- Email: [email protected]
Activin and inhibin have opposite effects on steroid 5α-reductase activity in genital skin fibroblasts
ELSEVIER
Molecular and Cellular Endocrinology 107 (1995) 99-104
Activin and inhibin have opposite effects on steroid Sa-reductase genital skin fibroblasts* I. Antonipillai*,
M. Wahe, J. Yamamoto,
University ofSouthern California,
Division qf Endocrinology,
activity in
R. Horton
Los Angeles, CA 90033, USA
Received 22 August 1994; accepted 3 1 October 1994
Abstract
The transforming growth factor B (TGF-/?) superfamily includes several closely related peptides including the activins and inhibins. Since we recently reported that TGF-Bl and 82 are potent inducers of steroid Sa-reductase (5aR), we have now studied the effects of these other peptides using primary cultures of human scrotal skin fibroblasts. Recombinant human activin A or inhibin A were added to cultured cells (2 x lo5 cells) for 2 days in a serum free media and 5aR activity was measured by the %-conversion of tracer [‘HItestosterone to dihydrotestosterone (DHT) over a 4-h period. Activin significantly stimulated 5aR activity in a dose related manner (control 3.0 f 0.4%. activin (1.2 x 10m9M) 6 f 0.746, P c 0.01, (2.4 x 10e9 M) 8.5 + 0.6%, P c 0.001). In comparison, androgen (DHT lo-’ M) induction of 5aR was 4.7 f 0.2%, P c 0.05. Combined exposure of fibroblasts to activin (1.2 x 10m9M) and androgen (lo-’ M) did not result in additive or synergistic effect on 5aR activity. In contrast, exposure of cells to an androgen (lo-’ M) and TGF-/I (2 x lo-lo M) led to synergistic effects on 5aR activity (control 1.5 f O.l%, DHT 2.6 f 0.2% TGF-Bl 4.8 * 0.5, TGF$l+DHT 9.2 f 1.2%). Finasteride, a Caza steroid inhibitor of 5aR (10m8M) inhibited both activin and TGF-p-induced 5aR activity suggesting that the type II isoenzyme is induced by these peptides. Activin mediated 5aR activity was abolished by the addition of cycloheximide, consistent with the proposition that enzyme induction requires new protein synthesis. Recombinant human inhibin alone did not alter basal 5aR activity but dose dependently inhibited DHT (lo-’ M)-induced 5aR activity (control 4.1 f 0.4%, DHT 7.5 k 0.78, DHT + inhibin (0.6 x 10s9 M) 5.7 f 0.5%, P < 0.05 DHT + inhibin (1.2 x 10m9M] 4.3 f 0.2%, P < 0.001). The effects of activin or inhibin were not associated with changes in cell number or thymidine uptake. These studies indicate that activin is 100 times more potent on a molar basis than androgen in induction of 5aR activity. Although both activin and TGF-/?I induce 5aR activity, the actions of the two peptides differ in the presence of an androgen. In contrast, inhibin significantly inhibits androgen induction of 5aR. Activin and inhibin, two closely related molecules, potentially play opposite roles in DHT formation in sexual tissue. Keywords:
Activin; Inhibin; Steroid Sa-reductase; Scrotal skin fibroblasts
1. Introduction Activin and inhibin are glycoprotein hormones composed of two related subunits linked by disulfide bridges. Inhibit is composed of an a-subunit and a B-subunit @-A or 8-B) and activin is the dimer of two &subunits. Activin and inhibin can respectively stimulate or inhibit pituitary FSH secretion (Ling et al., 1986; Vale et al., 1986). However, these peptides are also expressed in a variety of tis-
* Presented in part at the 75th Annual Meeting of the Endocrine Society, Vegas, NE, Abstract No. 1936.1993. Corresponding author, USC Medical Center, 1200 N. State Street, Unit I, Room 18-632, Los Angeles, CA 90033, USA. Tel.: +l 213 226 4635; Fax: +I 213 226 2652. e
sues, (Meunier et al., 1988; Roberts et al., 1989) and are members of the TGF-/? superfamily (Massague, 1990). Besides the effects on pituitary FSH release, they have been implicated in the regulation of various biological processes, including the production and release of hormones. Other potential actions include cell proliferation, glucose metabolism, and early embryonic development (DePaolo et al., 1991; Vale et al., 1990). Recently we have reported that in sexual skin, TGF-fil and 82, are potent modulators of steroid Sa-reductase (5aR) activity (Wahe et al., 1994). Regulation of 5aR activity and dihydrotestosterone (DHT) formation is central to prostate and sexual skin (hair) growth. 5aR is involved in differentiation of male external genitalia and prostate (Wilson et al., 1975). Two types of 5aR isoen-
0303-7207/95/$09.50 0 1995 Elsevier Science Ireland Ltd. All rights reserved SSDl 0303-7207(94)03430-2
100
I. Antonipillai et al. I Mnleculur and Cellular Endocrinology 107 (1995) 99-104
zymes have been described. Type I is expressed in much lower amounts in prostate, is relatively insensitive to the 5aR inhibitor finasteride and is not related to classical 5aR deficiency in humans. Type II is the predominant isoenzyme detectable in fetal genital skin, male accessory sex glands and in prostate (Thigpen et al., 1993; Griffin and Wilson, 1989). In tissues such as prostate and liver, the 5aR gene is regulated by androgens (Andersson et al., 1989). Using rat and human sexual skin fibroblasts, we have shown that not only androgens but also certain growth factors such as TGF-/I1 and 82 are potent inducers of 5aR activity (Wahe et al., 1994). Since activins/inhibins and TGF-/I are made up of several closely related proteins, we have evaluated the role of these other members of the TGF-/? superfamily. We find that although activin A stimulates 5aR activity, inhibin A has inhibitory effects on androgen induction of 5aR in sexual tissue.
All tracer steroids were purified after receipt by celite and paper chromatography as previously described, in order to guarantee radiochemical purity (Morimoto et al., 1981). 2.4. Culture techniques Biopsy specimen of human scrotal skin were cleaned from subcutaneous fat and minced to approximately lmm cubes and spread on lOO-mm Falcon dishes. RPMI1640 medium containing 10% fetal bovine serum (FBS) and 100 units/ml penicillin and 100 ,ug/ml streptomycin buffered with NaHC03 and 25 mM HEPES were added to each dish and incubated at 37°C in the presence of 5%CO* in a humidified atmosphere in a Stericult 200 Forma Scientific incubator (Marietta, OH). When cells reached confluence, they were routinely sub-cultured after trypsin dissociation. These cells were plated in 6-well dishes and used between 3 and 7 passages for the assay of 5areductase activity.
2. Materials and methods 2.1. Tissue sources Human scrotal skin was obtained from healthy male individuals undergoing bilateral vasectomy. All patients gave written consent and the protocol was approved by our Internal Review Board. 2.2. Reagents RPM1 1640 medium, streptomycin and penicillin were obtained from Gibco-BRL Corporation (Grand Island, NY). Testosterone, dihydrotestosterone, fetal bovine serum (FBS) and cycloheximide were purchased from Sigma Chemical Company (St. Louis, MO). Bovine serum album (BSA) fraction V was obtained from Boehringer and Mannheim (Indianapolis, Indiana). Human recombinant TGF-/31, was purchased from R & D systems (Minneapolis, MN). Recombinant human (rh) activin A (40pg/ml) was a gift from Genentech, Inc., San Francisco, CA. Recombinant human inhibin A (1OOO~gl ml) was a gift from Dr. Bremner and Dr. de Kretser. The 5aR inhibitor, 4-aza steroid, finasteride (17/3-N,N-diethylcarbamoyl-4-methyl-4-aza-5a-androstane-3one) was generously provided by Dr. Gary H. Rasmusson from Merck, Sharp and Dohme Research, Laboratories (Rahway, NJ). Solvents used for extraction and chromatography were all spectrograde. TGF-Bl was constituted in sterile 4 mM HCl containing 1 mg/ml BSA and stored at -70°C. Dilutions of activin A, inhibin A as well as further dilutions of TGF-/?I were made in RPMI-1640 media with 1% BSA, freshly prepared on ice just before use. 2.3. Labeled steroids [ la,2a-3H(N)]Testosterone (spec. act. 5 1 Ci/mmol), was purchased from New England Nuclear Corp. (Boston, MA). [ la,2a-3H(N)]DHT and [4-14C]DHT were purchased from Amersham Corp. (Arlington Heights, IL).
2.5. Sa-Reductase activity One day before the experiment, cells (200 000 cells/well) were made quiescent by serum starving for 48 h in RPMI-1640 medium containing 0.2% BSA. Cells were then treated with test substances (DHT, TGF-#?) in serum depleted RPM1 media with 0.2% BSA for 2 days. After 48 h, the medium was removed and the cells were again incubated with serum free medium containing [3H]testosterone (200 000 cpm, 4.8 pmol) at 37’C in a 5%CO2 incubator for 4 h. At the end of incubation, the cells were rapidly cooled on ice and the medium was transferred into ice cold tubes containing diethyl ether and i4C standards to monitor recovery. Each well was rinsed with 1 ml phosphate buffered saline (PBS), and the rinse was added to the medium for extraction. The separation of [3H]DHT was achieved by celite and paper chromatography. We have previously validated that radiochemical purity was achieved by sequential chromatography. Results are expressed as % conversion in 4 h/2 x lo5 cells. Cell number in each well was determined by counting an aliquot in a hemocytometer before and after 2 days treatment period with test substances. The fibroblast cell number did not change during this period. 2.6. 3a-Reductase activity 3a-Reductase enzyme activity was measured in the same overall manner as Sa-reductase activity except that [3H]DHT was added (200 000 cpm) with i4C standards. [3H]DHT and [3H]androstane-3,17-diol (3a-diol) were purified by celite and paper chromatography as previously reported and validated by us (Morimoto et al., 1981). 2.7. Thymidine incorporation Cells (105) were incubated in serum-free RPM1 media with 0.2% BSA for 48 h. They were then treated with
101
I. Antonipillai et al. I Molecular and Cellular Endocrinology 107 (1995) 99-104
3. Results **
*
1
T Activin A ngiml (Molar concentrations) Fig. I. The effects of recombinant human activin A on steroid Sareductase (5aR) activity. 2 x IO5 human sexual skin fibroblasts were incubated for 48 h in the absence or presence of the indicated concentrations of activin A in the serum depleted medium. 5aR activity was measured by the percent conversion of [3H]testosterone to dihydrotestosterone (DHT) over a 4-h period as indicated in Section 2. Values are the mean f SEM of 5-6 separate sets of experiments. A dose response effect was noted from 1.2 x 10e9 M to 2.4 x 10e9 M. *P-C 0.02, **P c 0.001 versus control.
3.1. Effects of activin A ana’ DHT on 5aR activity Activin A or DHT was added to fibroblasts and incubated for 48 h. Activin A 1.2 x 10e9 M (30 @ml) increased 5aR activity from 3.0 + 0.2% to 6.0 r?:0.7%, P < 0.01 (0.17 f 0.01 to 0.33 f 0.04 pmoV2 x lo5 cells/ 4 h). A higher concentration of activin A 2.4 x lO-9 M (60 nglml) further increased 5aR to 8.5 r 0.6%. P < 0.001 (Fig. 1). DHT also (10m7 M) increases 5aR activity (4.7 f 0.2%, P < 0.05). Activin A (1.2 x 10e9 M) did not alter the effect of DHT (lO-7 M) on 5aR (control 3.1 + 0.4%, DHT 4.6 f 0.3%, activin A 6.1 + 0.6%, activin A + DHT 6.3 f 0.2%, NS) (Fig. 2). In comparison, TGF$l produced a synergistic increase in 5aR (control 1.5 + O.l%, DHT 2.6 +0.2%, TGF-/Q 4.8 20.5, TGF$l + DHT 9.2* 1.2%). 3.2. Effects ofpnasteride on acdvin A and TGF$Iinduced 5aR Experiments were performed with a 5aR enzyme inhibitor, finasteride using 2-day incubations with activin A or TGF-fi 1. Finasteride (10-s-10” M) progressively blocked both activin A (1.2 x 10T9 M) and TGF$l (2 x lo-lo M) induced 5aR activity (Table 1).
** test substances (activin A or inhibin A, 2.4 x 10m9M) for 48 h as described above, followed by incubation with [3H]thymidine (1 ,Ki/well). Six hours later cells were washed twice with 1 ml phosphate buffered saline (PBS), twice with 10% ice cold trichloroacetic acid solution, followed by a wash with PBS. The cells were then solubilized with 1% sodium dodecyl sulfate in 0.3 N NaOH. An aliquot was then counted in a scintillation counter. 2.8. Data analysis The result are expressed as the mean + SEM percentage Sa-reductase activity. The number of experiments are presented in the figure legends. Statistical significance was determined using an unpaired t-test and analysis of variance (using either Duncan’s or Dunett’s test, wherever appropriate). 2.9. Baseline Sa-reductase activity The conversion of [3H]testosterone to DHT was linear as described previously (Horton et al., 1993). Percent conversions reached levels up to 4.4 + 0.5% (0.24 f 0.03 pmoll2 x IO5 cells/4 h) in human cells after 4 h incubation of the tracer.
Fig. 2. Separate and combined effects of DHT (10e7 M) and recombinant human activin A (1.2 x 10m9M) on 5aR activity of human skin fibroblasts under the same conditions as above. Values represent the mean f SEM of 5-7 different experiments. Activin A was more potent than DHT in inducing 5aR activity. Activin A added with DHT did not have synergistic or additive effect on 5aR. *I’< 0.01, **P
102
i. Anronipilfai
et ul. I Molecular
and Cellular
Table I Dose related effects of finasteride on activin A and TGF,BI induced 5aR activity 8 .5a-reductase activity (T-+DHT) Control Activin Activin Activin Activin
A (1.2 x 10e9 M) A + tinasteride (10e8 M) A + fmasteride (lo-’ M) A + finasteride ( lOA M)
TG#l (2 x lo-“M) TGF-/31 + finasteride ( 10v8 M) TGF_Bl + finasteride (lo-’ M)
3.3. Effects of cycloheximide on activin A Cells were incubated in the presence of activin and cycloheximide (10 pg/ml). Cycloheximide abolished activin-induced 5aR activity (control 2.8 + 0.2, activin A 4.3 f 0.2, activin A + cycloheximide 2.8 + 0.4, P < 0.05). 3.4. Effects of inhibin A and DHT Inhibin A at 1.2 x 10e9 M (30 ng/ml) and 2.4 x 10m9M (60 ng/ml) concentrations produced no significant effects on basal 5aR activity (control 4.4 + OS%, inhibin A (1.2x 10-9M) 3.920.5, (2.4x 10-9M) 4.8+0.5%, NS). However, inhibin A when added with DHT (lo-’ M) prevented the induction of 5aR activity by androgen in a dose related manner (control 4.1 + 0.4, DHT 7.5 f 0.7%, DHT + inhibin (0.6 x 1O-9 M) 5.7 k 0.5%) P < 0.05 DHT + inhibin (1.2 x 10m9M) 4.2 f 0.4%, P < 0.001 (Fig. 3). A higher concentration of inhibin did not further reduce 5aR, (DHT + inhibin (2.4 x 10e9M) 4.3 + 0.2%).
Endocrinology
107 (1995) 99-104
tivity after addition of 2.4 x 10e9 M concentration of activin Afinhibin A. This enzyme catalyses the reversible conversion of DHT to 3a-androstanediol. Neither activin A or inhibin A alone or inhibin A + DHT had any effect on the conversion of [3H]DHT to 3a-androstanediol (control 2.1 f 0.3%, activin A 1.5 k 0.2%, inhibin A 2.4 + 0.3%, inhibin A + DHT 2.6 + O.l%, NS). 4. Discussion Recent studies suggest that steroid synthesis may be subjected to further modulation by growth factors. For example, EGF increases cortisol secretion by sheep adrenocortical cells in vitro (Singh-Asa et al., 1983). IGFI can augment the steroidogenic effect of LH on testosterone production (Lin et al., 1986). TGF-/31 in adrenal glomerulosa cells can reduce the utilization of cholesterol by adrenal cortical cells and inhibit steroidogenesis at an early step (Hotta and Baird, 1986; Gupta et al., 1993). In contrast, steroids such as androgens and estrogens act at least in part on target tissues by generating growth factors. Examples are the action of androgens on nerve growth factor and EGF formation (Bellve and Zheng, 1989; Jacobs and Story, 1988). In sexual tissue, androgen also increase steroid 5aR activity and DHT formation via production of IGF-I (Horton et al., 1993). EGF has been shown to replace estrogens in the stimulation of female genital tract growth and differentiation (Nelson et al., 1991).
g1
3.5. Cell number and [3H]thymidine incorporation There was no significant change in cell number after addition of activin A or inhibin A (2.4 x 10e9 M) compared with controls (1.12 f 0.2 and 0.95 + 0.07 times control, both P = NS). Similarly [3H]thymidine incorporation was also not altered by these peptides in quiescent cells, deprived of FBS. [3H]Thymidine incorporation per lo5 cells with activin A was 0.96 + 0.06 times control and with inhibin A 0.88 + 0.08 times control, both P = NS. In comparison, cells grown in 10% FBS prior to and during the 48-h treatment period demonstrated significant increase in r3H]thymidine incorporation (1.64 + 0.1 times control, P c 0.05). 3.6. Actions of activin A and inhibin A on 3a-reduction of DHT (DHT + 3a-dial) To determine whether the actions of activin Alinhibin A are specific for 5aR, we also studied 3a-reductase ac-
Fig. 3. Dose related effects of recombinant human inhibin A on DHTinduction of 5aR in human skin fibroblast. Values am the mean f SEM of 5-8 separate sets of experiments. Inhibin A significantly inhibits androgen-induced 5aR. *PC 0.05 versus DHT, **P
1. Antnnipillai et al. I Molecular and Cellular Endocrinology 107 (1995) 99-104
We have recently reported that TGF-#I may play a role in androgen formation in sexual tissue since it stimulates 5aR activity and formation of DHT (Wahe et al., 1994). The TGF-/3 superfamily is made up of several closely related peptides. Among these are inhibins and activins. Inhibin and activin were originally considered as gonadal hormones which preferentially suppress or stimulate FSH secretion in the pituitary, and regulate Leydig cell steroidogenesis and spermatogenesis (Hsueh et al., 1987; Vale et al., 1988; Lin et al., 1989; Mauduit et al., 1991; Mathur et al., 1990). Newer studies suggest that these peptides can regulate diverse physiological functions including the secretion of GH and ACTH, oxytocin secretion and erythropoiesis as well as potentially regulating embryonic development (Massague, 1990; Vale et al., 1990; DePaolo et al., 1991). The genes encoding inhibin and activin subunits are widely expressed in many cells especially from sexual tissues (Vale et al., 1988). Inhibin and activin may act as paracrine/autocrine regulators of cellular function in tissues other than testis and ovary. In the present study we have made the novel observation that activin A is a potent and selective stimulator of 5aR activity in genital skin fibroblasts. Activin A was 100 times more potent than DHT in its action. This dose of activin is similar to the amount that was effective in activation of FSH receptors in rat granulosa cells (Hasegawa et al., 1988; Xiao et al., 1992; Nakamura et al., 1993) and osteoclast-like cell formation in bone marrow cell cultures (Sakai et al., 1993). Our new findings demonstrate that both TGF-/?l and activin A increase 5aR activity. However, the effects of these two peptides differ in the presence of an androgen. In contrast to the synergistic action of TGF-/31 and 82 with DHT on 5aR (Wahe et al., 1994) no additional effects of activin A in the presence of DHT on androgeninduction of 5aR were observed. It is of interest that TGF-/I and activin A similarly stimulate FSH-induced aromatase activity in rat granulosa cells. (Hutchinson et al., 1987; Ying et al., 1988). We also evaluated the effects of inhibin on 5aR. Exposure of cells to inhibin alone did not result in changes in basal 5aR, but inhibin significantly inhibited androgeninduced 5aR in a dose related manner. This is analogous to reports where gonadotropin-releasing hormone stimulated FSH and LH release was reduced by inhibin (Jakubowiak et al., 1991). Similarly FSH-mediated estrogen secretion is decreased by inhibin (Ying et al., 1986). The stimulatory/inhibitory effects of activin and inhibin on 5aR are not due to non-specific actions of these peptides as both altered 5aR activity but had no effect on 3aR-reduction of DHT. Cell number or [3H]thymidine uptake did not change after addition of these peptides indicating that these effects are not due to general growth or mitogenic effects. Since enzyme induction requires new protein synthesis and cycloheximide is known to reduce protein synthesis,
103
we examined the effects of cycloheximide on activin Ainduced 5aR activity. Induction of 5aR activity by activin A was completely abolished by treatment with cycloheximide. As indicated earlier, two isoenzymes of 5aR exist (types I and II) with distinct functions. A synthetic inhibitor such as 4-aza steroid, finasteride has been used to determine the biochemical behavior of these two isoenzymes. While the type I enzyme is relatively insensitive to this inhibitor (Ki - 300-600 nM), the type II enzyme is markedly inhibited by finasteride (Ki - 10 nM) (Jenkins et al., 1992). In our studies, finasteride at 10 nM significantly inhibited both activin A and TGF-bl-induced 5aR activity. These findings suggest that both activin A and TGF-/31 induce type II 5aR isoenzyme in this sexual tissue. The in vivo physiological significance of the interaction between TGF-/I family members and androgens is only speculative at this time. There is now ample evidence that TGF-#?l and 82 are not only formed in human and rat prostates but the expression of TGF-B;! mRNA is increased in benign prostatic hyperplasia (BPH) (Kyprianou and Isaacs, 1989; Mori et al., 1990). Similarly TGF-/3 has been detected in both human and rat prostate cancer and TGF-/ll mRNA levels are markedly elevated in prostate carcinoma sublimes (Steiner and Barrack, 1992; Thompson, 1990; Matuo et al., 1990) Therefore, not only androgens but members of the TGF-/3 family may play a role in normal as well as abnormal growth of sexual tissue and our work indicates that these peptides have the potential to modulate 5aR activity. It is now well established that androgen regulation of 5aR is mediated at the level of transcription of 5aR mRNA (Anderson et al., 1989), however it needs to be established if the mechanism(s) of action of TGF-b and activiniinhibin on 5aR is at the level of transcription, translation or enzyme turnover. The present results also emphasize the potential role of inhibin in the negative regulation of 5aR. Although androgens are known inducers of 5aR, these effects are completely abolished by the presence of low concentrations of inhibin. In view of the opposite actions of activin (TGF-b) and inhibin, the regulation of 5aR could depend on the balance between androgen and activin/inhibin interactions. Acknowledgements We wish to thank Thomas Jenkins and Lydia Baas for typing this manuscript. References Andersson, S., Bishop, R.W. and Russell, D.W. (1989) J. Biol. Chem. 264,16249-16255. Bellve, A.R. and Zheng, W. (1989) Reprod. Fertil. 85.771-793.
104
I. Anfonipillai et al. I Molecular and Cellular Endocrinology IO7 (1995) 99-104
DePaolo, L.V., Bicsak, T.A., Erickson, G.F., Shimasaki. S. and Ling, N. (1991) Proc. Sot. Exp. Biol. Med. 198,500-512. Griffin, J.E. and Wilson, J.D. (1989) pp. 1919-1944, McGraw-Hill, New York). Gupta, P., France-Saenz. R. and Mulrow, P.J. (1993) Endocrinology 32. 1184-I 188. Hasegawa, Y., Miyamoto, K., Abe, Y., Nakamura, T., Sugino, H., Elo, Y., Shibai, H. and Igarashi, M. (1988) Biochem. Biophys. Res. Commun. 156.668-674. Horton, R., Pasupuletti, V. and Antonipillai, 1. (1993) Endocrinology 133.447-45 1. Hotta, M. and Baird, A. (1986) Proc. Natl. Acad. Sci. USA 83, 77957799. Hsueh, A.J.W., Dahl, K.D., Vaughan, J., Tucker, E., Rivier, J., Bardin, C.W. and Vale, W. (1987) Proc. Natl. Acad. Sci. USA 84, 50825086. Hutchinson, L.A., Findly, J.K., de Vos, F.L. and Robertson, D.M. (1987) Biochem. Biophys. Res. Commun. 146, 1405-1412. Jacobs, S. and Story, M. (1988) Prostate 13,79-87. Jakubowiak, J., Janecki, A., Tong, D., Sanborm, B.M. and Steinberger. A. (1991) Endocrinology 82.265-273. Jenkins, E.P., Andersson, S., Imperato-McGinley, J., Wilson, J.D. and Russell, D.W. (1992) J. Clin. Invest. 89, 293-300. Kyprianou, N. and Isaacs, J.T. (1989) Mol. Endocrinol. 3, 1515-1522. Lin, T., Haskell, J., Vinson, N. and Terracio, L. (1986) Biochem. Biophys. Res. Commun. 137.950-956. Lin, T., Calkins, J.H., Morris, P.L., Vale, W., Bardin, C.W. (1989) Endocrinology 125,2134-2140. Ling, N., Ying, S.-Y., Ueno, N., Shimasaki, S., Esch, F., Hotta, M. and Guillemin, R. (1986) Nature 321.779-782. Massague, J., (1990) Annu. Rev. Cell. Biol. 6,597-641. Mather, J.P., Attie, K.M., Woodruff, T.K., Rice, G.C. and Phillips, D.M. (1990) Endocrinology 127.3206-3214. Matuo, Y., Nishi, N., Takasuka, H., Masuda, Y., Nishikawa, K., Isaacs. J.T., Adams, P.S., McKeehan, W.L. and Sato. G.H. (1990) Biothem. Biophys. Res. Commun. 166,840-847. Mauduit, C.. Chaurin, M.A., de Pretti, E., Morera, A.M. and Benahmed. M. (1991) Biol. Reprod. 45, 101-109.
Meunier, H., Rivier, C., Evans, R.M. and Vale, W. (1988) Proc. Nat]. Acad. Sci. USA 85,247-25 1. Mori, H., Maki, M.. Oishi, K., Jaye, M., Igarashi, K., Yoshida, 0. and Hatanaka, M. (1990) Prostate 16,71-80. Morimoto, I., Edminston, A., Hawks, D. and Horton, R. (1981) J. Clin. Endocrinol. Metab. 52,772-778. Nakamura, M., Minegishi, T., Hasegawa, Y., Nakamura. K., Igarashi, S., Ito, I., Shinozaki, H., Miyamoto, K., Eto, Y. and Ibuki, Y. (1993) Endocrinology 133.538-544. Nelson, K.G., Takahashi, T., Bosser, N.L., Walmer, D.K. and McLachian, J.A. (1991) Ptoc. Natl. Acad. Sci. USA 88.21-25. Roberts, V., Meunier, H., Vaughan, J., Rivier, J., Rivier, C., Vale, W. and Sawchenko, P. (1989) Endocrinology 124,552-554. Sakai, R., Eto, Y., Ohtsuka, M., Hirafaji, M. and Shinoda, H. (1993) Biochem. Biophys. Res. Commun. 195.39-46. Singh-Asa, P. and Waters, M.J. (1983) Mol. Cell. Endocrinol. 30, 189199. Steiner, M.S. and Barrack, E.R. (1992) Mol. Endocrinol. 6, 15-25. Thigpen, A.E., Silver, R.I., Guileyardo, J.M., Casey, M.L., McConnell. J.D. and Russell, D.M. (1993) J Clin. Invest. 92.903-910. Thompson, T.C. (1990) Cancer Cells 2,345-354. Vale, W., River, J., Vaughan, J., McClintock, R., Corrigan, A., Woo, W., Katr, D. and Spiess, J. (1986) Nature 321, 776-779. Vale, W., Hsueh, A., Rivier, C. and Yu, J. (1990) Handb. Exp. Pharmacol. (Berlin) 95,211-248. Vale, W., Rivier, C., Hsueh, A., Campen, C., Meunier, H.. Bicsak, T.. Vaughan, J., Corrigan, S., Bardin, W., Sawchenko, P., Petraglia, F., Yu, J., Plotsky, P., Spiess, J. and Rivier, J. (1988) Recent Prog. Horm. Pes. 44. I-34. Wahe, M., Antonipillai, I. and Horton, R. (1993) Mol. Cell. Endocrinol. 98.55-59. Wilson, J.D., (1975) Am. Physiol. Sot. 5.491-508. Xiao, S., Robertson, D.M. and Findlay, J.K. (1992) Endocrinology 131, 1009-1016. Ying, S and Shao-Yao (1988) Endocr. Rev. 9,267-292. Ying, S., Becker, A., Ling, N., Ueno, N. and Guillemin, R. (1986) Biochem. Biophys. Res Commun. 136.969-975.
Molecular and Cellular Endocrinology 107 (1995) 99-104
Activin and inhibin have opposite effects on steroid Sa-reductase genital skin fibroblasts* I. Antonipillai*,
M. Wahe, J. Yamamoto,
University ofSouthern California,
Division qf Endocrinology,
activity in
R. Horton
Los Angeles, CA 90033, USA
Received 22 August 1994; accepted 3 1 October 1994
Abstract
The transforming growth factor B (TGF-/?) superfamily includes several closely related peptides including the activins and inhibins. Since we recently reported that TGF-Bl and 82 are potent inducers of steroid Sa-reductase (5aR), we have now studied the effects of these other peptides using primary cultures of human scrotal skin fibroblasts. Recombinant human activin A or inhibin A were added to cultured cells (2 x lo5 cells) for 2 days in a serum free media and 5aR activity was measured by the %-conversion of tracer [‘HItestosterone to dihydrotestosterone (DHT) over a 4-h period. Activin significantly stimulated 5aR activity in a dose related manner (control 3.0 f 0.4%. activin (1.2 x 10m9M) 6 f 0.746, P c 0.01, (2.4 x 10e9 M) 8.5 + 0.6%, P c 0.001). In comparison, androgen (DHT lo-’ M) induction of 5aR was 4.7 f 0.2%, P c 0.05. Combined exposure of fibroblasts to activin (1.2 x 10m9M) and androgen (lo-’ M) did not result in additive or synergistic effect on 5aR activity. In contrast, exposure of cells to an androgen (lo-’ M) and TGF-/I (2 x lo-lo M) led to synergistic effects on 5aR activity (control 1.5 f O.l%, DHT 2.6 f 0.2% TGF-Bl 4.8 * 0.5, TGF$l+DHT 9.2 f 1.2%). Finasteride, a Caza steroid inhibitor of 5aR (10m8M) inhibited both activin and TGF-p-induced 5aR activity suggesting that the type II isoenzyme is induced by these peptides. Activin mediated 5aR activity was abolished by the addition of cycloheximide, consistent with the proposition that enzyme induction requires new protein synthesis. Recombinant human inhibin alone did not alter basal 5aR activity but dose dependently inhibited DHT (lo-’ M)-induced 5aR activity (control 4.1 f 0.4%, DHT 7.5 k 0.78, DHT + inhibin (0.6 x 10s9 M) 5.7 f 0.5%, P < 0.05 DHT + inhibin (1.2 x 10m9M] 4.3 f 0.2%, P < 0.001). The effects of activin or inhibin were not associated with changes in cell number or thymidine uptake. These studies indicate that activin is 100 times more potent on a molar basis than androgen in induction of 5aR activity. Although both activin and TGF-/?I induce 5aR activity, the actions of the two peptides differ in the presence of an androgen. In contrast, inhibin significantly inhibits androgen induction of 5aR. Activin and inhibin, two closely related molecules, potentially play opposite roles in DHT formation in sexual tissue. Keywords:
Activin; Inhibin; Steroid Sa-reductase; Scrotal skin fibroblasts
1. Introduction Activin and inhibin are glycoprotein hormones composed of two related subunits linked by disulfide bridges. Inhibit is composed of an a-subunit and a B-subunit @-A or 8-B) and activin is the dimer of two &subunits. Activin and inhibin can respectively stimulate or inhibit pituitary FSH secretion (Ling et al., 1986; Vale et al., 1986). However, these peptides are also expressed in a variety of tis-
* Presented in part at the 75th Annual Meeting of the Endocrine Society, Vegas, NE, Abstract No. 1936.1993. Corresponding author, USC Medical Center, 1200 N. State Street, Unit I, Room 18-632, Los Angeles, CA 90033, USA. Tel.: +l 213 226 4635; Fax: +I 213 226 2652. e
sues, (Meunier et al., 1988; Roberts et al., 1989) and are members of the TGF-/? superfamily (Massague, 1990). Besides the effects on pituitary FSH release, they have been implicated in the regulation of various biological processes, including the production and release of hormones. Other potential actions include cell proliferation, glucose metabolism, and early embryonic development (DePaolo et al., 1991; Vale et al., 1990). Recently we have reported that in sexual skin, TGF-fil and 82, are potent modulators of steroid Sa-reductase (5aR) activity (Wahe et al., 1994). Regulation of 5aR activity and dihydrotestosterone (DHT) formation is central to prostate and sexual skin (hair) growth. 5aR is involved in differentiation of male external genitalia and prostate (Wilson et al., 1975). Two types of 5aR isoen-
0303-7207/95/$09.50 0 1995 Elsevier Science Ireland Ltd. All rights reserved SSDl 0303-7207(94)03430-2
100
I. Antonipillai et al. I Mnleculur and Cellular Endocrinology 107 (1995) 99-104
zymes have been described. Type I is expressed in much lower amounts in prostate, is relatively insensitive to the 5aR inhibitor finasteride and is not related to classical 5aR deficiency in humans. Type II is the predominant isoenzyme detectable in fetal genital skin, male accessory sex glands and in prostate (Thigpen et al., 1993; Griffin and Wilson, 1989). In tissues such as prostate and liver, the 5aR gene is regulated by androgens (Andersson et al., 1989). Using rat and human sexual skin fibroblasts, we have shown that not only androgens but also certain growth factors such as TGF-/I1 and 82 are potent inducers of 5aR activity (Wahe et al., 1994). Since activins/inhibins and TGF-/I are made up of several closely related proteins, we have evaluated the role of these other members of the TGF-/? superfamily. We find that although activin A stimulates 5aR activity, inhibin A has inhibitory effects on androgen induction of 5aR in sexual tissue.
All tracer steroids were purified after receipt by celite and paper chromatography as previously described, in order to guarantee radiochemical purity (Morimoto et al., 1981). 2.4. Culture techniques Biopsy specimen of human scrotal skin were cleaned from subcutaneous fat and minced to approximately lmm cubes and spread on lOO-mm Falcon dishes. RPMI1640 medium containing 10% fetal bovine serum (FBS) and 100 units/ml penicillin and 100 ,ug/ml streptomycin buffered with NaHC03 and 25 mM HEPES were added to each dish and incubated at 37°C in the presence of 5%CO* in a humidified atmosphere in a Stericult 200 Forma Scientific incubator (Marietta, OH). When cells reached confluence, they were routinely sub-cultured after trypsin dissociation. These cells were plated in 6-well dishes and used between 3 and 7 passages for the assay of 5areductase activity.
2. Materials and methods 2.1. Tissue sources Human scrotal skin was obtained from healthy male individuals undergoing bilateral vasectomy. All patients gave written consent and the protocol was approved by our Internal Review Board. 2.2. Reagents RPM1 1640 medium, streptomycin and penicillin were obtained from Gibco-BRL Corporation (Grand Island, NY). Testosterone, dihydrotestosterone, fetal bovine serum (FBS) and cycloheximide were purchased from Sigma Chemical Company (St. Louis, MO). Bovine serum album (BSA) fraction V was obtained from Boehringer and Mannheim (Indianapolis, Indiana). Human recombinant TGF-/31, was purchased from R & D systems (Minneapolis, MN). Recombinant human (rh) activin A (40pg/ml) was a gift from Genentech, Inc., San Francisco, CA. Recombinant human inhibin A (1OOO~gl ml) was a gift from Dr. Bremner and Dr. de Kretser. The 5aR inhibitor, 4-aza steroid, finasteride (17/3-N,N-diethylcarbamoyl-4-methyl-4-aza-5a-androstane-3one) was generously provided by Dr. Gary H. Rasmusson from Merck, Sharp and Dohme Research, Laboratories (Rahway, NJ). Solvents used for extraction and chromatography were all spectrograde. TGF-Bl was constituted in sterile 4 mM HCl containing 1 mg/ml BSA and stored at -70°C. Dilutions of activin A, inhibin A as well as further dilutions of TGF-/?I were made in RPMI-1640 media with 1% BSA, freshly prepared on ice just before use. 2.3. Labeled steroids [ la,2a-3H(N)]Testosterone (spec. act. 5 1 Ci/mmol), was purchased from New England Nuclear Corp. (Boston, MA). [ la,2a-3H(N)]DHT and [4-14C]DHT were purchased from Amersham Corp. (Arlington Heights, IL).
2.5. Sa-Reductase activity One day before the experiment, cells (200 000 cells/well) were made quiescent by serum starving for 48 h in RPMI-1640 medium containing 0.2% BSA. Cells were then treated with test substances (DHT, TGF-#?) in serum depleted RPM1 media with 0.2% BSA for 2 days. After 48 h, the medium was removed and the cells were again incubated with serum free medium containing [3H]testosterone (200 000 cpm, 4.8 pmol) at 37’C in a 5%CO2 incubator for 4 h. At the end of incubation, the cells were rapidly cooled on ice and the medium was transferred into ice cold tubes containing diethyl ether and i4C standards to monitor recovery. Each well was rinsed with 1 ml phosphate buffered saline (PBS), and the rinse was added to the medium for extraction. The separation of [3H]DHT was achieved by celite and paper chromatography. We have previously validated that radiochemical purity was achieved by sequential chromatography. Results are expressed as % conversion in 4 h/2 x lo5 cells. Cell number in each well was determined by counting an aliquot in a hemocytometer before and after 2 days treatment period with test substances. The fibroblast cell number did not change during this period. 2.6. 3a-Reductase activity 3a-Reductase enzyme activity was measured in the same overall manner as Sa-reductase activity except that [3H]DHT was added (200 000 cpm) with i4C standards. [3H]DHT and [3H]androstane-3,17-diol (3a-diol) were purified by celite and paper chromatography as previously reported and validated by us (Morimoto et al., 1981). 2.7. Thymidine incorporation Cells (105) were incubated in serum-free RPM1 media with 0.2% BSA for 48 h. They were then treated with
101
I. Antonipillai et al. I Molecular and Cellular Endocrinology 107 (1995) 99-104
3. Results **
*
1
T Activin A ngiml (Molar concentrations) Fig. I. The effects of recombinant human activin A on steroid Sareductase (5aR) activity. 2 x IO5 human sexual skin fibroblasts were incubated for 48 h in the absence or presence of the indicated concentrations of activin A in the serum depleted medium. 5aR activity was measured by the percent conversion of [3H]testosterone to dihydrotestosterone (DHT) over a 4-h period as indicated in Section 2. Values are the mean f SEM of 5-6 separate sets of experiments. A dose response effect was noted from 1.2 x 10e9 M to 2.4 x 10e9 M. *P-C 0.02, **P c 0.001 versus control.
3.1. Effects of activin A ana’ DHT on 5aR activity Activin A or DHT was added to fibroblasts and incubated for 48 h. Activin A 1.2 x 10e9 M (30 @ml) increased 5aR activity from 3.0 + 0.2% to 6.0 r?:0.7%, P < 0.01 (0.17 f 0.01 to 0.33 f 0.04 pmoV2 x lo5 cells/ 4 h). A higher concentration of activin A 2.4 x lO-9 M (60 nglml) further increased 5aR to 8.5 r 0.6%. P < 0.001 (Fig. 1). DHT also (10m7 M) increases 5aR activity (4.7 f 0.2%, P < 0.05). Activin A (1.2 x 10e9 M) did not alter the effect of DHT (lO-7 M) on 5aR (control 3.1 + 0.4%, DHT 4.6 f 0.3%, activin A 6.1 + 0.6%, activin A + DHT 6.3 f 0.2%, NS) (Fig. 2). In comparison, TGF$l produced a synergistic increase in 5aR (control 1.5 + O.l%, DHT 2.6 +0.2%, TGF-/Q 4.8 20.5, TGF$l + DHT 9.2* 1.2%). 3.2. Effects ofpnasteride on acdvin A and TGF$Iinduced 5aR Experiments were performed with a 5aR enzyme inhibitor, finasteride using 2-day incubations with activin A or TGF-fi 1. Finasteride (10-s-10” M) progressively blocked both activin A (1.2 x 10T9 M) and TGF$l (2 x lo-lo M) induced 5aR activity (Table 1).
** test substances (activin A or inhibin A, 2.4 x 10m9M) for 48 h as described above, followed by incubation with [3H]thymidine (1 ,Ki/well). Six hours later cells were washed twice with 1 ml phosphate buffered saline (PBS), twice with 10% ice cold trichloroacetic acid solution, followed by a wash with PBS. The cells were then solubilized with 1% sodium dodecyl sulfate in 0.3 N NaOH. An aliquot was then counted in a scintillation counter. 2.8. Data analysis The result are expressed as the mean + SEM percentage Sa-reductase activity. The number of experiments are presented in the figure legends. Statistical significance was determined using an unpaired t-test and analysis of variance (using either Duncan’s or Dunett’s test, wherever appropriate). 2.9. Baseline Sa-reductase activity The conversion of [3H]testosterone to DHT was linear as described previously (Horton et al., 1993). Percent conversions reached levels up to 4.4 + 0.5% (0.24 f 0.03 pmoll2 x IO5 cells/4 h) in human cells after 4 h incubation of the tracer.
Fig. 2. Separate and combined effects of DHT (10e7 M) and recombinant human activin A (1.2 x 10m9M) on 5aR activity of human skin fibroblasts under the same conditions as above. Values represent the mean f SEM of 5-7 different experiments. Activin A was more potent than DHT in inducing 5aR activity. Activin A added with DHT did not have synergistic or additive effect on 5aR. *I’< 0.01, **P
102
i. Anronipilfai
et ul. I Molecular
and Cellular
Table I Dose related effects of finasteride on activin A and TGF,BI induced 5aR activity 8 .5a-reductase activity (T-+DHT) Control Activin Activin Activin Activin
A (1.2 x 10e9 M) A + tinasteride (10e8 M) A + fmasteride (lo-’ M) A + finasteride ( lOA M)
TG#l (2 x lo-“M) TGF-/31 + finasteride ( 10v8 M) TGF_Bl + finasteride (lo-’ M)
3.3. Effects of cycloheximide on activin A Cells were incubated in the presence of activin and cycloheximide (10 pg/ml). Cycloheximide abolished activin-induced 5aR activity (control 2.8 + 0.2, activin A 4.3 f 0.2, activin A + cycloheximide 2.8 + 0.4, P < 0.05). 3.4. Effects of inhibin A and DHT Inhibin A at 1.2 x 10e9 M (30 ng/ml) and 2.4 x 10m9M (60 ng/ml) concentrations produced no significant effects on basal 5aR activity (control 4.4 + OS%, inhibin A (1.2x 10-9M) 3.920.5, (2.4x 10-9M) 4.8+0.5%, NS). However, inhibin A when added with DHT (lo-’ M) prevented the induction of 5aR activity by androgen in a dose related manner (control 4.1 + 0.4, DHT 7.5 f 0.7%, DHT + inhibin (0.6 x 1O-9 M) 5.7 k 0.5%) P < 0.05 DHT + inhibin (1.2 x 10m9M) 4.2 f 0.4%, P < 0.001 (Fig. 3). A higher concentration of inhibin did not further reduce 5aR, (DHT + inhibin (2.4 x 10e9M) 4.3 + 0.2%).
Endocrinology
107 (1995) 99-104
tivity after addition of 2.4 x 10e9 M concentration of activin Afinhibin A. This enzyme catalyses the reversible conversion of DHT to 3a-androstanediol. Neither activin A or inhibin A alone or inhibin A + DHT had any effect on the conversion of [3H]DHT to 3a-androstanediol (control 2.1 f 0.3%, activin A 1.5 k 0.2%, inhibin A 2.4 + 0.3%, inhibin A + DHT 2.6 + O.l%, NS). 4. Discussion Recent studies suggest that steroid synthesis may be subjected to further modulation by growth factors. For example, EGF increases cortisol secretion by sheep adrenocortical cells in vitro (Singh-Asa et al., 1983). IGFI can augment the steroidogenic effect of LH on testosterone production (Lin et al., 1986). TGF-/31 in adrenal glomerulosa cells can reduce the utilization of cholesterol by adrenal cortical cells and inhibit steroidogenesis at an early step (Hotta and Baird, 1986; Gupta et al., 1993). In contrast, steroids such as androgens and estrogens act at least in part on target tissues by generating growth factors. Examples are the action of androgens on nerve growth factor and EGF formation (Bellve and Zheng, 1989; Jacobs and Story, 1988). In sexual tissue, androgen also increase steroid 5aR activity and DHT formation via production of IGF-I (Horton et al., 1993). EGF has been shown to replace estrogens in the stimulation of female genital tract growth and differentiation (Nelson et al., 1991).
g1
3.5. Cell number and [3H]thymidine incorporation There was no significant change in cell number after addition of activin A or inhibin A (2.4 x 10e9 M) compared with controls (1.12 f 0.2 and 0.95 + 0.07 times control, both P = NS). Similarly [3H]thymidine incorporation was also not altered by these peptides in quiescent cells, deprived of FBS. [3H]Thymidine incorporation per lo5 cells with activin A was 0.96 + 0.06 times control and with inhibin A 0.88 + 0.08 times control, both P = NS. In comparison, cells grown in 10% FBS prior to and during the 48-h treatment period demonstrated significant increase in r3H]thymidine incorporation (1.64 + 0.1 times control, P c 0.05). 3.6. Actions of activin A and inhibin A on 3a-reduction of DHT (DHT + 3a-dial) To determine whether the actions of activin Alinhibin A are specific for 5aR, we also studied 3a-reductase ac-
Fig. 3. Dose related effects of recombinant human inhibin A on DHTinduction of 5aR in human skin fibroblast. Values am the mean f SEM of 5-8 separate sets of experiments. Inhibin A significantly inhibits androgen-induced 5aR. *PC 0.05 versus DHT, **P
1. Antnnipillai et al. I Molecular and Cellular Endocrinology 107 (1995) 99-104
We have recently reported that TGF-#I may play a role in androgen formation in sexual tissue since it stimulates 5aR activity and formation of DHT (Wahe et al., 1994). The TGF-/3 superfamily is made up of several closely related peptides. Among these are inhibins and activins. Inhibin and activin were originally considered as gonadal hormones which preferentially suppress or stimulate FSH secretion in the pituitary, and regulate Leydig cell steroidogenesis and spermatogenesis (Hsueh et al., 1987; Vale et al., 1988; Lin et al., 1989; Mauduit et al., 1991; Mathur et al., 1990). Newer studies suggest that these peptides can regulate diverse physiological functions including the secretion of GH and ACTH, oxytocin secretion and erythropoiesis as well as potentially regulating embryonic development (Massague, 1990; Vale et al., 1990; DePaolo et al., 1991). The genes encoding inhibin and activin subunits are widely expressed in many cells especially from sexual tissues (Vale et al., 1988). Inhibin and activin may act as paracrine/autocrine regulators of cellular function in tissues other than testis and ovary. In the present study we have made the novel observation that activin A is a potent and selective stimulator of 5aR activity in genital skin fibroblasts. Activin A was 100 times more potent than DHT in its action. This dose of activin is similar to the amount that was effective in activation of FSH receptors in rat granulosa cells (Hasegawa et al., 1988; Xiao et al., 1992; Nakamura et al., 1993) and osteoclast-like cell formation in bone marrow cell cultures (Sakai et al., 1993). Our new findings demonstrate that both TGF-/?l and activin A increase 5aR activity. However, the effects of these two peptides differ in the presence of an androgen. In contrast to the synergistic action of TGF-/31 and 82 with DHT on 5aR (Wahe et al., 1994) no additional effects of activin A in the presence of DHT on androgeninduction of 5aR were observed. It is of interest that TGF-/I and activin A similarly stimulate FSH-induced aromatase activity in rat granulosa cells. (Hutchinson et al., 1987; Ying et al., 1988). We also evaluated the effects of inhibin on 5aR. Exposure of cells to inhibin alone did not result in changes in basal 5aR, but inhibin significantly inhibited androgeninduced 5aR in a dose related manner. This is analogous to reports where gonadotropin-releasing hormone stimulated FSH and LH release was reduced by inhibin (Jakubowiak et al., 1991). Similarly FSH-mediated estrogen secretion is decreased by inhibin (Ying et al., 1986). The stimulatory/inhibitory effects of activin and inhibin on 5aR are not due to non-specific actions of these peptides as both altered 5aR activity but had no effect on 3aR-reduction of DHT. Cell number or [3H]thymidine uptake did not change after addition of these peptides indicating that these effects are not due to general growth or mitogenic effects. Since enzyme induction requires new protein synthesis and cycloheximide is known to reduce protein synthesis,
103
we examined the effects of cycloheximide on activin Ainduced 5aR activity. Induction of 5aR activity by activin A was completely abolished by treatment with cycloheximide. As indicated earlier, two isoenzymes of 5aR exist (types I and II) with distinct functions. A synthetic inhibitor such as 4-aza steroid, finasteride has been used to determine the biochemical behavior of these two isoenzymes. While the type I enzyme is relatively insensitive to this inhibitor (Ki - 300-600 nM), the type II enzyme is markedly inhibited by finasteride (Ki - 10 nM) (Jenkins et al., 1992). In our studies, finasteride at 10 nM significantly inhibited both activin A and TGF-bl-induced 5aR activity. These findings suggest that both activin A and TGF-/31 induce type II 5aR isoenzyme in this sexual tissue. The in vivo physiological significance of the interaction between TGF-/I family members and androgens is only speculative at this time. There is now ample evidence that TGF-#?l and 82 are not only formed in human and rat prostates but the expression of TGF-B;! mRNA is increased in benign prostatic hyperplasia (BPH) (Kyprianou and Isaacs, 1989; Mori et al., 1990). Similarly TGF-/3 has been detected in both human and rat prostate cancer and TGF-/ll mRNA levels are markedly elevated in prostate carcinoma sublimes (Steiner and Barrack, 1992; Thompson, 1990; Matuo et al., 1990) Therefore, not only androgens but members of the TGF-/3 family may play a role in normal as well as abnormal growth of sexual tissue and our work indicates that these peptides have the potential to modulate 5aR activity. It is now well established that androgen regulation of 5aR is mediated at the level of transcription of 5aR mRNA (Anderson et al., 1989), however it needs to be established if the mechanism(s) of action of TGF-b and activiniinhibin on 5aR is at the level of transcription, translation or enzyme turnover. The present results also emphasize the potential role of inhibin in the negative regulation of 5aR. Although androgens are known inducers of 5aR, these effects are completely abolished by the presence of low concentrations of inhibin. In view of the opposite actions of activin (TGF-b) and inhibin, the regulation of 5aR could depend on the balance between androgen and activin/inhibin interactions. Acknowledgements We wish to thank Thomas Jenkins and Lydia Baas for typing this manuscript. References Andersson, S., Bishop, R.W. and Russell, D.W. (1989) J. Biol. Chem. 264,16249-16255. Bellve, A.R. and Zheng, W. (1989) Reprod. Fertil. 85.771-793.
104
I. Anfonipillai et al. I Molecular and Cellular Endocrinology IO7 (1995) 99-104
DePaolo, L.V., Bicsak, T.A., Erickson, G.F., Shimasaki. S. and Ling, N. (1991) Proc. Sot. Exp. Biol. Med. 198,500-512. Griffin, J.E. and Wilson, J.D. (1989) pp. 1919-1944, McGraw-Hill, New York). Gupta, P., France-Saenz. R. and Mulrow, P.J. (1993) Endocrinology 32. 1184-I 188. Hasegawa, Y., Miyamoto, K., Abe, Y., Nakamura, T., Sugino, H., Elo, Y., Shibai, H. and Igarashi, M. (1988) Biochem. Biophys. Res. Commun. 156.668-674. Horton, R., Pasupuletti, V. and Antonipillai, 1. (1993) Endocrinology 133.447-45 1. Hotta, M. and Baird, A. (1986) Proc. Natl. Acad. Sci. USA 83, 77957799. Hsueh, A.J.W., Dahl, K.D., Vaughan, J., Tucker, E., Rivier, J., Bardin, C.W. and Vale, W. (1987) Proc. Natl. Acad. Sci. USA 84, 50825086. Hutchinson, L.A., Findly, J.K., de Vos, F.L. and Robertson, D.M. (1987) Biochem. Biophys. Res. Commun. 146, 1405-1412. Jacobs, S. and Story, M. (1988) Prostate 13,79-87. Jakubowiak, J., Janecki, A., Tong, D., Sanborm, B.M. and Steinberger. A. (1991) Endocrinology 82.265-273. Jenkins, E.P., Andersson, S., Imperato-McGinley, J., Wilson, J.D. and Russell, D.W. (1992) J. Clin. Invest. 89, 293-300. Kyprianou, N. and Isaacs, J.T. (1989) Mol. Endocrinol. 3, 1515-1522. Lin, T., Haskell, J., Vinson, N. and Terracio, L. (1986) Biochem. Biophys. Res. Commun. 137.950-956. Lin, T., Calkins, J.H., Morris, P.L., Vale, W., Bardin, C.W. (1989) Endocrinology 125,2134-2140. Ling, N., Ying, S.-Y., Ueno, N., Shimasaki, S., Esch, F., Hotta, M. and Guillemin, R. (1986) Nature 321.779-782. Massague, J., (1990) Annu. Rev. Cell. Biol. 6,597-641. Mather, J.P., Attie, K.M., Woodruff, T.K., Rice, G.C. and Phillips, D.M. (1990) Endocrinology 127.3206-3214. Matuo, Y., Nishi, N., Takasuka, H., Masuda, Y., Nishikawa, K., Isaacs. J.T., Adams, P.S., McKeehan, W.L. and Sato. G.H. (1990) Biothem. Biophys. Res. Commun. 166,840-847. Mauduit, C.. Chaurin, M.A., de Pretti, E., Morera, A.M. and Benahmed. M. (1991) Biol. Reprod. 45, 101-109.
Meunier, H., Rivier, C., Evans, R.M. and Vale, W. (1988) Proc. Nat]. Acad. Sci. USA 85,247-25 1. Mori, H., Maki, M.. Oishi, K., Jaye, M., Igarashi, K., Yoshida, 0. and Hatanaka, M. (1990) Prostate 16,71-80. Morimoto, I., Edminston, A., Hawks, D. and Horton, R. (1981) J. Clin. Endocrinol. Metab. 52,772-778. Nakamura, M., Minegishi, T., Hasegawa, Y., Nakamura. K., Igarashi, S., Ito, I., Shinozaki, H., Miyamoto, K., Eto, Y. and Ibuki, Y. (1993) Endocrinology 133.538-544. Nelson, K.G., Takahashi, T., Bosser, N.L., Walmer, D.K. and McLachian, J.A. (1991) Ptoc. Natl. Acad. Sci. USA 88.21-25. Roberts, V., Meunier, H., Vaughan, J., Rivier, J., Rivier, C., Vale, W. and Sawchenko, P. (1989) Endocrinology 124,552-554. Sakai, R., Eto, Y., Ohtsuka, M., Hirafaji, M. and Shinoda, H. (1993) Biochem. Biophys. Res. Commun. 195.39-46. Singh-Asa, P. and Waters, M.J. (1983) Mol. Cell. Endocrinol. 30, 189199. Steiner, M.S. and Barrack, E.R. (1992) Mol. Endocrinol. 6, 15-25. Thigpen, A.E., Silver, R.I., Guileyardo, J.M., Casey, M.L., McConnell. J.D. and Russell, D.M. (1993) J Clin. Invest. 92.903-910. Thompson, T.C. (1990) Cancer Cells 2,345-354. Vale, W., River, J., Vaughan, J., McClintock, R., Corrigan, A., Woo, W., Katr, D. and Spiess, J. (1986) Nature 321, 776-779. Vale, W., Hsueh, A., Rivier, C. and Yu, J. (1990) Handb. Exp. Pharmacol. (Berlin) 95,211-248. Vale, W., Rivier, C., Hsueh, A., Campen, C., Meunier, H.. Bicsak, T.. Vaughan, J., Corrigan, S., Bardin, W., Sawchenko, P., Petraglia, F., Yu, J., Plotsky, P., Spiess, J. and Rivier, J. (1988) Recent Prog. Horm. Pes. 44. I-34. Wahe, M., Antonipillai, I. and Horton, R. (1993) Mol. Cell. Endocrinol. 98.55-59. Wilson, J.D., (1975) Am. Physiol. Sot. 5.491-508. Xiao, S., Robertson, D.M. and Findlay, J.K. (1992) Endocrinology 131, 1009-1016. Ying, S and Shao-Yao (1988) Endocr. Rev. 9,267-292. Ying, S., Becker, A., Ling, N., Ueno, N. and Guillemin, R. (1986) Biochem. Biophys. Res Commun. 136.969-975.