The physiological role of intraovarian activin (beta/beta) and inhibin (alpha/beta) dimers in humans in unclear. The identification of follistatin as a beta-subunit-specific high affinity binding protein has added complexities for the interpretation of in vitro studies concerning the functionalities of these ovarian peptides. We, therefore, have attempted to define in vivo compartmental distributions of gene expression and protein localization for inhibin and activin subunits (alpha, beta A, and beta B) concurrent with follistatin in ovarian follicles and corpus lutea obtained from a large number of human ovaries. In situ hybridization and immunohistochemistry were used for detecting the expression of genes encoding inhibin/activin subunits and follistatin and their gene products, the proteins. Granulosa cells of small antral follicles (1-8 mm) were found to express mRNA for alpha-, beta A-, and beta B-subunits as well as follistatin, and the strongest signals were localized in the cumulus granulosa cells. In the thecal cell layer, only alpha-subunit mRNA was detected. Proteins were localized in cellular compartments corresponding to their mRNA, but in addition, proteins for beta A-subunit and follistatin were detected in the thecal cell layers in the absence of gene expression, an observation compatible with a paracrine action. Thus, granulosa cells of the small antral follicle have the potential to form all dimers of inhibin and activin, and their autocrine and paracrine actions may be modulated by follistatin in both granulosa cell and thecal cell layers. With the development of a dominant follicle, remarkable switches in subunit gene expressions occurred; beta B-subunit mRNA was no longer detectable in any cell type, and beta A-subunit expression emerged in the thecal cells along with continued abundant expression of beta A-subunit and follistatin in the granulosa cells. Proteins were found only in granulosa cells corresponding to their mRNAs. In the corpus luteum, the inhibin/activin alpha- and beta A-subunits and follistatin mRNA and proteins were expressed exclusively in the luteinized granulosa cells. Luteinized thecal cells were devoid of detectable mRNA message or proteins. Thus, the inhibin-activin-follistatin system in the corpus luteum appears to function in an autocrine fashion.(ABSTRACT TRUNCATED AT 400 WORDS)

We recently demonstrated that inhibin/activin alpha-, beta A-, and beta B-subunit messenger RNAs (mRNAs) are localized in a variety of embryonic rat tissues from 12-20 days post coitum (pc) and reported localizations consistent with possible growth effects of activin during rat embryogenesis. In the present study, we examined the tissue-specific distribution of mRNAs encoding all known players of the inhibin/activin system. In situ hybridization with radiolabeled RNA probes specific for mouse activin receptors (ActRII and ActRIIB), rat follistatin, and rat inhibin/activin subunits was used to examine the spatiotemporal expression of these molecules in adjacent sections of postimplantation rat embryos (8-20 days pc) as well as in midgestation placenta and uterine tissues (8-12 days pc). With the exception of the dorsal root ganglion and salivary gland, alpha- and beta B-subunit mRNAs were found exclusively in reproductive tissues (brain, pituitary, and/or gonads). beta A-Subunit mRNA signal was observed in the brain and gonads as well as in a variety of other tissues during embryogenesis. ActRII mRNA was found exclusively in neuronal tissue from 14 days pc until birth. ActRIIB mRNA was also found in brain, spinal cord, and ganglion, but usually appeared earlier in development than the ActRII message. ActRIIB message was also expressed in a number of other tissues, in some cases along with beta A-subunit mRNA. In these tissues, ActRIIB expression was confined to epithelial and endothelial cell types. Follistatin message was observed in all tissues (except the heart and vessels) localizing beta A-subunit and/or ActRIIB but not in the same cell type. Outside the embryo, beta A-subunit mRNA was localized in the decidua capsularis during midgestation, whereas ActRIIB message was found in placenta as early as 9 days pc. Expression of follistatin message was apparent in decidua from 8-11 days pc, then disappeared from this tissue and was abundant in myometrium at 12 days pc. These data suggest that: 1) inhibin and activin regulate aspects of the fetal reproductive system, whereas activin A may regulate the growth and differentiation of many embryonic tissues; 2) ActRII and ActRIIB serve different roles during development of the rat embryo; and 3) follistatin is in a position to modulate the effects of activin during postimplantation rat embryogenesis.

Inhibin and activin are best known as gonadal glycoprotein hormones but have a broad anatomical distribution. We previously described the central distribution ofinhibin/activin beta A- and beta B-subunit proteins in some neuronal cell bodies, fibers, and nuclei of the rat brain and reported a possible role for central activin in suckling-induced oxytocin secretion and corticotropin releasing factor release. In the present report, we mapped the detailed immunohistochemical localization of inhibin/activin alpha-, beta A-, and beta B-subunits throughout the rat brain to further clarify their central distribution. In addition, the localization and distribution of their corresponding mRNAs was assessed. The results are summarized as follows: 1) Both beta A- and beta B-subunit immunoreactivity are found in neuronal cell bodies in the nucleus of the solitary tract and the dorsal and ventral medullary reticular nuclei, and in fibers and terminals of known projection sites for these nuclei. 2) beta B-subunit immunoreactivity is localized in a group of perifornical neurons in the hypothalamus. 3) beta A-subunit immunoreactivity is present in discrete populations of neuronal cell nuclei scattered throughout the CNS. 4) mRNAs encoding each of the inhibin/activin subunits are expressed in all major brain regions as determined by S1 nuclease assay and in a variety of specific neuroanatomical sites as shown by in situ hybridization. The results suggest that central inhibin and activin proteins are produced in the brain where they may potentially serve inter- and intracellular functions in multiple systems.

The role of inhibin, activin, and follistatin in the pathophysiology of polycystic ovary syndrome (PCOS) was investigated by examining the expression of human inhibin/activin subunit, follistatin, and type II activin receptor (ActRII and -IIB) messenger ribonucleic acid (mRNA) signals (via in situ hybridization) and encoded proteins (via immunocytochemistry) in ovarian follicles (n = 42) from 6 women diagnosed with PCOS. The localization patterns in cellular compartments were compared to those in small antral follicles of comparable size (3-7 mm; n = 40) from 17 normal human ovaries. In small antral follicles of both normal and PCOS ovaries, mRNA signals for all three subunits of inhibin and activin (alpha, beta a, and beta b) were expressed in granulosa cells, whereas in the thecal cell layer, only alpha-subunit mRNA was expressed. The relative intensity of the alpha-subunit mRNA signal was distinctly different in granulosa and thecal cells between PCOS and normal follicles; in small antral follicles of normal ovaries, the alpha-subunit mRNA signal was stronger in the granulosa cell layer than in the thecal cells, and the reverse was found in the polycystic follicles. A light follistatin mRNA signal was found in the granulosa cells of normal small antral follicles, but no follistatin mRNA was detected in any cell type of PCOS follicles. ActRII and -IIB mRNAs were not detected in any cell layer in either normal or PCOS follicles. There were no notable differences in the protein localization pattern of the inhibin/activin system between the PCOS and normal ovaries. In both types of follicles, follistatin and alpha-, beta a-, and beta b-subunit cytoplasmic staining were observed in granulosa cells, as were their corresponding messages, with the exception of the undetectable follistatin mRNA signal in the PCOS follicles. In both normal and PCOS follicles, follistatin and beta a-subunit cytoplasmic staining were occasionally found in thecal interna cells, with no corresponding localization of mRNA, and alpha-subunit protein was not detected in thecal cells despite the presence of the alpha-subunit mRNA. ActRII and -IIB protein localizations were not examined due to the lack of available antisera. These results suggest that granulosa cells of small antral follicles are less active in polycystic than in normal ovaries with respect to inhibin alpha-subunit and follistatin mRNA expression. A consequence of these differences could be an increase in the availability of activin, relative to inhibin, in the arrested follicles in PCOS.

Follistatin (FS), which binds to the inhibin/activin beta A- or beta B-subunit is localized with and modulates the biological actions of activin in many systems. However, in contrast to the wide distribution of the activin beta-subunit proteins and messenger RNAs (mRNA) in the brain, demonstration of FS mRNA signal has been limited to the olfactory tubercle and layer II of the frontal cortex. We have hypothesized a more extensive distribution of central FS gene expression and localization in regions coinciding with inhibin/activin beta-subunits and possible activin-mediated effects. In the present study, we examined the central distribution of FS mRNA expression in the normal adult male rat. With in situ hybridization analysis, using a 33P-labeled RNA probe specific for rat FS, gene expression is shown to be widely distributed throughout the brain. Abundant FS mRNA expression is localized in several areas of the olfactory bulb as well as the frontal cortex, a few thalamic nuclei, and in septal regions. Moderate FS mRNA is observed in the caudate putamen and various hypothalamic areas including the paraventricular, ventromedial, dorsomedial, and arcuate nuclei. Several brain stem regions are also found to express FS mRNA, including the medial vestibular and solitary tract nuclei. Notably, FS mRNA, including the medial vestibular and solitary septal/diagonal band region is localized in patterns that are highly correlative with those of GnRH gene expression and hence may serve to regulate possible activin-mediated effects in these areas. FS mRNA is also expressed in areas associated with the activin-oxytocin pathway (solitary tract nucleus and paraventricular nucleus) and is therefore in a position to modulate the role of activin in the solitary tract nucleus-paraventricular nucleus pathway (afferent system mediating the milk-ejection reflex). The results suggest that FS is centrally localized in sites compatible with a role in the regulation of central reproductive functions.