One of the first steps in nervous system development is the generation of neural progenitor cells that will subsequently give rise to the vast diversity of neurons present in the adult. How neuronal progenitor specification is achieved is a fundamental question that we are still far away from completely understanding. The Sox proteins are a highly conserved group of transcription factors and the expression of some of them (group B) correlates with the commitment of cells to a neural fate. The evolutionary conserved expression of sox genes in neuronal progenitors and the fact that they are regulated by signaling molecules involved in neuronal specification suggests a role for sox genes in establishing neural fate. However, the relevance of Sox proteins in vertebrate neuronal development has been difficult to prove genetically mainly because of presumed redundant functions. C. elegans has only 5 sox genes (sem-2, sox-2, sox-3, sox-4 and egl-13) and we have investigated their potential role in early neurogenesis. First, we have analyzed the expression pattern of all sox genes during embryogenesis by using fosmid-based reporter genes. sem-2 and sox-2 are the sox genes expressed earliest and in a more broad manner during embryogenesis, being expressed in neuron progenitors, although not exclusively neither in all of them. Their expression pattern would be be compatible with a role in early neuronal specification. On the other hand, sox-3, sox-4 and egl-13 are expressed in a few cells during late embryogenesis, when most neurons are already born. Hence, we have mainly focused our efforts in further characterizing sem-2 and sox-2 as potential regulators of early neuronal specification events. For that purpose we have generated a new sox-2 deletion allele taking advantage of the MosDel tecnique. Both sem-2 and sox-2 null mutants are embryonic or early larval lethal and show different degrees of morphogenesis defects. When we crossed sem-2 or sox-2 mutants with a panneuronal reporter we did not observe any obvious neuronal loss and the expression of neuron-type specific reporters was not generaly affected either. These results strongly suggest that, contrary to our initial hypothesis, sox genes are not required for correct neuronal specification. The possibility of sem-2 and sox-2 acting redundantly is unlikely, at least in a broad manner, since their co-expression during embryogenesis in rather limited. Potential redundancy or compensatory mechanisms between different sox genes are currently being investigated.

Affiliations:
- Howard Hughes Medical Institute
- Department of Biochemistry & Molecular Biophysics, Columbia University Medical Center, New York, NY.