Interleukin-1β-converting enzyme-related proteases (IRPs) and mammalian cell death: Dissociation of IRP-induced oligonucleosomal endonuclease activity from morphological apoptosis in granulosa cells of the ovarian follicle

The Caenorhabditis elegans death susceptibility gene, ced-3, has a number of homologs in vertebrate species, including interleukin-1β (IL-1β)-converting enzyme (ICE), Ich-1_long, and CPP32. These genes, which encode a family of related proteases, have been shown to induce apoptosis when transfected into eukaryotic cells. However, it remains to be determined whether these proteases are involved in apoptotic cell death under physiological conditions. The purpose of these studies was to examine the role of ICE-related proteases (IRPs) in apoptosis using a physiologically relevant model system, the ovarian follicle. Somatic granulosa cells within ovarian follicles undergo apoptosis during follicular atresia, a process responsible for the depletion of greater than 95% of the follicles established in the postnatal ovary. To accomplish these studies, we cloned partial rat complementary DNAs encoding ICE, Ich-1, and CPP32 and used these complementary DNAs to examine the gonadotropin regulation of ICE, Ich-1, and CPP32 gene expression in the immature rat ovary. We also examined levels of ICE activity in healthy and atretic rat follicles by monitoring the conversion of exogenous pro-IL-1β to the active cytokine, and then evaluated the actions of recombinant IL-1β on apoptosis in follicles incubated in vitro. Finally, we tested the requirement for IRP activity in granulosa cell apoptosis and follicular atresia by incubating follicles without and with IRP inhibitors. Northern blot analysis of total RNA samples indicated that gonadotropin-promoted follicular survival was associated with reduced ovarian expression of messenger RNAs encoding Ich-1 and CPP32. In contrast, ICE messenger RNA levels were extremely low and were not affected by gonadotropin treatment. We were also unable to detect ICE activity in proteins extracted from either healthy or atretic rat follicles, collectively suggesting that ICE per se may not function in granulosa cell death. As another approach to determine whether ICE is involved in atresia, healthy antral follicles were isolated from ovaries of gonadotropin-primed immature rats and incubated for 24 h in the absence or presence of 100 ng/ml transforming growth factor-α (TGFa) without and with 100 ng/ml IL-1β. Granulosa cells within follicles incubated in medium alone exhibited extensive levels of apoptosis, and this onset of apoptosis was prevented by the inclusion of TGFα. Addition of IL-1β did not alter basal levels of apoptosis nor did the cytokine antagonize TGP-α-promoted follicle survival, providing additional evidence that ICE activity is not required for atresia to occur. To further evaluate whether IRPs are involved in atresia, we analyzed the effects of several protease inhibitors on apoptosis in follicles incubated in vitro. Treatment of hormone-deprived follicles with either of two inhibitors of IRP activity, sodium aurothiomalate (0.01-1 mM) or iodoacetic acid (10 μM), effectively suppressed inter- nucleosomal DNA cleavage associated with apoptosis. However, histological analysis of follicles treated with these inhibitors revealed extensive cellular degeneration, dramatically contrasting with the biochemical results obtained through DNA analysis. In summary, these data have provided the first evidence that the expression of IRPs, but not ICE per se, is down-regulated during gonadotropin-promoted follicular survival. Furthermore, the activity of IRPs may be involved in activation of oligonucleosomal endonucleases, an event that can be dissociated from morphological indexes of apoptosis, in granulosa cells during atresia.