Origin, Induction, and Specification

According to the most recent data, the earliest stages of neural crest induction may occur as early as gastrulation, but according to the classical model, the neural crest arises as the result of inductive actions by the adjacent non-neural ectoderm and possibly nearby mesoderm on the neural plate (Fig. 12.1). The ectodermal inductive signals are bone morphogenetic proteins (BMPs) and Wnts. Fibroblast growth factor-8 (FGF-8) from mesoderm plays a role in neural crest induction in amphibians, and it seems to be involved in mammals as well. The role of BMPs is complex and relates to a concentration gradient along the ectodermal layer as neurulation proceeds. The highest concentrations of BMP are seen in the lateral ectoderm, and cells exposed to these concentrations remain ectodermal. Cells within the neural plate are exposed to the lowest concentrations of BMP because of the local inhibitory actions of noggin and chordin (see Fig. 5.8C), and, by default, they remain neural. Cells at the border of the neural plate are exposed to intermediate levels of BMP, and, in this environment, they are induced to form neural crest precursor cells.

In response to these inductive signals, cells at the border of the neural plate activate genes coding for several transcription factors, including Msx-1 and Msx-2, Dlx-5, Pax-3/Pax-7, and Gbx-2. These and other gene products turn on a network of genes that transform the epithelial neural crest precursor cells into mobile mesenchymal cells that break free from the neuroepithelium of the neural tube.

Epitheliomesenchymal Transformation and Emigration from the Neural Tube

Within the neural tube, neural crest precursor cells are epithelial and are tightly adherent to other neuroepithelial cells through a variety of intercellular connections. Prominent among them are the cadherins. Among the new transcription factors upregulated in induced neural crest precursor cells are snail-1 and snail-2 (formerly called slug) and Foxd-3, which are instrumental in allowing the neural crest cells to break free from the neural epithelium and then migrate away as mesenchymal cells.* Under the influence of snail-1 and snail-2, the profile of cadherins produced by the neural crest precursors changes from type I cadherins (e.g., N-cadherin and E-cadherin), which are strongly adhesive, to type II cadherins, which are less adhesive.

Neural crest cells break free from the neural tube in the trunk at the level of the last-formed somite or the neural plate in the head by changing their shape and properties from those of typical neuroepithelial cells to those of mesenchymal cells. Important to this process is the loss of cell-to-cell adhesiveness. This loss is effected by the loss of cell adhesion molecules (CAMs) characteristic of the neural tube (e.g., N-CAM, E-cadherin, and N-cadherin). These molecules remain downregulated during migration, but after neural crest cells have completed their migrations and have differentiated into certain structures (e.g., spinal ganglia), CAMs are often expressed again.

In the head, where closure of the neural plate has not yet occurred, neural crest cells must penetrate the basal lamina underlying the neural plate. This is accomplished by the production of enzymes that degrade components of the basal lamina and by sending out processes that penetrate the basal lamina. In the trunk, neural crest cells do not leave the neuroepithelium until after the neural tube has formed. They do not, however, have to contend with penetrating a basal lamina because the dorsal part of the neural tube does not form a basal lamina until after emigration of the crest cells.

Neural Crest Cell Migration

After leaving the neuroepithelium, the neural crest cells first encounter a relatively cell-free environment rich in extracellular matrix molecules (Fig. 12.2). In this environment, the cells undergo extensive migrations along several well-defined pathways. These migrations are determined by intrinsic properties of the neural crest cells and features of the external environment encountered by the migrating cells.

Neural crest migration is influenced by a variety of molecules residing in the extracellular matrix. Although the presence of a basal lamina can inhibit their emigration from the neural tube, neural crest cells often prefer to migrate along basal laminae, such as those of the surface ectoderm or neural tube, after they have left the neural tube. Components of the extracellular matrix permissive for migration include molecules found in basal laminae, such as fibronectin, laminin, and type IV collagen (Fig. 12.3). Attachment to and migration over these substrate molecules are mediated by the family of attachment proteins called integrins. Other molecules, such as chondroitin sulfate proteoglycans, are not good substrates for neural crest cells and inhibit their migration.

Neural crest cells emigrate from the neural tube or neural folds in streams, with each cell in contact with neighbors through filopodial contacts. During their migratory phase, neural crest cells are exquisitely sensitive to guidance molecules, most of which are inhibitory. Among the most important of these guidance molecules are the ligand/receptor pairs Robo/Slit, Neuropilin/Semaphorin and Ephrin/Eph (see Table 11.1). Much less is known about attractive influences on neural crest cell migration. During migration, neural crest cells extend protrusions that both test the environment and are part of the propulsive mechanism. If an inhibitory influence is encountered, the protrusions collapse through signals derived from a planar cell polarity pathway (see p. 87). This mechanism acts as a brake when the cells encounter an inhibitory environment, but it is also involved in their forward propulsion. In a migrating stream of neural crest cells, contact with the cells behind also results in the pulling of protrusions at the trailing edge of the cells, thus resulting in a net forward motion of the leading cells. Specific examples of the environmental control of neural crest cell migrations are given later in this chapter. Much remains to be learned about what causes neural crest cells to stop migrating, but often they stop migrating in areas where repulsive signals are low.

Differentiation of Neural Crest Cells

Neural crest cells ultimately differentiate into an astonishing array of adult structures (Table 12.1). What controls their differentiation is one of the principal questions of neural crest biology. Two opposing hypotheses have been proposed. According to one, all neural crest cells are equal in developmental potential, and their ultimate differentiation is entirely determined by the environment through which they migrate and into which they finally settle. The other hypothesis suggests that premigratory crest cells are already programmed for different developmental fates, and that certain stem cells are favored, whereas others are inhibited from further development during migration. More recent research indicates that the real answer can be found somewhere between these two positions. Increasing evidence suggests that among migrating neural crest cells is a mix of cells whose fate has been predetermined within the neural tube and cells whose ultimate phenotype depends on environmental influences.

A correlation exists between the time of migration of neural crest cells from the neural tube and their developmental potential. Many cells that first begin to migrate have the potential to differentiate into several different types of cells. Crest cells that begin to migrate later are capable of forming only derivatives characteristic of more dorsal locations (e.g., spinal ganglia), but not sympathetic neurons or adrenal medullary cells. Crest cells that leave the neural tube last can form only pigment cells.

Several experiments have shown that the fates of some neural crest cells are not irreversibly fixed along a single pathway. One type of experiment involves the transplantation of neural crest cells from one part of the body to another. For example, many neural crest cells from the trunk differentiate into sympathetic neurons that produce norepinephrine as the transmitter. In the cranial region, however, neural crest cells give rise to parasympathetic neurons, which produce acetylcholine. If thoracic neural crest cells are transplanted into the head, some cells differentiate into cholinergic parasympathetic neurons instead of the adrenergic sympathetic neurons normally produced. Conversely, cranial neural crest cells grafted into the thoracic region respond to their new environment by forming adrenergic sympathetic neurons.

A more striking example is the conversion of cells of the periocular neural crest mesenchyme, which in birds would normally form cartilage, into neurons if they are associated with embryonic hindgut tissue in vitro. Many of the regional influences on the differentiation of local populations of neural crest cells are now recognized to be interactions between the migrating neural crest cells and specific tissues that they encounter during migration. Examples of tissue interactions that promote the differentiation of specific neural crest derivatives are given in Table 12.2.

The plasticity of differentiation of neural crest cells can be shown by cloning single neural crest cells in culture. In the same medium, and under apparently the same environmental conditions, the progeny of the single cloned cells frequently differentiate into neuronal and non-neuronal (e.g., pigment cell) phenotypes. Similarly, if individual neural crest cells are injected in vivo with a dye, greater than 50% of the injected cells will give rise to progeny with two to four different phenotypes containing the dye. By exposing cloned neural crest precursor cells to specific environmental conditions in vitro, one can begin to understand the mechanisms that determine phenotype in vivo. In one experiment, rat neural crest cells grown under standard in vitro conditions differentiated into neurons, but when they were exposed to glial growth factor, they differentiated into Schwann cells because the glial growth factor suppressed their tendency to differentiate into neurons. Similarly, the growth factors BMP-2 and BMP-4 cause cultured neural crest cells to differentiate into autonomic neurons, whereas exposure of these cells to transforming growth factor-β causes them to differentiate into smooth muscle.

Not all types of transformations among possible neural crest derivatives can occur. Crest cells from the trunk transplanted into the head cannot form cartilage or skeletal elements, although this is normal for cells of the cranial neural crest. Most experiments suggest that early neural crest cells segregate into intermediate lineages that preserve the option of differentiating into several, but not all, types of individual phenotypes. In the chick embryo, some neural crest cells are antigenically different from others even before they have left the neural tube.

Many neural crest cells are bipotential, depending on signals from their local environment for cues to their final differentiation. Cultured heart cells secrete a protein that converts postmitotic sympathetic neurons from an adrenergic (norepinephrine transmitter) phenotype to a cholinergic (acetylcholine-secreting) phenotype (see Fig. 11.22). During normal development, the sympathetic neurons that innervate sweat glands are catecholaminergic until their axons actually contact the sweat glands. At that point, they become cholinergic.

Trunk Neural Crest

The neural crest of the trunk extends from the level of the sixth somite to the most caudal somites. Three pathways of migration are commonly described (Fig. 12.4). These pathways occur in different sequences and are subject to different controls. The first neural crest cells to leave the neural tube migrate around and between the somites, which are still in an epithelial configuration. Their migratory path follows the intersomitic blood vessels, and the cells rapidly reach the region of the dorsal aorta (see Fig. 12.4, pathway 1). It may be that at this early stage no other pathway is available to these migrating cells. These cells constitute the sympathoadrenal lineage.

Slightly later in development, the somites have become dissociated into sclerotomal and dermomyotomal compartments. At this stage, the neural crest cells preferentially enter the anterior compartment of the sclerotome. They are kept from entering the posterior compartment mainly through the repulsive action of semaphorinA3F (SEMA3F) in the posterior sclerotome, by acting through its receptor Neuropilin-2 (Nrp-2) on the neural crest cells. Other molecular repulsion mechanisms are also involved, but in mammals, this mechanism is the most influential. Passage through the anterior sclerotome is facilitated by extracellular matrix molecules, in particular thrombospondin. These cells constitute the ventrolateral pathway, and they ultimately form the dorsal root ganglia (see Fig. 12.4, pathway 2). These cells form the ganglia in concert with the outgrowth of the motor axons from the spinal cord, which follow similar environmental cues.

The last pathway (see Fig. 12.4, pathway 3) is the dorsolateral pathway, and the cells that follow it appear to be determined even before emigration from the neural tube to become pigment cells. Other neural crest cells are not able to use this pathway. In mammals, cells that follow this pathway depend on the Steel factor, produced by the dermomyotome, to be able to use this pathway. The cells that take this pathway migrate just beneath the ectoderm and ultimately enter the ectoderm as pigment cells (melanocytes).

Cranial Neural Crest

The cranial neural crest is a major component of the cephalic end of the embryo. Comparative anatomical and developmental research suggests that the cranial neural crest may represent the major morphological substrate for the evolution of the vertebrate head. Largely because of the availability of precise cellular marking methods, the understanding of the cranial neural crest has increased dramatically. Most studies on the cranial neural crest have been conducted on avian embryos; however, the properties and role of the neural crest in mammalian cranial development are quite similar to those in birds.

In the mammalian head, neural crest cells leave the future brain well before closure of the neural folds (Fig. 12.5). In the area of the forebrain, no neural crest arises rostral to the anterior diencephalon (anterior neural ridge [see Fig. 6.4B]), but from the region marked by prosomeres 1 to 3, a continuous sheet of neural crest cells migrates over much of the head (Fig. 12.6). Neural crest is inhibited from forming in the anterior neural ridge by the signaling molecule Dickkopf 1, a Wnt inhibitor that is secreted by the nearby prechordal mesoderm. Specific streams of neural crest cells emanating from the hindbrain populate the first three pharyngeal arches. Although the streams of migrating cranial neural crest appear at first glance to be not very discrete, there is an overall very specific spatiotemporal order in their pathways to their final destinations in the head and neck.

A major functional subdivision of cranial neural crest occurs at the boundary between rhombomeres 2 (r2) and 3 (r3). Neural crest cells emerging from the diencephalon posteriorly through r3 do not express any Hox genes, whereas the cells emerging from the hindbrain region from r4 and posteriorly express a well-ordered sequence of Hox genes (see Fig. 12.8).

There is remarkable specificity in the relationship among the origins of the neural crest in the hindbrain, its ultimate destination within the pharyngeal arches, and the expression of certain gene products (Figs. 12.7 and 12.8). Neural crest cells associated with r1 and r2 migrate into and form the bulk of the first pharyngeal arch; those of r4, into the second arch; and those of r6 and r7, into the third arch, as three separate streams of cells.

For many years, it was thought that neural crest cells did not migrate from r3 or r5 even though neural crest cells form in these areas. Some of the neural crest cells associated with r3 and r5 undergo apoptosis because of the presence of the apoptosis-inducing molecule BMP-4, but research has shown that semaphorins in the mesenchyme lateral to r3 and r5 exert a repulsive effect on neural crest cells that try to enter these areas. A few neural crest cells from r3 diverge into small streams that enter the first and second pharyngeal arches, and cells from r5 behave similarly, by merging with the streams of neural crest cells emanating from r4 and r6.

A close correlation exists between the pattern of migration of the rhombomeric neural crest cells and the expression of products of the Hoxb gene complex. Hoxb-2, Hoxb-3, and Hoxb-4 products are expressed in a regular sequence in the neural tube and the neural crest–derived mesenchyme of the second, third, and fourth pharyngeal arches. Hoxb is not expressed in r1 and r2 or in the first pharyngeal arch mesenchyme. Only after the pharyngeal arches become populated with neural crest cells does the ectoderm overlying the arches express a similar pattern of Hoxb gene products (see Fig. 12.8). These Hoxb genes may play a role in positionally specifying the neural crest cells with which they are associated. Interactions between the neural crest cells and the surface ectoderm of the pharyngeal arches may specify the ectoderm of the arches.

The Hox genes play an important role in determining the identity of the pharyngeal arches. The first arch develops independently from Hox influence, but Hoxa2 is critical in determining the identity of the second arch by repressing the elements that would turn it into a first arch. In the absence of Hoxa2 function, the second arch develops into a mirror image of the first arch. Overall, members of the Hox3 paralogous group are heavily involved in patterning the third arch and Hox4 paralogues, the fourth, although research has produced evidence of some overlap of functions.

Emigrating cranial neural crest cells consist of a mix of cells whose fate has already been fixed and those whose fate is largely determined by their environment. As they move away from the brain, cranial crest cells migrate as sheets rostrally or streams (in the pharyngeal area) in the dorsolateral pathway directly beneath the ectoderm. This is in strong contrast to migratory patterns in trunk neural crest, where the first two waves of migration head directly ventrally or ventrolaterally (see Fig. 12.4, pathways 1 and 2). As they approach the pharyngeal arches, especially the second arch, the lead cells in the streams of neural crest are attracted by vascular endothelial growth factor (VEGF), a chemoattractant produced by the distal ectoderm. The trailing cells in the stream are interconnected by long filopodia and follow the lead cells as they disperse into the pharyngeal arches themselves.

Cranial neural crest cells differentiate into a wide variety of cell and tissue types (see Table 12.1), including connective tissue and skeletal tissues. These tissues constitute much of the soft and hard tissues of the face (Fig. 12.9). (Specific details of morphogenesis of the head are presented in Chapter 14.)

Circumpharyngeal Neural Crest

The circumpharyngeal neural crest arises in the posterior rhombencephalic region at the levels of somites 1 to 7 (Fig. 12.10). This region of neural crest represents a transition between cranial and trunk neural crest. Cells arising at the levels of the first four somites behave more like cranial crest, whereas those emigrating at the levels of somites 5 to 7 follow pathways more characteristic of trunk crest. A prominent landmark in this area is the circumpharyngeal ridge, an arc-shaped aggregation of cells that passes behind the sixth pharyngeal arch (Fig. 12.11). Ventral to the pharynx, this ridge sweeps cranially and provides the pathway through which the hypoglossal nerve (XII) and its associated skeletal muscle precursors pass. Most neural crest cells from the somite 1 to 3 level pass into either the outflow tract of the heart or into the fourth and sixth pharyngeal arches (see Fig. 12.10). These cells are considered to constitute the cardiac crest. Other cells from this level, as well as those arising from the level of somites 4 to 7, are called the vagal crest. These cells migrate into the gut as precursors of the parasympathetic innervation of the digestive tract. They also form sensory neurons and glia, as well as making some contribution to sympathetic ganglia. Like the cranial crest cells, most cells of the cardiac crest migrate along the dorsolateral pathway between the somites and the ectoderm (see Fig. 12.10), whereas those of the vagal crest, like those of the trunk, initially migrate along the ventral pathways between the neural tube and the dermomyotome.

Cardiac Crest

The cardiac crest, arising at the level of somites 1 to 3, surrounds the endothelial precursors of the third, fourth, and sixth aortic arches, and it contributes massively to the truncoconal ridges that separate the outflow tract of the heart into aortic and pulmonary segments (see Chapter 17). Under the strong influence of semaphorins, cardiac crest cells migrate toward the heart and contribute to the leaflets of the semilunar valves at the base of the outflow tract, and in birds, at least, they may penetrate the interventricular septum. The cardiac neural crest may interact with pharyngeal endoderm to modify the signals leading to the normal differentiation of myocardial cells.

Although much of the cardiac crest contributes to the outflow tract of the heart and the great vessels, portions of the cardiac neural crest population become associated with the newly forming thymus, parathyroid, and thyroid glands. Two streams of cardiac neural crest cells leave the neural tube. The earlier stream contributes principally to the cardiac outflow tract and aortic arch arteries, whereas cells of the later stream become incorporated into pharyngeal glands. On their way to the heart and pharyngeal structures, cardiac crest cells migrate along the dorsolateral pathway and reach their destinations via the circumpharyngeal ridge.

Some neural crest cells migrate ventral to the pharynx in bilateral streams accompanying the somite-derived myoblasts that are migrating cranially to form the intrinsic muscles of the tongue and the hypopharyngeal muscles. This is the only known case in which somite-derived muscles are invested with neural crest–derived connective tissue. The cardiac neural crest also supplies the Schwann cells that are present in the hypoglossal and other cranial nerves.

A disturbance in this region of neural crest can result in cardiac septation defects (aorticopulmonary septum) and glandular and craniofacial malformations. DiGeorge’s syndrome, which is associated with a deletion on chromosome 22, is characterized by hypoplasia and reduced function of the thymus, thyroid, and parathyroid glands and cardiovascular defects, such as persistent truncus arteriosus and abnormalities of the aortic arches. Hoxa3 mutant mice show a similar spectrum of pharyngeal defects. The common denominator for this constellation of pathological features is a defect of the cardiac crest supplying the third and fourth pharyngeal arches and cardiac outflow tract. Similar defects have been described in human embryos exposed to excessive amounts of retinoic acid early in embryogenesis.

Vagal Crest

Within the gut, neural crest cells form the enteric nervous system, which in many respects acts like an independent component of the nervous system. The number of enteric neurons nearly matches the number of neurons in the spinal cord, and most of these neurons are not directly connected to either the brain or the spinal cord. This independence explains how the bowel can maintain reflex activity in the absence of direct input from the central nervous system.

The cells that form the neurons of the enteric nervous system come from the part of the circumpharyngeal crest known as the vagal crest. These cells exit from the levels of somites 1 to 7, follow a ventral pathway through the dorsal part of the circumpharyngeal ridge, and then exit this pathway caudal to the sixth pharyngeal arch. Most of these cells become closely associated with the embryonic gut, but some are involved in formation of sensory dorsal root ganglia and associated glia. At the level of somite 7, some cells even contribute to local sympathetic ganglia. Neural crest cells are not committed to form gut-associated nervous tissue before they leave the spinal cord. If vagal crest is replaced by neural crest of the trunk, which does not normally give rise to gut-associated derivatives, the gut is colonized by the transplanted trunk-level neural crest cells.

Despite the strong influence of the environment of the gut on the differentiation of neural crest cells exposed to its influences, neural crest cells retain a surprising degree of developmental flexibility. If crest-derived cells already in the gut of avian embryos are retransplanted into the trunk region of younger embryos, they seem to lose the memory of their former association with the gut. They enter the pathways (e.g., adrenal or peripheral nerve) common to trunk cells (except that they cannot enter the pigment cell pathway) and differentiate accordingly.

Under the influence of glial-derived neurotrophic factor (GNDF), vagal crest cells enter the anterior region of the foregut and begin to populate the gut. One potential reason that neural crest cells in the trunk are unable to enter the gut is that the cells in the mesentery near the gut express Slit-2, the molecule that also prevents neurons from crossing the midline of the central nervous system. Trunk neural crest cells express the Slit receptor Robo, thus causing them to avoid cells that express Slit. Vagal crest cells do not express Robo and are permitted access to the gut wall.

Within the gut wall, vagal crest cells undertake a major invasion that sweeps through the length of the gut and ultimately stops near the posterior end of the hindgut by the end of the seventh week of pregnancy. The vagal crest–derived enteric neuronal precursors, which later become parasympathetic neurons, advance down the gut at a rate of approximately 40 to 45 µm/hour. These cells advance as interconnected strands, and they undergo proliferation at the level of the wavefront. Advancement of the wavefront is apparently more the result of proliferation and the spilling over of cellular neural crest progeny into the unpopulated regions of the gut than of actual directed migration of individual cells. When the cellular wavefront arrives at the cecum, the cells pause for several hours because of the presence of local signaling factors. They then proceed into the future colon. Within the colon, the vagal crest cells ultimately meet with a smaller number of cells emigrating from the sacral neural crest, at which point invasive activity ceases, and further organization of the enteric ganglia continues. When they first colonize the gut, the neural crest cells express no neuronal markers, but under the influence of Hand-2, a wave of differentiation passes down the gut, and the cells synthesize neurofilament proteins and initially express catecholaminergic traits. These cells form myenteric plexuses.

Some common neurocristopathies are presented in Clinical Correlation 12.1.

Clinical Correlation 12.1   Neurocristopathies

Because of the complex developmental history of the neural crest, various congenital malformations are associated with its defective development. These malformations have commonly been subdivided into two main categories—defects of migration or morphogenesis and tumors of neural crest tissues (Box 12.1). Some of these defects involve only a single component of the neural crest; others affect multiple components and are recognized as syndromes.

Box 12.1   Major Neurocristopathies

Defects of Migration or Morphogenesis

Trunk Neural Crest

Hirschsprung’s disease (aganglionic megacolon)

Cranial Neural Crest

Aorticopulmonary septation defects of heart

Anterior chamber defects of eye

Cleft lip, cleft palate, or both

Frontonasal dysplasia

DiGeorge’s syndrome (hypoparathyroidism, thyroid deficiency, thymic dysplasia leading to immunodeficiency, defects in cardiac outflow tract and aortic arches)

Certain dental anomalies

Trunk and Cranial Neural Crest

CHARGE association (coloboma, heart disease, atresia of nasal choanae, retardation of development, genital hypoplasia in males, anomalies of the ear)

Waardenburg’s syndrome

Tumors and Proliferation Defects

Pheochromocytoma: tumor of chromaffin tissue of adrenal medulla

Neuroblastoma: tumor of adrenal medulla, autonomic ganglia, or both

Medullary carcinoma of thyroid: tumor of parafollicular (calcitonin-secreting) cells of thyroid

Carcinoid tumors: tumors of enterochromaffin cells of digestive tract

Neurofibromatosis (von Recklinghausen’s disease): peripheral nerve tumors

Neurocutaneous melanosis (multiple congenital pigmented nevi, melanocytic tumors of the central nervous system)

Differentiation Defect Involving Neural Crest Cells

Albinism

Several syndromes or associations of defects are understandable only if the wide distribution of derivatives of the neural crest is recognized. For example, one association called CHARGE consists of coloboma (see Chapter 13), heart disease, atresia of nasal choanae, retardation of development, genital hypoplasia in males, and anomalies of the ear.

Types I and III of Waardenburg’s syndrome, which is caused by Pax-3 mutations, involve various combinations of pigmentation defects (commonly a white stripe in the hair and other pigment anomalies in the skin), deafness, cleft palate, and ocular hypertelorism (increased space between the eyes). One variant (type I) of Waardenburg’s syndrome is also characterized by hypoplasia of the limb muscles; this is not surprising considering the important association between Pax-3 and myogenic cells migrating into the limb buds from the somites. Pax-3 is similarly expressed in migrating cardiac neural crest cells, but it is downregulated when the cells settle in the walls of the cardiac outflow tract or aortic arches. Cardiovascular defects in these areas are also seen in Pax-3 mutants.

DiGeorge syndrome, which is associated with a deletion on chromosome 22 that encompasses up to 15 genes, is characterized by hypoplasia and reduced function of the thymus, thyroid, and parathyroid glands and cardiovascular defects, including persistent truncus arteriosus and abnormalities of the aortic arches. Hoxa3 mutant mice show a similar spectrum of pharyngeal defects. The common denominator for this constellation of pathological features is a defect of the neural crest supplying the third and fourth pharyngeal arches and the cardiac outflow tract. Similar defects have been described in human embryos exposed to excessive amounts of retinoic acid early in embryogenesis.

Neurofibromatosis (von Recklinghausen’s disease) is a common genetic disease manifested by multiple tumors of neural crest origin. Common features are café au lait spots (light brown pigmented lesions) on the skin, multiple (often hundreds) neurofibromas (peripheral nerve tumors), occasional gigantism of a limb or digit, and various other conditions (Fig. 12.12). Neurofibromatosis occurs in approximately 1 of 3000 live births, and the gene is very large and subject to a high mutation rate. Some rare syndromes involving features such as disturbed pigmentation, cutaneous microvascular lesions, asymmetric enlargements of structures, and nerve lesions (including neurofibromatosis) are sometimes lumped together in the category of neural crest–based neurocutaneous syndromes.

Because of the massive contribution of the neural crest to the face and other parts of the head and neck, various malformations in the craniofacial region involve neural crest derivatives. Many different facial abnormalities lumped together under the term frontonasal dysplasia (see Chapter 14) heavily involve neural crest–derived tissues.