Oligodendrocytes are the myelinating cells of the central nervous system (CNS), and although they become widely distributed throughout the CNS, they seem to originate in a rather narrow zone in the neural tube. However, the precise location of this zone has been a matter of debate, with studies in rodent and chick embryos producing conflicting results. This has led some researchers to conclude that there are interspecies differences in the origin of oligodendrocytes. However, a new study by Fu et al. goes some way towards resolving the argument by showing that oligodendrocytes are in fact generated at two sites within the spinal cord.

The dorsoventral axis of the developing spinal cord is divided into several domains, each of which expresses a distinct combination of homeodomain proteins. Initial studies indicated that in the chick, oligodendrocyte progenitor cells (OLPs) arise in the Nkx2.2-expressing p3 domain, which generates V3 interneurons at an earlier stage of development. However, in the mouse, the OLPs seemed to originate in the more dorsal motor neuron precursor domain (pMN), which expresses the OLP marker Olig2. By examining the expression patterns of early oligodendrocyte markers, including Nkx2.2 and Olig2, the authors showed that, in fact, both zones could give rise to OLPs in both species. In addition, they found that the Olig2-expressing OLPs eventually acquire expression of Nkx2.2, although the timing of this event differs between species: chick OLPs already express both genes before they begin to migrate, whereas Olig2-expressing OLPs in the mouse, at least during the early phase of oligodendrogenesis, do not express Nkx2.2 until they have left the proliferative zone. Also, in the chick embryo, Nkx2.2-expressing OLPs eventually acquire Olig2 expression, and in vitro studies indicate that this might also happen in mice.

Olig2 and Nkx2.2 co-expression in OLPs coincides with the onset of myelin gene expression in these cells, so Fu et al. investigated whether Nkx2.2 and Olig2 might have an instructive role in oligodendrocyte maturation. They showed that if either of the genes was inactivated using antisense oligonucleotides, myelin gene expression was decreased and the OLPs failed to undergo terminal differentiation. This indicates that Nkx2.2 and Olig2 might cooperate to induce oligodendrocyte differentiation.

This study shows that in both mouse and chick, OLPs are generated in two domains in the spinal cord neural tube, and that they need to express both Nkx2.2 and Olig2 to attain a mature oligodendrocyte phenotype. However, the timing of this co-expression, and of myelin gene activation, differs between the mouse and the chick, and it will be interesting to investigate the functional significance of this difference in relation to CNS maturation in these species.