Programmed cell death is one mechanism, through which sexual dimorphism is created in the C. elegans nervous system. The hermaphrodite-specific neurons (HSNs), two serotonergic motor neurons required for egg laying in hermaphrodites, and the CEM neurons, four sensory neurons of unknown function in males, are formed in both sexes, hermaphrodites and males. Through programmed cell death, the HSNs are subsequently eliminated in males and the CEMs are eliminated in hermaphrodites (1). The cell-death activator gene egl-1 (egl, egg-laying defective) is the most upstream acting factor of the general cell-death pathway in somatic tissues. Loss-of-function (lf) mutations in egl-1 block most if not all somatic cell deaths. In contrast, gain-of-function (gf) mutations in egl-1 cause the HSNs to inappropriately undergo programmed cell death in hermaphrodites. These gf mutations disrupt a binding site for the zinc finger DNA-binding protein TRA-1A (tra, transformer), the global, terminal regulator of somatic sexual fate, downstream of the egl-1 transcription unit (2). It has therefore been proposed that TRA-1A represses egl-1 transcription in the HSNs (possibly by negatively regulating a HSN-specific activator of egl-1) to ensure the survival of these neurons (2). This model is supported by the finding that the tra-1 gene acts genetically as a negative regulator of egl-1 in the HSNs. We have now begun to analyse the role of the tra-1 and egl-1 genes in the specification of the deaths of the CEMs. The CEMs survive in egl-1(lf) hermaphrodites and also in tra-1(lf) XX animals, indicating that both egl-1 and tra-1 are required for the CEMs to undergo programmed cell death in hermaphrodites. Hence, TRA-1A might not act as a repressor of egl-1 in the CEMs, as it does in the HSNs, but as an activator of egl-1. However, the egl-1(gf) mutations, which disrupt the ability of TRA-1A to repress egl-1 transcription in the HSNs, do not affect the ability of TRA-1A to induce the deaths of the CEMs: in egl-1(gf) hermaphrodites, the CEMs still undergo programmed cell death. This suggests that, in the CEMs, TRA-1A does not function through the TRA-1A binding site downstream of the egl-1 transcription unit. We are currently determining the region of the egl-1 locus that is required to specify the death of the CEMs. To identify other components required for the specification of the CEM death, we are, in addition, screening for mutations that cause the CEMs to survive in hermaphrodites, using a CEM-specific reporter construct (pkd-2::gfp) (3). (1) Sulston, J.E. and H.R. Horvitz. (1977). Dev. Biol., : 110-156. (2) Conradt, B. and H.R. Horvitz. (1999). Cell, 98:317-327. (3) Barr, M.M. and P.W. Sternberg. (1999). Nature, :386-389.