During C. elegans hermaphrodite development, 131 of the 1090 somatic cells generated undergo programmed cell death. The gene
ced-9 appears to play a central role in controlling this process. A gain-of-function mutation that activates
ced-9 prevents the cell deaths that occur during normal C. elegans development. Conversely, mutations that inactivate
ced-9 cause cells that normally live to undergo programmed cell death. Mutations in
ced-3 and
ced-4 ,which block essentially all programmed cell deaths, suppress this
ced-9 (lf)phenotype, indicating that the loss of
ced-9 function results in an inappropriate activation of the pathway for programmed cell death. These observations show that
ced-9 activity is both sufficient and necessary to protect cells from programmed cell death. We proposed that the normal function of
ced-9 is to protect cells that are scheduled to survive from undergoing programmed cell death (Hengartner, Ellis, and Horvitz, Nature 356, 494-499, 1992). To understand better how
ced-9 prevents programmed cell death, we have cloned
ced-9 .RFLP mapping followed by micro-injection of cosmids and various subclones have localized
ced-9 rescuing activity to a 6 kb. We have sequenced this rescuing fragment and have isolated cDNAs hybridizing to this region from the Barstead library. Sequencing of these cDNAs revealed that the putative
ced-9 transcript contains an 846 bp ORF that could encode a 280 amino acid protein. The
ced-9 protein sequence was then used to search protein databases. The protein in the databases that showed the most similarity to
ced-9 was the product of the mammalian proto-oncogene
bcl-2 :the
ced-9 and human
bcl-2 proteins show 23% identity over their entire sequence.
ced-9 possesses the C-terminal hydrophobic tail thought to be important for
bcl-2 localization (S. Korsmeyer; personal communication) or function (Alnemri et al., PNAS 89, 7295-7299,1992). One of the three introns in
ced-9 is in the same position as one of the two introns in human
bcl-2 . Interestingly, the proposed function of the
bcl-2 oncogene is identical to the function we proposed for
ced-9 :
bcl-2 is thought to be required to prevent cells from undergoing programmed cell death. For example, overexpression of
bcl-2 prevents or delays the programmed death of immune cells following a variety of stimuli, such as removal of growth factors, treatment with glucocorticoids or irradiation with low doses of gamma rays. These similarities in both sequence and function suggest that
bcl-2 could be the vertebrate homologue of
ced-9 .We are currently testing this hypothesis in several ways. For example, can human
bcl-2 prevent programmed cell death in worms? Could
ced-9 function in mammals? The fact that human
bcl-2 appears to be able to delay cell death in insects (Alnemri et al., op. cit.) makes these ideas not as far-fetched as they may seem. It is tempting to speculate that not only
ced-9 but in fact the whole cell death pathway has been conserved throughout evolution, and that therefore the molecular mechanism underlying programmed cell death might be essentially the same in organisms as diverse as worms and humans.