might act indirectly, by affecting genes that are needed for specific cells to acquire
their proper developmental cell fate. Alternatively, genes might be directly involved
in the regulation of the cell-death machinery in specific target cells. The best-studied
mutation that affects cell type-specific cell death is the egl-1 gain-of-function mutation
that causes the inappropriate death of both hermaphrodite HSN neurons. These two
neurons are used in hermaphrodites for the innervation of vulval muscle cells to
regulate egg laying. Consequently, and hence the name, egg laying is defective in egl-1
(gf) animals (Desai et al., 1988). In males, which do not need to lay eggs, the HSNs
are also generated, but subsequently eliminated by programmed cell death.
Interestingly, it turned out that all egl-1 (gf) mutations are within a DNA binding
site of the tra-1 sex determination gene (Conradt and Horvitz, 1999). In
hermaphrodites, TRA-1 acts at the egl-1 promoter as a transcriptional repressor of
the induction of programmed cell death of the HSN neuron by inhibiting the
expression of egl-1 transcription (Conradt and Horvitz, 1999). In males, TRA-1 does
not bind to the transcriptional repressor; hence, HSNs die in male animals.
Another example of genes that are likely to control directly the induction of
apoptosis in a specific subset of cells are the ces (cell-death specification) genes, which
affect the apoptosis of the sister cells of the pharyngeal NSM neurons. Mutations in
the two ces genes (ces-1 and ces-2) define a linear pathway that leads to the cell type-
specific activation of apoptosis (Ellis and Horvitz, 1991). Epistasis analysis reveals
that within this pathway ces-2 seems to be a negative upstream regulator of ces-1. The
evidence for the epistatic relationship is derived from the analysis of various single
and double mutants. ces-1 gain-of-function mutants prevent death of the NSM sisters
that usually die (as well as preventing death of two additional cells, the sisters of the
12 neurons). NSM sister cell survival is also triggered by the loss of function of ces-2.
Since the cell survival caused by loss of ces-2 function is suppressed by loss of ces-1
function, ces-1 probably acts downstream of ces-2. ces-1 is likely to act upstream of
the core apoptotic machinery since the observed cell death in ces-1 (If) is still readily
suppressed by ced-3 mutants (Ellis and Horvitz, 1991; C. elegans Sequencing
Consortium, 1998). The ces-2 gene encodes a putative bZIP transcription factor,
suggesting that some nematode cell deaths might be regulated at a transcriptional
level (Ellis and Horvitz, 1991; Metzstein et al., 1996). ces-lencodes a Snail family
zinc-finger protein, most similar in sequence to the Drosophila neuronal
differentiation protein Scratch (Ellis and Horvitz, 1991; Metzstein and Horvitz,
1999). CES-2 may directly repress ces-1 transcription by binding to a transcriptional
element within the ces-2 promoter (Ellis and Horvitz, 1991; Metzstein and Horvitz,
1999). Transcription of ces-1 dependent genes may be required to prevent death of
the NSM sister cells (Ellis and Horvitz, 1991; Metzstein and Horvitz, 1999). It is
unclear whether egl-1 transcription is inhibited by ces-1 in NSM sister cells. Thus, a
transcriptional cascade may be needed to control death of specific cells such as the
NSM sister cells in C. elegans.
172 GENETICS OF APOPTOSIS