Boundary cap (BC) cell derivatives, labelled in red, migrating along the nerve root into the dorsal root ganglion. The star marks the position of the BC. The neuronal marker β-III-tubulin is immunolabelled in blue and nuclei are counterstained in green. NT, neural tube. Image courtesy of P. Charnay, INSERM U 368, Paris, France.

Boundary cap (BC) cells are a neural-crest-derived cell population that cluster at the peripheral-nerve entry and exit points of the developing spinal cord (the dorsal root entry zone and motor exit point, respectively). Here, they 'police' the CNS/PNS interface, where they allow axons, but not whole cells, to pass through. Until recently, little was known about their fate on completion of this task, but an intriguing new report in Nature Neuroscience provides evidence that they make a substantial cellular contribution to the somatic PNS.

In mice, there is a time window — between embryonic day (E) 10 and E15.5 — when expression of the Egr2 (Krox20) gene in the trunk region is confined to BC cells. Maro and colleagues exploited this fact to generate knock-in mice, in which Egr2 expression triggered a recombination event that resulted in permanent activation of a marker gene (either β-galactosidase or yellow fluorescent protein). This enabled them to track the fate of both the BC cells and their progeny.

The authors found that the labelled cells migrated along peripheral axons and colonized the spinal nerve roots and dorsal root ganglia (DRGs). They seemed to give rise to all of the Schwann cells in the dorsal root, as well as a subset of DRG neurons and satellite cells. Most of the BC-derived neurons in the DRG expressed TrkA, a marker of nociceptive neurons, and this observation provided a clue to the sequence of developmental events. Primary sensory neurons are known to arise in two waves — the first wave generates large-diameter proprioceptive and mechanoreceptive neurons, and the second wave generates small-diameter nociceptive neurons. The origin of the second wave was previously unclear, but these results indicate that BC neurons are at least partly responsible.

So, it seems that the neural crest can generate components of the PNS both directly, through an early-migrating cell population, and indirectly, through a BC-cell intermediate. It will be interesting to investigate whether the BC cells represent a true multipotent stem-cell population or a heterogeneous collection of precursors that are already committed to a neuronal or glial cell fate.