Gradients in the developing spinal cord – mitogens join forces with morphogens

Abstract

Developmental neurobiologists are growing accustomed to discoveries of signaling centers that help pattern the developing nervous system. A signaling center can be the focal source of a morphogen – a diffusible molecule whose concentration gradient across an embryonic field serves to specify different cell fates at different levels, or ‘thresholds’. A fine example is found in ventral spinal cord, where Sonic hedgehog (Shh) is released from the floor plate. Culture experiments confirm that neural progenitors adopt different phenotypes when confronted with different concentrations of Shh. Opposing the action of this ventral signaling center is a similar one at the dorsal midline of the spinal cord, the roof plate. Here we find another classical morphogen family, the bone morphogenetic proteins (BMPs). A recent paper by Megason and McMahon suggests that morphogens are not the only class of signaling molecules secreted by the roof plate: a mitogen gradient could exist alongside the morphogen gradient to coordinate neurogenesis along the dorsal–ventral (D–V) axis of the spinal cord [1xA mitogen gradient of dorsal midline Wnts organizes growth in the CNS. Megason, S.G. and McMahon, A.P. Development. 2002; 129: 2087–2098PubMedSee all References[1].The mitogens in question belong to the Wnt family. Although several Wnts are present in embryonic spinal cord, evidence implicates Wnt1 and Wnt3a in stimulating proliferation and inhibiting differentiation along the D–V axis. Their message levels peak at the roof plate. Abolishing this D–V asymmetry on one side (via unilateral electroporation of expression constructs) caused a drastic increase in cell division at the expense of differentiation. Wnts enhance cell division, at least in part, by upregulating D-type cyclins. It is notable that the D–V sequence of progenitor identities was not altered, as this might have been expected had a morphogen gradient been disturbed. By contrast, blocking the Wnt-signaling pathway reduced cell division without promoting premature differentiation, suggesting that there must be something more than a simple Wnt-regulated switch between proliferation and differentiation. In fact, loss-of-function experiments by Muroyama et al. could be interpreted as lowering the dose of a dorsal morphogen in Wnt1;Wnt3a double knockouts: two dorsal cell types were depleted, whereas the next most-ventral cell type was expanded [2xWnt signaling plays an essential role in neuronal specification of the dorsal spinal cord. Muroyama, Y. et al. Genes Dev. 2002; 16: 548–553Crossref | PubMed | Scopus (185)See all References[2]. In other words, these Wnts might be ‘double-dipping’, as both mitogens and morphogens.In the mitogen gradient model, ventral progenitors find themselves ever further from the dorsal growth organizer as the spinal cord enlarges, increasing the likelihood they will drop out of division to differentiate. Megason and McMahon have used computer modeling to explore how such a scenario plays out across a linear array of cells (http://hombiosys.com/downloads/MitogenGradient). They found that the size and shape of the responding tissue are regulated by the slope of the mitogen gradient. The next challenge will be to embellish the model to account for molecules that have both mitogenic and morphogenic activities. What might otherwise be exhausting mental gymnastics could happily be avoided through computational analysis.