- Email: [email protected]
Insulin Sensitizer, Troglitazone, Directly Inhibits Aromatase Activity in Human Ovarian Granulosa Cells
Biochemical and Biophysical Research Communications 271, 710 –713 (2000) doi:10.1006/bbrc.2000.2701, available online at http://www.idealibrary.com on
Insulin Sensitizer, Troglitazone, Directly Inhibits Aromatase Activity in Human Ovarian Granulosa Cells Yi-Ming Mu,* Toshihiko Yanase,* ,1 Yoshihiro Nishi,† Naoko Waseda,‡ Tanaka Oda,‡ Atsushi Tanaka,§ Ryoichi Takayanagi,* and Hajime Nawata* *Third Department of Internal Medicine, Faculty of Medicine, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582, Japan; †Division of Internal Medicine and ‡Division of Obstetrics and Gynecology, National Kokura Hospital, Harugaoka 10-1, Kokura-Minami-Ku, Kitakyushu, 802-0803, Japan; and §Department of Obstetrics and Gynecology, Saint Mother Hospital, 4-9-12, Yahata Nishi-ku, Kitakyushu City, Fukuoka, 807-0825, Japan
Received April 13, 2000
Ovarian granulosa cells synthesize estrogens from androgens, which are catalyzed by aromatase cytochrome P450 (P450arom). Troglitazone (Tro), one of the insulin-sensitizing compounds, thiazolidinediones (TZDs), is a ligand for peroxisome proliferator activated receptor ␥ (PPAR␥) and is effective in the treatment of both non-insulin-dependent diabetes mellitus (NIDDM) as well as polycystic ovary syndrome (PCOS). PPAR␥ exerts a transcriptional activity as a PPAR␥:RXR heterodimer. In this study, we investigated the effects of Tro and/or RXR ligand, LG100268 (LG) on the aromatase activity in cultured human ovarian granulosa cells obtained from patients who underwent in vitro fertilization. Human ovarian granulosa cells expressed PPAR␥ mRNA assessed by RTPCR. The treatment of the granulosa cells with Tro for 24 h resulted in a dramatic inhibition of the aromatase activity in a dose-dependent manner. While the treatment with LG alone also inhibited the aromatase activity, the combined treatment with both Tro and LG caused a much more reduction in the aromatase activity. The changes in the aromatase activity by Tro and/or LG were associated with comparable changes in P450arom mRNA assessed by RT-PCR. These results suggest that Tro directly inhibit the aromatase activity in human granulosa cells probably via nuclear receptor system PPAR␥:RXR heterodimer. The findings may provide a biochemical basis for the decrease in the blood concentrations of estrogens which is observed after the in vivo administration of Tro and may also possibly be useful as a novel therapy for estrogendependent diseases. © 2000 Academic Press
1 To whom correspondence should be addressed. Fax: 81-92-6425287 or 81-92-642-5297. E-mail: [email protected].
0006-291X/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.
Troglitazone (Tro) is one of the insulin-sensitizing compounds, thiazolidinediones (TZDs), which is known to be effective in the treatment of non-insulindependent diabetes (NIDDM) (1). TZDs are a high affinity ligand for the nuclear receptor called peroxisome proliferator-activated receptor ␥ (PPAR␥) (2). PPAR␥ forms a heterodimer with retinoid X receptor (RXR) and binds to the direct repeats of hormone response elements in gene promoters (1). Although the precise mechanism of action of TZD is still not fully understood, the insulin-sensitizing effects of TZDs have been suggested to be mediated through the activation of PPAR␥ because the rank order of the binding affinity of TZDs to PPAR␥ closely matches the order of potency of their antidiabetic action (1). Interestingly, Tro has also been proven to be effective in the treatment of metabolic and reproductive abnormalities in polycystic ovary syndrome (PCOS) (3), which is characterized by hyperandrogenism, chronic anovulation and the frequent occurrence of insulin resistance (4). Tro improves insulin resistance in PCOS, thus leading to decreased levels of androgens as well as estrogens (3). Another in vitro study demonstrated that Tro inhibits progestosterone production in cultured porcine granulosa cells (5). These findings suggest that TZDs may therefore have a direct or indirect effect on steroidogensis. Estrogens are mainly synthesized in ovarian granulosa cells before menopause, but also in peripheral tissues after menopause. The conversion of androgens to estrogens is catalyzed by an enzyme complex termed aromatase cytochrome P450 (P450arom) and a flavoprotein, NADPH-cytochrome P450 reductase. Aromatase is expressed in a variety of tissues, including ovary, brain, placenta, liver and adipose tissue. In the present study, to elucidate the mechanism of Tro action on ovarian steroidogenesis, we focused on the direct effects of Tro and/or RXR selective ligand
710
Vol. 271, No. 3, 2000
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
LG100268 (LG) on the aromatase activity in cultured human granulosa cells. MATERIALS AND METHODS Cell culture. Human ovarian granulosa cells were obtained from eight different individuals (20 – 42 years old) who underwent in vitro fertilization. The written informed consent was obtained from each subject before starting the study. The brief protocol of in vitro fertilization is as described below. Ovarian follicular growth was induced by the im administration of 150 –300 U hMG (Humegon, OrganonSankyo Co., Ltd.) per day for 7–14 days. Unexpected gonadotropin release was suppressed by the ongoing daily 900 g nasal administration of the GnRH analog (Sprccurc, Hoechst Co., Ltd.). Bolus im administration of 10,000 IU hCG (Gonatropin, Teikoku-Zouki Co., Ltd.) was performed after advantageous follicle had reached a diameter of 20 mm. Twenty-four hours after the hCG injection, oocytes with surrounding granulosa cells were picked up transvaginally under ultrasonographic guidance. The retrieved oocytes were preincubated for 5 h before insemination with spermatozoa. Other cells including granulosa cells were plated on a 96-well plate (Sumilon, Sumitomo Bakelite Co., Ltd.) and cultured with 200 l of DMEM/F12 medium (Gibco BRL, Life Technologies, Inc., Grand Island, NY) containing 5% DCS (dextran-coated, charcoal-treated FBS) for 12 h to facilitate the attachment of the cells to the plate. After the attachment of the cells, the plates were replaced with fresh medium and cultured for 24 h in the presence or absence of the following nuclear receptor ligands. Tro and LG (6) were kind gifts from Sankyo Chemical Industries, LTD, Tokyo and Ligand Pharmaceuticals Incorporated (San Diego, CA), respectively. Vitamin D, triiodethyronine (T3) and PPAR␣ ligand, fenofibrate were purchased from Sigma Chemical Co. (St. Louis, MO). All of these compounds were dissolved in dimethyl sulfoxide (DMSO) and applied to the cells in a 0.1% media volume. Aromatase assay. The aromatase activity was determined by measuring [ 3H]H 2O released upon the conversion of [1- 3H]androstenedion ( 3H-A-dione) to estrone as described before (7). Briefly, granulosa cells were plated on a 96-well plate and cultured with 200 l DMEM/F12 medium containing 5% DCS for 12 h. After the cells were treated with Tro and/or LG for 24 h, [1- 3H]androstenedione (New England Nuclear Corp., Boston, MA; specific activity, 27.5 Ci/mmol) was added, and the cells were then further incubated for 6 h. The following protocol for the extraction of the medium in order to measure the amount of radioactivity in [ 3H]H 2O was performed exactly as previously described (7). The cell protein content was determined using a micro BCA kit (Pierce Chemical Co., Rockford, IL) after the cells were dissolved in 1.0 N NaOH. The aromatase activity was expressed as pmol/mg protein/6 h. Reverse transcriptase-polymerase chain reaction (RT-PCR). The expression levels of P450arom and PPAR␥ mRNAs in human ovarian granulosa cells were examined by RT-PCR. The cells were treated with or without 10 ⫺5 M of Tro and/or 10 ⫺7 M LG for 24 h. Total RNA was prepared using a commercially available reagent, Isogen (Wako Pure Chemical Co., Osaka, Japan). First-strand complementary DNA (cDNA) was synthesized using 5 g of total RNA as a template, using a RT-PCR kit (Clontech, Palo Alto, CA). The sequences of the sense/antisense primers to amplify a 987 bp cDNA fragment for P450arom were 5⬘-CGGCCTTGTTCGTATGGTCA-3⬘/5⬘-GTCTCATCTGGGTGCAAGGA-3⬘ (7). The sequences of the sense/antisense primers to amplify a 276 bp of PPAR␥ cDNA fragment were 5⬘-ACAGAGATGCCATTCTGGCCC-3⬘/5⬘-CTTATTGTAGAGCTGAGTCTTCTC-3⬘) (8). The above two primers were designed to amplify a region common to both PPAR␥1 and ␥2. As a control 289 bp of -actin cDNA was amplified using the sense/antisense primers (5⬘TGGACTTCGAGCAAGAGATGG-3⬘/5⬘-ATCTCCTTCTGCATCCTGTCG-3⬘) (9). In order to achieve linear amplifying conditions, the PCR reactions for P450arom and PPAR␥ were performed with 35 cycles (denaturation at 94°C, annealing at 62°C and extension at
FIG. 1. The effect of various nuclear receptor ligands on the aromatase activity in cultured human ovarian granulosa cells. The Tro, fenofibrate, T 3, and vitamin D were used at a concentration of 10 ⫺6 M. LG used alone or combination with Tro was at a concentration of 10 ⫺7 M, and combinations with fenofibrate, Vit D or T 3 was at a concentration of 10 ⫺8 M. The cells were cultured in the presence or absence of various nuclear receptor ligands for 24 h. The aromatase activity in the medium was assayed as described under Materials and Methods. Bars represent mean ⫾ SD of three independent experiments with triplicate wells: a, P ⬍ 0.01 vs control cells treated with DMSO; b, P ⬍ 0.01 vs cells treated with Tro or LG. 72°C), while those for -actin cDNA were 28 cycles. The PCR products were then electrophoresed on 2% agarose gel and the intensity of the ethidium bromide luminescence was measured by a CCD image sensor (Densitograph AE6900F, Tokyo, Japan). We finally verified the nucleotide sequence of each PCR product by direct sequencing using the appropriate primers. Statistical analysis. Differences between the aromatase activity values for control cells and cells treated with the drugs were determined by Student’s t test. The statistical significance levels were P ⬍ 0.05. All data-points are presented as the mean ⫾ SD.
RESULTS Tro and/or LG inhibit the aromatase activity. Each basal aromatase activity in cultured human ovarian granulosa cells obtained from eight different individuals was about 16 to 24 pmol/mg protein/6 h (19.8 ⫾ 1.8 pmol/mg protein/6 h). At first, the effect of 10 ⫺6 M concentration of various nuclear receptor ligands including T 3, vitamin D, fenofibrate and Tro on the aromatase activity in cultured human ovarian granulosa cells were tested. Among them, 10 ⫺6 M Tro caused a 54% suppression of the aromatase activity in comparison to the control, while the treatment with 10 ⫺6 M of T 3, fenofibrate or vitamin D alone had no or little effect on the aromatase activity (Fig. 1). The strongest inhibitory effect of Tro on the aromatase activity was observed at 10 ⫺5 M (about 10% of the control) in all of the cultured human granulosa cells obtained from five different individuals (data not shown). In addition, 10 ⫺7 M of LG alone also suppressed the aromatase activity to 35% that of the control (Fig. 1). The combined treatment with 10 ⫺6 M Tro and 10 ⫺7 M LG for 24 h caused almost complete inhibition of the aromatase activity (2.5% of the control) (Fig. 1).
711
Vol. 271, No. 3, 2000
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. The dose-dependent effect of Tro and/or LG on the aromatase activity in cultured human granulosa cells. The human granulosa cells were cultured in the presence or absence of various concentrations of Tro and/or LG as indicated for 24 h. The aromatase activity in the medium was assayed as described under Materials and Methods. Bars represent mean ⫾ SD of three independent experiments with triplicate wells: *P ⬍ 0.05 vs control cells treated with DMSO; **P ⬍ 0.01 vs control cells treated with DMSO.
We next investigated the dose-dependency of Tro and/or the RXR ligand on the aromatase activity of human granulosa cells cultured for 24 h (Fig. 2). The inhibitory effect of Tro on the aromatase activity was dose-dependent and a 50% and 90% suppression of the activity was observed at concentrations of 10 ⫺6 M and 10 ⫺5 M, respectively. Although LG had little effect on the aromatase activity at concentrations of less than 10 ⫺8 M, it was effective in suppressing the aromatase activity at concentration of 10 ⫺7 M. Importantly, the inhibitory action was intensified by the combined treatment with Tro and LG at various concentrations. Even 10 ⫺10 M LG enhanced the suppressive effect of 10 ⫺8 M Tro on the aromatase activity (Fig. 2). However, combined treatment with 10 ⫺8 M LG and 10 ⫺6 M each of T 3, vitamin D and fenofibrate had little or no effect on the aromatase activity in cultured human granulosa cells (Fig. 1). Tro and LG decreased the P450arom mRNA expression. To investigate whether the changes in the aromatase activity were associated with comparable changes in the levels of mRNA encoding this enzyme, total RNA was extracted from the cells maintained in the absence or presence of 10 ⫺5 M Tro and/or 10 ⫺7 M LG for 24 h. The mRNA expression levels of P450arom and PPAR␥ were assessed by RT-PCR. Changes of P450arom mRNA levels were associated with comparable changes in the aromatase activity in granulosa cells. In contrast to the P450arom transcript, the mRNA levels of PPAR␥ remained unchanged after Tro and/or LG treatment (Fig. 3). DISCUSSION In the present study, we demonstrated for the first time that Tro suppresses the aromatase activity in
cultured human ovarian granulosa cells obtained from patients who underwent in vitro fertilization. Furthermore, while RXR selective ligand, LG alone also had a potent inhibitory effect on the aromatase activity, the combined treatment with Tro and LG caused a far more intensive reduction of the aromatase activity, thus suggesting that these actions are mediated through the nuclear system constituted by PPAR␥: RXR heterodimer. The suppressive effect of Tro and/or RXR ligand on the aromatase activity was reproducible in different cultures at least from eight different individuals. Little effect of other specific ligands for RXR heterodimer partners including fenofibrate (a ligand for PPAR␣), T 3 and vitamin D 3, may further support the notion that the remarkable effect of Tro in human granulosa cells may be mediated through the specific activation of PPAR␥. This is also supported by the detection of the gene transcript of PPAR␥ in the cells. The peak concentration of Tro in human plasma after the oral administration of clinically used dose of Tro (200-mg) is reported to range between 10 ⫺7–10 ⫺6 M (10). Thus, the inhibitory effect of 10 ⫺6 M Tro on the aromatase activity is well expected to be manifested as a reduced production of estrogens when Tro is clinically administered. Indeed, it has been reported that the oral administration of 200 – 400 mg Tro for 3 months in patients with PCOS caused a significant
FIG. 3. Expression of the P450arom, the PPAR␥, and the -actin genes in cultured human granulosa cells. The RT-PCR products were electrophoresed on 2% agarose gel containing 0.5 g/ml ethidium bromide. M indicates the DNA size marker, ⌽X174-Hinc II fragment. Negative indicates the PCR products without cDNA template (A). The relative expression levels of P450arom mRNA were determined by measurements of the intensity of the ethidium bromide. The data represent mean ⫾ SD of three independent experiments and are presented as the fold of -actin (B): a, P ⬍ 0.01 vs control cells treated with DMSO; b, P ⬍ 0.01 vs cells treated with Tro or LG.
712
Vol. 271, No. 3, 2000
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
decrease in the plasma concentration of E 2 and estrone (E 1) as well as of their precursors, androgens (3). Since insulin stimulates androgen production in vivo and in vitro (11, 12), the lowered circulating insulin levels by Tro in PCOS may result in a general decrease in the production of androgens as well as estrogens (3). However, in the same study (3), an increase in the T/E 2 ratio was also noted, thus suggesting a decreased conversion of androgens to estrogens (i.e., aromatization) by Tro. Since insulin can also stimulate aromatization both in vivo (13) and in vitro (12), the decrease in the plasma concentration of estrogens by Tro in PCOS may be due to the decreased aromatization as a result of the lowered circulating insulin level. In addition to these possible mechanisms, our present finding gives a third plausible explanation for the Tro-induced decrease in estrogen production. Namely, Tro can directly suppress estrogen production by inhibiting the aromatase activity in human ovarian granulosa cells. On the other hand, Tro has been demonstrated to have no effect on the aromatase activity in porcine granulosa cells (5). The major reason for this difference may be due to the stage of follicular development at which the experiments were done, since the development-related difference in the steroidogenesis of granulosa cell has been suggested in many previous studies (14). Estrogens play an important role in the growth and development of estrogen-dependent cancers, such as breast cancer (15). Our findings suggest that treatment with Tro or the combined treatment by Tro and selective RXR ligands could provide a new therapeutic tool in the treatment of estrogen-dependent diseases. The synergistic effect by both TZD and RXR ligand has also been noted in other systems such as the sensitization of diabetic mice to insulin (16) and the adipogenic differentiation of liposarcoma cells (17). In addition, the breast itself has a potential for estrogen biosynthesis, both in mammary adipose tissue and breast cancers displaying the aromatase activity necessary to convert androgen precursor into estrogen (15). Most recently, Tro has also been suggested to suppress such local production of estrogens by inhibiting the aromatase activity (18). TZD may therefore be useful in suppressing estrogen production from both ovarian and local origins. However, apart from the above described ben-
eficial aspects, we should also pay careful attention to the possibility that long-term treatment with TZD in patients with NIDDM might also cause unexpected side effects, such as ovarian dysfunction and climacteric problems. REFERENCES 1. Spiegelman, B. M. (1998) Diabetes 47, 507–514. 2. Lehmann, J. M., Moore, L. B., Smith-Oliver, T. A., Wilkison, W. O., Willson, T. M., and Kliewer, S. A. (1995) J. Biol. Chem. 270, 12953–12956. 3. Dunaif, A., Sxott, D., Finegood, D., Quintana, B., and Whitcomb, R. (1996) J. Clin. Endocrinol. Metab. 81, 3299 –3306. 4. Franks, S. (1995) N. Engl. J. Med. 333, 853– 861. 5. Gasic, S., Bodenburg, Y., Nagamani, M., Green, A., and Urban, R. J. (1998) Endocrinology 139, 4962– 4966. 6. Boehm, M. F., Zhang, L., Zhi, L., McClurg, M. R., Berger, E., Wagoner, M., Mais, D. E., Suto, C. M., Davies, J. A., and Heyman, R. A. (1995) J. Med. Chem. 38, 3146 –3155. 7. Tanaka, S., Haji, M., Takayanagi, R., Tanaka, S., Sugioka, Y., and Nawata, H. (1996) Endocrinology 137, 1860 –1869. 8. Yanase, T., Yashiro, T., Takitani, K., Kato, S., Taniguchi, S., Takayanagi, R., and Nawata, H. (1997) Biochem. Biophys. Res. Commun. 233, 320 –324. 9. Nakajima-Iijima, S., Hamada, H., Reddy, P., and Kakunaga, T. (1985) Proc. Natl. Acad. Sci. USA 82, 6133– 6137. 10. Nolan, J. J., Ludvik, B., Beerdsen, P., Joyce, M., and Olefsky, J. (1994) N. Engl. J. Med. 331, 1188 –1193. 11. Nestler, J. E., Barlascini, C. O., and Matt, D. W. (1989) J. Clin. Endocrinol. Metab. 68, 1027–1032. 12. Nestler, J. E., and Straus, J. F., III (1991) Endocrinol. Metab. Clin. North. Am. 20, 807– 823. 13. Dunaif, A., and Graf, M. (1989) J. Clin. Invest. 83, 23–29. 14. Amsterdam, A., and Selvarai, N. (1997) Endocr. Rev. 18, 435– 461. 15. Sasano, H., and Harada, N. (1998) Endocr. Rev. 19, 593– 607. 16. Mukherjee, R., Davies, P. J., Crombie, D. L., Bischoff, E. D., Cesario, R. M., Jow, L., Hamann, L. G., Boehm, M. F., Mondon, C. E., Nadzan, M., Paterniti, J. R., Jr., and Heyman, R. A. (1997) Nature 386, 407– 410. 17. Tontonoz, P., Singer, S., Forman, B. M., Sarraf, P., Fletcher, J. A., Fletcher, C. D., Brun, R. P., Mueller, E., Altiok, S., Oppenheim, H., Evans, R. M., and Spiegelman, B. M. (1997) Proc. Natl. Acad. Sci. USA 94, 237–241. 18. Rubin, G. L., Kalus, A. M., and Simpson, E. R. (1999) The Endocrine Society’s 81st Annual Meeting, San Diego, CA, 1999, p. 101.
713
Insulin Sensitizer, Troglitazone, Directly Inhibits Aromatase Activity in Human Ovarian Granulosa Cells Yi-Ming Mu,* Toshihiko Yanase,* ,1 Yoshihiro Nishi,† Naoko Waseda,‡ Tanaka Oda,‡ Atsushi Tanaka,§ Ryoichi Takayanagi,* and Hajime Nawata* *Third Department of Internal Medicine, Faculty of Medicine, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582, Japan; †Division of Internal Medicine and ‡Division of Obstetrics and Gynecology, National Kokura Hospital, Harugaoka 10-1, Kokura-Minami-Ku, Kitakyushu, 802-0803, Japan; and §Department of Obstetrics and Gynecology, Saint Mother Hospital, 4-9-12, Yahata Nishi-ku, Kitakyushu City, Fukuoka, 807-0825, Japan
Received April 13, 2000
Ovarian granulosa cells synthesize estrogens from androgens, which are catalyzed by aromatase cytochrome P450 (P450arom). Troglitazone (Tro), one of the insulin-sensitizing compounds, thiazolidinediones (TZDs), is a ligand for peroxisome proliferator activated receptor ␥ (PPAR␥) and is effective in the treatment of both non-insulin-dependent diabetes mellitus (NIDDM) as well as polycystic ovary syndrome (PCOS). PPAR␥ exerts a transcriptional activity as a PPAR␥:RXR heterodimer. In this study, we investigated the effects of Tro and/or RXR ligand, LG100268 (LG) on the aromatase activity in cultured human ovarian granulosa cells obtained from patients who underwent in vitro fertilization. Human ovarian granulosa cells expressed PPAR␥ mRNA assessed by RTPCR. The treatment of the granulosa cells with Tro for 24 h resulted in a dramatic inhibition of the aromatase activity in a dose-dependent manner. While the treatment with LG alone also inhibited the aromatase activity, the combined treatment with both Tro and LG caused a much more reduction in the aromatase activity. The changes in the aromatase activity by Tro and/or LG were associated with comparable changes in P450arom mRNA assessed by RT-PCR. These results suggest that Tro directly inhibit the aromatase activity in human granulosa cells probably via nuclear receptor system PPAR␥:RXR heterodimer. The findings may provide a biochemical basis for the decrease in the blood concentrations of estrogens which is observed after the in vivo administration of Tro and may also possibly be useful as a novel therapy for estrogendependent diseases. © 2000 Academic Press
1 To whom correspondence should be addressed. Fax: 81-92-6425287 or 81-92-642-5297. E-mail: [email protected].
0006-291X/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.
Troglitazone (Tro) is one of the insulin-sensitizing compounds, thiazolidinediones (TZDs), which is known to be effective in the treatment of non-insulindependent diabetes (NIDDM) (1). TZDs are a high affinity ligand for the nuclear receptor called peroxisome proliferator-activated receptor ␥ (PPAR␥) (2). PPAR␥ forms a heterodimer with retinoid X receptor (RXR) and binds to the direct repeats of hormone response elements in gene promoters (1). Although the precise mechanism of action of TZD is still not fully understood, the insulin-sensitizing effects of TZDs have been suggested to be mediated through the activation of PPAR␥ because the rank order of the binding affinity of TZDs to PPAR␥ closely matches the order of potency of their antidiabetic action (1). Interestingly, Tro has also been proven to be effective in the treatment of metabolic and reproductive abnormalities in polycystic ovary syndrome (PCOS) (3), which is characterized by hyperandrogenism, chronic anovulation and the frequent occurrence of insulin resistance (4). Tro improves insulin resistance in PCOS, thus leading to decreased levels of androgens as well as estrogens (3). Another in vitro study demonstrated that Tro inhibits progestosterone production in cultured porcine granulosa cells (5). These findings suggest that TZDs may therefore have a direct or indirect effect on steroidogensis. Estrogens are mainly synthesized in ovarian granulosa cells before menopause, but also in peripheral tissues after menopause. The conversion of androgens to estrogens is catalyzed by an enzyme complex termed aromatase cytochrome P450 (P450arom) and a flavoprotein, NADPH-cytochrome P450 reductase. Aromatase is expressed in a variety of tissues, including ovary, brain, placenta, liver and adipose tissue. In the present study, to elucidate the mechanism of Tro action on ovarian steroidogenesis, we focused on the direct effects of Tro and/or RXR selective ligand
710
Vol. 271, No. 3, 2000
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
LG100268 (LG) on the aromatase activity in cultured human granulosa cells. MATERIALS AND METHODS Cell culture. Human ovarian granulosa cells were obtained from eight different individuals (20 – 42 years old) who underwent in vitro fertilization. The written informed consent was obtained from each subject before starting the study. The brief protocol of in vitro fertilization is as described below. Ovarian follicular growth was induced by the im administration of 150 –300 U hMG (Humegon, OrganonSankyo Co., Ltd.) per day for 7–14 days. Unexpected gonadotropin release was suppressed by the ongoing daily 900 g nasal administration of the GnRH analog (Sprccurc, Hoechst Co., Ltd.). Bolus im administration of 10,000 IU hCG (Gonatropin, Teikoku-Zouki Co., Ltd.) was performed after advantageous follicle had reached a diameter of 20 mm. Twenty-four hours after the hCG injection, oocytes with surrounding granulosa cells were picked up transvaginally under ultrasonographic guidance. The retrieved oocytes were preincubated for 5 h before insemination with spermatozoa. Other cells including granulosa cells were plated on a 96-well plate (Sumilon, Sumitomo Bakelite Co., Ltd.) and cultured with 200 l of DMEM/F12 medium (Gibco BRL, Life Technologies, Inc., Grand Island, NY) containing 5% DCS (dextran-coated, charcoal-treated FBS) for 12 h to facilitate the attachment of the cells to the plate. After the attachment of the cells, the plates were replaced with fresh medium and cultured for 24 h in the presence or absence of the following nuclear receptor ligands. Tro and LG (6) were kind gifts from Sankyo Chemical Industries, LTD, Tokyo and Ligand Pharmaceuticals Incorporated (San Diego, CA), respectively. Vitamin D, triiodethyronine (T3) and PPAR␣ ligand, fenofibrate were purchased from Sigma Chemical Co. (St. Louis, MO). All of these compounds were dissolved in dimethyl sulfoxide (DMSO) and applied to the cells in a 0.1% media volume. Aromatase assay. The aromatase activity was determined by measuring [ 3H]H 2O released upon the conversion of [1- 3H]androstenedion ( 3H-A-dione) to estrone as described before (7). Briefly, granulosa cells were plated on a 96-well plate and cultured with 200 l DMEM/F12 medium containing 5% DCS for 12 h. After the cells were treated with Tro and/or LG for 24 h, [1- 3H]androstenedione (New England Nuclear Corp., Boston, MA; specific activity, 27.5 Ci/mmol) was added, and the cells were then further incubated for 6 h. The following protocol for the extraction of the medium in order to measure the amount of radioactivity in [ 3H]H 2O was performed exactly as previously described (7). The cell protein content was determined using a micro BCA kit (Pierce Chemical Co., Rockford, IL) after the cells were dissolved in 1.0 N NaOH. The aromatase activity was expressed as pmol/mg protein/6 h. Reverse transcriptase-polymerase chain reaction (RT-PCR). The expression levels of P450arom and PPAR␥ mRNAs in human ovarian granulosa cells were examined by RT-PCR. The cells were treated with or without 10 ⫺5 M of Tro and/or 10 ⫺7 M LG for 24 h. Total RNA was prepared using a commercially available reagent, Isogen (Wako Pure Chemical Co., Osaka, Japan). First-strand complementary DNA (cDNA) was synthesized using 5 g of total RNA as a template, using a RT-PCR kit (Clontech, Palo Alto, CA). The sequences of the sense/antisense primers to amplify a 987 bp cDNA fragment for P450arom were 5⬘-CGGCCTTGTTCGTATGGTCA-3⬘/5⬘-GTCTCATCTGGGTGCAAGGA-3⬘ (7). The sequences of the sense/antisense primers to amplify a 276 bp of PPAR␥ cDNA fragment were 5⬘-ACAGAGATGCCATTCTGGCCC-3⬘/5⬘-CTTATTGTAGAGCTGAGTCTTCTC-3⬘) (8). The above two primers were designed to amplify a region common to both PPAR␥1 and ␥2. As a control 289 bp of -actin cDNA was amplified using the sense/antisense primers (5⬘TGGACTTCGAGCAAGAGATGG-3⬘/5⬘-ATCTCCTTCTGCATCCTGTCG-3⬘) (9). In order to achieve linear amplifying conditions, the PCR reactions for P450arom and PPAR␥ were performed with 35 cycles (denaturation at 94°C, annealing at 62°C and extension at
FIG. 1. The effect of various nuclear receptor ligands on the aromatase activity in cultured human ovarian granulosa cells. The Tro, fenofibrate, T 3, and vitamin D were used at a concentration of 10 ⫺6 M. LG used alone or combination with Tro was at a concentration of 10 ⫺7 M, and combinations with fenofibrate, Vit D or T 3 was at a concentration of 10 ⫺8 M. The cells were cultured in the presence or absence of various nuclear receptor ligands for 24 h. The aromatase activity in the medium was assayed as described under Materials and Methods. Bars represent mean ⫾ SD of three independent experiments with triplicate wells: a, P ⬍ 0.01 vs control cells treated with DMSO; b, P ⬍ 0.01 vs cells treated with Tro or LG. 72°C), while those for -actin cDNA were 28 cycles. The PCR products were then electrophoresed on 2% agarose gel and the intensity of the ethidium bromide luminescence was measured by a CCD image sensor (Densitograph AE6900F, Tokyo, Japan). We finally verified the nucleotide sequence of each PCR product by direct sequencing using the appropriate primers. Statistical analysis. Differences between the aromatase activity values for control cells and cells treated with the drugs were determined by Student’s t test. The statistical significance levels were P ⬍ 0.05. All data-points are presented as the mean ⫾ SD.
RESULTS Tro and/or LG inhibit the aromatase activity. Each basal aromatase activity in cultured human ovarian granulosa cells obtained from eight different individuals was about 16 to 24 pmol/mg protein/6 h (19.8 ⫾ 1.8 pmol/mg protein/6 h). At first, the effect of 10 ⫺6 M concentration of various nuclear receptor ligands including T 3, vitamin D, fenofibrate and Tro on the aromatase activity in cultured human ovarian granulosa cells were tested. Among them, 10 ⫺6 M Tro caused a 54% suppression of the aromatase activity in comparison to the control, while the treatment with 10 ⫺6 M of T 3, fenofibrate or vitamin D alone had no or little effect on the aromatase activity (Fig. 1). The strongest inhibitory effect of Tro on the aromatase activity was observed at 10 ⫺5 M (about 10% of the control) in all of the cultured human granulosa cells obtained from five different individuals (data not shown). In addition, 10 ⫺7 M of LG alone also suppressed the aromatase activity to 35% that of the control (Fig. 1). The combined treatment with 10 ⫺6 M Tro and 10 ⫺7 M LG for 24 h caused almost complete inhibition of the aromatase activity (2.5% of the control) (Fig. 1).
711
Vol. 271, No. 3, 2000
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. The dose-dependent effect of Tro and/or LG on the aromatase activity in cultured human granulosa cells. The human granulosa cells were cultured in the presence or absence of various concentrations of Tro and/or LG as indicated for 24 h. The aromatase activity in the medium was assayed as described under Materials and Methods. Bars represent mean ⫾ SD of three independent experiments with triplicate wells: *P ⬍ 0.05 vs control cells treated with DMSO; **P ⬍ 0.01 vs control cells treated with DMSO.
We next investigated the dose-dependency of Tro and/or the RXR ligand on the aromatase activity of human granulosa cells cultured for 24 h (Fig. 2). The inhibitory effect of Tro on the aromatase activity was dose-dependent and a 50% and 90% suppression of the activity was observed at concentrations of 10 ⫺6 M and 10 ⫺5 M, respectively. Although LG had little effect on the aromatase activity at concentrations of less than 10 ⫺8 M, it was effective in suppressing the aromatase activity at concentration of 10 ⫺7 M. Importantly, the inhibitory action was intensified by the combined treatment with Tro and LG at various concentrations. Even 10 ⫺10 M LG enhanced the suppressive effect of 10 ⫺8 M Tro on the aromatase activity (Fig. 2). However, combined treatment with 10 ⫺8 M LG and 10 ⫺6 M each of T 3, vitamin D and fenofibrate had little or no effect on the aromatase activity in cultured human granulosa cells (Fig. 1). Tro and LG decreased the P450arom mRNA expression. To investigate whether the changes in the aromatase activity were associated with comparable changes in the levels of mRNA encoding this enzyme, total RNA was extracted from the cells maintained in the absence or presence of 10 ⫺5 M Tro and/or 10 ⫺7 M LG for 24 h. The mRNA expression levels of P450arom and PPAR␥ were assessed by RT-PCR. Changes of P450arom mRNA levels were associated with comparable changes in the aromatase activity in granulosa cells. In contrast to the P450arom transcript, the mRNA levels of PPAR␥ remained unchanged after Tro and/or LG treatment (Fig. 3). DISCUSSION In the present study, we demonstrated for the first time that Tro suppresses the aromatase activity in
cultured human ovarian granulosa cells obtained from patients who underwent in vitro fertilization. Furthermore, while RXR selective ligand, LG alone also had a potent inhibitory effect on the aromatase activity, the combined treatment with Tro and LG caused a far more intensive reduction of the aromatase activity, thus suggesting that these actions are mediated through the nuclear system constituted by PPAR␥: RXR heterodimer. The suppressive effect of Tro and/or RXR ligand on the aromatase activity was reproducible in different cultures at least from eight different individuals. Little effect of other specific ligands for RXR heterodimer partners including fenofibrate (a ligand for PPAR␣), T 3 and vitamin D 3, may further support the notion that the remarkable effect of Tro in human granulosa cells may be mediated through the specific activation of PPAR␥. This is also supported by the detection of the gene transcript of PPAR␥ in the cells. The peak concentration of Tro in human plasma after the oral administration of clinically used dose of Tro (200-mg) is reported to range between 10 ⫺7–10 ⫺6 M (10). Thus, the inhibitory effect of 10 ⫺6 M Tro on the aromatase activity is well expected to be manifested as a reduced production of estrogens when Tro is clinically administered. Indeed, it has been reported that the oral administration of 200 – 400 mg Tro for 3 months in patients with PCOS caused a significant
FIG. 3. Expression of the P450arom, the PPAR␥, and the -actin genes in cultured human granulosa cells. The RT-PCR products were electrophoresed on 2% agarose gel containing 0.5 g/ml ethidium bromide. M indicates the DNA size marker, ⌽X174-Hinc II fragment. Negative indicates the PCR products without cDNA template (A). The relative expression levels of P450arom mRNA were determined by measurements of the intensity of the ethidium bromide. The data represent mean ⫾ SD of three independent experiments and are presented as the fold of -actin (B): a, P ⬍ 0.01 vs control cells treated with DMSO; b, P ⬍ 0.01 vs cells treated with Tro or LG.
712
Vol. 271, No. 3, 2000
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
decrease in the plasma concentration of E 2 and estrone (E 1) as well as of their precursors, androgens (3). Since insulin stimulates androgen production in vivo and in vitro (11, 12), the lowered circulating insulin levels by Tro in PCOS may result in a general decrease in the production of androgens as well as estrogens (3). However, in the same study (3), an increase in the T/E 2 ratio was also noted, thus suggesting a decreased conversion of androgens to estrogens (i.e., aromatization) by Tro. Since insulin can also stimulate aromatization both in vivo (13) and in vitro (12), the decrease in the plasma concentration of estrogens by Tro in PCOS may be due to the decreased aromatization as a result of the lowered circulating insulin level. In addition to these possible mechanisms, our present finding gives a third plausible explanation for the Tro-induced decrease in estrogen production. Namely, Tro can directly suppress estrogen production by inhibiting the aromatase activity in human ovarian granulosa cells. On the other hand, Tro has been demonstrated to have no effect on the aromatase activity in porcine granulosa cells (5). The major reason for this difference may be due to the stage of follicular development at which the experiments were done, since the development-related difference in the steroidogenesis of granulosa cell has been suggested in many previous studies (14). Estrogens play an important role in the growth and development of estrogen-dependent cancers, such as breast cancer (15). Our findings suggest that treatment with Tro or the combined treatment by Tro and selective RXR ligands could provide a new therapeutic tool in the treatment of estrogen-dependent diseases. The synergistic effect by both TZD and RXR ligand has also been noted in other systems such as the sensitization of diabetic mice to insulin (16) and the adipogenic differentiation of liposarcoma cells (17). In addition, the breast itself has a potential for estrogen biosynthesis, both in mammary adipose tissue and breast cancers displaying the aromatase activity necessary to convert androgen precursor into estrogen (15). Most recently, Tro has also been suggested to suppress such local production of estrogens by inhibiting the aromatase activity (18). TZD may therefore be useful in suppressing estrogen production from both ovarian and local origins. However, apart from the above described ben-
eficial aspects, we should also pay careful attention to the possibility that long-term treatment with TZD in patients with NIDDM might also cause unexpected side effects, such as ovarian dysfunction and climacteric problems. REFERENCES 1. Spiegelman, B. M. (1998) Diabetes 47, 507–514. 2. Lehmann, J. M., Moore, L. B., Smith-Oliver, T. A., Wilkison, W. O., Willson, T. M., and Kliewer, S. A. (1995) J. Biol. Chem. 270, 12953–12956. 3. Dunaif, A., Sxott, D., Finegood, D., Quintana, B., and Whitcomb, R. (1996) J. Clin. Endocrinol. Metab. 81, 3299 –3306. 4. Franks, S. (1995) N. Engl. J. Med. 333, 853– 861. 5. Gasic, S., Bodenburg, Y., Nagamani, M., Green, A., and Urban, R. J. (1998) Endocrinology 139, 4962– 4966. 6. Boehm, M. F., Zhang, L., Zhi, L., McClurg, M. R., Berger, E., Wagoner, M., Mais, D. E., Suto, C. M., Davies, J. A., and Heyman, R. A. (1995) J. Med. Chem. 38, 3146 –3155. 7. Tanaka, S., Haji, M., Takayanagi, R., Tanaka, S., Sugioka, Y., and Nawata, H. (1996) Endocrinology 137, 1860 –1869. 8. Yanase, T., Yashiro, T., Takitani, K., Kato, S., Taniguchi, S., Takayanagi, R., and Nawata, H. (1997) Biochem. Biophys. Res. Commun. 233, 320 –324. 9. Nakajima-Iijima, S., Hamada, H., Reddy, P., and Kakunaga, T. (1985) Proc. Natl. Acad. Sci. USA 82, 6133– 6137. 10. Nolan, J. J., Ludvik, B., Beerdsen, P., Joyce, M., and Olefsky, J. (1994) N. Engl. J. Med. 331, 1188 –1193. 11. Nestler, J. E., Barlascini, C. O., and Matt, D. W. (1989) J. Clin. Endocrinol. Metab. 68, 1027–1032. 12. Nestler, J. E., and Straus, J. F., III (1991) Endocrinol. Metab. Clin. North. Am. 20, 807– 823. 13. Dunaif, A., and Graf, M. (1989) J. Clin. Invest. 83, 23–29. 14. Amsterdam, A., and Selvarai, N. (1997) Endocr. Rev. 18, 435– 461. 15. Sasano, H., and Harada, N. (1998) Endocr. Rev. 19, 593– 607. 16. Mukherjee, R., Davies, P. J., Crombie, D. L., Bischoff, E. D., Cesario, R. M., Jow, L., Hamann, L. G., Boehm, M. F., Mondon, C. E., Nadzan, M., Paterniti, J. R., Jr., and Heyman, R. A. (1997) Nature 386, 407– 410. 17. Tontonoz, P., Singer, S., Forman, B. M., Sarraf, P., Fletcher, J. A., Fletcher, C. D., Brun, R. P., Mueller, E., Altiok, S., Oppenheim, H., Evans, R. M., and Spiegelman, B. M. (1997) Proc. Natl. Acad. Sci. USA 94, 237–241. 18. Rubin, G. L., Kalus, A. M., and Simpson, E. R. (1999) The Endocrine Society’s 81st Annual Meeting, San Diego, CA, 1999, p. 101.
713