Development
Neural Crest Cells
The fourth germ layer — delamination, epithelial-mesenchymal transition, and the migratory cells that built the vertebrate head
The neural crest is a transient, multipotent, migratory population of cells — often called the fourth germ layer — that appears in every vertebrate embryo, detaches from the top of the newly formed neural tube, and streams across the body to build tissues no other ectoderm can make. After neural tube closure, crest cells undergo an epithelial-mesenchymal transition, delaminate from the dorsal neuroepithelium, and migrate along stereotyped corridors to become peripheral and enteric neurons, Schwann cells, melanocytes, adrenal chromaffin cells, smooth muscle, and most of the craniofacial cartilage and bone. First described by Wilhelm His in 1868 as the Zwischenstrang, and mapped in exhaustive detail by Nicole Le Douarin's quail-chick chimeras in the 1970s, the neural crest is regarded as a defining vertebrate innovation — the cell population behind Gans and Northcutt's 1983 "New Head" hypothesis. When its formation, migration, or differentiation fails, the result is a neurocristopathy: Hirschsprung disease, Waardenburg syndrome, DiGeorge syndrome, or neuroblastoma.
- Also calledthe fourth germ layer
- Originneural plate border → dorsal neural tube
- Exit mechanismepithelial-mesenchymal transition
- Enteric neurons made~500 million (whole gut)
- DiscoveredWilhelm His, 1868
- Fate mapLe Douarin quail-chick, 1970s
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
Why the neural crest matters
- It builds most of your face. Nearly the entire craniofacial skeleton — the jaws, the middle-ear ossicles, the frontal and nasal bones, the dentine of your teeth, and the connective tissue of the face — derives from cranial neural crest, not from mesoderm. In the head, crest cells replace mesoderm as the source of skeletal tissue, which is why an evolutionary biologist can call the vertebrate head "an ectodermal skeleton."
- It wires the peripheral nervous system. Every sensory neuron in your dorsal root ganglia, every sympathetic neuron, every Schwann cell wrapping a peripheral axon, and the roughly 500 million neurons of the enteric nervous system that run peristalsis are neural crest descendants. Damage or failure of this program leaves organs unwired.
- It pigments the body. Melanocytes — the pigment cells of skin, hair, iris, and the stria vascularis of the inner ear — are crest-derived. This is why pigment loss and deafness so often travel together: the same migratory cells that color the skin also support hearing.
- It divides the heart's outflow. Cardiac neural crest cells migrate into the developing outflow tract and form the aorticopulmonary septum that separates the aorta from the pulmonary artery. Ablating this population in chick produces persistent truncus arteriosus, and the same failure underlies the conotruncal heart defects of DiGeorge syndrome.
- It makes stress hormones. The chromaffin cells of the adrenal medulla, which secrete adrenaline and noradrenaline in the fight-or-flight response, are modified sympathetic neurons of neural crest origin — a paraneuron lineage shared with the sympathetic ganglia.
- Its failures are a whole disease family. Because one cell population seeds so many tissues, "neurocristopathies" span pigment-deafness syndromes, aganglionic megacolon, cleft palate with heart defects, and pediatric cancers such as neuroblastoma and melanoma — all traceable to a shared embryonic origin.
- It is a window into vertebrate origins. Because invertebrate chordates lack a true migratory crest, comparing its gene-regulatory network across amphioxus, lamprey, and jawed vertebrates reveals how a single new cell type helped launch the vertebrate body plan.
Common misconceptions
- "The neural crest is part of the neural tube." Premigratory crest cells sit at the dorsal-most tip of the neural tube, but they are a distinct population specified at the neural plate border before the tube closes. Their defining act is to leave the tube. The central nervous system neurons stay behind; crest cells become the peripheral nervous system and much of the non-neural body.
- "Neural crest only makes nervous tissue." The name misleads. Crest cells make cartilage, bone, dentine, smooth muscle, corneal endothelium, meninges, adipocytes, and endocrine cells — mesenchymal and skeletal fates that are elsewhere the monopoly of mesoderm. This ectoderm-making-skeleton is precisely why it earns "fourth germ layer" status.
- "Delamination is just the cells falling out." Delamination is an active, tightly regulated EMT. Cells must switch cadherins, dismantle apical tight junctions, lose apicobasal polarity, degrade the basal lamina, and time their exit to the cell cycle. Blocking any step — for example, forcing continued N-cadherin expression — traps cells in the neuroepithelium.
- "Every crest cell is equally potent everywhere." Individual premigratory crest cells are broadly multipotent, but axial level constrains the realistic repertoire. Only cranial crest readily forms cartilage and bone; trunk crest does not normally make skeleton in situ. Hox-gene expression along the body axis is a major reason cranial and trunk crest behave differently.
- "Migration routes are random." Trunk crest follows two stereotyped paths: an early ventromedial route through the anterior half of each somite (making neurons and glia) and a later dorsolateral route between ectoderm and somite (making melanocytes). Cells are steered by attractants and repellents — ephrins, semaphorins, and the anterior/posterior somite boundary — not by chance.
- "The crest is a vertebrate-only invention with no precursor." The crest is a vertebrate novelty, but its gene-regulatory network is a redeployed, elaborated version of border programs present in the last common chordate ancestor. Amphioxus and tunicates express homologs of the specifier genes at the neural plate border without producing a fully migratory, multipotent crest.
How the neural crest works, step by step
1. Border induction. The neural crest is specified at the border between the neural plate (future CNS) and the non-neural ectoderm (future epidermis). Intermediate levels of BMP signaling, combined with Wnt and FGF inputs from adjacent mesoderm and ectoderm, switch on the neural plate border specifier genes: Pax3, Pax7, Msx1/2, Zic1, and Dlx factors. This is a classic morphogen-gradient readout — too much BMP makes epidermis, too little makes neural plate, and the intermediate zone becomes crest and placode.
2. Neural crest specification. The border cells then express a second tier of neural crest specifier transcription factors — FoxD3, Sox9, Sox10, Snail2 (Slug), Ets1, Twist, Id, and c-Myc. These define crest identity and prime the cells for both migration and their later multipotent fates. FoxD3 and the SoxE factors Sox9/Sox10 are the master regulators; Sox10 in particular is required for glia and melanocytes and is mutated in several neurocristopathies.
3. Epithelial-mesenchymal transition and delamination. After neural tube closure, specifier factors — Snail2 and Zeb2 above all — repress the cell-adhesion genes that hold cells in the epithelium. N-cadherin and cadherin-6B are transcriptionally downregulated and proteolytically cleaved, while migratory cadherins (cadherin-7, cadherin-11) come on. Apical tight and adherens junctions dissolve, apicobasal polarity is lost, the actin cytoskeleton is remodeled through Rho-GTPase signaling, and matrix metalloproteinases (MMP-2) and ADAM proteases degrade the basal lamina. Delamination is coupled to the cell cycle at the G1/S transition. The cell emerges as a free mesenchymal cell — the same EMT logic that drives gastrulation and, pathologically, cancer metastasis.
4. Migration along stereotyped corridors. Freed crest cells crawl through extracellular-matrix-rich spaces guided by contact inhibition of locomotion, co-attraction, and molecular signposts. In the trunk, an early ventromedial stream passes through the anterior half of each somite (the posterior half repels it via ephrin/Eph and semaphorin/neuropilin signaling), forming a segmented pattern of ganglia; a later dorsolateral stream between ectoderm and somite carries prospective melanocytes. Cranial crest migrates in three broad streams into the pharyngeal (branchial) arches.
5. Differentiation to final fate. Once at their destinations, crest cells read local signals and differentiate. BMPs in the dorsal aorta region drive sympathetic and chromaffin fates via Mash1/Phox2b; endothelin-3 and GDNF-RET signaling drive enteric neuron colonization of the gut; Wnt and Kit signaling drive melanocyte fate; and in the pharyngeal arches, endothelin-1 and a nested Dlx code pattern the crest-derived jaw skeleton. The result is the full derivative catalog — neurons, glia, pigment, skeleton, and endocrine cells — assembled from one migratory founder population.
Neural crest vs central nervous system (neural tube)
| Feature | Neural crest | Neural tube / CNS |
|---|---|---|
| Position of origin | Dorsal-most tip, neural plate border | The tube itself (floor, walls, roof) |
| Fate of the cells | Leave the tube (delaminate) | Stay and form brain + spinal cord |
| Adhesion behavior | Undergo EMT, become mesenchymal | Remain epithelial (neuroepithelium) |
| Migration | Extensive, long-range | Local (radial migration within tube) |
| Key specifiers | FoxD3, Sox9/10, Snail2, Ets1 | Sox1/2/3, Pax6, Nkx, Olig2 |
| Main derivatives | PNS, melanocytes, craniofacial skeleton | CNS neurons and glia |
| Skeletal potential | Yes (cranial crest → bone, cartilage) | None |
Cranial vs trunk neural crest
| Property | Cranial (cephalic) crest | Trunk crest |
|---|---|---|
| Axial level | Head, before somites; Hox-negative rostrally | Along somites; Hox-positive |
| Skeletal fate | Yes — jaws, skull, ossicles, dentine | No skeleton in situ |
| Migration route | Streams into pharyngeal arches | Ventromedial + dorsolateral through somites |
| Neural fates | Cranial sensory & parasympathetic ganglia | Dorsal root & sympathetic ganglia |
| Pigment | Melanocytes | Melanocytes (dorsolateral stream) |
| Cardiac contribution | Yes — cardiac crest → aorticopulmonary septum | No |
| Signature disorder | Treacher Collins, DiGeorge, cleft palate | Hirschsprung, neuroblastoma |
Famous experiments and history
- Wilhelm His (1868). Working on chick embryo serial sections, His identified a cord of cells wedged between the closing neural tube and the overlying ectoderm and named it the Zwischenstrang — the intermediate cord. This is the first description of the neural crest, though its remarkable fates would take a century to establish.
- Julia Platt (1893). Studying the mudpuppy Necturus, Platt made the then-heretical claim that the visceral cartilages of the head and the dentine of teeth came from ectodermal neural crest, not mesoderm. Contemporaries rejected it because it violated the germ-layer doctrine; she was proved right decades later, and the "ectomesenchyme" she inferred is now textbook.
- Nicole Le Douarin's quail-chick chimeras (from 1969). Le Douarin exploited a permanent natural cell marker — the condensed heterochromatin nucleolus of quail cells, absent in chick — to graft quail neural tube into chick hosts and follow every crest-derived cell through the entire life of the animal. The chimeras produced the complete crest fate map and axial-level derivative catalog still used today, and established that the gut's enteric nervous system comes from vagal and sacral crest.
- Bronner-Fraser and Fraser single-cell lineage tracing (1988). By injecting a fluorescent dye into single premigratory crest cells in the chick, they showed that one cell could give rise to multiple derivative types — direct proof that individual neural crest cells are multipotent, not a pre-sorted mixture of committed precursors.
- Kirby's cardiac crest ablation (1983). Margaret Kirby and colleagues surgically removed a defined region of premigratory cranial crest in chick and produced persistent truncus arteriosus — a failure of outflow-tract septation. This defined the "cardiac neural crest" and explained the heart defects seen when this population is disrupted in humans.
- Gans and Northcutt's "New Head" hypothesis (1983). Carl Gans and R. Glenn Northcutt argued that the vertebrate head is a major evolutionary novelty built largely from neural crest and cranial placodes, and that this new head enabled the transition from passive filter-feeding to active predation. It reframed the neural crest as a driver of vertebrate evolution, not merely an embryological curiosity.
Frequently asked questions
Why is the neural crest called the fourth germ layer?
Classical embryology recognizes three germ layers laid down at gastrulation: ectoderm, mesoderm, and endoderm. The neural crest is called the fourth germ layer because it behaves like an independent lineage with its own vast fate repertoire. It arises from ectoderm at the neural plate border, yet it produces derivatives that other ectodermal tissue never makes — cartilage, bone, dentine, smooth muscle, and adipocytes — cell types that are otherwise the exclusive domain of mesoderm. This mesenchymal, skeleton-forming ectoderm is so unusual that Nicole Le Douarin and Brian Hall argued it deserves germ-layer status. The label is a conceptual honorific rather than a claim that the crest forms before or independently of the three canonical layers: it is induced from ectoderm, not from the primitive streak, but its breadth of potential and its early, coherent segregation justify treating it as a fourth founding population of the vertebrate body.
How do neural crest cells leave the neural tube?
Neural crest cells leave by undergoing an epithelial-mesenchymal transition (EMT) at the dorsal neural tube, a process also called delamination. While still epithelial, premigratory crest cells are held in the neuroepithelium by tight junctions and by cadherins — mainly N-cadherin and cadherin-6B. Signals including BMP, Wnt, and FGF from the surrounding tissue activate the transcription factors Snail2 (Slug), FoxD3, Sox9, and Sox10, plus the zinc-finger factor Zeb2. These repress cell-cell adhesion genes: N-cadherin and cadherin-6B are downregulated and cleaved, while migratory cadherins such as cadherin-7 and cadherin-11 come on. The cells break down their apical junctions, lose apicobasal polarity, remodel their actin cytoskeleton, degrade the basal lamina with matrix metalloproteinases like MMP-2 and ADAM proteins, and detach from the epithelium as free mesenchymal cells. Cell-cycle timing matters too — delamination is coupled to the G1-to-S transition. Once detached, they crawl into the embryo along extracellular-matrix corridors.
What do neural crest cells become?
Neural crest cells are famously multipotent and produce an enormous range of derivatives, which is why they are grouped by axial level. Cranial crest builds most of the facial and skull bones and cartilage, the dentine-forming odontoblasts of teeth, and connective tissue of the face. Cardiac crest, a subset of cranial-vagal crest, forms the septum that divides the outflow tract into the aorta and pulmonary artery. Trunk crest generates the sensory neurons of dorsal root ganglia, sympathetic neurons, Schwann cells, and the catecholamine-secreting chromaffin cells of the adrenal medulla. Vagal and sacral crest colonize the gut to form the entire enteric nervous system — the roughly 500 million neurons that run peristalsis. And crest cells from every axial level give rise to melanocytes, the pigment cells of skin, hair, and the inner ear. The crest also makes smooth muscle, corneal endothelium, the meninges around the forebrain, and endocrine cells such as the calcitonin-producing C-cells of the thyroid.
What is a neurocristopathy?
A neurocristopathy is a birth defect or disease caused by abnormal formation, migration, differentiation, or survival of neural crest cells — a term coined by Robert Bolande in 1974. Because the crest builds so many tissues, its failures are strikingly diverse. Hirschsprung disease is congenital aganglionosis of the distal gut: enteric crest fails to colonize the hindgut, often through mutations in RET or EDNRB, so the bowel cannot relax and megacolon results, affecting about 1 in 5,000 births. Waardenburg syndrome combines patchy pigment loss and deafness from melanocyte deficiency, linked to PAX3, MITF, and SOX10. DiGeorge (22q11.2 deletion) syndrome disrupts cardiac and pharyngeal crest, causing outflow-tract heart defects, thymic and parathyroid hypoplasia, and cleft palate. Familial dysautonomia, Treacher Collins syndrome, CHARGE syndrome, and the childhood cancers neuroblastoma and melanoma are also neurocristopathies. They can be isolated or combined, reflecting which crest population and which developmental step was hit.
Who discovered the neural crest and when?
The Swiss anatomist Wilhelm His described the neural crest in 1868 in chick embryos, naming the cord of cells lying between the neural tube and the overlying ectoderm the Zwischenstrang, or intermediate cord. For nearly a century its fate could only be inferred from serial sections. The definitive fate map came from Nicole Le Douarin in the 1970s, who exploited a natural marker: the nucleus of the Japanese quail carries a dense heterochromatin condensation absent in chick. By grafting quail neural tube into chick embryos, she could follow every crest-derived cell for the animal's whole development, tracing where it went and what it became. These quail-chick chimeras, published from 1969 onward, revealed the full derivative catalog and the axial-level fate map still used today. Single-cell lineage dye injections by Marianne Bronner-Fraser and Scott Fraser in the late 1980s confirmed that individual premigratory crest cells are multipotent.
Why is the neural crest considered a key vertebrate innovation?
The neural crest is one of the defining novelties that separate vertebrates from their invertebrate chordate relatives such as amphioxus and tunicates, which lack a bona fide migratory crest. In 1983 Carl Gans and R. Glenn Northcutt argued in their New Head hypothesis that the vertebrate head is a fundamentally new structure built largely from neural crest and cranial placodes, and that this new head enabled the shift from passive filter-feeding to active predation. Crest-derived jaws, cranial sensory ganglia, pigment, and the elaborated skull gave early vertebrates the ability to seize, sense, and pursue prey. The crest's ancestral gene-regulatory network — the border specifiers Pax3/7, Msx1, Zic1, followed by the neural-crest specifiers FoxD3, Sox9/10, Snail2, Ets1, and Twist — appears to be a redeployed and elaborated version of programs present in the last common chordate ancestor. Its evolution is a textbook case of how a new cell type can drive a major body-plan transition.