In the last Newsletter [7(1)] we described the identification of the gene
ced-3, which appears to be necessary for the initiation of programmed cell death in C. elegans. We have constructed double mutants involving
ced-3 and three previously identified genes known to affect cell deaths:
ced-1 and
ced-2 mutants prolong the highly refractile stage of cell death (Ed Hedgecock, personal communication), and
nuc-1 mutants fail to degrade the DNA of dying cells and result in the persistence of pycnotic nuclei (John Sulston, personal communication). The epistasis of
ced-1 and
ced-2 over
nuc-1 and of
ced-3 over all three other genes suggests that these genes act sequentially in a developmental pathway for programmed cell death: [See Figure 1] We believe that this program is expressed by most cells that undergo programmed cell death, but there may be exceptions. For example, the death of the male linker cell, which requires the presence of a 'killer cell' (Sulston et al., Develop. Biol. 78: 542, 1980), can occur in
ced-3 mutants. Perhaps
ced-3 is required only for deaths that are cell autonomous. At least some cells that survive in
ced-3 mutants differentiate. We previously noted that the postdeirids of
ced-3 animals contain supernumerary dopaminergic neurons (presumably the cells that die in wild type) and that the male-specific cephalic companion neurons (the homologues of which die in wild-type hermaphrodites) appear to survive and express normal characteristics in
ced-3 hermaphrodites. More recently we have obtained evidence that two additional classes of surviving cells differentiate in
ced-3 animals. First,
ced-3 mutants contain supernumerary pharyngeal serotonergic neurons: in
ced-3 animals four pharyngeal cells have a serotonin-uptake system, whereas in wild-type animals only two neurons, known as the NSM's, accumulate exogenous serotonin (Horvitz et al., Science 216: 1012, 1982). John Sulston has found that the embryonically-generated NSM's are the sisters of cells that die in N2, which suggests that the extra serotonin cells in
ced-3 animals may be the sisters of the NSM's. Second, the hermaphrodite-specific neurons (HSN's), the homologues of which die in wild-type males, appear to survive and express normal characteristics in
ced-3 males. Carol Trent has recently isolated a number of Egl (egg-laying defective) mutants that lack HSN function (see Trent, Tsung and Horvitz, this Newsletter). Two of these mutants,
egl-1(
n487) V and
n695 V, have no HSN's at hatching. Both of these mutations are dominant and are suppressed by
ced-3.
ced-3; lay eggs and have HSN's, suggesting that the Egl phenotype of
egl-1 reflects the expression in hermaphrodites of the normally male-specific program for HSN cell death. That
ced-1; odite embryos have supernumerary cell deaths when and where the HSN-homologues die in wild-type males supports this interpretation (we used
ced-1 to facilitate the identification of these cells).
ced-3; 5) animals have HSN's; however, other phenotypes associated with
n695 (see Trent and Horvitz, this Newsletter) are not suppressed by
ced-3. We are currently generating Egl+ revertants of
eg1-1 to isolate additional
ced-3 alleles as well as mutations in other genes that block programmed cell death.