| Ovarian physiologists have long recognized that granulosa cells are extraordinary in their ability to divide exponentially during hormone-dependent growth of the ovarian follicle in vivo. However, the molecular mechanisms that link hormone signals from the cell membrane to the nuclear processes that govern DNA replication and cell division have remained mysterious. During follicular development, follicle-stimulating hormone (FSH) and insulin-like growth factor (IGF-I) are believed to act both independently and in coordination with one another to promote granulosa cell replication and differentiation. Recent studies have indicated that the pathways mediating the individual, classic effects of these hormones may participate in "cross-talk" events, which could result in synergism or coordinated timing of events. The goal of these studies was to investigate the ability of FSH to stimulate the PI3K/Akt pathway, an established IGF-I-activated pathway, and to impact on the transcription factor FoxO1a, which is involved in promoting cell cycle arrest and apoptosis in multiple cell systems. We hypothesized that FSH and IGF-I stimulation of granulosa cells results in the inactivation of this transcription factor, resulting in increased proliferation. Experiments utilizing granulosa cell cultures demonstrated that FoxO1a is expressed and hormonally regulated in granulosa cells. FSH and IGF-I cause phosphorylation and nuclear exclusion of FoxO1a in a PI3K-dependent manner. Additionally, we find that overexpressing wild-type (WT)-FoxO1a in granulosa cells results in increased p27kip protein, a cyclin-dependent kinase inhibitor involved in cell cycle regulation. In contrast, overexpression of a dominant negative (DN) truncation mutant of Foxo1a results in decreased levels of p27kip protein. Cell cycle analysis of these DN-Foxo1a-expressing cells indicated an 8--10-fold increase in the % S-phase cells vs. null-infected cells, suggesting that inhibition of endogenous FoxO1a function removes a level of control at the G1/S transition. Co-expression of WT-p27kip reduced the effect of the DN-Foxo1 by 50% (4-fold increase in % S-phase), suggesting that p27kip indeed plays a role in the observed FoxO1a effects. Perhaps most interestingly, co-expression of WT-p27kip with the DN-Foxo1a also altered the subcellular localization of p27kip, causing it to be localized to the cytosol (where it cannot act to inhibit cyclinE/cdk2). Thus, FoxO1a modulates both p27kip abundance and activity to restrict the cell cycle. Together, these data suggest that the nuclear exclusion and inactivation of FoxO1 by FSH and IGF-I may result in the initiation or support of proliferation, in part, through regulation of p27kip. |
