Proc. Nati. Acad. Sci. USA Vol. 86, pp. 5114-5117, July 1989 Medical Sciences Inhibin and activin modulate the release of gonadotropin-releasing hormone, human chorionic gonadotropin, and progesterone from cultured human placental cells (foliculostatin/follicle-stimulating hormone-releasing protein/follitropin/foliicle-stimulating hormone) FELICE PETRAGLIA*, JOAN VAUGHAN, AND WYLIE VALEt Clayton Foundation Laboratories for Peptide Biology, The Salk Institute, 10010 North Torrey Pines Road, La Jolla, CA 92037 Communicated by C. H. Sawyer, March 20, 1989 (receivedfor review September 23, 1988) ABSTRACT Although it is clear that human chorionic cotropin], supports the hypothesis that placental hormono- gonadotropin (hCG) and progesterone play fundamental roles genesis may be regulated in part by locally produced peptides in pregnancy, the regulation of placental production of these (13-22). hormones remains to be defined. Recent evidence suggests that A recent report (11) showed that the addition of inhibin the human placenta expresses proteins related to inhibin (afi antiserum increased hCG production in primary human pla- subunits) or activin (8P subunits). Inhibin and activin (follicle- cental cultures, suggesting that endogenous inhibin might stimulating hormone-releasing protein) possess opposing ac- tonically inhibit secretion of the placental gonadotropin. To tivities in several biological systems including pituitary follicle- evaluate the possible roles of inhibin-related proteins in the stimulating hormone (follitropin) secretion, erythroid differ- regulation of placental hormones more directly, we have entiation, and gonadal sex-steroid production. The actions of investigated the effects-ofinhibin and activin on the secretion purified inhibin and activin on hormonogenesis by primary ofGnRH, hCG, and progesterone by cultured placental cells. cultures of human placental cells were studied. The addition of Moreover, because inhibin and activin are structurally re- activin increased gonadotropin-releasing hormone (GnRH) lated to the transforming growth factor /3 (TGF-,3) family of and progesterone production and potentiated the GnRH- growth factors (23) and because TGF-,8 and its receptors are induced release of hCG. Inhibin by itself did not modify present in human placenta (24, 25), we also evaluated the placental immunoreactive GnRH, hCG, and progesterone se- possible action of TGF-f3 on placental hormone release. cretion but reversed the activin-induced changes. Neither inhibin nor activin influenced the release of human placental lactogen. Furthermore, transforming growth factor (3, struc- MATERIALS AND METHODS turally related to inhibin/activin, did not signifcantly influ- Preparation of Placental Cell Cultures. Placentae were ence hormone release from cultured placental cells. These obtained from pregnant women (n = 7) undergoing elective results support the hypothesis that inhibin and activin may play caesarean section at term. Permission to obtain the tissue was a role in regulating the release of GnRH, hCG, and proges- granted by the Human Investigation Committee ofUniversity terone from placenta and implicate inhibin-related proteins in of California, San Diego and The Salk Institute. the endocrine physiology of human pregnancy. Immediately after the collection, a placenta was placed on ice and chunks of tissue were minced (50-70 g). After being Inhibins are heterodimeric proteins, consisting of an a sub- rinsed three times in cold Hepes dissociation buffer (HDB; unit and one oftwo ,8 subunits (13A or PB), that were originally ref. 26), chunks were dissected free of membranes, and identified based upon their abilities to selectively suppress connective tissue and the soft tissues were minced. Cells follicle-stimulating hormone (FSH) secretion (1-6). FSH- were dissociated with 50 ml of HDB containing collagenase releasing proteins or activins were subsequently isolated and solution [0.1% bovine serum albumin/0.4% collagenase (type characterized as dimers comprised of inhibin (3 subunits II, Cooper Biomedical)/3000 Kunitz units of DNase II (type APA or AJB) (7, 8). Although inhibin and activin were first IV, Sigma)] in a water-jacketed Spinner suspension flask isolated from the gonads, various tissues including the pla- (Wheaton Scientific). The dissociation mixture was main- centa contain inhibin subunit mRNAs (9-12). Whereas the tained at 370C and continuously stirred at 200-300 rpm for 1 gonads contain a large excess ofa- over (B-chain mRNAs, the hr. Trypsin (0.4%; Sigma) was then added, and the solution placenta has an excess of p-subunit mRNAs; such observa- was stirred for an additional 10 min. The tissue fragments tions raise the possibility that the placenta might have a were filtered through a 53-gum mesh filter (Spectrum Medical preponderance of,,8 dimers (12). However, there is not yet Industries). Filtered cells were transferred to a sterile plastic conclusive evidence that intact activin is synthesized by tube and centrifuged at 300 x g for 10 min. The resultant human placenta. supernatant was further centrifuged at 400 x g for 10 min. Placental inhibin a-subunit immunoreactivity is localized The pelleted cells from both centrifugations were resus- in the cytotrophoblast layer of the villi (11), a region shown pended in culture medium (J3-PJ containing 10%o fetal bovine earlier to contain other regulatory peptides including gonad- serum; ref. 16). The remaining unfiltered cells and tissue otropin-releasing hormone (GnRH; refs. 13-20), somato- fragments were further digested with trypsin for 10 min and statin (21), and corticotropin-releasing factor (22). The evi- subjected to the above procedure. All cells obtained were dence that the placenta produces these hypophysiotropic peptides, which modulate the secretion of various hormones Abbreviations: hCG, human chorionic gonadotropin; GnRH, gonad- related to those in the pituitary [human chorionic gonado- otropin-releasing hormone; FSH, follicle-stimulating hormone; tropin (hCG), human placental lactogen (hPL), and corti- TGF-p, transforming growth factor ,B; hPL, human placental lacto- gen; irGnRH, immunoreactive GnRH. *Present address: Department of Obstetrics and Gynecology, Uni- The publication costs ofthis article were defrayed in part by page charge versity of Modena School of Medicine, Via del Pozzo 71, 41100 payment. This article must therefore be hereby marked "advertisement" Modena, Italy. in accordance with 18 U.S.C. §1734 solely to- indicate this fact. tTo whom reprint requests should be addressed. 5114 Downloaded by guest on September 30, 2021 Medical Sciences: Petraglia et al. Proc. Natl. Acad. Sci. USA 86 (1989) 5115 resuspended in the culture medium and plated in 35-mm followed by the multiple-range test ofDuncan and by Tukey's six-well multiwell dishes (Costar) previously coated with test. poly(D-lysine) (Sigma). This procedure yielded 3 x 106 cells per g of tissue, and 5.0-6.5 X 105 cells were plated per well. RESULTS Experimental Procedures. The cultures were maintained at The addition ofpurified ovine inhibin to the cultured placen- 370C in a water-saturated atmosphere containing 5% C02; the tal cells did not alter the basal secretion rate of hCG but culture medium was changed every 2 days. The secretion reduced the GnRH-induced release of hCG (Fig. 1). Activin, experiments were conducted between 5 and 10 days after which only slightly and not significantly increased the release plating. The incubation medium consisted of fi-PJ medium of hCG when added alone, strongly potentiated the stimula- supplemented with 0.1% bovine serum albumin. Treatments tory action of GnRH on hCG secretion (Fig. 2). This effect were added in a small volume (<50 1.d) to triplicate wells was dose-related (Fig. 2) and was reversed by inhibin (Fig. 1). containing 1.0 ml ofincubation medium. The incubation time While inhibin did not modify the immunoreactive GnRH was 48 hr. (irGnRH) concentrations in placental cell cultures, the high- Peptide and Protein Hormones. Purified ovine inhibin (in- est doses of activin (105 pM and 350 pM) significantly hibin aPA heterodimer) was isolated from rete testis fluid by increased the secretion ofirGnRH (from 12.2 ± 2.5 to 42.2 ± immunoaffinity chromatography using an antiserum to the 3.1 and 53.3 ± 4.6 pg ofirGnRH per ml, respectively). Inhibin amino-terminal portion of the porcine inhibin a subunit (27). (90 pM) completely reversed the activin-induced irGnRH On NaDodSO4/PAGE, purified nonreduced ovine inhibin increase. Moreover, activin stimulated progesterone release shows a single band of Mr 32,000; after reduction, two bands from placental cultures in a dose-related manner (Fig. 3). of Mr 21,000 and Mr 14,000 are observed. Porcine activin, a Ihhibin did not significantly change progesterone concentra- homodimer of inhibin PAPA, was purified to homogeneity tion but completely reversed (90 pM) the effect of activin on from porcine follicularfluid as described by Vale et al. (7). On progesterone release (Fig. 3). NaDodSO4/PAGE, purified porcine activin shows a single The addition of moderate doses of TGF-,j (from 4 to 400 band ofMr 28,000 nonreduced and a single band ofMr 14,000 pM) neither altered the release of hCG, irGnRH, and pro- upon reduction. GnRH was prepared by solid-phase meth- gesterone from cultured placental cells nor influenced the odology (28). hCG, a-subunit hCG, and the antiserum against action of activin and/or of inhibin on the secretion of these hCG were gifts of the National Hormone and Pituitary