On the crest of becoming vertebrate

The discovery of cells in an invertebrate that share several features with vertebrate neural-crest cells provides insights into how this vital vertebrate cell population might have evolved. See Letter p.371

The evolution of vertebrates is intimately linked to the advent of an embryonic cell population called the neural crest. Neural-crest cells arise in the central nervous system (CNS) and then invade the periphery of the vertebrate embryo, where they differentiate to form a wide range of cell types, from facial cartilage and bone, to pigment cells of the skin, to neurons and glial cells of the peripheral nervous system. The evolution of these features is thought to have imbued vertebrates with their predatory ability1, facilitating their success on Earth. Although all vertebrates have neural-crest cells, how this population evolved has remained a mystery. In this issue, Stolfi et al.2 (page 371) report that a type of neuron in an ascidian (sea squirt) — an invertebrate filter feeder that is a close relative of vertebrates — has intriguingly similar features to neurons derived from the vertebrate neural crest.

The neural crest is characterized by its origin in the CNS, its migratory behaviour and its ability to differentiate into many cell types (multipotency). Although all animals of the chordate phylum, which includes vertebrates, have a similar body plan (including a dorsal CNS; a structure called the notochord that runs down the midline of the embryo; and a segmented trunk), invertebrates lack cells that have all the characteristics of the neural crest. So far, no intermediate cell type — which would be expected to originate from the CNS and become migratory — has been identified in invertebrates.

Stolfi and colleagues investigated the origin of bipolar tail neurons (BTNs) in the ascidian Ciona intestinalis. They found that BTNs arise from precursor cells that originate in the developing CNS and migrate through adjacent tissues before differentiating to form mature neurons. Furthermore, the authors discovered that BTNs have a similar function to sensory neurons in vertebrates.

This study is a good complement to previous work3 demonstrating that precursors to pigment cells are also present in ascidians. These precursors normally remain in the CNS, but can be induced to migrate through misexpression of just one gene, Twist. Taken together, these two studies point to the intriguing hypothesis that the evolution of vertebrates might have involved precursor cells in the CNS of a chordate ancestor gaining the ability to form many cell types — having already developed differentiation programs for forming some neural-crest derivatives, including sensory neurons and pigment cells. Moreover, by demonstrating that BTN precursors migrate from the CNS before differentiating, Stolfi et al. provide evidence that these cells have analogous traits to two major characteristics of the vertebrate neural crest.

The authors also found that BTN precursors have similar genes to those that encode the vertebrate transcription factors Neurogenin and Islet, which are both required for the formation of sensory neurons. However, the cells do not express all the genes expressed by the neural crest, raising the possibility that BTN precursors represent an intermediate cell population capable of some, but not all, neural-crest-like behaviours.

In contrast to ascidians, more-primitive chordates such as amphioxus, a cephalochordate, lack any precursor cells with neural-crest-like characteristics. These species do, however, have neurons and pigment cells4. Moreover, the genomes of all invertebrate chordates, including amphioxus5, harbour genes that are similar to those involved in neural-crest formation in vertebrates. Thus, vertebrate evolution did not require the invention of new genes. Rather, progressive changes that rewired the regulation of gene circuits by using existing factors in new ways probably permitted each transition, from cephalochordates that have no neural crest, to urochordates such as ascidians that have an intermediate cell population with some neural-crest-like characteristics, to the base of vertebrate evolution, with the advent of bona fide neural-crest cells (Fig. 1).

Figure 1: Evolution of the neural crest.

This simplified phylogenetic tree depicts the evolution of the chordate lineage. An embryonic cell population called the neural crest, which migrates extensively and forms many cell types, arose with the advent of vertebrates. Whereas jawless and jawed vertebrates possess neural-crest cells, this population is not present in cephalochordates or urochordates. Stolfi et al.2 have demonstrated that a cell population in the urochordate Ciona intestinalis shares some characteristics with vertebrate neural-crest cells. This may be an intermediate cell population from which the neural crest evolved. (Drawings taken from ref. 8.)

Key questions remain to be answered. For example, it is not clear whether the gene-regulatory networks in the sensory-neuron precursors of ascidians are similar to those of vertebrates. Because signalling pathways that mediate differentiation are thought6 to be crucial for conferring multipotency on the neural crest, it will be interesting to determine how these signalling pathways came under the regulatory control of the genes that induce the formation of neural-crest cells in vertebrates. Finally, although C. intestinalis is useful for experimental analysis, it is highly derived — its genome is much simpler than that of other chordates. Thus, analogous studies should be performed in other, less-derived ascidians, and in other chordates at different positions on the evolutionary tree.

The idea that cell-differentiation programs have come under the progressive control of neural-crest genes during vertebrate evolution is not new7. It has long been thought likely that evolutionary precursors of neural-crest cells lacked the multipotency that defines this cell population in vertebrates. Stolfi and colleagues' study neatly demonstrates the existence of an intermediate cell type that might have gained this ability, enabling the evolution of the neural crest.Footnote 1


  1. 1.

    See all news & views


  1. 1

    Gans, C. & Northcuttt, R. G. Science 220, 268–273 (1983).

    ADS  CAS  Article  Google Scholar 

  2. 2

    Stolfi, A., Ryan, K., Meinertzhagen, I. A. & Christiaen, L. Nature 527, 371–374 (2015).

    ADS  CAS  Article  Google Scholar 

  3. 3

    Abitua, P. B., Wagner, E., Navarrete, I. A. & Levine, M. Nature 492, 104–107 (2012).

    ADS  CAS  Article  Google Scholar 

  4. 4

    Holland, L. Z. & Holland, N. D. Curr. Opin. Neurobiol. 9, 596–602 (1999).

    CAS  Article  Google Scholar 

  5. 5

    Yu, J.-K., Meulemans, D., McKeown, S. J. & Bronner-Fraser, M. Genome Res. 18, 1127–1132 (2008).

    CAS  Article  Google Scholar 

  6. 6

    Anderson, D. J. Neuron 3, 1–12 (1989).

    CAS  Article  Google Scholar 

  7. 7

    Green, S. A., Simoes-Costa, M. & Bronner, M. E. Nature 520, 474–482 (2015).

    ADS  CAS  Article  Google Scholar 

  8. 8

    Sansom, R. S., Gabbott, S. E. & Purnell, M. A. Nature 463, 797–800 (2010).

    ADS  CAS  Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Marianne E. Bronner.

Related links

Related links

Related links in Nature Research

Evolutionary developmental biology: Head start

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Bronner, M. On the crest of becoming vertebrate. Nature 527, 311–312 (2015). https://doi.org/10.1038/nature15645

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.