In summary, the results presented underscore the importance of intracellular potassium homeostasis in suppressing activity of effector molecules required for the execution phase of apoptosis (i.e., caspases and nucleases) and suggest that loss of potassium from oocytes and granulosa cells may be involved in generating the cascade of events leading to their demise.
Because of the fact that changes in cellular morphology and internucleosomal DNA cleavage associated with apo-ptosis are events that reportedly depend upon the activity of caspases, we next employed substrate cleavage assays to assess a role for potassium in the regulation of caspases during apoptosis of oocytes and granulosa cells. We observed that KCl, NaCl, and LiCl effectively blocked caspase-1-mediated processing of pro-IL1p and caspase-3-catalyzed cleavage of cellular actin. These data differ from those of Hughes et al. using thymocytes and those of Bortner et al. using S49-Neo lymphoma cells; those studies showed that potassium prevented procaspase processing (i.e., generation of the active enzyme) but did not inhibit active caspases, as assessed by cleavage of a tetra-peptide (DEVD) known to be a substrate for several members of this family of enzymes.
To determine whether the effect of potassium on inter-nucleosomal DNA cleavage in ovarian cells occurs at the level of a direct suppression of endonucleolytic activity, we used a number of previously characterized cell-free assays to monitor activity of nuclear extract-derived or purified nucleases. The results from these experiments demonstrated that potassium directly suppresses activity of the nucle-ase(s) responsible for internucleosomal DNA cleavage that is present in granulosa cell nuclei, a finding consistent with data derived from comparable analyses of potassium effects on internucleosomal DNA-cleaving nucleases involved in thymocyte or S49-Neo lymphocyte apoptosis.
The analysis of the potassium-independent chromatin cleavage patterns in oocytes proved intriguing. Although potassium apparently did not completely prevent DNA degradation, it did alter its pattern of electrophoretic movement as assessed by the comet assay. On the basis of the results of the ISEL analysis, we conclude that potassium prevents activation and/or activity of the endonuclease or endonucleases responsible for final degradation of oocyte chromatin to low-molecular weight fragments (as revealed by the absence of a plume of DNA outside of the oocyte when assessed by the comet assay) but that higher order DNA cleavage still occurs (as assessed by ISEL). The latter would only permit accumulation of the cleaved high-molecular weight chromatin against the oocyte plasma membrane.
Interestingly, suppression of oocyte condensation and fragmentation could also be reproduced with LiCl, but not with NaCl, indicating some level of ion selectivity for these effects. Although previous in vitro studies with thymocytes indicated that the cell death endpoints studied were equally affected by all monovalent cations, the general similarity in the results across these two studies suggest that ionic strength, as opposed to ionic specificity, is the main modifier. The importance of potassium in vivo is therefore probably due to its high ionic concentration, as opposed to the presence of potassium per se, in the cytoplasm of viable cells. The finding that LiCl was as effective as KCl in preventing some of the biochemical and morphological events associated with apoptosis may be related to the fact that lithium can pass through some types of potassium channels, resulting in an exchange of potassium for lithium inside the cell.
Recent evidence, derived primarily from studies of cultured thymocytes, suggests that a major efflux of intracellular potassium occurs in the early stages of apoptosis and that this efflux is required for the activation of key components of the cell death machinery. In the present study, we undertook a series of experiments to assess potassium levels during apoptosis of ovarian germ cells and granulosa cells and the role of this ion in controlling several morphological and biochemical characteristics of cell death. Using anti-cancer drug-treated murine oocytes as a model for female germ cell apoptosis, we observed a consistent decrease in intracellular levels of potassium as the oocytes died.
The direct suppressive effects of these ions could be overcome, however, because all three ions tested were unable to suppress plasmid DNA degradation catalyzed by high amounts of nuclease (0.1 U/reaction; Fig. 9C; data not shown for 0.1 U DNase-I) whereas SAM remained an effective inhibitor of nuclease activity (Fig. 9C). The ability of 150 mM KCl to inhibit plasmid DNA degradation catalyzed by increasing amounts of DNase-I or DNase-II is shown in Figure 9, B and D, respectively.
Direct Inhibition of Nuclease Activity by KCl
Because it was possible that the potassium-mediated inhibition of internucleosomal DNA cleavage in isolated granulosa cell nuclei could be indirect via a suppression of nuclear protease activity, two final experiments were conducted to further examine the possible direct actions of KCl on nuclease activity. Consistent with data recently reported using a cell-free assay to monitor granulosa cell nuclease activity, nuclear protein extracts prepared from nona-poptotic granulosa cells catalyzed cleavage of linearized plasmid DNA (Fig. 8). However, inclusion of 150 mM KCl, NaCl, or LiCl in the reaction mixture completely suppressed nuclease activity present in the granulosa cell nuclear protein extracts (Fig. 8). Because all nuclease activity in this assay was suppressed by potassium, it may be that the large-fragment cleavage enzyme was not efficiently extracted from the granulosa cell nuclei or that the large-frag-ment cleaving nuclease requires DNA to be in a chromatin conformation in order to be active.
However, PFGE analysis revealed that none of the ions tested prevented high molecular weight DNA fragmentation in cultured follicles, although an attenuation of 50-kb fragment accumulation was consistently noted in KCl- and LiCl-treated follicles compared with the case of controls (Fig. 6A). These results suggest that a potassium-independent pathway or pathways of DNA degradation is also present in ovarian somatic (granulosa) cells, findings confirmed and extended by fluorescence analysis of nuclear morphology that revealed that pyknosis (brightly fluorescent, condensed nuclei) remained widespread in the granulosa and theca-interstitial cell layers of follicles incubated with 150 mM KCl (Fig. 6D).
Potassium and Ovarian Somatic (Granulosa)
To explore the possibility that this novel potassium-independent pathway for nuclear destruction occurs in non-germline ovarian cell types as well, we next examined potassium movement during granulosa cell death induced by serum-free culture in vitro. After 24 h in culture, healthy (nonapoptotic, noncondensed) granulosa cells stained brightly with PBFI, whereas shrunken (apoptotic) or necrotic (trypan blue-positive) granulosa cells stained only lightly with PBFI (Fig. 4). To extend these observations, we next examined a follicle culture model that has been extensively used to characterize the pathways responsible for the regulation of granulosa cell death during atresia.