Molecular and Cellular Endocrinology, 75 (1991) R1-R6 © 1991 Elsevier Scientific Publishers Ireland, Ltd. 0303-7207/91/$03.50
R1
MOLCEL 02462
Rapid Paper Effect of recombinant inhibin on androgen synthesis in cultured human thecal cells S.G. Hillier a, E.L. Yong
a, p.j.
Illingworth a, D.T. Baird 1, R.H. Schwall 2 and A.J. Mason 2
1 Department of Obstetrics and Gynaecology, University of Edinburgh Centre for Reproductive Biology, Edinburgh EH3 9E W, U.K., and 2 Genentech Inc., South San Francisco, CA 94080, U.S.A.
(Received 13 December 1990; accepted 13 December 1990)
Key words: Ovary; Follicle; Luteinizing hormone; Insulin-like growth factor I
Summary Effects of inhibin (recombinant human inhibin-A) on ovarian androgen synthesis were tested in vitro using serum-free monolayer cultures of human thecal cells. Treatment for 4 days with inhibin alone at doses between 10 and 100 n g / m l caused modest (approximately 2-fold) increases in production of androgen (androstenedione and dehydroepiandrosterone): similar to the maximal level of stimulation caused by luteinizing hormone (LH) (10 n g / m l ) alone but only about one-third of that caused by insulin-like growth factor I (IGF-I) (30 n g / m l ) alone. Combined treatment with L H and inhibin elicited additive effects on androgen production whereas LH and IGF-I were synergistic, giving rise to androgen production rates at least 40 times greater than control. Additional presence of inhibin caused up to 10-fold augmentation of the response to LH + IGF-I. Activin (recombinant human activin-A) was previously shown to inhibit L H + IGF-I-induced androgen synthesis in this human thecal cell culture system. In the present study we found that the additional presence of inhibin ( > 1 n g / m l ) completely neutralized this inhibitory action of activin (10 ng/ml). These effects of inhibin were dose-dependent (EDs0 1-10 n g / m l ) and maximal at - 1 0 0 ng/ml. Inhibin stimulation of androgen synthesis occurred in the absence of measurable effects on progesterone production, and cell numbers in cultured cell monolayers were unaltered by the protein. It is concluded that inhibin exerts potent and selective stimulation of human thecal cell androgen synthesis in vitro. These results a paracrine role for inhibin(s) in modulating follicular androgen biosynthesis in the human ovary.
Introduction Address for correspondence: Dr. Stephen G. Hillier, Reproductive Endocrinology Laboratory, University of Edinburgh, Centre for Reproductive Biology, 37 Chalmers Street, Edinburgh EH3 9EW, Scotland, U.K. * Supported by MRC Programme Grant PG8929583 (S.G.H. and D.T.B.).
Inhibin is an - 32 kDa glycoprotein which has been isolated from ovarian follicular fluid as a heterodimer of a common a-subunit inhibin with one of two fl-subunits: flA (inhibin-A) or fib (in-
R2
hibin-B) (Vale et al., 1988; Ying et al., 1988). The three subunits are encoded by separate mRNAs (Forage et al., 1985; Mason et al., 1985, 1986; Esch et al., 1987) which are expressed in granulosa cells under the control of follicle-stimulating hormone (FSH) (Woodruff et al., 1987, 1988; Turner et al., 1989). An endocrine function for ovarian inhibin(s) has been proposed, based on the ability of the protein to inhibit FSH release by pituitary cells (Vale et al., 1988; Ying et al., 1988). However, an intraovarian function(s) for inhibin is also likely, supported by evidence that inhibin purified from porcine follicular fluid stimulates luteinizing hormone (LH)-stimulated androgen production by rat ovarian thecal/interstitial tissues in vitro (Hsueh et al., 1987). To determine if inhibin might regulate thecal function in the adult human ovary, we tested the effect of recombinant human inhibin-A (inhibin) on steroid production by primary monolayer cultures of human thecal cells. Here, we report that when thecal cell steroid synthesis is induced by treatment with LH and (or) insulin-like growth factor I (IGF-I), presence of femtomolar concentrations of inhibin in serum-free culture medium causes a selective stimulation of androgen (androstenedione and dehydroepiandrosterone (DHA)) production. We also show that inhibin overcomes the previously described (Hillier et al., submitted) inhibitory action of activin on human thecal androgen synthesis. These data therefore provide in vitro evidence that inhibin(s) may function as a paracrine modulator of androgen synthesis in the human ovary. Patients and methods
Patients. Ovarian tissue was obtained from three women who underwent surgery (hysterectomy with uni- or bilateral oophorectomy) to treat severe uterine fibroids, endometriosis, a n d / o r heavy and painful menstrual bleeding (Table 1). The study had local ethics committee approval. Isolation of thecal cells. Resected ovarian tissue was placed in ice-cold physiological saline, transported to the laboratory and transferred to culture medium which was also used for follicular dissection: medium 199 containing Earle's salts, 25 mM Hepes buffer, extra (2 mM) L-glutamine,
TABLE 1 DETAILS OF PATIENTS A N D OVARIAN FOLLICLES STUDIED Experiment Patient Age LMP a No. and diameter (mm) (years) (day) of follicles studied
I
F
38
27
17 (3-5)
b
II
G
48
24
3 (7-8)
b
III IV
H
34
8
3 (10-11) b 2 (14-15) b
a Last menstrual period; b Pooled.
50 U / m l penicillin, 50 /~g/ml streptomycin (all from Gibco, Paisley, Renfrewshire, U.K.) and 0.1% (w/v) bovine serum albumin (from ICN Biomedicals, High Wycombe, Bucks, U.K.). Dissection of follicles (Table 1) and isolation of thecae were carried out with visualization using zoom-stereomicroscopic optics, as described previously (Hillier et al., 1981). Only thecae from apparently healthy follicles were studied, i.e. yielding strawcoloured follicular fluid, granulosa cells and (usually) a cumulus-enclosed oocyte. After scraping all visible granulosa cells from the lamina basalis, hemisected thecae from individual or pooled follicles were rinsed in fresh medium and transferred to more of the same medium (2 ml) containing 0.1% (w/v) collagenase type II from C1. histolyticure and 0.01% (w/v) deoxyribonuclease I (both from Sigma Cemical Company, Poole, Dorset, U.K.). Complete dispersal into a cellular suspension was achieved by incubating the tissue in the enzyme solution for two successive 15 rain periods at 37°C in a shaking water bath, with repeated pipetting after each incubation. The cells were sedimented by centrifugation (5 min at 800 x g), resuspended in fresh culture medium containing 5.0% (v/v) donor calf serum (Gibco) and counted in a haemocytometer. Cell viability, determined by staining with 0.4% (w/v) trypan blue, was consistently > 90%. Thecal cell culture. Twenty-four-well plastic dishes (Linbro Space Savers from Flow Laboratories, Rickmansworth, Herts, U.K.) were inoculated with replicate 500 ~l portions of thecal cell suspension (2.0-5.0 x 104 viable cells per well) in culture medium containing 5% donor calf serum.
R3
Preincubation in serum-containing medium to allow cell anchorage and recovery from the enzymatic dispersal procedure was for 24 h at 37 o C in a humidified tissue-culture incubator gassed with 95% air/5% CO 2. After removal of serumcontaining medium and washing with 1 ml prewarmed (37°C) Dulbecco's phosphate-buffered saline, the cultures received 500 /~1 serum-free medium containing human L H (LER-1972: donated by Dr. L.E. Reichert, Jr.), recombinant human IGF-I (CGP-35'-126: donated by Dr. H.H. Peter and Dr. K. Scheibli), recombinant human activin-A (Schwall et al., 1989) a n d / 0 r recombinant human inhibin-A (Mason, 1988). Incubation at 3 7 ° C was then for 4 days with a medium change on day 2. Media collected on days 2 and 4 of culture were stored frozen at - 2 0 ° C until assayed for steroid content, as described below. In some experiments the cultured cell monolayers were trypsinized (Hillier and de Zwart, 1982) on day 4 and cell number was estimated with a Coulter counter. Assay of steroids in thecal cell culture medium. Androstenedione and progesterone in unextracted culture medium were measured individually by radioimmunoassay, using standard procedures and antisera as described previously (Hillier et al., 1981; McNatty et al., 1983). The D H A radioimmunoassay was similar, using a sheep antiserum against DHA-3-hemisuccinyl-ovalbumin purchased from Steranti Research, St. Albans, Hefts, U.K. (product code A005). The inter- and intraassay precision for each steroid was less than 15%. Steroid production rates are expressed as pmol steroid/1000 cells/48 h, related to the number of viable cells used to inoculate each culture well. Statistics. Analysis of variance with the Neuman-Keuls test was used to analyze differences between experimental and control observations. Differences assigned a P value of < 0.05 were regarded as statistically significant. Results
Production of androstenedione and D H A was increased approximately 2-fold by inhibin at concentrations between 10 and 100 n g / m l (Fig. 1). Inhibin was compared with other factors previously shown to stimulate androgen synthesis in
0 A4 o.oo.
~= =08 o gg 2~
-o -~
*
• DHA
~
**
0.04
----4
0.02
40 Inhibin concentration (ng/ml) Fig. 1. Stimulation of androgen synthesis in cultured human thecal cells by inhibin. Thecal cells (5.0 × 104 viable cells per culture well) pooled from 17 follicles between 3 and 5 m m in diameter (Experiment I, Table 1) were incubated for 4 days in serum-free culture medium containing inhibin at the concentrations shown. Androstenedione (A4) and dehydroepiandrosterone (DHA) accumulation in medium collected on day 4 are shown. Data are mean + SEM, n = 3. * P < 0.05, * * P < 0.01 vs. treatment without inhibin.
the same thecal culture system. As shown in Fig. 2, the stimulatory effect of inhibin was approximately equivalent to a maximal stimulatory dose of L H (10 n g / m l ) and the effects of inhibin and L H were additive. However, the response to a maximal stimulatory dose of inhibin was only a fraction (approximately one-third) of that to IGF-I
(A)
0.50 G.c: .o ¢o
(B)
*
r l - LH
- r
E~I+LH
~L~
0.15 0.12 0.09
t o
~
0.06
-~o 0.20
R1
MOLCEL 02462
Rapid Paper Effect of recombinant inhibin on androgen synthesis in cultured human thecal cells S.G. Hillier a, E.L. Yong
a, p.j.
Illingworth a, D.T. Baird 1, R.H. Schwall 2 and A.J. Mason 2
1 Department of Obstetrics and Gynaecology, University of Edinburgh Centre for Reproductive Biology, Edinburgh EH3 9E W, U.K., and 2 Genentech Inc., South San Francisco, CA 94080, U.S.A.
(Received 13 December 1990; accepted 13 December 1990)
Key words: Ovary; Follicle; Luteinizing hormone; Insulin-like growth factor I
Summary Effects of inhibin (recombinant human inhibin-A) on ovarian androgen synthesis were tested in vitro using serum-free monolayer cultures of human thecal cells. Treatment for 4 days with inhibin alone at doses between 10 and 100 n g / m l caused modest (approximately 2-fold) increases in production of androgen (androstenedione and dehydroepiandrosterone): similar to the maximal level of stimulation caused by luteinizing hormone (LH) (10 n g / m l ) alone but only about one-third of that caused by insulin-like growth factor I (IGF-I) (30 n g / m l ) alone. Combined treatment with L H and inhibin elicited additive effects on androgen production whereas LH and IGF-I were synergistic, giving rise to androgen production rates at least 40 times greater than control. Additional presence of inhibin caused up to 10-fold augmentation of the response to LH + IGF-I. Activin (recombinant human activin-A) was previously shown to inhibit L H + IGF-I-induced androgen synthesis in this human thecal cell culture system. In the present study we found that the additional presence of inhibin ( > 1 n g / m l ) completely neutralized this inhibitory action of activin (10 ng/ml). These effects of inhibin were dose-dependent (EDs0 1-10 n g / m l ) and maximal at - 1 0 0 ng/ml. Inhibin stimulation of androgen synthesis occurred in the absence of measurable effects on progesterone production, and cell numbers in cultured cell monolayers were unaltered by the protein. It is concluded that inhibin exerts potent and selective stimulation of human thecal cell androgen synthesis in vitro. These results a paracrine role for inhibin(s) in modulating follicular androgen biosynthesis in the human ovary.
Introduction Address for correspondence: Dr. Stephen G. Hillier, Reproductive Endocrinology Laboratory, University of Edinburgh, Centre for Reproductive Biology, 37 Chalmers Street, Edinburgh EH3 9EW, Scotland, U.K. * Supported by MRC Programme Grant PG8929583 (S.G.H. and D.T.B.).
Inhibin is an - 32 kDa glycoprotein which has been isolated from ovarian follicular fluid as a heterodimer of a common a-subunit inhibin with one of two fl-subunits: flA (inhibin-A) or fib (in-
R2
hibin-B) (Vale et al., 1988; Ying et al., 1988). The three subunits are encoded by separate mRNAs (Forage et al., 1985; Mason et al., 1985, 1986; Esch et al., 1987) which are expressed in granulosa cells under the control of follicle-stimulating hormone (FSH) (Woodruff et al., 1987, 1988; Turner et al., 1989). An endocrine function for ovarian inhibin(s) has been proposed, based on the ability of the protein to inhibit FSH release by pituitary cells (Vale et al., 1988; Ying et al., 1988). However, an intraovarian function(s) for inhibin is also likely, supported by evidence that inhibin purified from porcine follicular fluid stimulates luteinizing hormone (LH)-stimulated androgen production by rat ovarian thecal/interstitial tissues in vitro (Hsueh et al., 1987). To determine if inhibin might regulate thecal function in the adult human ovary, we tested the effect of recombinant human inhibin-A (inhibin) on steroid production by primary monolayer cultures of human thecal cells. Here, we report that when thecal cell steroid synthesis is induced by treatment with LH and (or) insulin-like growth factor I (IGF-I), presence of femtomolar concentrations of inhibin in serum-free culture medium causes a selective stimulation of androgen (androstenedione and dehydroepiandrosterone (DHA)) production. We also show that inhibin overcomes the previously described (Hillier et al., submitted) inhibitory action of activin on human thecal androgen synthesis. These data therefore provide in vitro evidence that inhibin(s) may function as a paracrine modulator of androgen synthesis in the human ovary. Patients and methods
Patients. Ovarian tissue was obtained from three women who underwent surgery (hysterectomy with uni- or bilateral oophorectomy) to treat severe uterine fibroids, endometriosis, a n d / o r heavy and painful menstrual bleeding (Table 1). The study had local ethics committee approval. Isolation of thecal cells. Resected ovarian tissue was placed in ice-cold physiological saline, transported to the laboratory and transferred to culture medium which was also used for follicular dissection: medium 199 containing Earle's salts, 25 mM Hepes buffer, extra (2 mM) L-glutamine,
TABLE 1 DETAILS OF PATIENTS A N D OVARIAN FOLLICLES STUDIED Experiment Patient Age LMP a No. and diameter (mm) (years) (day) of follicles studied
I
F
38
27
17 (3-5)
b
II
G
48
24
3 (7-8)
b
III IV
H
34
8
3 (10-11) b 2 (14-15) b
a Last menstrual period; b Pooled.
50 U / m l penicillin, 50 /~g/ml streptomycin (all from Gibco, Paisley, Renfrewshire, U.K.) and 0.1% (w/v) bovine serum albumin (from ICN Biomedicals, High Wycombe, Bucks, U.K.). Dissection of follicles (Table 1) and isolation of thecae were carried out with visualization using zoom-stereomicroscopic optics, as described previously (Hillier et al., 1981). Only thecae from apparently healthy follicles were studied, i.e. yielding strawcoloured follicular fluid, granulosa cells and (usually) a cumulus-enclosed oocyte. After scraping all visible granulosa cells from the lamina basalis, hemisected thecae from individual or pooled follicles were rinsed in fresh medium and transferred to more of the same medium (2 ml) containing 0.1% (w/v) collagenase type II from C1. histolyticure and 0.01% (w/v) deoxyribonuclease I (both from Sigma Cemical Company, Poole, Dorset, U.K.). Complete dispersal into a cellular suspension was achieved by incubating the tissue in the enzyme solution for two successive 15 rain periods at 37°C in a shaking water bath, with repeated pipetting after each incubation. The cells were sedimented by centrifugation (5 min at 800 x g), resuspended in fresh culture medium containing 5.0% (v/v) donor calf serum (Gibco) and counted in a haemocytometer. Cell viability, determined by staining with 0.4% (w/v) trypan blue, was consistently > 90%. Thecal cell culture. Twenty-four-well plastic dishes (Linbro Space Savers from Flow Laboratories, Rickmansworth, Herts, U.K.) were inoculated with replicate 500 ~l portions of thecal cell suspension (2.0-5.0 x 104 viable cells per well) in culture medium containing 5% donor calf serum.
R3
Preincubation in serum-containing medium to allow cell anchorage and recovery from the enzymatic dispersal procedure was for 24 h at 37 o C in a humidified tissue-culture incubator gassed with 95% air/5% CO 2. After removal of serumcontaining medium and washing with 1 ml prewarmed (37°C) Dulbecco's phosphate-buffered saline, the cultures received 500 /~1 serum-free medium containing human L H (LER-1972: donated by Dr. L.E. Reichert, Jr.), recombinant human IGF-I (CGP-35'-126: donated by Dr. H.H. Peter and Dr. K. Scheibli), recombinant human activin-A (Schwall et al., 1989) a n d / 0 r recombinant human inhibin-A (Mason, 1988). Incubation at 3 7 ° C was then for 4 days with a medium change on day 2. Media collected on days 2 and 4 of culture were stored frozen at - 2 0 ° C until assayed for steroid content, as described below. In some experiments the cultured cell monolayers were trypsinized (Hillier and de Zwart, 1982) on day 4 and cell number was estimated with a Coulter counter. Assay of steroids in thecal cell culture medium. Androstenedione and progesterone in unextracted culture medium were measured individually by radioimmunoassay, using standard procedures and antisera as described previously (Hillier et al., 1981; McNatty et al., 1983). The D H A radioimmunoassay was similar, using a sheep antiserum against DHA-3-hemisuccinyl-ovalbumin purchased from Steranti Research, St. Albans, Hefts, U.K. (product code A005). The inter- and intraassay precision for each steroid was less than 15%. Steroid production rates are expressed as pmol steroid/1000 cells/48 h, related to the number of viable cells used to inoculate each culture well. Statistics. Analysis of variance with the Neuman-Keuls test was used to analyze differences between experimental and control observations. Differences assigned a P value of < 0.05 were regarded as statistically significant. Results
Production of androstenedione and D H A was increased approximately 2-fold by inhibin at concentrations between 10 and 100 n g / m l (Fig. 1). Inhibin was compared with other factors previously shown to stimulate androgen synthesis in
0 A4 o.oo.
~= =08 o gg 2~
-o -~
*
• DHA
~
**
0.04
----4
0.02
40 Inhibin concentration (ng/ml) Fig. 1. Stimulation of androgen synthesis in cultured human thecal cells by inhibin. Thecal cells (5.0 × 104 viable cells per culture well) pooled from 17 follicles between 3 and 5 m m in diameter (Experiment I, Table 1) were incubated for 4 days in serum-free culture medium containing inhibin at the concentrations shown. Androstenedione (A4) and dehydroepiandrosterone (DHA) accumulation in medium collected on day 4 are shown. Data are mean + SEM, n = 3. * P < 0.05, * * P < 0.01 vs. treatment without inhibin.
the same thecal culture system. As shown in Fig. 2, the stimulatory effect of inhibin was approximately equivalent to a maximal stimulatory dose of L H (10 n g / m l ) and the effects of inhibin and L H were additive. However, the response to a maximal stimulatory dose of inhibin was only a fraction (approximately one-third) of that to IGF-I
(A)
0.50 G.c: .o ¢o
(B)
*
r l - LH
- r
E~I+LH
~L~
0.15 0.12 0.09
t o
~
0.06
-~o 0.20
