Paul G. Farnworth

Follistatins : more than follicle-stimulating hormone suppressing proteins

The purification and characterization of inhibins and activins, inhibitors and stimulators of pituitary follicle-stimulating hormone (FSH) secretion and synthesis, opened new perspectives in reproductive biology and substantially enhanced our understanding of the hypothalamus-pituitary-gonadal axis (De Jong et al., 1990; De Kretser, 1990; Ling et al., 1990). These hormones belong to the transforming growth factor (TGF)-@ family: inhibins consist of a common a-subunit which is covalently linked to one of the two different p-subunits (inhibin A: (Y/P*; inhibin B: a/P,,); activins are covalently linked dimers of the two p-subunits and therefore exist in three different forms (activin A: PA/PA; activin AB: PA//3n; activin B: Pn/Pn) (Ying, 1989). Inhibin subunits are expressed in gonadal and extragonadal tissues (Meunier et al., 19881, and activins have multiple biological effects in a wide variety of target tissues which are involved in reproductive and non-reproductive functions (for reviews see above, and Findlay et al., 1990; Wylie, 1990; De Paolo et al., 1991a; Slack, 1991). In 1987, two laboratories independently purified members of another protein family from porcine (Ying et al., 1987) and bovine (Robertson et al., 1987) follicular fluids (FF) based on their inhibin-like ability to suppress pituitary FSH secretion. These proteins were found to be structurally different from inhibins/activins and were termed follistatins (FS) or FSH-suppressing proteins. In vivo (DePaolo et al., 1991b) and in vitro (Ying et al., 1987; Robertson et al., 1987, 1990; Wang et al., 1990b) experiments showed that the potency of FS in suppressing pituitary FSH secretion was lo-30%

Growth Differentiation Factor 9 Is a Germ Cell Regulator of Sertoli Cell Function

Oocyte-secreted growth differentiation factor (GDF) 9 and bone morphogenetic protein (BMP) 15 are critical regulatory factors in female reproduction. Together, they promote granulosa cell proliferation and stimulate the maturation of preovulatory follicles. Despite their importance in female fertility, GDF9 and BMP15 expression patterns and function during spermatogenesis have not been investigated. In this study we show that the expression and stage-specific localization of both factors are limited to the germ cells of the rat seminiferous epithelium, with GDF9 being principally localized in round spermatids and BMP15 in gonocytes and pachytene spermatocytes. To identify potential cellular targets for GDF9 actions, cells of the seminiferous tubule were isolated and screened for the expression of signaling receptors [activin-like kinase (ALK) 5, ALK6, and BMP receptor, type II)]. Individual receptor types were expressed throughout the seminiferous epithelium, but coexpression of ALK5 and BMP receptor, type II was limited to Sertoli cells and round spermatids. Based on the reproductive actions of related TGFbeta ligands in the ovary and testis, GDF9 was assessed for its ability to regulate tight junction function and inhibin B production in rat Sertoli cell cultures. When recombinant mouse GDF9 was added to immature Sertoli cell cultures, it inhibited membrane localization of the junctional proteins claudin-11, occludin, and zonula occludens-1, thereby disrupting tight junction integrity. Concomitantly, GDF9 up-regulated inhibin subunit expression and significantly stimulated dimeric inhibin B protein production. Together, these results demonstrate that GDF9 and BMP15 are germ cell-specific factors in the rat testis, and that GDF9 can modulate key Sertoli cell functions.

Loss of Betaglycan Contributes to the Malignant Properties of Human Granulosa Tumor Cells

Betaglycan is a type III TGFbeta receptor that modulates cellular sensitivity to inhibins and TGFbeta. Previous studies have suggested that betaglycan acts as a tumor suppressor in certain human epithelial cancers. However, the roles of betaglycan in ovarian granulosa cell tumors (GCTs) are poorly understood. The objective of this study was to determine whether human GCTs exhibit betaglycan expression and, if so, what impact this receptor has on tumor biology. Real-time PCR was used to quantify betaglycan transcripts in human GCTs (n = 17) and normal premenopausal ovaries (n = 11). This analysis established that GCTs exhibited a significant 2-fold lower mean betaglycan mRNA level as compared with the normal ovary (P < 0.05). Similarly, two human GCT cell lines, KGN and COV434, exhibited low betaglycan expression and poor responsiveness to TGFbeta and inhibin A in luciferase reporter assays, which was restored by stable transfection of wild-type betaglycan. Betaglycan significantly increased the adhesion of COV434 (P < 0.05) and KGN (P < 0.0001) cells, decreased cellular invasion through Matrigel, and inhibited wound healing. Expression of mutant forms of betaglycan that are defective in TGFbeta and/or inhibin binding in each GCT cell line revealed that the inhibitory effects of betaglycan on wound healing were most strongly linked to the inhibin-binding region of betaglycan. Furthermore, knockdown of INHA mRNA expression abrogated the betaglycan-mediated inhibition of wound healing and invasion, whereas both INHA silencing and TGFbeta neutralization abolished the betaglycan-mediated increase in adhesion to substrate. These data suggest that loss of betaglycan contributes to the pathogenesis of GCTs.

Inhibin binding sites and proteins in pituitary, gonadal, adrenal and bone cells

Activin signals via complexes of type I (50-55 kDa) and II (70-75 kDa) activin receptors, but the mechanism of inhibin action is unclear. Proposed models range from an anti-activin action at the type II activin receptor to independent actions involving putative inhibin receptors. Two membrane-embedded proteoglycans, betaglycan and p120, have recently been implicated in inhibin binding, but neither appears to be a signalling receptor. The present studies on primary cultures of rat pituitary and adrenal cells, and several murine and human cell lines were undertaken to characterise inhibin binding to its physiological targets. High affinity binding of inhibin to the primary cultures and several of the cell lines, like that previously described for ovine pituitary cells, was saturable and reversible. Scatchard analysis revealed two classes of binding sites (K(d) of 40-400 and 500-5000 pM, respectively). Affinity labelling identified [125I]inhibin binding proteins with apparent molecular weights of 41, 74, 114 and >170 kDa in all cell types that displayed high affinity, high capacity binding of inhibin. Additional labelling of a 124 kDa species was evident in gonadal TM3 and TM4 cell lines. In several cases, activin (> or =20 nM) competed poorly or not at all for binding to these proteins. The 74, 114 and >170 kDa inhibin binding proteins in TM3 and TM4 cells were immunoprecipitated by an anti-betaglycan antiserum. These three proteins correspond in size to the activin receptor type II and the core protein and glycosylated forms of betaglycan, respectively, that have been proposed to mediate anti-activin actions of inhibin, but the identity of the 74 kDa species is yet to be confirmed. Studies of [125I]inhibin binding kinetics and competition for affinity labelling of individual binding proteins in several cell lines suggest these three species and the 41 and 124 kDa proteins form a high affinity inhibin binding complex. In summary, common patterns of inhibin binding and affinity labelling were observed in inhibin target cells. Novel inhibin binding proteins of around 41 and 124 kDa were implicated in the high affinity binding of inhibin to cells from several sources.

Acute inhibitory effect of follicle-stimulating hormone-suppressing protein (FSP) on gonadotropin-releasing hormone-stimulated gonadotropin secretion in cultured rat anterior pituitary cells

Follicle-stimulating hormone (FSH)-suppressing protein (FSP) or follistatin, a novel gonadal glycoprotein hormone, has been shown to have chronic inhibitory effects on the secretion of both FSH and luteinizing hormone (LH) in response to gonadotropin-releasing hormone (GnRH) in vitro. The present study was designed to investigate the acute effects of bovine FSP on GnRH-stimulated gonadotropin secretion and to examine the potential subcellular sites of this action of FSP using cultured pituitary cells. Anterior pituitaries from adult male Sprague-Dawley rats were enzymatically dispersed and cultured for 48 h, after which the cells were treated with bovine FSP for 6 h, followed by a 4 h stimulation with secretagogues in the continued presence of FSP. Results showed that the 35 kDa form of bovine FSP (0.1-3 nM) dose-dependently suppressed GnRH-stimulated FSH and LH secretion, with inhibition of 38 and 25%, respectively, at 3 nM. In addition, FSP suppressed gonadotropin secretion in response to activators of protein kinase C (phorbol 12-myristate 13-acetate (PMA) and mezerein) and a calcium ionophore (A23187). However, FSP had no effect on gonadotropin secretion evoked by melittin, an activator of phospholipase A2. Furthermore, 35 kDa bovine FSP did not compete with GnRH for GnRH binding sites in a direct competition study and treatment of cultured pituitary cells with FSP (0.1-3 nM) for 10 h did not alter the number of GnRH binding sites on the cell membranes. Finally, similar inhibitory effects on gonadotropin secretion in response to GnRH, PMA and mezerein were obtained with the 31 and 39 kDa forms of bovine FSP, each at a concentration of 1 nM. We conclude from the present study that FSP acutely inhibits GnRH-stimulated gonadotropin secretion in cultured pituitary cells, and that FSP exerts its action beyond the GnRH receptor, possibly by affecting the protein kinase C and/or the calcium-calmodulin systems.

Bombesin Regulation of Adrenocorticotropin Release From Ovine Anterior Pituitary Cells

Mammalian members of the bombesin-like peptide family (gastrin releasing peptides; GRP) have been localized in the ovine median eminence and in hypophysial-portal blood, suggesting a role in the regulation of anterior pituitary function. In this study we have shown that although bombesin cannot stimulate ACTH secretion alone, it potentiates release by ovine CRF, an effect blocked by the GRP receptor antagonist D-Tyr6bombesin (6-13) propylamide. Bombesin did not potentiate AVP-stimulated ACTH release; instead release was attenuated when bombesin was given at a 10-fold or greater molar excess over AVP, with no interaction seen at lower concentrations. We conclude that ovine corticotrophs express bombesin receptors, and that GRP may act in concert with other hypothalamic releasing factors to regulate ACTH secretion.

Extra-ovarian expression and activity of growth differentiation factor 9

Growth differentiation factor 9 (GDF9) produced within the ovary plays an essential role during follicle maturation through actions on granulosa cells, but extra-ovarian expression, signalling and actions of GDF9 are less well characterised. The present studies confirm GDF9 expression in the mouse testis, pituitary gland and adrenocortical cancer (AC) cells, and establish its expression in L beta T2 gonadotrophs, and in mouse adrenal glands, particularly foetal and neonatal cortical cells. AC, L beta T2, TM3 Leydig and TM4 Sertoli cells express the requisite GDF9 binding signalling components, particularly activin receptor-like kinase (ALK) 5 and the bone morphogenetic protein (BMP)/GDF type II receptor, BMPRII (BMPR2). We therefore compared GDF9 activation of these potential extra-ovarian target cell types with its activation of granulosa cells. Recombinant mouse GDF9 stimulated expression of activin/transforming growth factor-beta-responsive reporters, pGRAS-luc or pAR3-lux, in TM4 and AC cells (IC50=145 ng/ml in the latter case), and two granulosa cell lines, KGN and COV434. The ALK4/5/7 inhibitor, SB431542, blocked GDF9 activity in each case. By contrast, GDF9 lacked specific effects on TM3 cells and rat primary pituitary and mouse L beta T2 gonadotrophs. Our findings show that GDF9 regulates the expression of R-SMAD2/3-responsive reporter genes through ALK4, 5 or 7 in extra-ovarian (adrenocortical and Sertoli) cells with similar potency and signalling pathway to its actions on granulosa cells, but suggest that expression of BMPRII, ALK5 (TGFBR1) and R-SMADs 2 and 3 may not be sufficient for a cell to respond to GDF9.

Growth hormone-releasing peptide-2 (GHRP-2) does not act via the human growth hormone-releasing factor receptor in GC cells

Effect of growth hormone-releasing peptide-2 (GHRP-2) on ovine somatotrophs is abolished by a growth hormone-releasing factor (GRF) receptor antagonist, which raises the possibility that GHRP-2 may act on GRF receptors. In the present study, we used rat pituitary GC cells with or without stable transfection of cDNA coding for the human GRF receptor (GC/R+ or GC/R−) to determine whether or not GHRP-2 acts via the GRF receptor. Northern blot analysis indicated that GRF receptor mRNA was undetectable in GC/R− cells, whereas a high level of expression occurred in GC/R+ cells that were transfected by GRF receptor cDNA. In GC/R− cells, incubation with up to 10−7M of either hGRF or GHRP-2 did not alter the intracellular cAMP, [Ca2+]i, or GH secretion. In GC/R+ cells, hGRF (10−11–10−7M) increased cAMP levels in a concentration-dependent manner up to 20-fold. This increase in cAMP levels was blocked by a GRF receptor antagonist, [Ac-Tyr1, d-Arg2]-GRF 1–29, but not by a Ca2+ channel blocker, NiCl2 (0.5 mM). GH secretion and [Ca2+]i were, however, not increased by hGRF. Incubation of the transfected cells with 10−11–10−8M GHRP-2 did not modify intracellular cAMP levels. This result suggests that GHRP-2 does not act through the GRF receptor.

Production and Actions of Inhibin, Activin, and Follistatin in the Pituitary and Ovary

The overall aim of our research is to elucidate control of the production and mechanisms of action of inhibin (INH), activin (ACT), and follistatin (FS) in the pituitary and ovary. The broad hypothesis being tested is that, apart from the endocrine actions of INH on the pituitary, these substances exert autocrine or paracrine actions on the gonadotrophs in the pituitary and the follicular cells of the ovary to regulate follicle-stimulating homone (FSH) secretion and folliculogenesis, respectively. To test this hypothesis, it is necessary to demonstrate (a) that the cells of the pituitary and ovary are capable of secreting INH, ACT, and FS in a timely manner, and (b) that these substances can modify the functions of the target cells in a manner consistent with their roles in regulating pituitary and ovarian function. We also need to know the forms in which INH, ACT, and FS are secreted and measure the relative biological activities of these forms. To this end, we have concentrated on the following specific questions: a. What are the relative biological activities of the high molecular weight (MW) forms of INH and ACT identified in biological fluids? b. Can expression of the FS and INH-α and -β mRNA or protein be detected in pituitary and ovarian cells at times consistent with their proposed roles? c. Is there a role for ACT in the acquisition of responsiveness of follicular granulosa cells to FSH during the early stages of folliculogenesis?