The mammalian interleukin-1 beta-converting enzyme (ICE) has sequence similarity to the C. elegans cell death gene ced-3. We show here that overexpression of the murine ICE (mICE) gene or of the C. elegans ced-3 gene causes Rat-1 cells to undergo programmed cell death. Point mutations in a region homologous between mICE and CED-3 eliminate the ability of mICE and ced-3 to cause cell death. The cell death caused by mICE can be suppressed by overexpression of the crmA gene, a specific inhibitor of ICE, as well as by bcl-2, a mammalian oncogene that can act to prevent programmed cell death. Our results suggest that ICE may function during mammalian development to cause programmed cell death.

Interleukin-1 beta converting enzyme (ICE) is a mammalian homolog of CED-3, a protein required for programmed cell death in the nematode Caenorhabditis elegans. The activity of ICE can be specifically inhibited by the product of crmA, a cytokine response modifier gene encoded by cowpox virus. Microinjection of the crmA gene into chicken dorsal root ganglion neurons was found to prevent cell death induced by deprivation of nerve growth factor. Thus, ICE is likely to participate in neuronal death in vertebrates.

Apoptosis is a type of cell death that plays an important role in early development and growth of normal adult tissues. It is regulated by physiological stimuli and is present in many species and tissues. The main morphological characteristics are nuclear fragmentation and cellular breakdown in apoptotic vesicles. Internucleosomal DNA fragmentation is an important biochemical feature that is the result of a yet to be isolated endonuclease activity. Experiments using RNA and protein synthesis inhibitors suggest the presence of either intracellular repressors or inducers of apoptosis.

Summary is the only known physiologic substrate of the interleukin-l[3 (IL-l[3)-converting enzyme (ICE), the founding member of the ICE/ced-3 cell death gene family. Since secreted mature IL-113 has been detected after apoptosis, we investigated whether this cytokine, when produced endogenously, plays a role in ceil death. We found that hypoxiainduced apoptosis can be inhibited by either the IL-1 receptor antagonist (IL-11ka) or by neutralizing antibodies to IL-1 or to its type 1 receptor. IL-1Ra also inhibits apoptosis induced by trophic factor deprivation in primary neurons, as well as by tumor necrosis factor ix in fibroblasts. In addition, during the G1/S phase arrest, mature IL-l[3 induces apoptosis through a pathway independent of CrmA-sensitive gene activity. We also demonstrate that Ice, when expressed in COS cells, requires the coexpression ofpro-IL-l~ for the induction of apoptosis, which is inhibited by IL-1Ra. Interestingly, we found that mature IL-113 has antiapoptotic activity when added exogenously before the onset of hypoxia, which we found is caused in part by its ability to downregulate the IL-1 receptor. Our findings demonstrate that pro-IL-l[3 is a substrate of ICE relevant to cell death, and depending on the temporal cellular commitment to apoptosis, mature IL-1 [3 may function as a positive or negative mediator of cell death.