Neural patterning: the role of Nkx genes in the ventral spinal cord.

The mammalian nervous system is extremely complex, consisting of many thousands of distinct neuronal cell types, any one of which may require multiple cell interactions to generate a functional circuit. The elaboration of this circuitry is dependent in no small part on the logic of the developmental program that ensures that appropriate cell types are generated at specific positions in the brain or spinal cord, in the correct numbers, and at the correct stage of development. Given the relative simplicity of the spinal cord circuitry, it has become the model of choice for addressing these problems. Only 7 years ago our understanding of neural development in the spinal cord was rudimentary. A number of studies had highlighted the importance of regional organizing centers, both outside and within the developing neural plate and neural tube, in generating the early pattern (for review, see Edlund and Jessell 1999). However, none of these putative inductive signals had been identified. Consequently, there was no molecular framework within which to construct a plausible model for generating cell type diversity in the developing spinal cord. As we fast-forward to the present, the situation is dramatically different. The first major inroad came with the discovery of Sonic hedgehog (Shh; Echelard et al. 1993; Krauss et al. 1993; Riddle et al. 1993; Roelink et al. 1994). Shh encodes a ventralizing signal produced by two ventral midline structures, the notochord, which underlies the ventral neural plate, and the floor plate, a specialized population of glial cells at the ventral midline of the developing hindbrain and spinal cord. It is now clear that Shh is both necessary and sufficient for the induction of the floor plate (notwithstanding evidence that, in some organisms, cell lineage and not simply induction may account for some floor plate cell fates; Le Douarin and Halpern 2000; Placzek et al. 2000), motor neurons, and a number of distinct ventral interneurons (for review, see Jessell 2000). Moreover, there is strong evidence that the concentration of Shh dictates the position where distinct ventral cell identities arise; developmentally naive intermediate neural plate explants respond to small incremental changes in the external concentration of Shh protein to give rise to distinct cell fates (Ericson et al. 1997). At least five distinct cell identities are thought to arise as a primary response to Shh signaling (Fig. 1). These are (from ventral to dorsal) floor plate, V3 interneurons, motor neurons, V2 interneurons, and V1 interneurons. This general model leads to three central questions. How is the Shh concentration gradient generated? How is this gradient translated into distinct transcriptional responses? And finally, what are the transcriptional regulators that lock in a particular neuronal or glial fate? Little progress has been made in addressing the first question. Shh undergoes a unique processing event that appends a cholesterol moiety to the carboxy-terminal amino acid of the amino-terminal signaling moiety (Porter et al. 1996). As a result, the modified protein appears to be retained at the cell surface. However, antibody blocking experiments demonstrate that at least some protein does travel beyond Shh-expressing cells and into the ventral field of Shh-responsive cells (Ericson et al. 1996). This conclusion is also supported indirectly by the observed transcriptional up-regulation of the gene encoding a multipass membrane protein, Patched-1 (Goodrich et al. 1996; Marigo and Tabin 1996), which serves as a general Hedgehog (Hh) receptor and is a general transcriptional target of the pathway. Many lines of evidence indicate that transduction of an Hh signal modulates the activity and/or expression of the Gli/Ci family of zinc-finger DNA binding proteins (for review, see Ruiz i Altaba 1999). In the mouse, which has three Gli family members, Gli2 loss-of-function mutants have a complete absence of floor plate development, consistent with a specific role for Gli2 in mediating the role of Shh in floor plate induction (Ding et al. 1998; Matise et al. 1998). The simplest model would then indicate that other Gli members may direct the development of other ventral cell identities, either uniquely or in combination. However, this does not appear to be the case, at least as far as current publications on the analysis of combinatorial mutants in Gli genes indicate. If this result holds, this leaves open two opposing interpretations. In the first, Shh plays a less direct role in inducing all ventral cell fates, that is, there may be secondary signals. Evidence has been presented for 1E-MAIL amcmahon@mcb.harvard.edu; FAX (617) 496-3763. Article and publication are at www.genesdev.org/cgi/doi/10.1101/ gad.840800.