.4, the sinus is adjacent to extensions from the occipital myotomes that contribute to sternocleidomastoid, trapezius and platysma. Fusion of the hyoid arch with the cardiac elevation closes the cervical sinus, excluding the third, fourth and sixth arches from contributing to the skin of the neck and also results in platysma, which lies within the superficial fascia, extending along the neck to the ventral thoracic wall.

EED Recognizes a Number of Histone Motifs

The EED subunit of PRC2 is characterized by a prominent C-terminal WD40 beta-propeller, a closed solenoid of seven blades, each containing a four-stranded antiparallel beta sheet. Most beta-propellers possess remarkably similar folds despite low sequence homology and are typically used as scaffolds on which multisubunit protein complexes are assembled [50]. In addition, EED contains a small 80-residue subdomain at its N-terminus, and studies in Drosophila reveal that this subdomain interacts directly with histone H3 to ensure the trimethylation of H3K27 [51].

Developmental studies have shown that missense mutations in one of the WD40 repeats of EED result in an embryonic lethal phenotype, and reconstitution of these mutations in both mammalian and yeast two-hybrid systems disrupted binding between EED and Ezh2 [52]. Interestingly, the structural mediators of EED–Ezh2 interaction are conserved in Ezh2, as well as in the Drosophila homolog E(Z) [48].

There are currently 13 EED structures registered in the Protein Data Bank (PDB), many of them bound to ligands such as trimethylated nucleosomal peptides. These co–crystal structures provide valuable information on the contributions of EED to PRC2 biochemistry. For instance, a structure of EED bound to a 30-residue fragment of Ezh2 reveals that the Ezh2 peptide is wedged into a narrow binding groove along the “front” face of EED. This interaction is maintained by a stable network of hydrogen bonding and van der Waals interactions and contributes to the EED-Ezh2 pairing that is necessary for PRC2 methyltransferase activity [48].

In comparison, several trimethylated nucleosomal histone peptides (H1K26, H3K4, H3K9, H3K27, H4K20) are observed to bind across the rear face of the beta-propeller, with the trimethylated lysine side chains honing in on an aromatic cage at the EED surface [13]. The cage consists of three aromatic residues (Phe97, Tyr148, Tyr365) that interact with the terminal trimethyl ammonium group of the lysine through van der Waals and cation-π interactions. An additional fourth aromatic residue (Trp364) stabilizes the aliphatic portion of the lysine side chain. Of note, the conformations of the side chains in the EED residues involved in these different interactions are essentially the same in the various co–crystal structures (Fig. 8.3).

Interestingly, the Drosophila homolog of EED (ESC) is not required for PRC2 recruitment to the nucleosome. However, once bound, ESC appears to interact with histone H3 through its N-terminal domain and may be required to present the lysine substrate to the catalytic subunit in EZH2 [51]. A related model was proposed for the human complex, in which EED binding to trimethylated histone tails leads to an allosteric activation of PRC2 activity. Specifically, H3K27me3 was found to stimulate PRC2 methyltransferase activity against purified nucleosomes by approximately threefold [20].

In contrast, the addition of H1K26me3 to the reaction mixture inhibited the reaction, but resulted in the heterogeneous (mono-, di-, and tri-) methylation of four surface lysines on EED [20]. It is possible that H1K26me3 affects the overall PRC2 structure to preferentially accept EED as a substrate, but confirmation of this hypothesis will necessitate the structural determination of the PRC2 holoenzyme in complex with the requisite peptide ligands. The interplay between H3K27me3 and H1K26me3 binding comprises an attractive model for an enzymatic switch, and will likely act alongside the activities of histone lysine demethylases for the sensitive and specific control of PRC2 methyltransferase activity [53].

Surface Ectoderm Forms Epidermis

The surface ectoderm covering of the embryo consists initially of a single layer of cells. After neurulation in the fourth week, the surface ectoderm produces a new outer layer of simple squamous epithelium called theperiderm (Fig. 7.2A). The underlying layer of cells is now called thebasal layer and is separated from the dermis by the basement membrane containing proteins such as collagen, laminin, and fibronectin. The cells of the periderm are gradually sloughed into the amniotic fluid. The periderm is normally shed completely by the 21st week, but in some fetuses it persists until birth, forming a “shell” or “cocoon” around the newborn infant that is removed by the physician or shed spontaneously during the first weeks of life. These babies are calledcollodion babies.

In the 11th week, proliferation of the basal layer produces a newintermediate layer just deep to the periderm (Fig. 7.2B). This layer is the forerunner of the outer layers of the mature epidermis. The basal layer, now called thegerminative layer orthestratum germinativum, constitutes the layer of stem cells that will continue to replenish the epidermis throughout life. The cells of the intermediate layer contain thekeratin proteins characteristic of differentiated epidermis; therefore, these cells are calledkeratinocytes.

During the early part of the fifth month, at about the time that the periderm is shed, the intermediate layer is replaced by the three definitive layers of the outer epidermis: the innerstratum spinosum (orspinous layer), the middlestratum granulosum (orgranular layer), and the outerstratum corneum (orhorny orcornified layer) (Figs. 7.3, 7.4). This transformation begins at the cranial end of the fetus and proceeds caudally. The layers of the epidermis represent a maturational series: presumptive keratinocytes are constantly produced by the stratum germinativum, they differentiate as they pass outward to the stratum corneum, and, finally, they are sloughed from the surface of the skin.

The cells of the stratum germinativum are the only dividing cells of the normal epidermis. These cells contain a dispersed network of primary keratin (Krt) filaments specific to this layer, such as Krt5 and Krt14, and are connected by cell-to-cell membrane junctions calleddesmosomes. Together with adherens junctions, desmosomes provide a tight, impervious barrier resistant to water uptake or loss and infection. In addition, desmosomes help to distribute force evenly over the epidermis.

As the cells in the stratum germinativum move into the overlying stratum spinosum (4 to 8 cells thick; seeFig. 7.4), the Krt5 and Krt14 intermediate filaments are replaced by two secondary keratin proteins, Krt1 and Krt10. These are cross-linked bydisulfide bonds to provide further strength. In addition, cells in the stratum spinosum produce theenvelope protein, involucrin.

Embryonic ectoderm

When the ingression of cells through the primitive streak is complete, the epithelial cells remaining in the epiblast layer are termed embryonic 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 rostral half of the disc when the primitive streak first appeared, at which time the neural-fatedcells were closest to the streak, and the surface ectoderm cells were most rostral (seeFig. 10.3). The process of primary neurulation relocates most of the neuroepithelial cells (see below).

2.1 Extraembryonic Ectoderm

The ExE forms from the polar trophectoderm, and while this tissue has an extraembryonic fate, it also has a key role in patterning the adjacent epiblast and VE cell layers during early development. The role of this patterning was revealed through dissecting the early E5.5 embryo into its embryonic and abembryonic halves, which leads to the Epi-VE expressing markers of the DVE in an abnormally unrestricted manner (Rodriguez et al., 2005). This suggests that an inhibitory signal from the ExE normally prevents the differentiation of proximal VE cells toward this fate. BMP signaling from the ExE has been suggested to be the repressive signal that prevents DVE-like fate in the VE. Thus, in this model the DVE only forms at the distal tip of the egg cylinder, furthest from the repressive ExE signals (Rodriguez et al., 2005; Yamamoto et al., 2009).

In support of this idea, the ExE expresses BMPs such as BMP4 and BMP8b (Winnier et al., 1995; Ying & Zhao, 2001), and knockout experiments show that BMP4 plays an important role as mutants show a loss of expression of the mesodermal marker Brachyury and fail to gastrulate (Ben-Haim et al., 2006; Winnier et al., 1995). BMP in the ExE is maintained by NODAL signaling from the epiblast, and in a reciprocal manner BMP from the ExE is required for the correct development of the epiblast and formation of the primitive streak (Ben-Haim et al., 2006). NODAL is required for correct development of the epiblast as well as to induce the formation of the DVE (Brennan et al., 2001). The VE also expresses NODAL, and loss of its expression exclusively in this tissue, as shown through VE-specific knockout mice (Kumar et al., 2015), leads to a reduction of NODAL in the epiblast and subsequent developmental arrest (Kumar et al., 2015). Subsequently the DVE feeds back on NODAL in a negative manner through secreting inhibitors of the pathway including CER1 and LEFTY1 (Brennan et al., 2001). Thus, there is an initial reciprocal interaction between the VE and epiblast to maintain NODAL signaling in the epiblast, followed by repressive interaction when the epiblast is patterned.

Interestingly, the autoregulatory loop between the ExE and epiblast has been shown to require another interactive tissue, the ExE-VE (Uez et al., 2008), as knockouts of the Sall4-a isoform, a Spalt-like zinc finger transcription factor expressed exclusively in ExE-VE at E5.0, leads to loss of Bmp4 expression in the underlying ExE. This loss of Bmp4 then results in a loss of NODAL expression in the epiblast and a failure of DVE induction (Uez et al., 2008). Thus, a complex series of interactions exists between these tissues acting to both induce and maintain expression of Bmp4, and Nodal, and to induce the DVE at the distal tip that subsequently feeds back to pattern the epiblast by inhibiting Nodal and WNT.

Choristomas and cysts of the orbitocranial region

Ectoderm sequestered within the orbit during fetal development, or after injury, may lead to epithelial-lined cysts within the orbit. Most dermoid or epidermoid cysts (the latter without dermal elements) are noted in infancy, lie superotemporally on the orbital rim (Fig. 17.8A), enlarge slowly due to accumulation of epithelial debris within the lumen, and may become acutely inflamed due to leakage of sebaceous oil into the surrounding tissues; in contrast, deep orbital dermoids present later in life, often with an inflammatory episode (Fig. 17.8B) (Shields and Shields, 2004). Implantation cysts simulate congenital dermoid cysts, but do not respect the characteristic anatomical sites of the latter and generally present in patients with a past history of periocular trauma or surgery. All epithelial cysts should be removed (Fig. 17.8C), as this is curative and the cysts become extremely difficult to remove when chronic inflammation causes adherence to normal orbital structures.

Dermolipomas comprise adnexal structures sequestered on the conjunctival surface (Fig. 17.8D) and may be associated with outer canthal cleft, hemifacial microsomia, or Goldenhar syndrome (Shields and Shields, 2007). The abnormal epithelium and underlying fat may be cautiously excised, as the cutaneous hairs and sebaceous glands tend to cause a chronic conjunctivitis and are frequently unsightly.

Sinus mucoceles, most commonly frontoethmoidal in origin, may expand into the orbit, with leaking debris leading to orbital inflammation (Fig. 17.8E). Infrequently, the mucocele will discharge spontaneously through an upper-lid fistula (Rossman et al., 2007), or intracranially to form an extradural collection of mucus or pus. Severe compressive optic neuropathy may result both from fulminant acute orbital cellulitis and from chronic sphenoidal mucoceles. Acute orbital cellulitis due to sinus mucopyoceles requires high-dose intravenous antibiotic therapy and all retention mucoceles should be drained; this is often achievable via an endonasal route, although a superolateral orbital collection may require simultaneous direct orbital drainage. Where there is intracranial extension of the abscess, it is likely that neurosurgical intervention will be required.

Microphthalmos with cyst arises from incomplete closure of the fissure in the optic vesicle, with formation of a cyst below, or as part of, a microphthalmic globe. Small cysts can be left, but large cysts should be removed before they cause excessive expansion of the developing orbit.

Orbitocranial clefts may be associated with various degrees of intracranial herniation, forming meningoceles, encephaloceles (Fig. 17.9A), or (if separated from its cranial source) ectopic brain. Such anomalies usually present in childhood with a superomedial soft mass, may occur with neurofibromatosis as sphenoid wing dysplasia (Fig. 17.9B), and can be associated with optic disc colobomas in the “morning glory syndrome” (Fig. 17.9C) (Mahapatra and Agrawal, 2006).

1.4 Signals and transcriptional regulators involved in the development of the first and second BAs

The ectoderm and endoderm surrounding the arches produce multiple signaling factors, including FGFs, BMPs, Endothelins (EDNs), and Sonic Hedgehog, that communicate to the underlying mesenchyme—a mixture of mesoderm and neural crest cells (Fig. 3). Fgfs are expressed in the epithelial tissues of the BAs and are particularly enriched in the pouch endoderm and overlying ectoderm in the clefts between the arches (Crossley & Martin, 1995; Wall & Hogan, 1995). In the developmental context of the BAs, FGFs function as survival factors for neural crest entering the arches and as signals that induce expression of arch patterning transcription factors in mesenchyme and ectoderm.

One important function of signaling factors from the arch ectoderm is defining the dorsal–ventral signals that pattern postmigratory neural crest cells in the arches (reviewed in Medeiros & Crump, 2012). In order to establish the dorsal–ventral axis in BA1, dorsally expressed Fgf in the ectoderm is restricted by bone morphogenetic protein 4 (BMP4), expressed in the ventral portion of the BA1 ectoderm. Ectopic Bmp4 expression or implantation of a BMP4-soaked bead reduces Fgf8 expression in the BAs of chick embryos (Shigetani, Nobusada, & Kuratani, 2000). Interestingly, in the arches, BMP4 induces expression of its own antagonists, Noggin and Chordin (Stottmann, Anderson, & Klingensmith, 2001). In addition to maintaining the boundaries of ectodermal Fgf8 expression, BMP4 also plays important roles in dorsal–ventral patterning of BA1. A signaling network in which BMP4 and Endothelin 1 (EDN1) overlap to establish the dorsal–ventral axis is conserved from fish to mammals (Alexander et al., 2011; Ozeki, Kurihara, Tonami, Watatani, & Kurihara, 2004; Ruest, Xiang, Lim, Levi, & Clouthier, 2004). BMP4 promotes expression of ventral territory transcription factors in the arch mesenchyme and represses expression of dorsal transcription factors (Liu et al., 2005). Conditional deletion of Bmp4 from both mouse arch ectoderm and endoderm results in almost complete absence of the mandible and shift toward the midline of the tympanic ring and Meckel's cartilage (Liu et al., 2005).

There is evidence in zebrafish that Bmp4 induces ectodermal expression of Edn1 (Alexander et al., 2011). As the arches develop, factors induced by Bmp4 come to define the most ventral region of BA1, whereas Edn1 targets occupy a more intermediate domain. One key role of Edn1 signaling is to establish the nested expression pattern of Dlx transcription factors that delineate the dorsal–ventral axis of BA1 (Ozeki et al., 2004). Zebrafish and mouse Edn1-null mutants transform mandibular arch structures into structures normally derived from the maxillary process, with a mirror duplication observable in skeletal staining (Miller, Schilling, Lee, Parker, & Kimmel, 2000; Ozeki et al., 2004; Ruest et al., 2004).

A fourth signaling pathway, the Hedgehog pathway is also of critical importance for patterning the pharyngeal region. Sonic hedgehog (Shh) is expressed in the foregut endoderm prior to arch outgrowth (Wall & Hogan, 1995). Ablation of Shh-expressing ectoderm in chick embryos leads to the absence of BA1-derived structures; the defects can be rescued by implantation of Shh-soaked beads in the approximate location of the missing endoderm (Brito, Teillet, & Le Douarin, 2006). These ablation and bead rescue experiments demonstrate a role for Shh in establishing much of the ectodermal patterning required for arch development, including Fgf8 and Bmp4 expression (Haworth et al., 2007).

C. Ectoderm

Cranial ectoderm is a critical component of the craniogenesis machinery, contributing both directly and indirectly to the development of craniofacial structures. It has been suggested that the ectodermal field can be segregated into ectomeres. though the functional significance to these divisions remains to be determined.

The cranial surface ectoderm has been shown to play a vital role in the differentiation of CNC into skeletal tissue. In amphibian and chick studies. isolated CNC does not appear to self-differentiate in vitro to form primary cartilage, dentine or intramembranous bone (Thorogood, 1993; Hall, 1999). and there is compelling extirpation and tissue recombination evidence that focal (localized) epitheliomesenchymal tissue interactions have a fundamental role in leading ectomesenchyme to differentiate into these hard tissues (re-viewed by Hall, 1987, 1999; Lumsden, 1988; Thorogood, 1993). For example. extirpation of surface or neural ectoderm leads to an absence of associated dermal bones (Schowing, 1968a,b). Tissue recombination studies conducted by Hall and colleagues suggest that the basal lamina of a mitotically active ectoderm is necessary for CNC osteogenesis: the ectoderm need not be cephalic because limb bud, dorsal trunk, and periscleral epithelium can induce osteogenesis in mandibular ectomesenchyme (Tyler and Hall 1977; Bradamante and Hall, 1980; Hall, et al., 1983; Tyler, 1983). Thorogood and colleagues have extensively studied the chondrogenic promoting capacities of the collagen II-rich ECM secreted by the neurepithelium at the condensation sites of the chondrocranium (reviewed by Thorogood, 1993). Though a capacity as a direct inductor has been ruled out, this collagen II-rich ECM may play a role as part of a three-dimensional repository for ECM-sequestered chondrogenic factors (Bissell and Barcellos-Hoff, 1987; Thorogood, 1993; Hall, 1999). The potential to form cartilage, but apparently not intramembranous bone, may be informed while the CNC is still associated with the NT or soon thereafter (Hall and Tremaine, 1979; Bee and Thorogood, 1980; Hall, 1980a,b, 1999). It has further been demonstrated that the skeletogenic factors associated with the craniofacial ectoderm can be either matrix mediated or diffusible (Bee and Thorogood, 1980; Thorogood and Smith, 1984).

These studies further address an acknowledged but perhaps underappreciated aspect of craniofacial development: the temporal delay between the advent of the chondrocranium (the initial, cartilaginous embryonic cranial skeletal structures) and the dermatocranium (the perinatal skull that arises with the advent of the nascent dentition and the intramembranously ossified elements of the skull that develop around the chondrocranium). This is likely to be under the control of the ectoderm (Hall, 1987). It may be that the one is necessary for the development of the other, as is suggested by the work of Corsin (1966, 1975, 1977) showing the need of amphibian B.A dermatocranium for the presence of the viscerocranium. The embryonic skull has an unique set of functional demands for which the cartilaginous skull is best fit to fulfill, and likewise for the perinatal dermatocranium built around the chondrocranium (Hanken and Thorogood, 1993; Presley, 1993). Clearly it is advantageous to temporally regulate the onset of each (Presley, 1993). How this is achieved is not yet clear. Thus, the skeletogenic promoting factors of the embryonic skull are not uniform with regard to mechanism, and the diversity in cellular fate is due to the influences of other cells and the extracellular environment.

Significantly, regulation of these events in BA1—where cartilage, membrane bone, and teeth all form—must be highly regulated spatially and temporally. For example, teeth develop on the rostral (oral) surface of the developing mandibular primordium, whereas bone and cartilage develop more caudally (aborally). The different positional fates of these CNC cells must be determined early in the formation of the primordium. Expression of the closely related LIM domain homeobox genes Lhx6 and Lhx7 is restricted to oral ectomesenchyme of the mandibular and maxillary processes and complements that of Gsc, which is expressed in aboral ectomesenchyme (Tucker et al., 1999). The ectoderm appears to be involved in inducing both oral and aboral mesenchymal gene expression. The ectoderm expresses a wide range of signaling molecules, including FGFs. BMPs, WNTs, and HHs. and it is the restriction of Fgf8 expression to the oral (and pericleftal) ectoderm that appears to set up the anterior-posterior (AP) axis of BA1 (Grigoriou, et al., 1998; Trumpp, et al., 1999). The restriction of Gsc expression to aboral mesenchyme involves repression by Lhx6/7 expressing cells. although the mechanism that restricts Lhx6/7 expression to oral mesenchyme is independent of Gsc and is more probably related to the distance from the source of FGF8. Targeted mutations in Lhx6 or Lhx7, however, do not result in dental defects: defects may only be revealed when these mutations are combined (Zhao et al., 1999; V. Pachnis, personal communication). Mutations in Gsc do, however, lead to mandibular bone defects (see below) but the teeth develop normally (Rivera-Pérez, et al., 1995; Yamada et al., 1995). Endothelin-1 expression in the entire mandibular epithelium appears to act as a maintenance factor for Gsc expression (Tucker et al., 1999) and it is regulated in part by FGF8 (Trumpp et al., 1999): as with Gsc, both Endothelin-1 and Endothelin receptor A knock-outs have mandibular defects where bone is affected but dental development essentially is not (Kurihara et al., 1994; Clouthier et al., 1998). More specific roles for the ectoderm in dental development are discussed later in this chapter.

The expression of signaling molecules in the BA ectoderm has been shown to be independent of the NC, such that when the NC is ablated the expression of signaling molecules still comes on in a defined spatially restricted pattern (Veitch et al., 1999). Thus, it would appear that although the skeletal identity of an arch might be determined by the NC cells, the AP polarity within an arch is determined by the ectoderm. This perhaps explains why, when R1 and R2 (which both produce crest destined for the BA1) are rotated, no change in first arch AP pattern is seen (Noden, 1983b; unpublished observations). This may also explain why, when mesencephalic CNC is grafted caudally (Noden, 1983b), or in the Hoxa2 knock-out (Rijli et al., 1993; Gendron Maguire et al., 1993), the duplicated first arch elements have a mirror image symmetry. In addition to expressing signaling molecules involved in local epithelial-mesenchymal interactions. BA ectoderm also expresses a range of transcription factors, including Hox genes. Hoxa2, for example is expressed in the ectoderm of BA2 as well as its CNC. Such ectodermal expression comes on after the NC has migrated and was initially thought to be induced by the crest (Hunt et al., 1991a). However. the ectoderm's Hox code has now been shown to be independent of the NC: When non-Hox-expressing crest replaces Hox-expressing crest, the ectoderm still turns on its normal Hox code (Couly et al., 1998). Arch identity may therefore come from a combination of patterning information from the CNC and patterning information in the tissues into which it migrates, possibly explaining why Hox- expressing NC appears unable to form cartilage elements when placed in a normally Hox-devoid setting (Couly, et al., 1998; Grammatopoulos et al., 2000). Hence, the ectoderm may impart positional information and the CNC its interpretation, necessitating rigorous investigation of the proximate regulators of ectodermal gene expression.

Ectoderm

The ectoderm gives rise to the skin, the brain, the spinal cord, subcortex, cortex and peripheral nerves, pineal gland, pituitary gland, kidney marrow, hair, nails, sweat glands, cornea, teeth, the mucous membrane of the nose, and the lenses of the eye (see Fig. 5.3). The part of the ear where these organs are found represented is served by a branch of plexus cervicalis which has indirect connections with the cortex. In the ear the organs are represented in the lobe and the tail of the helix.