Inhibin, activin, and follistatin. Potential roles in ovarian physiology.

The physiological significance of these results will not become clear until patterns of activin and inhibin protein production and the expression of their receptors have been more thoroughly characterized in relation to follicular development. Meanwhile, in situ hybridization studies on rat and monkey ovaries suggest that inhibin/activin beta-subunit mRNA (favoring synthesis of activin) is relatively abundant in granulosa cells of immature antral follicles, whereas alpha-subunit mRNA (favoring synthesis of inhibin) predominates in Graafian follicles. The increased production of follistatin associated with advanced preovulatory development would serve to further reduce the activin "tone" relative to inhibin (Fig. 1). At the level of protein action in vitro, the pattern emerging is that inhibin minimally affects granulosa cell steroidogenesis at any stage of follicular development, whereas activin has pronounced modulatory effects that alter with follicular maturity. As suggested previously,60 the ability of activin to enhance gonadotropin-responsive aromatase activity and simultaneously suppress progesterone production by mature granulosa cells has physiological implications in that it hints at a mechanism for promoting estrogen synthesis and simultaneously suppressing progesterone synthesis, which is precisely what occurs in the preovulatory follicle. The effects of inhibin and activin on human thecal androgen synthesis observed in vitro suggest how these proteins might act locally to modulate preovulatory follicular growth and estrogen synthesis in vivo (Fig. 2).57 In essence, we propose that activin acting at early stages of antral follicular development plays a role in follicular recruitment through sensitizing immature granulosa cells to the cytodifferentiative action of FSH. On the other hand, inhibin is more likely to play a role in preovulatory follicular selection and maintenance of follicular dominance. Studies of follicular fluid levels of androgen and estrogen in relation to granulosa cell aromatase activity indicate that the capacity of the theca interna to generate aromatase substrate (androstenedione) increases hand in hand with aromatase activity in the human preovulatory follicle. It has therefore been suggested that a positive feedback loop (granulosa on theca) exists that promotes thecal androgen synthesis and hence estrogen synthesis in this follicle.64 The discovery that inhibin production in vitro is greatest by granulosa cells isolated from preovulatory follicles strongly implicates inhibin as a component of this feedback loop.