Identification, Characterization, and Biological Activity of Endothelin

0021-972X/97/$03.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1997 by The Endocrine Society

Vol. 82, No. 12 Printed in U.S.A.

Identification, Characterization, and Biological Activity of Endothelin Receptors in Human Ovary* ROSA MANCINA, TULLIO BARNI, ALDO E. CALOGERO, SANDRA FILIPPI, SANDRA AMERINI, ALESSANDRO PERI, TOMMASO SUSINI, GABRIELLA B. VANNELLI, NUNZIATINA BURRELLO, GIANNI FORTI, AND MARIO MAGGI Department of Clinical Physiopathology, Andrology Unit (M.M., R.M., G.Fo.) and Endocrinology Unit (A.P.), Department of Human Anatomy and Histology (T.B., G.B.V.), Department of Pharmacology (S.A., S.F.), Department of Obstetrics and Gynecology (T.S.), University of Florence, 50139 Florence; and the First Department of Internal Medicine, Division of Andrology, University of Catania (A.C., N.B.), Catania 95123, Italy ABSTRACT We previously reported the expression of endothelin-1 (ET-1) in granulosa cells (GCs) of the human ovary and the presence of ET-1-like immunoreactivity in human follicular fluid obtained from women in an in vitro fertilization program. In follicular fluid, but not in plasma, the levels of ET-1-like immunoreactivity were higher in gonadotropin-stimulated vs. spontaneous cycles, suggesting hormonal regulation of follicular ET-1. To identify and characterize ET receptors in human ovary, we performed autoradiography, radioligand binding, and functional studies. Mathematical analysis of families of self- and cross-competition curves among [125I]ET-1, [125I]ET3, and selective analogs indicates that human ovary expresses both subtypes of ET receptors, i.e. ETA and ETB receptors. However, the concentration of the ETB site was 100-fold lower than that of the ETA one. By using [125I]ET-1, we demonstrated that the density of binding sites in human ovary is not affected by the hormonal milieu (similar concentrations in normal cycling, postmenopausal, and combined oral

T

HE ENDOTHELINS (ETs) and sarafotoxin (SRTXs) gene family comprises a series of 21 amino acid peptides that activate 2 distinct, but homologous, G protein-coupled receptors, designated ETA and ETB (1–3). Although these two receptors show approximately 63% amino acid homology and are both coupled to inositol phosphate hydrolysis, they bind with different affinity ET isopeptides and analogs, show different patterns of expression, and mediate distinct or even opposite functions (i.e. smooth muscle contraction/relaxation). It is generally accepted (4 – 6) that the ETA subtype recognizes with high affinity ET-1 but not ET-3. In addition, ETA-mediated cellular effects are blocked by the peptide antagonist BQ123 and the nonpeptide antagonist SB209670. Conversely, the ETB receptor binds ET-1 and ET-3 with virtually equal affinity and is selectively stimulated by SRTX6c and IRL-1620. Although expression of SRTXs is limited to the snake venom of Atractapsis engaddensis, ETs are widely exReceived March 14, 1997. Revision received August 14, 1997. Accepted August 21, 1997. Address all correspondence and requests for reprints to: Mario Maggi, M.D., Andrology Unit, Viale Pieraccini 6, 50139 Florence, Italy. * This work was supported by a grant from Consiglio Nazionale delle Ricerche Targeted Project “Prevention and Control of Disease Factors” (Contract 93.00097.PF41) and a grant from University of Florence.

contraceptive-treated women). In situ binding studies indicate that the majority of ETA and ETB receptors are expressed in the blood vessels of the ovary. In particular, ETA receptors are abundant in the ovulatory follicles and localized in the theca interna, in close proximity to the granulosa layer. Few GCs of the ovulatory follicle were specifically labeled. Conversely, in the rat ovary, used as a control, ETB receptors were predominantly expressed and localized in GCs. Accordingly, ETB receptors negatively regulated estrogen accumulation in rat GCs. In human granulosa-luteal cells, neither ET-1 (unselective ligand) nor ET-3 or sarafotoxin 6c (ETB ligands) affected estrogen or progesterone secretion. ET-1 was 2.5-fold more potent than noradrenaline in eliciting contraction of ovarian artery, acting through the ETA receptor. Our results indicate that in human ovary, at variance with rat ovary, the endothelin system is primarily involved in the regulation of ovarian blood flow and not steroidogenesis. (J Clin Endocrinol Metab 82: 4122– 4129, 1997)

pressed in mammals, acting as neurotransmitters or autocrine/paracrine hormones (4 – 6). The bioactive ET peptides ET-1, ET-2, and ET-3 are coded from three distinct genes and cleaved from their relative precursors (big ETs) by endothelin-converting enzymes. One of these enzymes, ECE-1, as well as ET-1 (7), is particularly abundant in gonadal tissues (8), suggesting an active processing of ET-1 in these tissues. We have recently provided evidences that this is the case also in human gonads (9, 10). In particular, we found that human ovarian granulosa cells (GCs) express ET-1 gene and protein and that immunoreactive ET-1 is present in human follicular fluid and is regulated by gonadotropin (10). Little information is available on the presence and function of ET receptors in human ovary. A recent study (11) indicated that in luteinized human granulosa cells (L-HGCs), ET-1, but not ET-3 or the selective ETB agonist IRL-1620, stimulates an increase in intracellular calcium concentration and cell proliferation, suggesting the presence of the ETA receptor subtype on L-HGCs. In the same study (11), both ET-3 and IRL-1620 were as potent as ET-1 in inhibiting basal and gonadotropin-stimulated progesterone (P4) accumulation. These findings taken together indicate that L-HGCs contain ETA and ETB receptors. Accordingly, L-HGCs express specific transcripts for both receptor genes; the ETA receptor messenger ribonucleic

4122

ETR IN OVARY

acid (mRNA) is more highly expressed than the ETB one (11). In contrast, in rat ovary, ETB mRNA was definitively more abundant than ETA mRNA and is localized, by in situ hybridization, in the GCs of the developing follicles (12). Radioligand binding and functional studies confirmed the predominant presence of ETB receptors in cultured rat GCs (13). Indeed, in these cells ET-1 was as potent as ET-3 in displacing [125I]ET-1 and in decreasing FSH-induced P4 accumulation (13). Similar results were reported in porcine GCs (14 –16), suggesting the predominant expression of ETB receptors in swine ovary. However, the pharmacological characterization of [125I]ET-1-binding sites in either porcine GCs (17) or bovine luteal cells (18) strongly indicates the presence of ETA, not ETB, receptors. Although conflicting results on the effect of ETs on GCs might reflect species-related differences, they might also be related to the artefactual condition of cell isolation and culturing. The only report on the localization of ET receptor protein in intact swine ovary showed a virtual absence of specific [125I]ET-1-binding sites in the GCs of the developing follicles and a specific signal only in the GCs of the periovulatory follicles (17). Interestingly, the highest specific autoradiographic signal was detected in the blood vessel of the corpus luteum, suggesting, for the first time, a role for ETs in the regulation of ovarian blood flow (17). To investigate the subtype of ET receptors expressed by intact human ovary and their localization we performed selfand cross-competition binding curves among labeled ETs and selective analogs and autoradiographic studies using a blocking design aimed to selectively label ETA and ETB receptors. In addition, to evaluate the function of these receptors we also studied the effect of ETs on steroid production from purified human GCs and on contractility of ovarian blood vessels. Results were compared with those obtained in rat ovary and GCs. Our study indicates that the human ovary expresses a relative abundance of ETA receptors, mostly localized in the blood vessels, that mediate vasoconstriction. Subjects and Methods Subjects For radioligand binding studies ovarian tissues were obtained from 15 women who underwent hysterectomy with salpingo-oophorectomy for benign uterine diseases, including leyomyoma, genital prolapse, and intractable uterine bleeding. Their informed consent was obtained, and the study was approved by the local ethical committee of the Department of Clinical Physiopathology (University of Florence, Florence, Italy). Each subject was interviewed concerning menstrual history, date of the last menstruation, use of oral contraceptive (OC) or other hormonal agents, and chronic consumption of any kind of drug. Further information was obtained by histological evaluation of the endometrium by a gynecological pathologist. On the basis of the aforementioned elements, these women were divided into 3 groups to investigate the relationship between ET receptor levels in the ovary and the hormonal milieu of the host. Group 1: normal cycling women. Eight women were included in this group (mean age, 45.6 yr; range 19 –57 yr). They had a history of regular menses and a proliferative or secretive pattern of the endometrium. According to the age, they were further divided into two subgroups: one included women of up to 45 yr of age (group 1a), and the other included women over 45 years (group 1b). Group 2: menopausal. Four women were included in this group (mean age, 63 yr; range, 57–71 yr). They had a history of at least 5 yr of amenorrhea,

4123

high levels of gonadotropins (FSH, .35 mIU/mL; LH, .25 mIU/mL), and/or a hypotrophic-atrophic pattern of the endometrium. Group 3: OC. Three women were included in this group (mean age, 48.6 yr; range 46 –52 yr). The OC used contained 20 mg ethinyl estradiol and desogestrel. For autoradiographic, in situ hybridization, and contractility studies, 14 ovarian specimens from fertile women (aged 19 – 42 yr) were obtained after informed consent was given. Patients underwent ovarian resection or oophorectomy for benign diseases of the reproductive system. For human granulosa-luteal cell culture, cells were obtained from women undergoing in vitro fertilization programs. All patients were treated by a long protocol with GnRH agonist (GnRH-a; 3.75 mg; d-Trp6-decapeptyl, Ipsen, Milan, Italy). GnRH-a was administered on day 21 of the menstrual cycle preceding oocyte retrieval. After attainment of pituitary desensitization, indicated by serum 17b-estradiol (E2) levels below 125 pmol/L and a follicle diameter ,10 mm, ovarian stimulation was initiated with human menopausal gonadotropin, corresponding to 150 IU FSH and 70 IU LH (Pergogreen, Serono, Milan, Italy). The daily dose of exogenous gonadotropin was continued on an individual basis, depending upon follicular growth. When the two largest follicles reached a mean diameter of 16 mm and the concentration of E2 was consistent, 10,000 IU hCG (Profasi HP, Serono) were administered. Oocytes were retrieved 36 h later.

Chemicals [125I]ET-1 (2000 Ci/mmol), [125I]ET-3 (2000 Ci/mmol), and [aP]CTP (3000 Ci/mmol) were purchased from Amersham (Amity PG, Milan, Italy). ET-1, ET-3, sarafotoxin 6C (SRTX6C), and cyclo-[d-Trpd-Asp-Pro-d-Val-Leu] (BQ123) were obtained from NovaBiochem (Laufelfingen, Switzerland). IRL-1620 was purchased from Alexis (Laufelfingen, Switzerland), and SB209670 was obtained from SmithKline Beecham (King of Prussia, PA). BSA, bovine insulin, bacitracin, benzamidine, diethylstilbestrol (DES), methylisobutylxanthine, and soybean trypsin inhibitor were obtained from Sigma Chemical Co. (St. Louis, MO). McCoy’s 5a medium, l-glutamine, and fungizone were obtained from Life Technologies (Paisley, UK). Insulin was purchased from Eli Lilly Co. (Indianapolis, IN). 32

Binding studies Ovarian samples obtained at surgery were stored in liquid nitrogen. Membranes were prepared as described previously (19). The protein concentration was determined using the Bio-Rad protein assay kit (BioRad Laboratories, Munich, Germany). Binding studies on ovarian membranes were carried out as described previously (19). Aliquots of membranes (0.075 mg/mL) were incubated for 60 min at 22 C with increasing concentrations (10 –70 pmol/L) of [125I]ET-1 or [125I]ET-3 with or without increasing concentrations of unlabeled compounds (10211–1024 mol/L) in buffer 2 in the presence of 0.1% BSA. Self- and cross-displacement curves using ET-1, ET-3, BQ123, IRL-1620, SB209670, and SRTX6C were performed using pooled membranes. All measurements were obtained in triplicate. After incubation, membranes or cells were filtered through Whatman GF/B (membranes) or GF/C (cells) filters (Whatman, Clifton, NJ), presoaked in ice-cold 50 mmol/L Tris, pH 7.4, in 0.1% BSA, using the Brandel M-48R 48-well cell harvester (Gaithersburg, MD). Filters were washed twice with 3 mL ice-cold 50 mmol/L Tris, pH 7.4. Radioactivity retained by filters was measured in a g-counter at 70% efficiency.

Autoradiography Autoradiographic studies in human or rat ovary were carried out according to the method of Maggi et al. (9). For studies in rat ovary, intact female Sprague-Dawley rats (Charles River, Calco, Italy) were used. Ovarian specimens were quickly frozen in isopentane cooled with liquid nitrogen. Serial sections (6 – 8 mm thick) were obtained using a cryostat at 218 C and mounted on gelatin-coated glass slides. The slides were air-dried at room temperature for 10 min and then incubated with 50 pmol/L [125I]ET-1 (in the presence of 100 nmol/L IRL-1620 to mask ETB receptors) or [125I]ET-3 (in the presence of 100 nmol/L BQ123 to mask ETA receptors) in buffer 2 and 0.2% BSA for 60 min at 22 C. To determine nonspecific binding, an excess (100 nmol/L) of unlabeled ET-1 or ET-3,

4124

JCE & M • 1997 Vol 82 • No 12

MANCINA ET AL.

TABLE 1. Concentrations of receptors (femtomoles per mg protein) and affinity (Kd; nanomolar) for ET-1, ET-3, BQ 123, IRL 1620, SRTX 6c, VIC, and SB209670 in membranes from human ovary ETA

Binding capacity (fmol/mg protein) Ligand affinity (Kd, nmol/L) ET-1 ET-3 BQ123 IRL1620 SRTX6c SB209670

5309 6 743

0.88 6 0.14 344.00 6 51.6 4.94 6 1.53 .10,000 4017.1 6 2370 1.29 6 0.36

ETB

50 6 12.5

0.017 6 0.007 0.013 6 0.004 .10,000 13.97 6 8.38 0.44 6 0.11 9.79 6 3.72

Values 6 SEM were obtained from computer modeling using the program Ligand (23) of 2 families of self- and cross-displacement curves performed in 3 different pools of ovarian membranes, derived from normal cycling women. A total of 24 curves were analyzed simultaneously. A model involving 2 classes of sites was significantly better than a model involving a single class of sites (P , 0.0001).

FIG. 1. Two groups of competition curves for [125I]ET-1 (upper panel) and [125I]ET-3 (lower panel) with unlabeled ET-1 (closed triangles), ET-3 (open boxes), SRTX6c (closed boxes), BQ123 (open triangles), SB209670 (closed circles), and IRL-1620 (open circles) obtained in pooled human ovarian membranes. Ordinate: B/T, Bound to total ratio for [125I]ET-1 (upper panel) and [125I]ET-3 (lower panel). Abscissa: [Ligand], Total concentration (molar) of the varying ligand. Values are means of triplicate determinations in a typical experiment. Smooth curves show the predicted relationships for the two-site model shown in Table 1. respectively, was added to the mixture. The sections were washed three times with Tris-HCl buffer, pH 7.4; rinsed in deionized water; air-dried; and fixed by exposure to formaldehyde vapor at 80 C for 60 min. Coverslips coated with nuclear emulsion Illford K5 (Illford, Mobberley, UK) diluted 1:1 with distilled water were applied to the sections, which were exposed at 4 C in the dark for 6 – 8 days. The slides were then developed in Illford Phenisol, fixed in Illford Hypam, stained with hematoxylin, and observed with dark- and brightfield microscopy.

Granulosa cell culture Rat and human granulosa cells were cultured as previously described (20, 21). For rat GCs, intact female immature (25 days old) SpragueDawley rats (Charles River, Calco) were implanted with a 10-mm SILASTIC brand capsules (Dow Corning, Midland, MI) filled with DES. Four days after implantation, rats were killed by cervical dislocation, ovaries were punctured with 27-gauge hypodermic needles, and GCs were carefully expressed into McCoy’s 5a medium. An aliquot was diluted with trypan blue stain, and viable cells were counted with a hemacytometer. Cells were plated into 16-mm, 24-multiwell plates (Corning, Milan, Italy) at a density of 50,000 cells/wellz0.5 mL at 37 C in a humidified, 95% air-5% CO2 water-jacketed incubator in the presence of graded concentrations of ET-1, ET-3, or SRTX6C, for 6 days. Each experiment was run in triplicate in at least three different cultures. Incubation was carried out using McCoy’s 5a medium supplemented with 2 mmol/L l-glutamine, 200 U/mL penicillin, 200 mg/mL strep-

tomycin sulfate, 0.5 mg/mL fungizone, 1 mmol/L androstenedione, 0.1 mmol/L DES, 37.5 ng/mL hCG (NIDDK, CR-127), 35 mg/mL methylisobutylxanthine, and 1.25 mg/mL insulin. At the end of the culture period, medium samples were frozen at 220 C until assayed for estrogen. For L-HGCs, after identification and removal of the oocyte, the follicular fluid containing L-HGCs was centrifuged in 15-mL conical tubes at 300 3 g for 5 min, and the pellet was resuspended in 1 mL McCoy’s 5a medium supplemented with 2 mmol/L l-glutamine, 100 U/mL penicillin, 100 mg/mL streptomycin sulfate, 0.5 mg/mL fungizone, 25 mmol/L HEPES, and 0.1% BSA. This cell suspension was layered onto a 3-mL 50% Percoll gradient diluted with the above medium and centrifuged at 400 3 g for 20 min to pellet the red cells. The band of granulosa-luteal cells was collected by aspiration and washed in 10 mL tissue culture medium, and the pellet was resuspended in 1 mL tissue culture medium for cell counting; cells were plated into 11-mm 48multiwell plates (Costar, Cambridge, MA) at a density of 20,000 cells/ wellz0.2 mL and incubated at 37 C in a humidified, 95% air-5% CO2 water-jacketed incubator for 2 days in the presence of graded concentrations of ET-1, ET-3, and SRTX6C. Androstenedione (1 mmol/L) was added to the incubation medium. The effects of ETs were not tested on FSH-stimulated estrogen production because of the inconsistent response of these cells to FSH. Each experiment was performed in triplicate in at least three different cultures.

Estrogen RIA The concentration of estrogen in the incubation medium was measured directly, without extraction, by RIA, as previously described (21). Aliquots (200 mL) of medium, diluted 1:40 to 1:100, or standard solutions were incubated with 100 mL antiserum (final dilution, 1:10,000) and 100 [3H]E2 (;12,000 cpm) at 4 C for 18 –20 h. Ice-cold charcoal-dextran was added to achieve separation of bound from free labeled hormone. Tubes were centrifuged at 1,500 3 g at 4 C for 11 min, and the supernatants were collected and counted in a b-counter after the addition of scintillation fluid. Total and nonspecific binding values were 36.5 6 2.3% and 2.4 6 0.07%, respectively. Intra- and interassay coefficients of variation were 4.33 6 0.29% and 15.5%, respectively.

P4 RIA The concentration of P4 in the incubation medium was measured directly, without extraction, by RIA, as previously described (22). Aliquots (100 mL) of medium (diluted 1:10) or standard solutions were incubated with 100 mL anti-P4 serum (CU 413/1; Analytical Antibodies, Segrate, Milan, Italy; final dilution, 1:6000) and 100 mL [3H]P4 (;6000 cpm) at 4 C for 18 –20 h. Ice-cold charcoal-dextran was added to achieve separation of bound from free labeled hormone. Tubes were centrifuged at 1500 3 g at 4 C for 11 min, and the supernatants were collected and

ETR IN OVARY

4125

FIG. 2. Autoradiographic localization of ETA and ETB receptors in human ovary using a blocking experimental design. We employed 50 pmol/L [125I]ET-1 (in the presence of 100 nmol/L IRL-1620 to mask ETB receptors) to study the ETA site and 50 pmol/L [125I]ET-3 (in the presence of 100 nmol/L BQ123 to mask ETA receptors) to study the ETB site. A and B are darkfield images of total binding for ETB and ETA sites at the hilum of the ovary. L, Lumen of the vessels. Bars 5 25 mm. Silver grains represent binding sites. Note in A (ETB site) the intense grain localization over vascular endothelial cells. In B (ETA site), binding sites are mostly concentrated in the vascular smooth muscle cells. counted in a b-counter after the addition of scintillation fluid. Total and nonspecific binding were 42.5% and 6.7%, respectively. Intra- and interassay coefficients of variation were 8.9 6 1.2% and 12.4%, respectively.

In vitro contractility of ovarian artery At the time of surgery, the artery entering the hilum of the ovary was carefully dissected and cut into four or five rings (5 mm long). Ring segments were vertically mounted under 1.5 g resting tension in organ chambers containing 10 mL Tyrode’s solution (118 mmol/L NaCl, 25 mmol/L NaHCO3, 4.7 mmol/L KCl, 1.2 mmol/L KH2PO4, 1.2 mmol/L MgSO4, 2.5 mmol/L CaCl2, 10 mmol/L glucose) at 37 C, gassed with 95% O2 and 5% CO2, pH 7.4. The rings were allowed to equilibrate for at least 90 min; during this period the bath medium was replaced every 15 min. Changes in isometric tension was recorded on a chart polygraph. Noradrenaline (NA) increased the tonic tension in a concentration-dependent manner. The maximal effect was obtained at 100 mmol/L. The response to ETs was expressed as a percentage of the maximal effect of NA to normalize the data.

Analysis of experimental results The binding data were evaluated quantitatively with nonlinear least squares curve fitting using the computer program Ligand (23). The computer program Allfit (24) was used for the analysis of sigmoidal dose-response curves obtained in estrogen accumulation and contractility studies. Data were analyzed by one-way ANOVA followed by Duncan’s multiple range test. Each data point represents the mean 6 se.

Results

To identify and pharmacologically characterize ET-binding sites in human ovary, we performed binding studies in pooled ovarian membranes from normal cycling women. Preliminary results indicated that [125I]ET-1 and [125I]ET-3 binding to ovarian membranes reached apparent equilibrium within 60 min at 22 C (data not shown). Subsequent studies were, therefore, performed using this experimental condition. To characterize ET-binding sites in human ovary, we performed three families of self- and cross-competition curves using two labeled ligands ([125I]ET-1 and [125I]ET-3)

and several unlabeled ETA and ETB ligands, as previously described (19). Experiments were carried out using three different pools of ovarian membranes from normal cycling women. Mathematical analysis using the program Ligand (23) strongly indicated the presence of two distinct populations of sites in human ovary with different concentrations and affinities for the ET peptides. Indeed, the fit for the two-site model was consistently better than that for the onesite model (P , 0.0001). Figure 1 graphically represents binding curves and the predicted relationships for the two-site model. Binding parameters are given in Table 1. ET-1 and SB209670 bound with virtually equal affinities to both sites. Conversely, ET-3, and SRTX6c were relatively selective for the lowest capacity site. BQ123 and IRL-1620 showed high selectivity for the highest and the lowest capacity sites, respectively. Hence, the high capacity, BQ123-selective site corresponds to the ETA receptor, whereas the very low capacity, IRL-1620-selective site corresponds to the ETB receptor. To study the individual distribution of ET-1-binding sites in ovarian membranes derived from normal cycling [n 5 8; group 1, further divided into two subgroups: up to 45 yr (group 1a; n 5 3) and .45 yr (group 1b; n 5 5)], postmenopausal (n 5 4, group 2) and OC-treated women (n 5 3; group 3), we performed homologous competition curves for ET-1 in each membrane preparation. To analyze differences among the different groups in ET-1 binding affinities, we analyzed the experimental results using the program Ligand (23), as previously described (25). By using a three ligands by three binding sites model (25), we tested the hypothesis that the affinity (Kd) was different among groups. The imposed constraints of shared Kd values did not significantly degrade the fits compared to that obtained with free Kd (P 5 0.1; Kd 5 0.79 6 0.13 nmol/L). In addition, ANOVA did not show significant differences in the derived binding capacities for the four groups (group 1a, 5.115 6 1.084 pmol/mg protein;

4126

MANCINA ET AL.

group 1b, 4.906 6 0.653 pmol/mg protein; group 2, 3.422 6 0.471 pmol/mg protein; group 3, 5.114 6 1.033 pmol/mg protein; P 5 0.38). Hence, different hormonal milieux do not affect the Kd and binding capacity of ET receptors in human ovary. When the density of ET receptors was analyzed as a function of age in the population of hormonal untreated women (groups 1 and 2), a tendency for a significant, negative relationship was found (P 5 0.058). To localize the ET-binding sites in the human ovary we performed in situ binding studies using a blocking experimental design. We employed 50 pmol/L [125I]ET-1 (in the presence of 100 nmol/L IRL-1620 to mask ETB receptors) to study the ETA site and 50 pmol/L [125I]ET-3 (in the presence of 100 nmol/L BQ123 to mask ETA receptors) to study the ETB site. Experiments were carried out in five different ovarian specimens from normal cycling women at different stages of

JCE & M • 1997 Vol 82 • No 12

the menstrual cycle. Rat ovary was analyzed as a control. In all experiments a positive signal for ETB and ETA receptors was found in the largest ovarian blood vessels (Fig. 2, A and B). In particular, ETB sites were predominantly located in the endothelial cells, whereas ETA sites were more abundant in the muscular wall. As expected, in the rat ovary, an intense positivity for ETB receptors was found in GCs (data not shown). Conversely, in human ovary, GCs that do express ET-1 mRNA (10) were absolutely negative for either ETA- or ETB-binding sites. Figure 3 shows results obtained in a tertiary follicle. Although an evident hybridization signal of ET-1 mRNA was localized in the majority of GCs and in some endothelial cells of the follicular vessels (Fig. 3A), GCs were definitively negative for ETA (Fig. 3B) or ETB (Fig. 3D) receptors. On the other hand, almost all the smooth muscle cells of the ovarian vessels were positive for ETA receptors

FIG. 3. In situ hybridization of prepro-ET-1 and autoradiographic localization of ETA and ETB receptors in a tertiary follicle of the human ovary. For in situ hybridization, we performed experiments as described in Ref. 10, using a specific antisense 35S-labeled RNA probe. For in situ binding studies, we employed a blocking experimental design. Frozen sections were incubated with 50 pmol/L [125I]ET-1 (in the presence of 100 nmol/L IRL-1620 to mask ETB receptors) to study the ETA site and 50 pmol/L [125I]ET-3 (in the presence of 100 nmol/L BQ123 to mask ETA receptors) to study the ETB site. A, Darkfield microphotograph showing the distribution of ET-1 mRNA expression over the GCs (G) of the follicle and endothelial cells of the vessels (V; bar 5 25 mm). B and C, Darkfield images of total (B) and nonspecific (C) in situ binding for ETA receptors in adjacent sections of the same tertiary follicle as that in A (bars 5 25 mm). In B, silver grains are present in the blood vessels (V) and are virtually absent in GCs (G). In C, silver grains are almost absent in the field when sections were incubated in the presence of an excess (100 nmol/L) of ET-1. D represents autoradiographic results (total binding) for ETB receptors. Binding sites were virtually absent in both GCs (G) and blood vessels (V). Bar 5 50 mm.

ETR IN OVARY

(Fig. 3B). Silver grains for ETA receptors were virtually absent in control sections incubated in the presence of an excess (100 nmol/L) of ET-1 (Fig. 3C). Similar results were obtained in follicles at different stages of development (not shown). The autoradiographic results show the effects of ET- 1 and ET-3 on steroid accumulation in human and rat GCs. Figure 4 shows the effect of ETs on estrogen accumulation. We found that both ET-1 and ET-3 significantly inhibited estrogen production in rat GC (P , 0.05), with half-maximal inhibitory concentrations (IC50) of 74 and 2.2 pmol/L, respectively. Analysis of the sigmoidal dose-response relationships using the program Allfit indicated that constraining both curves to share identical IC50 values had no deleterious effect on the goodness of fit (P 5 0.961). This finding indicates that rat GCs are equally responsive to ET-1 and ET-3, suggesting expression of ETB receptors. Conversely, in L-HGCs both ETs as well as SRTX6c (data not shown) failed to affect estrogen accumulation. As a recent report (11) indicated that in L-HGCs, 100 nmol/L ET-1, ET-3, and IRL-1620 did not affect estrogen secretion, but slightly inhibited P4 accumulation (25%), we tested the effects of different concentration of ET-1, ET-3, and SRTX6c on P4 accumulation; the results are reported in Table 2. As shown, basal P4 production from L-HGCs was not affected by any of the peptides tested. As we found that the majority of ET receptors were localized in the ovarian blood vessels, we investigated the effects of ETs on ovarian artery contractility. Figure 5 shows traces of the dose-response relationship for cumulative applied ET-1 (upper panel) and IRL-1620 (lower panel) in isolated ovarian artery rings in a typical experiment. Results from four independent experiments are reported in Table 3. ET-1 caused dose-dependent vasoconstriction, with an ED50 of 3.9 6 1.3 nmol/L. This value is in good agreement with the affinity constant for ET-1 derived from binding studies (0.8 nmol/L). The maximum response to ET-1 was 2.7 6 0.16-fold greater than that to NA (100 mmol/L), indicating that ET-1 is one of the most potent vasoconstrictors of the ovarian artery. The increased tone of the arterial rings was hardly

FIG. 4. Effects of increasing concentration of ET-1 (triangles) and ET-3 (boxes) on estrogen production by human (open symbols) and rat (closed symbols) GCs. *, P , 0.05 vs. control.

4127

reduced and did not return to the resting state even after repeated washings with fresh solution. Mathematical analysis, using the program Allfit, of families of ET-1 dose-response curves in the presence or absence of different concentrations of the ETA antagonist BQ123 indicates that BQ123 significantly increased the ED50 for ET-1 (P , 0.0001; see Table 3), without affecting the maximum response (P 5 0.96). These findings indicate that the majority of ET receptors present in human ovarian artery belong to the ETA subtype and that BQ123 is a competitive antagonist for these receptors. Accordingly, increasing concentrations of the selective ETB agonist IRL-1620 were almost ineffective in stimulating ovarian artery contractility compared to ET-1 (see Fig. 5, lower panel). Discussion

In this study we report for the first time the presence, distribution, and biological activity of ET receptors in intact human ovary. Our autoradiographic studies indicate that in rat ovary, which we used as a control, ETB receptors are present in the developing follicular wall. Hence, the ETB receptor protein is expressed in the same cells where previous in situ hybridization studies found specific signal for ETB mRNA (12). Based on previous reports (13, 26) and the present study, it is conceivable that in the rat, ETB receptor negatively regulates steroidogenesis. A recent report indicates that this is the case also in L-HGCs (11). Indeed, a slight (25%), yet significant, inhibition of basal P4 secretion was found with 100 nmol/L ET-1, ET-3, and IRL-1620 (11). However, in a Percollpurified preparation of L-HGCs (20), we were unable to demonstrate any effect of a wide range of concentrations of SRTX6c and ET-3 (selective ETB agonists) or ET-1 (unselective ETA/ETB agonist) on estrogen or P4 accumulation, whereas the same preparation of L-HGCs was still responsive to other physiological stimuli (21). These results are in good agreement with those derived from in situ binding on

4128

JCE & M • 1997 Vol 82 • No 12

MANCINA ET AL.

TABLE 2. Effects of increasing concentration of ET-1, ET-3, and SRTX 6c on P4 accumulation by L-hGCs

ET-1 ET-3 SRTX6c

Control

10211

10210

1029

1028

1027

100 6 6.4 (16) 100 6 6.6 (14) 100 6 10.4 (9)

82.9 6 8.2 113.66 6 16 (10)

107.7 6 8.2 (8) 108.9 6 12 (15) 89.3 6 8.3 (9)

99.7 6 7.8 (17) 107.2 6 10.8 (15) 127 6 10.5 (9)

113.9 6 13 (18) 87.6 6 6.8 (12) 106.3 6 12.9 (9)

98.9 6 14 (10) 101.2 6 18 (10)

The number of determinations is given in parentheses.

FIG. 5. Isometric tension recordings in two rings of the same human ovarian artery induced by ET-1 (upper trace) and IRL-1620 (lower trace). ET-1 dosedependently increased tension, whereas IRL-1620 was without effect. After maximal stimulation with IRL1620 (300 nmol/L), the arterial ring was still responsive to ET-1 (100 nmol/L), indicating that ETB receptors are not involved in the contractile effect of ET-1.

TABLE 3. Effects of different concentrations of BQ 123 on ET-1induced vasoconstriction of human ovarian artery BQ 123 (mol/L)

ET-1 log (ED50) ET-1 Emax (% NA Emax)

None

1027

1026

28.4 6 0.14

27.9 6 0.25

27.1 6 0.3a

274 6 21

160 6 20

181 6 54

a

Results (mean 6 SE) are derived from the average of three or four separate experiments. We used constrained simultaneous curve fitting (program Allfit) (24) for testing the equality of parameters (ED50 and Emax) in the different experimental conditions, by inspecting the consequences of forcing them to be equal. Constraining the curves to share identical Emax values had no deleterious effect in the goodness of fit (mean Emax 5 267 6 16; P 5 0.96). In contrast, constraining the curves to share identical values of ED50 significantly worsened the fit. Hence, BQ 123 significantly blunted the ED50 without changing the Emax for ET-1. a P , 0.0001.

intact human ovary. Indeed, we did not find any specific binding sites for [125I]ET-3 (in the presence of 100 nmol/L BQ123) in human GCs. By using this tracer, binding sites were only identified in the endothelial cells and the muscular wall of large ovarian vessels. Accordingly, quantitative analysis of self- and cross-competition curves among [125I]ET-1, [125I]ET-3, and several selective analogs indicates that the

concentration of ETB sites in intact human ovary is 100-fold lower than the estimated density of ETA sites. In conclusion, we do not believe that ETB receptors play any important role in human GCs, at variance with the situation in rat ovary. A similar discrepancy between primate and rat follicular physiology has been reported for estrogen receptor. Indeed, although in rodents estrogens regulate preantral follicle growth and sensitivity to FSH, compelling evidence that this is also the case in the primate follicle is lacking (27). Although the concentration of ETB receptor in intact human ovary is negligible, the density of ETA receptor is relatively high (i.e. 5-fold higher than that in human testis) (9). However, we were unable to demonstrate significant differences in the concentration of [125I]ET-1-binding sites in ovarian membranes derived from normal cycling, postmenopausal, or OC-treated women. In addition, we did not find specific signal for ETA receptor in the large majority of GCs examined by in situ binding. Similar results were reported in porcine ovary (17). In particular, we did not find ET receptors in GCs expressing a positive signal for ET-1 mRNA. Our results are in apparent contrast with those of Kamada et al. (11), showing that ETA receptor is involved in the regulation of L-HGC proliferation and calcium mobilization. However, their data have been generated in cultured luteinized GCs, whereas our morphological data reflect the scenario of the

ETR IN OVARY

intact human follicles. It is indeed possible that the cell culture conditions altered the number and subtype of ET receptors expressed, as we have recently demonstrated for human hepatic stellate cells (28). As human GCs contain ET-1 mRNA and protein (10) without expressing an abundance of the corresponding receptor (present study), an autocrine action for follicular ET-1 is unlikely. Hence, we sought other ovarian targets. By autoradiography, we found an intense positivity for ETA (but not ETB) receptors in the vessels of the innermost part of the theca, immediately adjacent to the granulosa layer. This finding was more apparent in the thecal compartment of the ovulatory follicles and suggests a possible role for ET-1 in the regulation of ovarian vasomotion. To investigate the function of ETA and ETB receptors in ovarian vessels, we performed in vitro contractility studies using arterial rings prepared from the arterial vessels at the hilum of the ovary, where in situ binding demonstrated the presence of both subtypes of receptors. We found that ET-1 stimulated vasoconstriction (ED50 5 4 nmol/L) acting through ETA receptors. As the effect of ET-1 was 2.5-fold more potent than that of NA, this peptide should be considered the most important vasoconstrictor for ovarian vessels. Hence, we speculate that ET-1 produced by granulosa cells (10) may act in a paracrine fashion on the theca interna, which is extremely rich in blood vessels. Studies with radioactive microspheres in the ewe showed great variations in the follicular blood supply near the time of ovulation (29, 30). In particular, Murdoch reported that the LH surge induced an initial phase of intense hyperemia in the ovulatory follicle, followed by vasoconstriction. As we found higher levels of ET-1 in the follicular fluid of gonadotropin-stimulated than in spontaneous in vitro fertilization protocols (10), it is possible that the LH surge at midcycle stimulates ET-1 production from GCs, which, in turn, plays a role in regulating blood supply to the ovulatory follicle. Similar results have been recently provided for rat testis by Collins et al. (31). These researchers demonstrated that in the testis, hCG induced a time-dependent increase in ET-1 accumulation and that relatively low concentrations of ET-1 regulate testicular blood flow, acting through the ETA receptors (31). In conclusion, we demonstrated that human ovary expresses an elevated concentration of ETA receptors (4 –5 pmol/mg protein), mostly located on the muscular wall of the ovarian vessels. As human GCs produce ET-1, and receptors are present in arterioles of theca interna in close proximity to the granulosa layer, a paracrine role for follicular ET-1 is suggested. Acknowledgments We thank Dr. Elisabetta Baldi, Andrology Unit, University of Florence, for helpful suggestions during the course of the study.

References 1. Bax WA, Saxena PR. 1994 The current endothelin receptor classification: time for reconsideration? Trends Pharmacol Sci. 15:379 –386. 2. Davenport AP, Maguire JJ. 1995 Endothelin receptor sub-types and their function. In: Baldi E, Maggi M, Cameron IT, Dinn MJ, eds. Frontiers in endocrinology: new advances. Rome: Ares-Serono Symposia; vol 15:35– 45.

4129

3. Sokolovsky M. 1995 Endothelin receptor subtypes and their role in transmembrane signaling mechanism. Pharmacol Ther. 68:435– 471. 4. Masaki T. 1993 Endothelins: homeostatic and compensatory actions in the circulatory and endocrine systems. Endocr Rev. 14:256 –268. 5. Rubanyi GM, Polokoff MA. 1994 Endothelins: molecular biology, biochemistry, pharmacology, physiology, and pathophysiology. Phramacol Rev. 46:325– 415. 6. Levin ER. 1995 Endothelins. N Engl J Med. 6:356 –362. 7. Matsumoto H, Suzuki N, Onda H, Fujino M. 1989 Abundance of endothelin-3 in rat intestine, pituitary gland and brain. Biochem Biophys Res Commun. 164:74 – 80. 8. Xu D, Emoto N, Ciaid A, et al. 1994 ECE-1: a membrane-bound metalloprotease that catalyzes the proteolytic activation of big endothelin-1. Cell. 78:473– 485. 9. Maggi M, Barni T, Orlando C, et al. 1995 Endothelin-1 and its receptor in human testis. J Androl. 16:213–224. 10. Magini A, Granchi S, Orlando C, et al. 1996 Expression of endothelin-1 gene and protein in human granulosa cells. J Clin Endocrinol Metab. 81:1428 –1433. 11. Kamada S, Blackmore PF, Kubota T, et al. 1995 The role of endothelin-1 in regulating human granulosa cell proliferation and steroidogenesis in vitro. J Clin Endocrinol Metab. 80:3708 –3714. 12. Iwai M, Hori S, Shigemoto R, et al. 1993 Localization of endothelin receptor messenger ribonucleic acid in the rat ovary and fallopian tube by in situ hybridization. Biol Reprod. 49:675– 680. 13. Tedeschi C, Lohman C, Ittoop O, et al. 1994 Rat ovarian cell as site of endothelin reception and action: attenuation of gonadotropin-stimulated steroidogenesis via perturbation of the A-kinase signaling pathway. Biol Reprod. 51:1058 –1065. 14. Kamada S, Kubota T, Taguchi M, Eguchi S, Marumo F, Aso T. 1992 Direct effect of endothelin-1 on the granulosa cells of the porcine ovary. J Endocrinol. 134:59 – 66. 15. Iwai M, Hasegawa M, et al. 1991 Endothelins inhibit luteinization of cultured porcine granulosa cells. Endocrinology. 129:1909 –1914. 16. Kamada S, Kubota T, Hirata Y, et al. 1993 Endothelin 1 is an autocrine/ paracrine regulator of porcine granulosa cells. J Endocrinol Invest. 16:425– 431. 17. Flores JA, Winters TA, Knight JW, Veldhuis JD. 1995 Nature of endothelin binding in the porcine ovary. Endocrinology. 136:5016 –5019 18. Girsh E, Milvae RA, Wang W, Meidan R. 1996 Effect of endothelin-1 on bovine luteal cell function: role of prostaglandin F2a-induced antisteroidogenic action. Endocrinology. 137:1306 –1312. 19. Maggi M, Vannelli GB, Peri A, et al. 1991 Immunolocalization, binding and biological activity of endothelin in rabbit uterus: effect of ovarian steroids. Am J Physiol 260:E292–E305. 20. Erickson GF, Garzo VG, Magoffin DA. 1989 Insulin-like growth factor-I regulates aromatase activity in human granulosa and granulosa luteal cells. J Clin Endocrinol Metab. 69:716 –724. 21. Calogero AE, Burrello N, Negri-Cesi P, et al. 1996 Effects of corticotrophinreleasing hormone on ovarian estrogen production in vitro. Endocrinology. 137:4161– 4166. 22. Cianci A, Calogero AE, Palumbo MA, et al. 1996 Relationship between tumor necrosis factor-alpha and sex steroid levels in the follicular fluid of women with immunological infertility. Hum Reprod. 11:265–268. 23. Munson PJ, Rodbard D. 1980 LIGAND: a versatile computerized approach for characterization of ligand-binding systems. Anal Biochem. 107:220 –239. 24. De Lean A, Munson PJ, Rodbard D. 1978 Simultaneous analysis of families of sigmoidal curves: application to bioassay, radioligand assay, and physiological dose-response curves. Am J Physiol 235:E97–E102. 25. Maggi M, Peri A, Baldi E, et al. 1996 Interferon alpha downregulates expression of the oxytocin receptor in cultured human myometrial cells. Am J Physiol. 271: E840 –E846. 26. Tedeschi C, Hazum E, Kokia E, et al. 1992 Endothelin-1 as a luteinization inhibitor: inhibition of rat granulosa cell progesterone accumulation via selective modulation of key steroidogenic steps affecting both progesterone formation and degradation. Endocrinology. 131:2476 –2478. 27. Zelinski-Wooten MB, Stouffer RL. 1996 Steroid receptors and action in the primate follicle. Trends Endocrinol Metab. 5:177–183. 28. Pinzani M, Milani S, De Franco R. 1996 Endothelin-1 is overexpressed in human cirrothic liver and exerts effects on activated hepatic stellate cells. Gastroenterology. 110:534 –548. 29. Murdoch WJ, Nix KJ, Dunn TG. 1983 Dynamics of ovarian blood supply to periovulatory follicles of the ewe. Biol Reprod. 28:1001–1006. 30. Murdoch WJ, Myers DA. 1983 Effect of treatment of estrous ewes with indomethacin on the distribution of ovarian blood to the periovulatory follicle. Biol Reprod. 28:1229 –1232. 31. Collins A, Damber J-A, Bergh A. 1996 effects of endothelin-1 on the rat testicular vasculature. J Androl. 17:360 –366.