We have identified a new suppressor of apoptosis, ICD-1 (Inhibitor of Cell Death-1), that is necessary to protect cells from inappropriate apoptosis and sufficient to block apoptosis during development.
icd-1(RNAi) leads to a dramatic increase in apoptosis. While elimination of detectable ICD-1 causes very early embryonic arrest, reduction in ICD-1 levels results in a large increase in cell corpse number in embryos and larvae. While some of these corpse-like structures are abnormal in size or shape, all corpses are eliminated by a
ced-4 mutation, implying that they arise by apoptosis. This CED-4-dependent death results in elimination of numerous differentiated cell types, including neurons, rays in the male tail, and gut cells. ICD-1 is not only necessary to protect cells from undergoing apoptosis, but is sufficient, when overexpressed, to prevent CED-3-activated cell death during embryogenesis. While CED-4 is required for the inappropriate apoptosis seen in
icd-1(RNAi) mutants, surprisingly CED-3 is not. These results functionally separate CED-3 from CED-4 and may reflect the redundant action of CED-3 and another caspase, perhaps encoded by one of the three described csp genes (Shaham, J. Biol. Chem. 273, 1998). Indeed, we found that the apoptosis seen in
icd-1(-) embryos is blocked by baculovirus
p35, a potent caspase inhibitor, implicating a non-CED-3 caspase in this event. ICD-1 is the homologue of human NAC, one of two subunits of NAC, the nascent polypeptide-associated complex. We also found that RNAi of the gene encoding the C. elegans NAC homologue results in similar effects to those we have seen for
icd-1. ICD-1 contains a putative caspase cleavage site and caspase recruitment domain (CARD) that may facilitate interactions between ICD-1 and CED-3 and/or CED-4. It is expressed in most or all cells, apparently associating with mitochondria. This observation is consistent with the prominent role played by mitochondria in apoptosis and the mitochondrial localization of CED-4 and CED-9, through which it may function. We propose that ICD-1 is a novel apoptotic suppressor that acts to inhibit the function, or CED-4-dependent activation, of CED-3 and other caspases. Further, the caspase-inhibitory activity of ICD-1 may be eliminated by caspase-mediated cleavage. Such a system would provide a positive feedback loop that creates a strong OFF-ON switch during activation of apoptosis.