Patterning of the Dentition

Each human tooth has a distinctive morphology, and each type of tooth forms in a characteristic location. For many years, virtually nothing was known about the patterning of the dentition, but analysis of certain types of genetically modified mice has provided some concrete clues to the molecular basis of dental patterning. Both the dental field overall and the patterning of the dentition take shape very early in craniofacial development, before any overt indication of tooth formation. The expression of Pitx-2 (a bicoid-related transcription factor; see Fig. 4.1) outlines first the entire ectodermal dental field, and later the epithelium of the individual tooth germs. The homeobox-containing genes Dlx-1 and Dlx-2 are expressed in the maxillary arch and in the proximal part of the mandibular arch. When both these genes are knocked out in mice, the upper jaw develops without molar teeth, although molars develop in the lower jaws. The incisor teeth in both jaws develop normally. Another homeobox-containing gene, Barx-1, is induced by FGF-8 in proximal ectoderm of the mandibular process, and in the formation of molar teeth in knockouts it may compensate for the absence of Dlx-1 and Dlx-2 in the lower jaw. FGF-8 acts proximally to restrict Barx-1 and Dlx-2 to guide the molar-producing domain, and BMP-4 acts distally to activate Msx-1 and Msx-2 in guiding the formation of incisor teeth.

Not only the location, but also the type of tooth is under tight developmental control. The difference in morphology between an incisor tooth, possessing a single cusp, and a molar tooth, which contains several cusps, is related to the number of enamel knots (see later) in the developing tooth. A striking example of the molecular control of dental patterning is the conversion of incisor teeth to molars in mice. In the distal part of the mandibular arch, ectodermal BMP-4 signals normally repress the expression of Barx-1, but when BMP-4 signaling is inhibited by the implantation of noggin beads, Barx-1 expression is induced in the dental mesenchyme, and the developing incisor teeth are transformed into molars. The transcription factor Islet-1 is expressed only in the oral surface ectoderm in the area where incisors will form. In contrast, Pitx-1 is expressed only in the molar region of the mandibular mesenchyme where it acts upstream of Barx-1.

Mammals have only a single row of teeth in each jaw. This is controlled by two overlapping gradients of opposite polarity along the lingual-buccal axis (Fig. 14.20). On the buccal side of the jaw, a high concentration of BMP-4 stimulates the expression of Msx-1 in the dental mesenchyme in the normal process of tooth development, thus accounting for the development of the normal teeth. On the lingual side of the jaw, a high concentration of the transcription factor Osr-2 (the mammalian equivalent of the Drosophila pair-rule gene odd skipped; see Fig. 4.1) inhibits the BMP-Msx axis and consequently tooth formation in that area. When Osr-2 is inactivated, BMP-4–Msx-1 activity on the lingual side of the jaw is not inhibited, and supernumerary teeth form on the lingual side of the normal row of teeth (see Fig. 14.20B). Enhancing or inhibiting the function of many of the other genes involved in tooth development can also lead to the formation of supernumerary teeth, but these genes do not cause expansion of the entire dental field as does Osr-2.

Stages of Tooth Development

Tooth development begins with the migration of neural crest cells into the regions of the upper and lower jaws. The overlying oral ectoderm thickens into C-shaped bands (dental laminae) in the upper and lower jaws. The appearance of the dental laminae during the sixth week is the first manifestation of a series of ectodermal-mesenchymal interactions that continues until tooth formation is virtually completed.

Although each tooth has a specific time sequence and morphology of development, certain general developmental stages are common to all teeth (Fig. 14.21). As the dental lamina grows into the neural crest mesenchyme, epithelial primordia of the individual teeth begin to take shape as tooth buds. In keeping with their interactive mode of development, the tooth buds are associated with condensations of mesenchymal cells. The tooth bud soon expands, passing through a mushroom-shaped cap stage before entering the bell stage (see Fig. 14.21D).

By the bell stage, the tooth primordium already has a complex structure, even though it has not formed any components of the definitive tooth. The epithelial component, called the enamel organ, is still connected to the oral epithelium by an irregular stalk of dental lamina, which soon begins to degenerate. The enamel organ consists of an outer sheath of epithelium, a mesenchymelike stellate reticulum, and an inner epithelial ameloblast layer. Ameloblasts are the cells that begin to secrete the enamel of the tooth. The initial formation of ameloblasts depends on the actions of the transcription factor Tbx-1. Within the concave surface of the enamel organ is a condensation of neural crest mesenchyme called the dental papilla. Cells of the dental papilla opposite the ameloblast layer become transformed into columnar epithelial cells called odontoblasts (Fig. 14.22). These cells secrete the dentin of the tooth. Attached to the dental lamina close to the enamel organ is a small bud of the permanent tooth (see Fig. 14.21E and F). Although delayed, it goes through the same developmental stages as the deciduous tooth.

Late in the bell stage, the odontoblasts and ameloblasts begin to secrete precursors of dentin and enamel, beginning first at the future apex of the tooth. Over several months, the definitive form of the tooth takes shape (see Fig. 14.21). Meanwhile, a condensation of mesenchymal cells forms around the developing tooth. Cells of this structure, called the dental sac, produce specialized extracellular matrix components (cementum and the periodontal ligament) that provide the tooth with a firm attachment to the jaw. While these events are occurring, the tooth elongates and begins to erupt through the gums (gingiva).

Tissue Interactions in Tooth Development

Teeth are formed by a series of inductive interactions. Tissue recombination experiments have shown that the thickened ectoderm of the dental lamina initiates tooth formation. Early in development, dental ectoderm can induce nondental cranial neural crest mesenchyme to participate in the formation of a tooth, but predental neural crest mesenchyme cannot induce nondental ectoderm to form a tooth. Research suggests that the transcription factor Lef-1 may stimulate the predental surface ectoderm to secrete FGF-8, which induces the underlying mesenchyme to express Pax-9 (Fig. 14.23). In the absence of Pax-9 expression, tooth development does not proceed past the bud stage. BMP-2 and BMP-4, also produced by the surface ectoderm, inhibit the action of FGF-8. Investigators have suggested that such inhibition is the basis for the non–tooth-forming spaces between the developing teeth. More recent research has shown that lateral inhibition through the Delta/Notch system is also involved in dental spacing. Another characteristic transcription factor induced in the mesenchyme beneath the thickened dental lamina is Msx-1.

Slightly later in dental development, BMP-4, instead of functioning as an inhibitor, acts along with FGF-8 and shh to stimulate the mesenchyme of the tooth bud to express a variety of characteristic molecules, including the following: the transcription factors Msx-1, Msx-2, and EGR-1; the extracellular matrix molecules tenascin and syndecan; and BMP-4. If the overlying ectoderm is separated from the neural crest mesenchyme, the predental mesenchyme will not develop into a dental papilla, and a tooth will not form. The role of BMP-4 as an inducer was shown by adding a small bead soaked in BMP-4 to a small mass of cultured predental neural crest mesenchyme (see Fig. 14.23B). Under the influence of the BMP-4 released from the bead, the mesenchyme began to express Msx-1, Msx-2, Egr-1, and BMP-4. The mesenchyme did not produce tenascin or syndecan, however, thus showing that other signals in addition to BMP-4 are required to achieve the full inductive response.

With the initial induction of the dental mesenchyme, that tissue itself becomes the next prime mover in tooth development. Inductive signals emanating from the dental mesenchyme next act on the ectoderm of the dental ledge, now in the late bud to early cap stage. Recombination experiments have shown that the dental mesenchyme determines the specific form of the tooth. When molar mesenchyme is combined in vitro with incisor epithelium, a molar tooth takes shape, whereas combining incisor mesenchyme with molar epithelium results in the formation of an incisor.

Under the inductive influence of the dental mesenchyme, now called the dental papilla, a small group of ectodermal cells at the tip of the dental papilla ceases dividing. This mass of cells, called the enamel knot (Fig. 14.24), serves as a signaling center that regulates the shape of the developing tooth. Through the production of several signaling molecules, including shh, FGF-4, and BMP-2, BMP-4, and BMP-7, the enamel knot stimulates the proliferation of cells in the dental cap down and away from itself. By serving as a fixed point in this process, the enamel knot determines the site of the tip of a cusp in the developing tooth. In the case of molar teeth, which have multiple cusps, secondary enamel knots form—one for each cusp. The location and spacing of secondary enamel knots are determined by two molecules, both induced by BMPs. p21 is strongly expressed at sites where secondary enamel knots form, and ectodin is expressed in the intervening spaces. In the absence of ectodin, the secondary enamel knots and the resulting cusps of the molar teeth become massive because of the lack of its restraining influence. Ultimately, the cells of the enamel knot undergo apoptosis, possibly under the influence of BMP-4, which stimulates cell death in several other developmental systems. The appearance of apoptosis terminates the inductive signaling from this structure.

Formation of Dentin and Enamel

Late in their differentiation, odontoblasts withdraw from the cell cycle, elongate, and start secreting predentin from their apical surfaces, which face the enamel organ. The production of predentin signals a shift in the patterns of synthesis from type III collagen and fibronectin to type I collagen and other molecules (e.g., dentin phosphoprotein, dentin osteocalcin) that characterize the dentin matrix. The first dentin is laid against the inner surface of the enamel organ at the apex of the tooth (see Fig. 14.22B). With the secretion of additional dentin, the accumulated material pushes the odontoblastic epithelium from the odontoblast-ameloblast interface.

Terminal differentiation of the ameloblasts occurs after the odontoblasts begin to secrete predentin. In response to inductive signals from the odontoblasts, the ameloblasts withdraw from the cell cycle and begin a new pattern of synthesis, thus producing two classes of proteins: amelogenins and enamelins. About 5% of enamel consists of organic matrix, and amelogenins account for about 90% of this, with enamelins constituting most of the remainder. Enamelins are secreted before the amelogenins, and they may serve as nuclei for the formation of crystals of hydroxyapatite, the dominant inorganic component of enamel.

The amelogenin genes are located on the X and Y chromosomes in humans. Enamel genes have been highly conserved during vertebrate phylogeny. Investigators have suggested that, in early vertebrates, enamel once served as part of an electroreceptor apparatus.

Tooth Eruption and Replacement

Each tooth has a specific time of eruption and replacement (Table 14.1). With the growth of the root, the enamel-covered crown pushes through the oral epithelium. The sequence of eruption begins with the central incisor teeth, usually a few months after birth, and continues generally stepwise until the last of the deciduous molars forms at the end of the second year. A total of 20 deciduous teeth are formed.

Meanwhile, the primordium of the permanent tooth is embedded in a cavity extending into the bone on the lingual side of the alveolar socket in which the tooth is embedded (see Fig. 14.21F). As the permanent tooth develops, its increasing size causes resorption of the root of the deciduous tooth. When a sufficient amount of the root is destroyed, the deciduous tooth falls out, leaving room for the permanent tooth to take its place. The sequence of eruption of the permanent teeth is the same as that of the deciduous teeth, but an additional 12 permanent teeth (for a total of 32) are formed without deciduous counterparts.

The formation and eruption of teeth are important factors in midfacial growth, much of which occurs after birth. Tooth development and the corresponding growth of the jaw to accommodate the teeth, along with the development of the paranasal sinuses, account for much of the tissue mass of the midface.

Clinical Correlation 14.2 presents malformations of the teeth and dentition.

Clinical Correlation 14.2

Dental Anomalies

Abnormal Tooth Number

Hypodontia

Hypodontia or congenital absence of teeth can be isolated (nonsyndromic) or associated with certain genetic syndromes. Isolated hypodontia can be caused by mutations in the MSX1, EDA, AXIN2, PAX9, and WNT10A genes. Mutations in the EDA gene, which encodes ectodysplasin (see p. 162), cause alterations in its signaling pathway, which subsequently results in X-linked hypohidrotic ectodermal dysplasia (XLHED) or isolated hypodontia. Characteristic features of patients with XLHED consist of missing or malformed teeth, missing and sparse hair, and absent or dysfunctional exocrine glands. The normally soluble ligand ectodysplasin binds to the membrane-bound receptor Edar, and through its signaling pathway it activates transcription of its target genes. EDA signaling is known to suppress bone morphogenetic protein-4 (BMP-4) activity and upregulate sonic hedgehog (shh), both of which have crucial roles in tooth formation. Therefore, it is understandable that mutations of EDA cause missing or malformed teeth.

This disorder is X-linked; thus, most patients are male (Fig. 14.25A). Sporadic cases have been reported, but frequently the mutations are inherited from mothers who are heterozygous. The heterozygous mothers are usually normal or have mild manifestations, including microdontia and hypodontia (Fig. 14.25B).

Hyperdontia (Supernumerary Teeth)

Hyperdontia, an excess number of teeth, can be isolated or associated with several genetic syndromes, including cleidocranial dysplasia (see Fig. 9.24) and trichorhinophalangeal syndrome (TRPS 1), which can be caused by mutations in TRPS1. TRPS 1 is a strong transcriptional repressor protein, and its mutation has been hypothesized to cause gain of function and to be responsible for supernumerary teeth (Fig. 14.26) and mandibular prognathism.

TRPS 1 binds to the promoter of RUNX2. All lines of evidence suggest that TRPS 1 and RUNX-2 share the same pathway and that TRPS 1 acts as a repressor of RUNX-2. In mice, the coexpression of Runx2 and Trps 1 has been demonstrated in developing bone (see Fig. 9.17) and in dental mesenchyme during early tooth development.

Abnormal Tooth Size and Shape

Smallest Teeth

Microdontia (small teeth) can be isolated or associated with genetic syndromes. The most commonly affected teeth are the maxillary permanent lateral incisors. They may be of normal shape or peg-shaped. One of the most common causes of microdontia appears to be mutations of the MSX1 gene. Such mutations can also lead to hypodontia and orofacial clefting. The smallest known teeth have been reported in patients with microcephalic osteodysplastic primordial dwarfism type II (MOPD II; Fig. 14.27). Patients with this syndrome have severe prenatal and postnatal growth retardation, with a relatively proportionate head size at birth, but extreme microcephaly in adulthood. The striking dental anomalies consist of extremely small teeth, opalescent and abnormally shaped teeth, and rootless molars. Teeth are spontaneously exfoliated because the alveolar bone is severely hypoplastic. MOPD II, a very rare autosomal recessive disorder, is caused by mutations in PCNT. PCNT encodes the centrosomal protein pericentrin, a protein crucial for cell division. This gene is expressed in both the epithelium and the mesenchyme during early tooth development. The clinical phenotype, including the small size of the teeth, is the consequence of loss of microtubule integrity, which leads to defective centrosome function.

Largest Teeth

Macrodontia, the condition of having larger than normal teeth, is an extremely rare condition. Macrodontia of the maxillary permanent central incisors is typical in patients with KBG syndrome, which is characterized by intellectual disability, skeletal malformation, and macrodontia. It is caused by mutations in ANKRD11, whose protein plays an important role in neural plasticity. Generalized macrodontia is extremely rare. This syndrome has been reported in patients with Ekman-Westborg-Julin syndrome or multiple macrodontic multituberculism. Patients with this syndrome have the largest teeth ever reported (Fig. 14.28). The molecular origin of this syndrome is unknown.

Molarized Incisors

True transformation of incisors to molars is an extremely rare condition (Fig. 14.29). Misexpression of Barx-1 in mice by the inhibition of BMP signaling can result in the formation of molar teeth instead of incisors (see p. 310). This finding provides strong support for a role of homeobox genes in controlling tooth type.

Abnormal Tooth Structure

Abnormal Dentin: Dentinogenesis Imperfecta

Abnormalities of dentin include dentinogenesis imperfecta and dentin dysplasia. Teeth with dentinogenesis imperfecta appear blue-gray or amber brown and are opalescent. Dentinogenesis imperfecta can be nonsyndromic or associated with osteogenesis imperfecta (Fig. 14.30). Nonsyndromic dentinogenesis imperfecta is caused by mutations in the DSPP (dentin sialophosphoprotein) gene, which encodes dentin sialoprotein, a noncollagenous protein of dentin.

Dentinogenesis imperfecta, which is associated with osteogenesis imperfecta, is caused by mutations in the type I collagen genes COL1A1 or COL1A2. Osteogenesis imperfecta is a heterogeneous group of heritable connective tissue disorders, caused by abnormal type I collagen synthesis. Characteristic features include increased bone fragility, bone deformities (see Fig. 14.30A and B), joint hyperextensibility, blue sclerae, hearing loss, and dentinogenesis imperfecta.

Abnormal Enamel: Amelogenesis Imperfecta

Amelogenesis imperfecta is a group of clinically and genetically heterogeneous disorders that affect the development of enamel and result in abnormalities of the amount, composition, and/or structure of enamel. These disorders are caused by mutations in a variety of genes that are important for enamel formation. The enamel may be hypoplastic, hypomature, or hypocalcified (Fig. 14.31). Mutations in several genes, including ENAM, AMEL, DLX3, and P63, are known to cause isolated or syndromic amelogenesis imperfecta.

Dental Fluorosis

Excessive fluoride consumption during tooth formation can cause enamel fluorosis, which ranges from white spots or lines in the enamel to enamel hypoplasia (Fig. 14.32). The white opaque appearance of fluorosed enamel is caused by a hypomineralized enamel subsurface. The changes in enamel are related to cell-matrix interactions when the teeth are forming. At an early stage of enamel maturation, the amount of amelogenin in the enamel matrix of fluorosed enamel is very high, thus resulting in a delay in the removal of amelogenin by proteinases as the enamel matures. This subsequently causes abnormal mineralization of the enamel.

Investigators have demonstrated a genetic component to dental fluorosis susceptibility. This means that some people are more prone to have enamel fluorosis than others. Some people with excessive fluoride consumption may have skeletal fluorosis, but fluorosed bone disappears later in life because bones remodel, whereas enamel does not. Therefore, enamel fluorosis is permanent.

Tetracycline-Stained Teeth

Ingestion of tetracyclines during tooth development causes abnormal tooth formation (see Table 8.7). The discoloration of teeth is permanent and ranges from yellow or gray to brown (Fig. 14.33B). It appears fluorescent under ultraviolet light (Fig. 14.33A). The discoloration is permanent, and the degree of discoloration depends on the dose and type of drug. Tetracyclines are incorporated into tissues that are calcifying at the time of their administration. They are able to chelate with calcium ions and form a tetracycline-calcium orthophosphate complex in teeth, cartilage, and bone that results in discoloration of the primary and permanent teeth.

By Piranit N. Kantaputra, Chiang Mai University, Thailand.

Pharyngeal Arches

As mentioned earlier, the endoderm of the foregut is the main driver in organizing development of the pharynx. Under the influence of various concentrations of retinoic acid (see p. 297), combinations of Hox genes determine the craniocaudal identity of the segments of the pharynx. The first pharyngeal arch develops independently of Hox genes, whereas expression of Hoxa-2 and Hoxa-3 is required for formation of the second and third pharyngeal arches. Similar to the pharyngeal arches, the pharyngeal pouches, as well as having individual identities, are characterized by highly regionalized molecular patterns in the dorsoventral and the craniocaudal axes. Signals from the pharyngeal endoderm pattern the pharyngeal arches even before cranial neural crest cells enter the arches (see Fig. 14.5). Tbx-1 expression in the early pharyngeal endoderm influences FGF-8 signaling, and in its absence pharyngeal pouches do not form normally, and a sequence of malformations reminiscent of DiGeorge’s syndrome occurs.

In addition to being packed with mesenchyme (mainly of neural crest origin except for the premuscle mesoderm, which migrates from the somitomeres), each pharyngeal arch is associated with a major artery (aortic arch) and a cranial nerve (see Fig. 14.34). Each also contains a central rod of precartilaginous mesenchyme, which is transformed into characteristic adult skeletal derivatives. Understanding the relationship between the pharyngeal arches and their innervation and vascular supply is very important, because tissues often maintain their relationship with their original nerve as they migrate out or become displaced from their site of origin in the pharyngeal arch system.

The first pharyngeal arch (mandibular) contributes mainly to structures of the face (both mandibular and maxillary portions) and ear (see Fig. 14.34). Its central cartilaginous rod, Meckel’s cartilage, is a prominent component of the embryonic lower jaw until it is surrounded by locally formed intramembranous bone, which forms the definitive jaw. During later development, the distal part of Meckel’s cartilage undergoes resorption because of extensive apoptosis of the chondrocytes. More dorsally, Meckel’s cartilage forms the sphenomandibular ligament, the anterior ligament of the malleus, and the malleus (Fig. 14.36). In addition, the incus arises from a primordium of the quadrate cartilage. The first-arch musculature is associated with the masticatory apparatus, the pharynx, and the middle ear. A common feature of these muscles is their innervation by the trigeminal nerve (cranial nerve V).

The molecular basis for development of the first pharyngeal arch is quite different from that of the other arches, starting with the origin of the neural crest that populates it. The neural crest cells that populate the first arch are derived from rhombomeres 1 and 2 and midbrain, which are anterior to the expression domain of the Hox genes. Cellular precursors of the first-arch mesenchyme are instead associated with the expression of Otx-2. The role of signaling molecules and transcription factors, such as Dlx and Msx, is discussed earlier (see pp. 299-301).

The second pharyngeal arch (hyoid) also forms a string of skeletal structures from the body of the hyoid bone to the stapes of the middle ear. These structures are derived from Reichert’s cartilage, which, instead of being a single rod, is actually two cartilaginous rodlike elements with a strand of mesenchymal tissue connecting them. Although much of the second-arch mesoderm migrates to the face to form the muscles of facial expression, additional muscles become associated with other second-arch skeletal derivatives; a good example is the stapedius muscle, which is associated with the stapes bone. These second-arch muscles are innervated by the facial nerve (cranial nerve VII).

Patterning of the second pharyngeal arch is strongly influenced by the homeobox gene Hoxa2. When this gene is knocked out in mice, skeletal derivatives of the second arch fail to form. The second arch in such mutant animals contains mirror-image duplicates of many of the proximal bones of the first-arch skeleton. This may be a result of the response of the mutant second-arch mesenchyme to an ectodermal signal from the first pharyngeal cleft that plays a role in patterning the first arch. The formation of mirror-image first-arch structures is analogous to the formation of mirror-image supernumerary limbs after transplantation of the zone of polarizing activity in the limb bud (see p. 198). The finding that only proximal structures are affected reflects the different genetic control of proximal and distal segments of the arches.

The third and fourth pharyngeal arches are otherwise unnamed. The third arch gives rise to structures related to the hyoid bone and upper pharynx. The third-arch skeleton becomes the greater horn of the hyoid bone. The one muscular derivative (stylopharyngeus) of the third arch is innervated by the glossopharyngeal nerve (cranial nerve IX). The fourth arch gives rise to certain muscles and cartilages of the larynx and lower pharynx. The muscles are innervated by the vagus nerve (cranial nerve X), which also grows into the thoracic and abdominal cavities.

Pharyngeal Grooves

The first pharyngeal groove is the only one that persists as a recognizable adult structure: the external auditory meatus. Grooves II and III become covered by the enlarged external portion of the second arch (a phylogenetic homologue of the operculum [gill cover] of fish). The enlargement of the second arch is caused by the presence of a signaling center in the ectoderm at its tip; such a signaling center is not present in arches 3 or 4. As in the facial primordia, this signaling center produces shh, FGF-8, and BMP-7, which stimulate growth of the underlying mesenchyme. During the period of their overshadowing by the hyoid (second) arch, grooves II and III are collectively known as the cervical sinus (Fig. 14.37; see Fig. 13.27). As development progresses, the posterior ectoderm of the second arch fuses with ectoderm of a swelling (cardiac swelling) just posterior to the fourth arch, thus causing the cervical sinus to disappear and the external contours of the neck to become smooth.

Thyroid

Development of the thyroid gland begins with local mesodermal inductive signals acting on the ventral endoderm of the foregut. This process results in the specification of a small number of endodermal cells (as few as 60 in the mouse) to be destined to a thyroid fate. These founder cells, the thyroid anlage, increase in number to form a thickened placode that soon begins to extend into the surrounding mesodermal mesenchyme, at which point it is called a thyroid bud. These cells are characterized by the expression of four transcription factors (Hhex, Nkx2-1, Pax-8, and Foxe-1), which operate together in a complex interacting pattern and all of which are required for further thyroid development.

The unpaired primordium (thyroid bud) of the thyroid gland appears in the ventral midline of the pharynx between the first and second pouches (see Fig. 14.37). Starting during the fourth week as an endodermal thickening just caudal to the median tongue bud (tuberculum impar), the thyroid primordium soon elongates to form a prominent downgrowth called the thyroid diverticulum. The pathway of caudal extension of the bilobed thyroid diverticulum is determined by the pattern of arteries in the neck and extension and continues during pharyngeal development. During its caudal migration, the tip of the thyroid diverticulum expands and bifurcates to form the thyroid gland itself, which consists of two main lobes connected by an isthmus. For some time, the gland remains connected to its original site of origin by a narrow thyroglossal duct.

By about the seventh week, when the thyroid has reached its final location at the level of the second and third tracheal cartilages, the thyroglossal duct has largely regressed. Nevertheless, in almost half the population, the distal portion of the thyroglossal duct persists as the pyramidal lobe of the thyroid. The original site of the thyroid primordium persists as the foramen cecum, a small blind pit at the base of the tongue.

The thyroid gland undergoes histodifferentiation and begins functioning early in embryonic development. By 10 weeks of gestation, follicles containing some colloid material are evident, and a few weeks thereafter, the gland begins to synthesize noniodinated thyroglobulin. Secretion of triiodothyronine, one of the forms of thyroid hormone, is detectable by late in the fourth month.

Hypophysis

The hypophysis (pituitary gland) develops from two initially separate ectodermal primordia that secondarily unite. One of the primordia, called the infundibular process, forms as a ventral downgrowth from the floor of the diencephalon. The other primordium is Rathke’s pouch, a midline outpocketing from the stomodeal ectoderm that extends toward the floor of the diencephalon as early as the fourth week. An inductive event from the overlying diencephalon, mediated first by BMP-4 and then by FGF-8, stimulates the formation of a Rathke’s pouch primordium in the dorsal stomodeal ectoderm and provides cues for molecular events that stimulate cell proliferation. Through the action of Hesx-1 (homeobox gene expressed in embryonic stem cells, formerly called Rpx) and the Lim-type homeobox-containing genes Lhx3 and Lhx4, Rathke’s pouch primordium forms the definitive Rathke’s pouch (Fig. 14.38). In the early embryo, the cells in Rathke’s pouch originate in the anterior ridge of the neural plate.

The infundibular process is intimately related to the hypothalamus (see Fig. 1.15), and certain hypothalamic neurosecretory neurons send their processes into the infundibular process, which ultimately becomes the neural lobe of the hypophysis. Throughout development, the histological structure of the infundibulum retains a neural character.

As development proceeds, Rathke’s pouch elongates toward the infundibulum (see Fig. 14.38). While its blind end partially enfolds the infundibulum like a double-layered cup, the stalk of Rathke’s pouch begins to regress. The outer wall of the cup thickens and assumes a glandular appearance in the course of its differentiation into the pars distalis (anterior lobe) of the hypophysis. The inner layer of the cup, which is closely adherent to the neural lobe, becomes the pars intermedia (intermediate lobe). It remains separated from the anterior lobe by a slitlike residual lumen, which represents all that remains of the original lumen of Rathke’s pouch.

With the progression of pregnancy, the hypophysis undergoes a phase of cytodifferentiation. Late in the fetal period, specific cell types begin to produce small amounts of hormones. Molecular cascades underlying the differentiation of specific cell types in the pituitary are being discovered (Box 14.1).

Although Rathke’s pouch normally begins to lose its connections to the stomodeal epithelium by the end of the second month, portions of the tissue may occasionally persist along the pathway of the elongating stalk. If the tissue is normal, it is called a pharyngeal hypophysis. Sometimes, however, the residual tissue becomes neoplastic and forms hormone-secreting tumors called craniopharyngiomas.

Thymus and Lymphoid Organs

The paired endodermal thymic primordia begin to migrate from their third pharyngeal pouch origins during the sixth week. Their path of migration takes them through a substrate of mesenchymal cells until they reach the area of the future mediastinum behind the sternum. By the end of their migration, the two closely apposed thymic lobes are still epithelial structures. Soon, however, they become invested with a capsule of neural crest–derived connective tissue, which also forms septa among the endodermal epithelial cords and contributes to the thymic vasculature. In the absence of neural crest, the thymus fails to develop. An interaction between the neural crest and endodermal components of the thymic primordia conditions the latter for subsequent differentiation of thymic structure and function.

At about 9 to 10 weeks’ gestation, blood-borne thymocyte precursors (prothymocytes), which originate in the hematopoietic tissue, begin to invade the epithelial thymus, in response to the secretion of the chemokine CC121 by the thymus. Just before the prothymocytes invade the thymus, the thymic epithelium begins to express the transcription factor WHN, which is necessary for colonization of the thymic epithelium by the prothymocytes. In homozygous WHN mutants, the absence of such colonization results in the absence of functional T cells, thus leaving the individual severely immunocompromised and unable to reject foreign cells and tissue. Within the thymus, the prothymocytes force apart the epithelial cells and cause them to form a spongy epithelial reticulum. Responding to signals from the thymic epithelium, the prothymocytes proliferate and become redistributed, forming the cortical and medullary regions of the thymus.

By 14 to 15 weeks of gestation, blood vessels grow into the thymus, and a week later, some epithelial cells aggregate into small, spherical Hassall’s corpuscles. At this point, the overall organization of the thymus is the same as that of adults. Functionally, the action of various thymic hormones causes the thymus to condition or instruct the prothymocytes migrating into it to become competent members of the T-lymphocyte family. The T lymphocytes leave the thymus and populate other lymphoid organs (e.g., lymph nodes, spleen) as fully functional immune cells.

The T lymphocytes are principally involved in cellular immune responses. Another population of lymphocytes that also originates in the bone marrow is instructed to become B lymphocytes, which are the mediators of humoral immune responses. B-lymphocyte precursors (pro-B cells) also must undergo conditioning to become fully functional, but their conditioning does not occur in the thymus. In birds, the pro-B cells pass through a cloacal lymphoid organ known as the bursa of Fabricius, where conditioning occurs. Humans do not possess a bursa, but its functional equivalent, although still undefined, is assumed to exist. B-lymphocyte conditioning is thought to occur in the bone marrow; in early embryos, conditioning possibly occurs in the liver.

The thymus and the bursa or mammalian equivalent are commonly referred to as central lymphoid organs. The lymphoid structures that are seeded by B and T lymphocytes are called peripheral lymphoid organs. (Figure 14.39 shows the development and function of the lymphoid system.) Small cervical thymus glands have been discovered in the neck in mice. How their function fits into that of the overall lymphoid system remains to be determined.

Formation of the Tongue

The tongue begins to take shape from a series of ventral swellings in the floor of the pharynx at about the same time as the palate forms in the mouth. Major shifts in the positions of tissues of the tongue occur, thus making the characteristics of the adult form difficult to comprehend without knowledge of the basic elements of its embryonic development.

In 5-week-old embryos, the tongue is represented by a pair of lateral lingual swellings in the ventral regions of the first pharyngeal arches and two median unpaired swellings. The tuberculum impar is located between the first and second arches, and the copula (yoke) unites the second and third arches (Figs. 14.40A and B and 14.41). The foramen cecum, which marks the original location of the thyroid primordium, serves as a convenient landmark delineating the border between the original tuberculum impar and the copula. Caudal to the copula is another swelling that represents the epiglottis.

Growth of the body of the tongue is accomplished by a great expansion of the lateral lingual swellings, with a minor contribution by the tuberculum impar (Fig. 14.40C and D). The root of the tongue is derived from the copula, along with additional ventromedial tissue between the third and fourth pharyngeal arches.

The dorsal surface of the tongue is covered with a large number of papillae. Development of the filiform papillae, which constitute the bulk of the papillae, follows a course that is remarkably similar to that of hair follicles. The ectoderm, which surrounds a mesenchymal core, expresses Hoxc-13 and the signaling factors shh, BMP-2 and BMP-4, and FGF-8. These molecular components characterize the inductive signaling system in almost all ectodermal derivatives, including hairs, feathers, and teeth.

Corresponding to its innervation by the hypoglossal nerve (cranial nerve XII), the musculature of the tongue migrates from a considerable distance (the occipital [postotic] myotomes). Similar to their counterparts in the limb, the myoblasts express Pax-3 during their migration to the tongue. The general sensory innervation of the tongue accurately reflects the pharyngeal arch origins of the epithelium. The lingual epithelium over the body of the tongue is innervated by the trigeminal nerve (cranial nerve V), in keeping with the first-arch origins of the lateral lingual swellings. Correspondingly, the root of the tongue is innervated by the glossopharyngeal nerve (cranial nerve IX—third arch) and the vagus nerve (cranial nerve X—fourth arch). The epithelium of the second arch is overgrown by that of the third arch; there is no general sensory innervation of the tongue by the seventh nerve.

Cranial nerves VII (facial) and IX innervate the taste buds. The contribution of cranial nerve VII is facilitated by its chorda tympani branch, which joins with the lingual branch of the trigeminal nerve and has access to the body of the tongue. Taste buds appear during the seventh week of gestation. In contrast to an earlier hypothesis that they are induced by fibers of the visceral afferent cranial nerves VII and IX, which innervate taste buds postnatally, more recent evidence suggests that taste bud formation is independent of cranial nerves. An intrinsic signaling mechanism within the endodermal lingual epithelium, possibly mediated by a mechanism involving shh, Gli-1, and patched, seems to be the basis for the initial formation of taste buds. When formed, taste buds become strongly dependent on innervation for their maintenance. Considerable evidence indicates that the fetus is able to taste, and it has been postulated that the fetus uses the taste function to monitor its intra-amniotic environment.

Clinical Correlation 14.3 presents anomalies and syndromes involving the pharynx and pharyngeal arches.

Clinical Correlation 14.3   Anomalies and Syndromes Involving the Pharynx and Pharyngeal Arches

Syndromes Involving the First Pharyngeal Arch

Several syndromes involve hypoplasia of the mandible and other structures arising from the first pharyngeal arch. Various mechanisms can account for these syndromes. Many have a genetic basis, and others result from exposure to environmental teratogens. The numerous studies on genetically manipulated mice are beginning to produce malformations that may have human counterparts. With the fine balance of signaling molecules and transcription factors that are involved in the development of the pharyngeal arches, it is not surprising that disturbances in the function of single genes, whether through mutation or the action of teratogens, can result in a visible morphological anomaly. Hypoplasia of the lower face has been associated with the ingestion of isotretinoin (a vitamin A derivative used for the treatment of acne) during early pregnancy.

Pierre Robin syndrome involves extreme micrognathia (small mandible), cleft palate, and associated defects of the ear. An imbalance often exists between the size of the tongue and the very hypoplastic jaw, which can lead to respiratory distress caused by mechanical interference of the pharyngeal airway by the large tongue. Although many cases of Pierre Robin syndrome are sporadic, others have a genetic basis.

Treacher Collins syndrome (mandibulofacial dysostosis) is typically inherited as an autosomal dominant condition. The responsible gene, called TCOF1, has been identified. Operating through the Treacle protein, it affects the survival and proliferation of cranial neural crest cells. In mutations of this gene, neural crest cell migration is normal, but increased apoptosis and decreased proliferation result in a much reduced population of neural crest cells in the first pharyngeal arch. This syndrome includes various anomalies, not all of which are found in all patients. Common components of the syndrome include hypoplasia of the mandible and facial bones, malformations of the external and middle ears, high or cleft palate, faulty dentition, and coloboma-type defects of the lower eyelid (Fig. 14.42).

The most extreme form of first-arch hypoplasia is agnathia, in which the lower jaw basically fails to form (Fig. 14.43). In severe agnathia, the external ears remain in the ventral cervical region and may join in the ventral midline.

Lateral Cysts, Sinuses, and Fistulas

Structural malformations, such as lateral cysts, sinuses, and fistulas, can be related directly to the abnormal persistence of pharyngeal grooves, pharyngeal pouches, or both. A cyst is a completely enclosed, epithelially lined cavity that may be derived from the persistence of part of a pharyngeal pouch, a pharyngeal groove, or a cervical sinus. A sinus is closed on one end and open to the outside or to the pharynx. A fistula (Latin for “pipe”) is an epithelially lined tube that is open at both ends—in this case, to the outside and to the pharynx.

The postnatal location of these structures accurately marks the location of their embryonic precursors. External openings of fistulas are typically found anterior to the sternocleidomastoid muscle in the neck (Fig. 14.44). Fistulas from remnants of pharyngeal grooves II or III may result from incomplete closure of the cervical sinus by tissue of the hyoid arch. Although present from birth, cervical cysts are often not manifested until after puberty. At that time, they expand because of increased amounts of secretions by the epithelium lining the inner surface of the cyst, changes that correspond to maturational changes in the normal epidermis.

Preauricular sinuses or fistulas, which are found in a triangular region in front of the ear, are also common. These structures are assumed to represent persistent clefts between preauricular hillocks on the first and second arches. True fistulas (cervicoaural fistulas) represent persisting ventral portions of the first pharyngeal groove. These extend from a pharyngeal opening to somewhere along the auditory tube or even the external auditory meatus.

Thyroglossal Duct Remnants

Various abnormal structures can persist along the pathway of the thyroglossal duct. Ectopic thyroid tissue can be found anywhere along the pathway of migration of the thyroid primordium from the foramen cecum in the tongue to the isthmus of the normal thyroid gland (Fig. 14.45). This fact must be considered in the clinical diagnosis or surgical treatment of carcinomas and other conditions affecting thyroid tissue. Less common are midline cysts or sinuses involving the former thyroglossal duct (Fig. 14.46). Because of their location, these entities can usually be easily distinguished from their lateral cervical counterparts.

Malformations of the Tongue

The most common malformation of the tongue is ankyloglossia (tongue-tie). This condition is caused by subnormal regression of the frenulum, the thin midline tissue that connects the ventral surface of the tongue to the floor of the mouth. Nonsyndromic ankyloglossia is caused by a mutation in the T-box transcription factor TBX22. Less common malformations of the tongue are macroglossia and microglossia, which are characterized by hyperplasia and hypoplasia of lingual tissue. Although sometimes associated with macroglossia, furrowed tongue (Fig. 14.47) is not normally associated with major functional disturbances.

Ectopic Parathyroid or Thymic Tissue

Because of their extensive migrations during early embryogenesis, parathyroid glands and components of the thymus gland are often found in abnormal sites (Fig. 14.48). Typically, this displacement is not accompanied by functional abnormalities, but awareness of the possibility of ectopic tissue or even supernumerary parathyroid glands is important for the surgeon.

DiGeorge’s Syndrome

DiGeorge’s syndrome is a cranial neural crest deficiency and is manifested by immunological defects and hypoparathyroidism (see Clinical Correlation 12.1). The underlying pathological condition is failure of differentiation of the thymus and parathyroid glands. Associated anomalies are malformations of first-arch structures and defects of the outflow tract of the heart, which contains an important cranial neural crest contribution as well. Hoxa3 mutant mice exhibit many of the characteristics of human DiGeorge’s syndrome.