Soluble Guanylyl Cyclase: DISCUSSION
The present study provides new information regarding the cellular localization and regulation of sGC a1 and p1 subunit levels in ovarian cells. Given the transient expression of sGC proteins in granulosa cells during follicle growth and the detection of these subunits in theca cells and corpora lutea, our findings suggest that sGC (and the associated second-messenger cGMP) may play significant roles in the process of follicular development and/or survival and in as-yet-undefined thecal and luteal functions.
Our previous reports indicate that treatment of cultured rat granulosa cells with a NO donor or specific activator of sGC results in marked stimulation of cGMP production, which is consistent with expression of sGC in granulosa cells. Similarly, previous immunoblot analysis (using a different primary antisera that detects both sGC a and p subunits) also indicated expression of both a and p subunits in granulosa cells (unpublished results). The present results confirm and extend these findings. Immunoblot analysis of whole-ovarian lysates did not reveal an effect of gonadotropin treatment, likely reflecting the cell-specific nature of sGC expression. However, immunofluorescence studies suggest an inverse relationship between sGC subunit levels and granulosa cell proliferation in postnatal and gonadotropin-treated rats. Whereas sGC subunit expression is strong in granulosa cells of primordial and primary follicles, a marked decline in sGC levels occurs in granulosa cells at very early stages of follicle growth, with no detectable sGC a1 or p1 protein in proliferating granulosa cells of larger developing follicles. Similarly, the absence of sGC down-regulation in cultured granulosa cells in response to FSH may be related to the fact that granulosa cell proliferation is minimal in this serum-free culture system. Together, these findings indicate an inverse relationship between sGC levels and the rate of granulosa cell proliferation. Potentially, inhibition of sGC expression may result in activation of mechanisms facilitating granulosa cell proliferation. In this regard, previous findings demonstrate a role of NO as an inhibitor of cellular proliferation, decreasing activity of cell cycle-regulating factors in a wide variety of cell types. Thus, a decline in sGC expression may result in relief from the NO-mediated cytostasis previously proposed to exist in developing follicles. The identities of factors responsible for inhibition of sGC expression during early follicular development remain unknown.
The decrease in sGC a1 and p1 expression in maturing granulosa cells of growing and preovulatory follicles is consistent with the reported antagonistic effects of NO and cGMP on the functional maturation of granulosa cells. However, this decrease in sGC does not appear to be a required component of FSH-induced cell maturation, because no change in sGC subunit levels was observed during FSH-induced maturation of cultured granulosa cells. Similarly, FSH-induced increases of E2 production in cultured granulosa cells might also have been predicted to inhibit sGC expression because of the reported inhibitory effects of E2 on sGC expression in the uterus. However, FSH treatment did not significantly affect sGC subunit levels in granulosa cells despite anticipated induction of E2 synthesis. These unexpected results may reflect, in part, the fact that the majority of granulosa cells obtained for culture appear to exhibit down-regulation of sGC subunit levels (Fig. 8), resulting from granulosa cell proliferation and/or E2-implant treatment. Furthermore, the population of granulosa cells used in this culture system may include those from follicles undergoing atresia and, thus, may not completely represent the function of the granulosa cells of healthy follicles. Nevertheless, these findings do indicate that the down-regulation of sGC subunits in granulosa cells of growing follicles is not directly related to FSH-stimu-lated granulosa cell maturation, because FSH-induced differentiation of cells in this culture system clearly occurred but was not associated with a change in sGC a1 and p1 subunit levels.
Our findings may also provide insight regarding the relative importance of NO-derived cGMP as an inhibitor of apoptosis during different stages of follicular development. Previous studies indicate that relatively little apoptosis occurs in granulosa cells of preantral follicles of the ovary and that treatment of cultured preantral follicles with a cGMP analog inhibits programmed cell death. These observations are consistent with the high expression of sGC in such preantral follicles, supporting a potential role for cGMP as an inhibitor of follicular demise. In contrast, studies from the same laboratory also indicate that cGMP effectively inhibits apoptosis in cultured preovulatory follicles, which do not appear to express significant levels of sGC a1 or p1 subunit. Given our findings of strong sGC expression during early, but not middle and later, stages of folliculogenesis, the role of NO-stimulated cGMP may be as an inhibitor of early follicular demise, with other factors (e.g., FSH) playing roles in survival of follicles at later stages. This proposed importance of NO-derived cGMP as an inhibitor of early follicular demise could explain recent observations in which mice lacking endothelial NO synthase (NOS) expression exhibited decreased numbers of developing antral follicles and smaller ovulation rates.
The patterns of sGC regulation observed in the present study are consistent with the reported pattern of NOS expression. Reports of NOS expression in the ovary are not completely consistent, but it has been reported that inducible NOS and endothelial NOS levels decrease during follicular development. This is consistent with the expression of sGC, a target of NO, in early, but not in late, developing follicles. Similarly, our finding that sGC is expressed in corpora lutea is consistent with findings demonstrating luteal NOS expression and action, although the involvement of sGC in the effects of NO on luteal function is not well established. Similar patterns of expression for NOS and the target receptor for NO are evident, but the molecular mechanisms resulting in such regulated temporal coexpression require further study.
In addition to NO, sGC is also activated by CO, which is generated by heme oxygenases. A recent paper demonstrates expression of heme oxygenases in the ovary, suggesting a possible role for CO as an activator of sGC in ovarian cells. The observation that E2 increases heme oxygenase levels and cGMP production from human endothelial cells raises further intriguing possibilities regarding CO-mediated feedback mechanisms through which E2 may act on the ovary.
It is important to note that the present study utilizes antibodies that are specific for sGC aj and p! subunits. Thus, we cannot rule out the possibility that whereas larger developing follicles display low levels of sGC a! and p! subunits, granulosa cells of such follicles do express the a2 and p2 isoforms of sGC. This would allow the possibility that NO and/or CO may act on these target enzymes in larger developing follicles, influencing follicular development and/or steroidogenesis. Future studies utilizing a2 and p2 isoform-specific antisera or selective mRNA probes are required to examine the potential expression of these isoforms in the ovary.