6.
Four genes, egl-1, ced-3, ced-4, and ced-9, define a genetic
pathway needed for (almost) all programmed cell deathsThe original isolation by Ellis and Horvitz (1986) of mutants defective in all cell
deaths during nematode development was successful because all cells that are fated
to die use a common pathway to execute programmed cell death (Ellis and Horvitz,
1986) (Figure 3 ). Searching for the desired mutants was simplified by screening in a
genetic background where dead cells are easily recognized because they cannot be
engulfed by neighboring cells (see section on mutants defective in the engulfment
process). Later screens also focused on the identification of cell-death mutants with
extra undead cells (Ellis and Horvitz, 1991; Hengartner et al., 1992). As a result,
these screens, initially numerous loss-of-function (If) alleles of both ced-3 (cell death
abnormal) and ced-4, and a single gain-of-function (gf) allele of ced-9 were identified
(Figure 3 ) (Ellis and Horvitz, 1991; Hengartner et al., 1992). The loss of function
alleles of ced-3 and ced-4 can be classified into allelic series, the strongest alleles of
which presumably result in a complete loss of gene function that leads to the complete
inhibition of all programmed cell death. ‘Undead cells’ that fail to die tend to acquire
a fate similar to their ancestors, but, interestingly, they lose the capacity for further
division.
The isolation of ced-9 loss-of-function (If) mutants was accomplished by looking
for intragenic revertants of a rare ced-9 gain-of-function allele (Hengartner et al.,
1992). The analysis of these revertants revealed extensive apoptosis of cells that
normally do not die within homozygote ced-9 loss-of-function animals. As a
consequence of these extensive deaths, the affected animals die during embryogenesis
(Hengartner et al., 1992). These observations suggest that ced-9 normally functions
to prevent programmed cell death (Figures 2 and 3 ). Genetically, ced-9 acts upstream
of both ced-3 and ced-4, as loss-of-function alleles of these genes can suppress the
extensive cell-death phenotype of ced-9 (lf) animals (Hengartner et al., 1992) (Figures
2 and 3 ). Interestingly, embryos resulting from homozygous strong ced-9 (If) mutants
can die at a very early developmental stage without any overt morphologic indication
of programmed cell death, suggesting that ced-9 might have an essential function in
addition to its antiapoptotic role (Hengartner et al., 1992). The fourth gene that
affects all developmental cell death but that evaded discovery in the initial genetic
screens is egl-1 (egg laying defective) (Conradt and Horvitz, 1998) (Figure 3).
Interestingly, it turned out that egl-1 loss-of-function alleles are defective in all
programmed cell deaths occurring during somatic development. egl-1 function was
determined to be upstream of the other known cell-death components because of the
suppression of programmed cell death caused by the ectopic overexpression of egl-1
in ced-3 (lf), ced-4 (lf), and ced-9 (lf) mutants (Conradt and Horvitz, 1998; 1999).
Mosaic analysis suggests that ced-3 and ced-4 are likely to act cell-autonomously,
indicating that they are needed to act within the cells that die (Yuan and Horvitz,
1990). This conclusion is further supported by experiments showing that ced-3 and
ced-4 overexpression, in neuronal cells that usually do not die, is sufficient to induce
PROGRAMMED CELL DEATH IN C.ELEGANS 167