Effects of Flavonoid Phytochemicals on Cortisol Production and on Activities

Journal of Steroid Biochemistry & Molecular Biology 80 (2002) 355–363

Effects of ﬂavonoid phytochemicals on cortisol production and on activities of steroidogenic enzymes in human adrenocortical H295R cells Shuji Ohno a, Satoshi Shinoda a, Satoshi Toyoshima a, Hiroyuki Nakazawa b, Tsunehisa Makino c, Shizuo Nakajin a,∗ a Department of Biochemistry, Faculty of Pharmaceutical Sciences, Hoshi University, 2-4-41 Ebara, Shinagawa-ku, Tokyo 142-8501, Japan b Department of Analytical Chemistry, Faculty of Pharmaceutical Sciences, Hoshi University, 2-4-41 Ebara, Shinagawa-ku, Tokyo 142-8501, Japan c Department of Obstetrics and Gynecology, School of Medicine, Tokai University, Bohseidai, Isehara City, Kanagawa 259-1143, Japan Received 3 May 2001; accepted 7 December 2001

Abstract Inhibitory effects of flavonoid phytochemicals, flavones, flavonols and isoflavones on cortisol production were examined in human adrenal H295R cells stimulated with di-buthylyl cAMP. In addition, the inhibitory effects of these chemicals on the activity of P450scc, 3␤-HSD type II (3␤-HSD II), P450c17, P450c21 and P45011␤, steroidogenic enzymes involved in cortisol biosynthesis, were examined in the same cells. Exposure to 12.5 ␮M of the flavonoids 6-hydroxyflavone, 4-hydroxyflavone, apigenin, daidzein, genistein and formononetin significantly decreased cortisol production (by 6.3, 69.6, 47.5, 26.6, 13.8 and 11.3%, respectively), and biochanin A significantly decreased cortisol production (by 47.3%) at a concentration of 25 ␮M without any significant cytotoxic effects or changes in cell number. Daidzin, the 7-glucoside of daidzein, did not alter cortisol production by H295R cells at concentrations over 10 ␮g/ml (24 ␮M). Daidzein-induced reduction of cortisol production by H295R cells was not inhibited by the estrogen receptor antagonist ICI 182,780. The flavonoids 6-hydroxyflavone, daidzein, genistein, biochanin A and formononetin strongly and significantly inhibited microsomal 3␤-HSD II activity at concentrations from 1 to 25 ␮M, and I50 values were estimated to be 1.3, 2, 1, 0.5 and 2.7 ␮M, respectively. In addition, these flavonoids significantly inhibited microsomal P450c21 activity at 12.5 and/or 25 ␮M. In addition, 6-hydroxyflavone inhibited activity of microsomal P450c17 and mitochondrial P45011␤ at 12.5 and/or 25 ␮M. Results of Lineweaver–Burk’s plot analysis indicate that daidzein is a competitive inhibitor of the activity of 3␤-HSD II and P450c21. Km and Vmax values of 3␤-HSD II for DHEA were estimated to be 6.6 ␮M and 328 pmol/min mg protein, respectively. Km and Vmax values of P450c21 for progesterone were estimated to be 2.8 ␮M and 16 pmol/min mg protein, respectively. Ki values of 3␤-HSD II and P450c21 for daidzein were estimated to be 2.9 and 33.3 ␮M, respectively. © 2002 Elsevier Science Ltd. All rights reserved.

Keywords: Adrenocortical H295R cell; Steroidogenesis; Phytoestrogen; Flavonoids; Enzyme inhibition

1. Introduction estrogens to estrogen receptors. Many studies have, therefore, investigated estrogen receptor binding and subsequent The endocrine system is indispensable in sustaining downstream signaling events. Endocrine-disrupting chem- biological homeostasis. In particular, steroid hormone es- icals are exogenous agents that interfere with synthesis, trogens play an important role in female and male repro- production, transport, binding, action, and elimination of ductive systems and influence growth, differentiation, and the hormones responsible for maintenance of homeostasis, function of many target cells. Recent reports indicate that reproduction, development and/or behavior (White House environmental exposure to endocrine-disrupting chemicals Workshop, January 1997). Inhibition of the biosynthesis or adversely affect human and wildlife reproductive systems production of steroid hormones during fetal, perinatal, and [1]. Many environmental chemicals that are suspected neonatal periods can yield serious irreversible changes in endocrine-disrupting agents display native estrogen-like human and wildlife reproductive systems. structures and exhibit estrogenic activity, and have thus Phytoestrogen flavonoids are plant chemicals that are been dubbed environmental estrogens. Initial estrogenic structurally analogous to estrogen and are known to affect activity is mediated by binding of these environmental estrogenic activity [2]. Leguminous plants generally contain isoflavones, and soybeans are also rich in isoflavones such ∗ Corresponding author. Tel.: +81-3-5498-5775; fax: +81-3-3787-0036. as daidzein, genistein, glycitein and their glucosides, which E-mail address: [email protected] (S. Nakajin). are consumed in traditional diets containing soy-derived

0960-0760/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S0960-0760(02)00021-3 356 S. Ohno et al. / Journal of Steroid Biochemistry & Molecular Biology 80 (2002) 355–363 foods [3,4]. It has been reported that serum concentration were maintained in D-MEM/F-12 medium (1:1 mixture of total phytoestrogens (daidzein and genistein) is greater in of Dulbecco’s-modified Eagle’s and Ham’s F-12 medium) Japanese (0.16–0.89 ␮M, mean 0.4 ␮M) than Finnish men containing pyridoxine hydrochloride, l-glutamine, and (7–25 nM, mean 12.5 nM) [5], and reach still higher levels 15 mmol/l HEPES supplemented with insulin (6.25 ␮g/ml), in infants who consume large amounts of soy-derived foods transferrin (6.25 ␮g/ml), selenium (6.25 ng/ml), and linoleic (2.2–7 ␮M, mean 3.8 ␮M) [6]. The estrogen-like activity acid (5.35 ␮g/ml). Antibiotics (penicillin: 50 IU/ml and of phytoestrogens has been determined by assaying prolif- streptomycin: 50 ␮g/ml), 1% ITS Plus (Collaborative Re- eration of human breast cancer cells, competitive binding search, Bedford, MA), and 2% ultrose G (Sepracore, assays with estrogen receptor or transient gene expression France) were also added. Cells were maintained as mono- assays [7–13]. However, these reports did not consider the layer cultures in 75 cm2 flasks at 37 ◦C in an atmosphere effects of phytoestrogens on biosynthesis or production of of 5% CO2–95% air. Cells were subcultured onto 24-well steroid hormones. The aim of the present study was to in- dishes for experiments. After 48 h, medium was removed vestigate the inhibitory effects of phytoestrogen flavonoids and cells were treated as described below. on steroidogenesis. The recently developed human adrenocortical tumor cell line H295R can secrete steroids charac- 2.3. Stimulation of steroid production and analysis teristic of the three adrenocortical zones and, thus, appear to of steroids be pluripotent [14–17]. Therefore, we used adrenocortical H295R cells as a model of human steroidogenic cells to in- Cells subcultured onto 24-well plates were maintained vestigate the inhibitory effects of flavonoid phytochemicals for 24 h in D-MEM/F-12 medium containing 1% ITS Plus, on steroidogenesis. 0.01% bovine serum albumin (Bovuminar®: Intergen, New York) and antibiotics. Medium was then renewed (0.5 ml per well) and phytoestrogen flavonoid chemicals dissolved 2. Materials and methods in ethanol were added. The final concentration of alcohol solvents in the assay mixture did not exceed 1% and 2.1. Chemicals was confirmed to not obstruct cortisol production. At the same time, dibutyryl cyclic AMP (1 mM) was added to the 5-Hydroxyflavone, 6-hydroxyflavone, 6-methoxyflavone, medium to stimulate steroidogenesis because the H295R 7-hydroxyflavone, 4 -hydroxyflavone, 4 ,5,7-trihydroxyfla- cell line lacks sensitivity to ACTH treatment [15]. To assess vone (apigenin), 7-hydroxy-4 -methoxyisoflavone (for- the effects of phytoestrogen flavonoids, cells were exposed mononetin), and epidermal growth factor (EGF, human for 48 h except as otherwise noted. Cortisol content of each recombinant) were purchased form Funakoshi Co. Ltd. well was determined by radioimmunoassay with the DPC (Tokyo, Japan). Dibutyryl cAMP, 4 ,7-dihydroxyisoflavone cortisol kit (Diagnostic Product Corporation, Los Angels, (daidzein), 4 ,7-dihydroxyisoflavone–7-glucoside (daidzin) CA). Cytotoxicity of each chemical was tested using the 4 ,5,7-trihydroxyisoflavone (genistein), 4 ,5,7-trihydroxyiso- CytoTox96 non-radioactive cytotoxicity assay kit that deter- flavone–7-glucoside (genistin) and 5,7-dihydroxy-4 -metho- mines LDH activity (Promega Corp., Madison, WI). Cells xyisoflavone (biochanin A) were purchased form Wako were washed in phosphate buffered saline and solubilized Pure Chemical Industries Ltd. (Tokyo, Japan). Radioac- in Tris–HCl (50 mM, pH 7.4) containing NaCl (150 mM), 14 tive [4- C] steroids, cholesterol (1887 MBq/mmol), sodium dodecyl sulfate (SDS, 1%), EGTA (5 mM), MgCl2 dehydroepiandrosterone (2.05 GBq/mmol), progesterone (0.5 mM), MnCl2 (0.5 mM) and phenylmethylsulfonylflu- (2.05 GBq/mmol) and deoxycorticosterone (2.22 GBq/mmol) oride (PMSF, 0.2 mM). The protein content of each sam- were purchased from New England Nuclear Corp. (Boston, ple was determined by BCA assay (Pierce Chemical Co., + MA). Non-radioactive steroids, NAD , NADPH, glucose-6- Rockford, IL). phosphate, glucose-6-phosphate dehydrogenase and 3,3,4, 5,7-pentahydroxyflavone (quercetin) were purchased from 2.4. Preparation of H295R mitochondrial Sigma–Aldrich, K.K., Japan (Tokyo, Japan). Estrogen re- and microsomal fractions ceptor antagonist, ICI 182,780 was purchased form Nacalai Tesque, Inc. (Kyoto, Japan). All reagents were of the high- Confluent H295R cells were treated with dibutyryl cAMP est commercially available grade. Adrenodoxin, the redox (1 mM) and EGF (10 ng/ml) for 24 h. Cells were washed partner of cytochrome P450s in adrenal mitochondria, was three times with ice-cold PBS, harvested and collected by purified from bovine adrenal gland following the method of centrifugation at 350 × g for 5 min. All procedures were Kimura et al. [18]. carried out at 4 ◦C or on ice. Cells (1–2.1 g wet weight) were homogenized in a Dounce tissue grinder (Wheaton, 2.2. Cell culture Millville, NJ) in 3 ml of ice-cold 0.25 M sucrose containing 5 mM potassium phosphate (pH 7.4), 0.5 mM EDTA, H295R cells were generously provided by Prof. J.I. Mason 1 mM DTT and 0.1 mM PMSF (buffer A), then centrifuged (University of Edinburgh, Edinburgh, Scotland). Cells at 800 × g for 10 min. Supernatant was decanted and S. Ohno et al. / Journal of Steroid Biochemistry & Molecular Biology 80 (2002) 355–363 357 centrifuged at 9000 × g for 20 min. The resulting mito- pregnenolone, corticosterone and androstenedione pro- chondrial pellet was washed with ice-cold buffer A and duced, respectively. P450c17 activity was expressed as the 0.21 M mannitol containing 70 mM sucrose, 5 mM MgCl2, sum of 17␣-hydroxyprogesterone and 11-deoxycortisol pro- 0.5 mM CaCl2 and 20 mM Tris–HCl (pH 7.4) (buffer B) duction, while that of P450c21 was expressed as the sum of and resuspended in 3 ml buffer B containing 1 mM DTT 11-deoxycortisol and 11-deoxycorticosterone production. and 0.1 mM PMSF. Mitochondria were sonicated with 10 All enzyme activity was expressed as pmol product/min mg cycles of 15 s, 50 W (Model VC50T, Sonic & Materials protein. Inc., Danbury, CT) to assay P450scc (side-chain cleavage enzyme of cholesterol, CYP11A gene product) and P45011␤ 2.6. Statistical analysis (11␤-hydroxylase, CYP11B1 gene product). The supernatant produced by the above centrifugation at 9000 × g Statistical analysis of data was performed with Student’s for 20 min was incubated with 8 mM calcium chloride for t-test. The point of minimal statistical significance was set 1 h in an ice bath. Microsomes were harvested by centrifu- at P ≤ 0.05. gation at 10,000 × g for 10 min, followed by resuspension in 2 ml of buffer A containing 20% glycerol. Mitochondrial and microsomal samples contained 5 and 5.6 mg/ml of pro- 3. Results tein, respectively. Both preparations were stored at −80 ◦C until use. 3.1. Effects of flavonoid phytochemicals on cortisol production by H295R cells 2.5. Enzyme assay Inhibitory effects of various concentrations of flavonoids P450scc and P45011␤ activities were measured by incu- including some flavones, flavonols and isoflavones (Fig. 1), bating [4-14C] cholesterol (500 Bq/1 nmol/2 ␮l ethanol) and on cortisol production were examined in H295R cells stim- [4-14C] deoxycorticosterone (500 Bq/1 nmol/2 ␮l ethanol), ulated with di-buthylyl cAMP. Upon exposure to 12.5 ␮M respectively with 50 ␮g sonicated mitochondria in 0.2 ml of the flavonoids 6-hydroxyflavone, 4 -hydroxyflavone, of buffer B containing 25 mM glucose-6-phosphate, apigenin, daidzein, genistein and formononetin, cortisol 0.1 unit glucose-6-phosphate dehydrogenase, 0.5 mM production significantly decreased (6.3, 69.6, 47.5, 26.6, NADP+,2␮M adrenodoxin, flavonoid sample solution, 13.8 and 11.3% reduction, respectively) (Fig. 2). Expo- and steroid substrate, for 2 h at 37 ◦C. 3␤-HSD II ac- sure to other flavones and isoflavones at this concentration tivity (3␤-hydroxysteroid dehydrogenase coupled with did not affect cortisol production. In addition to these 5–4-isomerase, type II) was measured by incubating flavonoids, biochanin A also significantly decreases cor- 56 ␮g microsomes in 0.2 ml of 100 mM potassium phos- tisol production (47.3% reduction) at a concentration of phate (pH 7.4) containing 0.5 mM NAD+, sample solution 25 ␮M (data not shown). Exposure to even 10 ␮g/ml of and [4-14C] dehydroepiandrosterone (500 Bq/1 nmol 2 ␮l these flavones or isoflavones (corresponding to concentra- ethanol) for 1 h at 37 ◦C. The activities of P450c17 tions of 23.1–41.6 ␮M) produced no significant cytotoxic (17␣-hyroxylase/C17,20-lyase, CYP17 gene product) and effects on cell numbers, as estimated from LDH activity P450c21 (21-hydroxylase, CYP21B gene product) were and total protein of solubilized cells. Increasing concentra- measured by incubation of 28 ␮g microsomes in 0.2 ml of tions of 6-hydroxyflavone, apigenin, daidzein, genistein and 100 mM potassium phosphate (pH 7.4) containing 25 mM formononetin compounds were observed to reduce cortisol glucose-6-phosphate, 0.1 unit glucose-6-phosphate dehy- production in a dose-dependent manner (Fig. 3). Daidzin + drogenase, 5 mM MgCl2, 0.5 mM NADP , sample solution is the 7-glucoside of daidzein, and did not alter cortisol and [4-14C] progesterone (500 Bq/1 nmol/2 ␮l ethanol) at production by H295R cells at concentrations over 10 ␮g/ml 37 ◦C for 1 h. After incubation, steroids were extracted with (24 ␮M/ml). 0.6 ml of ethyl acetate/2,2,4-trimethylpentane (1/1, v/v). Af- ter concentration by evaporation, steroids were analyzed by 3.2. Effect of estrogen receptor antagonist ICI 182,780 TLC (Kieselgel 60F254, Merck, Darmstadt, Germany) in on inhibition of cortisol production by daidzein benzene/acetone (8/1, v/v) for P450scc and 3␤-HSD II, in benzene/ethyl acetate (2/1, v/v) for P450c17 and P450c21, Daidzein-induced reduction of cortisol production by in benzene/acetone (7/3, v/v) for P45011␤ activity. Steroids H295R cells is not inhibited by the estrogen receptor an- were detected by autoradiography (Scientific Imaging Film, tagonist ICI 182,780 (Fig. 4). Exposure to 5 ␮g/ml daidzein Biomax MR, Eastman Kodak, Rochester, NY). Radioactivity is sufficient to suppress cortisol production by H295R of relevant TLC bands was measured in ACS II liquid scin- cells. The addition of ICI 182,780 (50 and 500 nM) did tillation fluid (Amersham Pharmacia Biotech, Tokyo, Japan) not restore the daidzein-induced reduction in cortisol pro- in a Tri–Carb 2050CA liquid scintillation analyzer (Packard duction. This result suggests that reduction in cortisol Japan Co., Tokyo, Japan). The activities of P450scc, production by daidzein is not mediated through estrogen P45011␤ and 3␤-HSD II were expressed as amounts of receptors. 358 S. Ohno et al. / Journal of Steroid Biochemistry & Molecular Biology 80 (2002) 355–363

Fig. 1. Substances used in the present study. Structures of the ﬂavonoids are shown.

Fig. 2. Effect of exposure to phytoestrogen flavonoids on dibutyryl cAMP-induced cortisol secretion by H295R cells. Abbreviations: apigenin, 4,5,7-trihydroxyflavone; quercetin, 3,3,4,5,7-pentahydroxyflavone; daidzein, 4,7-dihydroxyisoflavone; daidzin, 4,7-dihydroxyisoflavone–7-glucoside; genistein, 4,5,7-trihydroxyisoflavone; genistin, 4,5,7-trihydroxyisoflavone-7-glucoside; biochanin A, 5,7-dihydroxy-4-methoxyisoflavone; formononetin, 7-hydroxy-4-methoxyisoflavone. Each column represents mean with S.D. (n = 3–6). Asterisks denote significant difference from control value, ∗P ≤ 0.05 and ∗∗P ≤ 0.01. S. Ohno et al. / Journal of Steroid Biochemistry & Molecular Biology 80 (2002) 355–363 359

Fig. 3. Effects of increasing concentrations of 6-hydroxyﬂavone (A), apigenin (B), daidzein (C), daidzin (D), genistein (E) and formononetin (F) on cortisol secretion from adrenocortical H295R cells induced by dibutyryl cAMP. Each bar represents mean with S.D. (n = 3). Asterisks denote signiﬁcant difference from the control (0 ␮g/ml) value from untreated cells, ∗∗P ≤ 0.01.

3.3. Effects of cortisol secretion-inhibiting flavonoid 6-Hydroxyflavone, daidzein, genistein, biochanin A and phytochemicals on steroidogenic enzymes formononetin strongly and significantly inhibited 3␤-HSD II activity at concentrations of 1–25 ␮M, and I50 values To elucidate the inhibitory effects of 6-hydroxyflavone, were estimated to be 1.3, 2, 1 and 0.5 and 2.7 ␮M. Further- daidzein, genistein, biochanin A and formononetin on cor- more, these flavonoids significantly inhibited P450c21 acti- tisol production of H295R cells, the effects of these com- vity at 12.5 and/or 25 ␮M. In addition, 6-hydroxyflavone, pounds on the steroidogenic enzymes P450scc, 3␤-HSD which inhibited cortisol production relatively effectively, II, P450c17, P450c21 and P45011␤, were investigated significantly inhibited P450c17 and P45011␤ activity at (Table 1). Enzyme activities were expressed relative 12.5 and/or 25 ␮M. Daidzin, the 7-glucoside of daidzein, to the control experiment, which was set at 100%. did not inhibit the activity of steroidogenic enzymes. 360 S. Ohno et al. / Journal of Steroid Biochemistry & Molecular Biology 80 (2002) 355–363

Neither daizein nor any of the isoﬂavonoids tested were able to alter P45011␤ activity.

3.4. Kinetic analysis of 3β-HSD II and P450c21 inhibition by daidzein

Inhibition of 3␤-HSD II and P450c21 activities by phytoestrogen ﬂavonoids was additionally examined using daidzein. Lineweaver–Burk’s plots for 3␤-HSD II and P450c21 activities in the presence of daidzein are shown in Fig. 5. Both plots indicate competitive inhibition. Km and Vmax values of 3␤-HSD II for DHEA were estimated to be 6.6 ␮M and 328 pmol/min mg protein, respectively. Mean- while, Km and Vmax values of P450c21 activity for progesterone were estimated to be 2.8 ␮M and 16 pmol/min mg protein, respectively. Ki values of 3␤-HSD II and P450c21 activities for daidzein were estimated to be 2.9 ␮M and 33.3 ␮M, respectively. Fig. 4. Effect of estrogen receptor antagonist ICI 182,780 on reduction of H295R cells cortisol production by exposure of daidzein (5 ␮g/ml). Each column represents mean with S.D. (n = 3). Asterisks denote signiﬁcant difference from control (untreated), ∗∗P ≤ 0.01. 4. Discussion

The present study examined the ability of several ﬂavonoid phytochemicals to inhibit steroidogenesis by human adrenocortical H295R cells through the alteration

Table 1 Inhibition of enzyme activities of steroid lyase, hydroxylases and dehydrogenase involved in cortisol synthesis Addition (␮M) Enzyme activity (% of control)

P450scc 3␤-HSD II P450c17 P450c21 P45011␤

None 100.0 100.0 100.0 100.0 100.0 6-Hydroxyflavone 1.0 – 57.3 ± 0.7∗∗ –– – 12.5 153.5 ± 10.2 8.5 ± 0.4∗∗ 85.3 ± 2.0 45.3 ± 2.3∗∗ 77.4 ± 2.6∗∗ 25.0 121.4 ± 6.3 4.2 ± 0.1∗∗ 68.3 ± 1.8∗ 34.7 ± 1.1∗∗ 61.5 ± 3.5∗∗ Daidzein 3.1 – 40.8 ± 1.7∗∗ –– – 12.5 81.6 ± 4.8 17.4 ± 1.5∗∗ 95.2 ± 13.5 63.3 ± 4.4∗∗ 159.2 ± 16.4 25.0 101.6 ± 16.1 9.0 ± 0.6∗∗ 101.3 ± 16.1 57.3 ± 2.4∗∗ 143.2 ± 13.1 Daidzin 12.5 115.6 ± 20.1 94.9 ± 1.4 112.6 ± 9.0 117.9 ± 19.1 110.3 ± 3.6 25.0 122.2 ± 15.1 93.1 ± 1.2 111.8 ± 6.2 105.7 ± 10.4 115.5 ± 7.4 Genistein 1.0 – 47.3 ± 0.6∗∗ –– – 12.5 134.7 ± 11.4 7.9 ± 0.0 ∗∗ 109.6 ± 2.4 82.7 ± 9.2∗∗ 123.1 ± 22.3 25.0 110.0 ± 6.9 4.6 ± 0.7∗∗ 95.0 ± 2.6 51.2 ± 1.9∗∗ 94.3 ± 17.5 Biochanin A 1.0 – 33.2 ± 1.0∗∗ –– – 12.5 78.1 ± 12.7 5.1 ± 0.1∗∗ 73.7 ± 1.1 71.3 ± 2.3∗ 99.5 ± 10.0 25.0 72.0 ± 1.0 2.5 ± 0∗∗ 71.6 ± 4.9 69.5 ± 3.2∗ 96.1 ± 23.7 Formononetin 1.0 – 75.7 ± 2.5∗∗ –– – 12.5 116.4 ± 8.2 18.1 ± 0.1∗∗ 88.7 ± 6.5 87.9 ± 5.6 92.3 ± 2.6 25.0 115.4 ± 8.8 9.4 ± 0.2∗∗ 103.4 ± 2.8 86.0 ± 1.3∗ 93.0 ± 1.5 Enzyme activities are expressed relative to control experiment. Values are mean ± S.E.M. (n = 3). Specific enzyme activity of a typical preparation of mitochondria and microsomal enzyme from H295R cells was 9.4 ± 1.1 pmol/min mg protein for P450scc, 103.2 ± 7.7 pmol/min mg protein for 3␤-HSD II, 100.5 ± 6 pmol/min mg protein for P450c17, 24.4 ± 1.5 pmol/min mg protein for P450c21 and 28.8 ± 2.6 pmol/min mg protein for P45011␤. Methods for determination of enzyme activities are given in Section 2. Statistical analysis was carried out using the actual values. Asterisks denote significant difference from control (no addition). ∗ Significant difference from control (no addition), P ≤ 0.05. ∗∗ Significant difference from control (no addition), P ≤ 0.01. S. Ohno et al. / Journal of Steroid Biochemistry & Molecular Biology 80 (2002) 355–363 361

Fig. 5. Inhibitory effect of daidzein on activities of 3␤-HSD II and P450c21. (A) Lineweaver–Burk’s plots of 3␤-HSD II activity are shown in absence and presence of various concentration of daidzein ((᭜) no addition; (᭿) 1.5 ␮M; ()3␮M). (B) Lineweaver–Burk’s plots of P450c21 are shown in absence and presence of various concentration of daidzein ((᭜) no addition; (᭿) 12.5 ␮M; ()25␮M). Each point represents the mean of three determinations. of cortisol secretion. It was found that 6-hydroxyflavone, strongly inhibit 3␤-HSD II activity at low concentrations 4 -hydroxyflavone, apigenin, daidzein, genistein, biochanin (IC50 = 0.5–2.7 ␮M). Furthermore, these isoflavones sig- A and formononetin significantly inhibit cortisol secretion nificantly inhibit P450c21 activity at relatively higher from H295R cells at concentrations of 12.5 ␮M. These com- concentration of 12.5 and/or 25 ␮M. 6-Hydroxyflavone in- pounds are flavones and isoflavones. The hydroxy group at hibited P450c17 and P45011␤ activity at comparatively position 6 of the pyran ring or the 4 position of the benzene higher concentrations. In addition, kinetic experiments re- ring of flavones seems to effect inhibition of cortisol pro- vealed that daidzein competitively inhibits 3␤-HSD II and duction. In isoflavones, a hydroxy or methoxy group in the P450c21activities at Ki values 2.9 and 33.3 ␮M, respectively. 4 position of the benzene ring effects inhibition of cortisol Mesiano et al. [21] reported that the phytoestrogens genis- production, but the glucosyl group in the 7 position on the tein and daidzein suppress ACTH or cAMP-stimulated glu- pyran ring does not. cocorticoid production by cultured human adrenocortical We investigated whether these phytoestrogen isoflavones cells. They reported that although genistein and daidzein that reduce cortisol secretion act through the estrogen re- did not alter steroidogenic enzyme expression, genistein in- ceptor. Estradiol and the synthetic estrogen, diethyl stillbe- hibited the activity of P450c21. This inhibitory effect was strol (DES), have been reported to inhibit forskolin-induced observed by exposing ACTH-stimulated fetal adrenocortical cortisol production by adrenocortical H295R cells, and cells to genistein in the presence of specific steroid precur- the inhibitory action of estradiol on steroid production by sors of cortisol synthesis. H295R cells is not affected by estrogen antagonists [19]. To assess the effect of isoflavones on steroidogenic en- Kuiper et al. [20] demonstrated that phytoestrogens, in- zyme activities, we scrutinized their direct effect on all cluding daidzein, genistein, formononetin and coumestrol, steroidogenic enzymes involved in cortisol synthesis. These compete with estradiol for binding to both estrogen receptor enzymes, cytochrome P450 monoxygenases and 3␤-HSD ␣ and ␤ and regulate estrogen-responsive gene expression. II, can be induced in H295R cells by stimulation of cAMP Therefore, we predicted that daidzein would interact with and EGF, respectively [22,23]. In the present experiments, the estrogen receptor. Our results indicate that the inhibitory mitochondrial and microsomal enzymes were prepared effects of daidzein on cortisol production from H295R cells from H295R cells that were stimulated by dibutyryl cAMP is not affected by the estrogen antagonist ICI 182,780, sug- and EGF for 24 h. Mitochondria were used to determine gesting that inhibitory effects are not mediated via estrogen P450scc and P45011␤ activities and microsomes were used receptors. to determine 3␤-HSD II, P450c17 and P450c21 activity. The inhibitory effect of cortisol production-suppressing These enzyme activities were specifically measured with isoflavones on the steroidogenic enzymes involved in corti- radioisotope-labeled substrate and NAD+ (for 3␤-HSD) or sol biosynthesis was also investigated. These enzymes are NADPH (for P450s) as the cofactor. 3␤-HSD II and four kinds of cytochrome P450, P450scc, It has been reported that isoflavones, including daidzein, P450c17, P450c21 and P45011␤. The activity of each selectively inhibit the ␥-isozyme of mammalian alcohol enzyme can be determined from mitochondria and micro- dehydrogenase (␥␥-ADH), which catalyzes oxidation of somes prepared from H295R cells because these enzymes 3␤-hydroxysteroids [24]. These isoflavones have also been and electron carriers are bound to organelle membrane. shown to inhibit 3␤-HSD from bovine adrenal glands [25]. Our data indicate that 6-hydroxyflavone, daidzein, genis- Recently, Le Bail et al. reported the effects of several phy- tein, biochanin A and formononetin significantly and toestrogens on aromatase, 3␤ and 17␤-HSD activities of 362 S. Ohno et al. / Journal of Steroid Biochemistry & Molecular Biology 80 (2002) 355–363 human placenta microsomes and the estrogenic activity of [7] L. Markiewicz, J. Garey, H. Adlercreutz, E. Gurpide, In vitro these phytoestrogens in human breast cancer cells [26]. bioassays of non-steroidal phytoestrogens, J. Steroid Biochem. Mol. The report indicates that some isoflavones preferentially Biol. 45 (1993) 399–405. ␤ ␤ [8] R.J. Miksicek, Interaction of naturally occurring non-steroidal inhibit 3 -HSD activity. 3 -HSD catalyzes the oxidation estrogens with expressed recombinant human estrogen receptor, J. and isomerization of 5-ene-3␤-hydroxy steroids (e.g. preg- Steroid Biochem. Mol. Biol. 49 (1994) 153–160. nenolone or dehydroepiandrosterone) to the corresponding [9] D.T. Zava, G. Duwe, Estrogenic and antiproliferative properties of 4-ene-3-ketosteroids (e.g. progesterone or androstenedione). genistein and other flavonoids in human breast cancer cells in vitro, ␤ Nutr. Cancer 27 (1997) 31–40. In humans, two types of 3 -HSD (I and II) clones and their [10] B.M. Collins, J.A. McLachlan, S.F. Arnold, The estrogenic and corresponding genes have been characterized, encoding de- antiestrogenic activities of phytochemicals with the human estrogen duced proteins of 371 and 372 amino acids, respectively receptor expressed in yeast, Steroids 62 (1997) 365–372. [27,28]. 3␤-HSD I enzyme is exclusively expressed in the [11] D.T. Zava, M. Blen, G. Duwe, Estrogenic activity of natural and placenta and skin, while type II is predominantly expressed synthetic estrogens in human breast cancer cells in culture, Environ. Health Perspect. 105 (1997) 637–645. in the adrenals, ovaries and testis [28,29]. Consequently, [12] G.G. Kuiper, J.G. Lemmen, B. Carlsson, J.C. Corton, S.H. Safe, 3␤-HSD II is the key enzyme in biosynthesis of not only cor- P.T. van der Saag, B. van der Burg, J.A. Gustafsson, Interaction of tisol (glucocorticoids), but also other steroid hormones such estrogenic chemicals and phytoestrogens with estrogen receptor ␤, as progestin, mineralocorticoids, androgens and estrogens. Endocrinology 139 (1998) 4252–4263. In addition, 3␤-HSD expression has been demonstrated in [13] J.C. Le Bail, F. Varnat, J.C. Nicolas, G. Habrioux, Estrogenic and antiproliferative activities on MCF-7 human breast cancer cells by the brain, where the enzyme is believed to play an essential flavonoids, Cancer Lett. 130 (1998) 209–216. role in the biosynthesis of neurosteroids [30]. We, therefore, [14] A.F. Gazdar, H.K. Oie, C.H. Shackleton, T.R. Chen, T.J. Triche, have some misgivings about the possibility that these phyto- C.E. Myers, G.P. Chrousos, M.F. Brennan, C.A. Stein, R.V. La chemicals can affect various aspects of steroidogenesis. As Rocca, Establishment and characterization of a human adrenocortical carcinoma cell line that expresses multiple pathways of steroid yet, few studies have investigated the direct effects of these biosynthesis, Cancer Res. 50 (1990) 5488–5496. isoflavone phytoestrogens in humans or other subjects. At [15] W.E. Rainey, I.M. Bird, C. Sawetawan, N.A. Hanley, J.L. McCarthy, this stage, in vivo studies for verification of this possibility E.A. McGee, R. Wester, J.I. Mason, Regulation of human adrenal are important. Within recent days, in vivo study data have carcinoma cell (NCI-H295) production of C19 steroids, J. Clin. reported that dietary soy-phytoestrogens decrease plasma Endocrinol. Metabol. 77 (1993) 731–737. [16] W.E. Rainey, I.M. Bird, J.I. Mason, The NCI-H295 cell line: a testosterone levels in adult male Sprague–Dawley rats [31]. pluripotent model for human adrenocortical studies, Mol. Cell. We are examining this report with keen interest. Endocrinol. 100 (1994) 45–50. [17] I.M. Bird, M.M. Pasquarette, W.E. Rainey, J.I. Mason, Differential control of 17␣-hydroxylase and 3␤-hydroxysteroid dehydrogenase in human adrenocortical H295R cells, J. Clin. Endocrinol. Metabiol. Acknowledgements 81 (1996) 2171–2178. [18] T. Kimura, J.H. Parcells, H.-P. Wang, Purification of adrenodoxine The authors would like to thank Prof. J. Ian Mason, reductase, adrenodoxin, and cytochrome P450 from adrenal cortex, University of Edinburgh, Royal Infirmary of Edinburgh, for in: S. Fleischer, L. Packer (Eds.), Methods in Enzymology, Vol. LII, Academic Press, New York, 1978, pp.132–142. providing the H295R cells used in this study. This work [19] J.S. Gell, J. Oh, W.E. Rainey, B.R. Carr, Effect of estradiol on was supported by Health Sciences Research grants from the DHEAS production in the human adrenocortical cell line, H295R, Ministry of Health, Labour and Welfare of Japan, and by a J. Soc. Gynecol. Invest. 5 (1998) 144–148. Grant-in-Aid for Scientific Research from the Ministry of [20] G.G. Kuiper, J.G. Lemmen, B. Carlsson, J.C. Corton, S.H. Safe, P.T. van der Saag, B. van der Burg, J.-A. Gustafsson, Interaction of Education, Culture, Sports, Science and Technology. estrogenic chemicals and phytoestrogens with estrogen receptor ␤, Endocrinology 139 (1998) 4252–4263. [21] S. Mesiano, S.L. Katz, J.Y. Lee, R.B. Jaffe, Phytoestrogens References alter adrenocortical function: genistein and daidzein suppress glucocorticoid and stimulate androgen production by cultured adrenal cortical cells, J. Clin. Endocrinol. Metabiol. 84 (1992) 2443–2448. [1] T. Colborn, F.S. von Seal, A.M. Sato, Developmental effects of + [22] I.M. Bird, J.M. Mathis, J.I. Mason, W.E. Rainey, Ca2 -regulated endocrine-disrupting chemicals in wildlife and human, Environ. expression of steroid hydroxylases in H295R human adrenocortical Health Perspect. 101 (1993) 379–384. cells, Endocrinology 136 (1995) 5677–5684. [2] K.D.R. Setchell, Naturally occurring non-steroidal estrogens of [23] J. Doi, H. Takemori, M. Ohta, Y. Nonaka, M. Okamoto, Diffe- dietary origin, in: J. McLachlan (Ed.), Estrogens in the Environment, rential regulation of 3␤-hydroxysteroid dehydrogenase type II and Elsevier, New York, 1985, pp. 69–85. 17␣-hydroxylase/lyase P450 in human adrenocortical carcinoma cells [3] H. Wang, P.A. Murphy, Isoflavone content in commercial soybean by epidermal growth factor and basic fibroblast growth factor, J. foods, J. Agric. Food Chem. 42 (1994) 1666–1673. Endocrinol. 168 (2001) 87–94. [4] H. Wang, P.A. Murphy, Isoflavone composition of variety, crop year, [24] W.M. Keung, Biochemical studies of new class of alcohol dehydro- and location, J. Agric. Food Chem. 42 (1994) 1674–1677. genase inhibitors from Radix puerariae, Alcohol Clin. Exp. Res. 17 [5] H. Adlercreutz, H. Markkanen, S. Watanabe, Plasma concentrations (1993) 1254–1260. of phyto-oestrogens in Japanese men, Lancet 342 (1993) 1209–1210. [25] C.K. Wong, W.M. Keung, Bovine adrenal 3␤-hydroxysteroid [6] K.D. Setchell, L. Zimmer-Nechemias, J. Cai, J.E. Heubi, Exposure dehydrogenase (E.C. 1.1.1.145)/5-ene-4-ene isomerase (E.C. 5.3.3.1): of infants to phyto-oestrogens from soy-based infant formula, Lancet characterization and its inhibition by isoflavones, J. Steroid Biochem. 350 (1997) 23–27. Mol. Biol. 71 (1999) 191–202. S. Ohno et al. / Journal of Steroid Biochemistry & Molecular Biology 80 (2002) 355–363 363

[26] J.-C. Le Bail, Y. Champavier, A.-J. Chulia, G. Habrioux, Effect M. Dumont, E. Dupont, E. Rhéaume, C. Martel, J. Couët, C. of phytoestrogens on aromatase, 3␤- and 17␤-hydroxysteroid Trudel, Structure and tissue-speciﬁc expression of 3␤-hydroxysteroid dehydrogenase activities and human breast cancer cells, Life Sci. 66 dehydrogenase/5–4-isomerase genes in human and rat classical (2000) 1281–1291. and peripheral steroidogenic tissues, J. Steroid Biochem. Mol. Biol. [27] V. Luu-The, Y. Lachance, C. Labrie, G. Leblanc, J.L. Thomas, R.C. 41 (1992) 421–435. Strickler, F. Labrie, Full length cDNA structure and deduced amino [30] R. Guennoun, R.J. Fiddes, M. Ggouezou, M. Lombes, E.E. acid sequence of human 3␤-hydroxy-5-ene steroid dehydrogenase, Baulieu, A key enzyme in the biosynthesis of neurosteroids, Mol. Endocrinol. 3 (1989) 1310–1312. 3␤-hydroxysteroid dehydrogenase/5–4-isomerase (3␤-HSD), is [28] E. Rhéaume, Y. Lachance, H.F. Zhao, N. Breton, M. Dumont, Y. expressed in rat brain, Brain Res. Mol. Brain Res. 30 (1995) 287– de Launoit, C. Trudel, V. Luu-The, J. Simard, F. Labrie, Structure 300. and expression of a new complementary DNA encoding the almost [31] K.S. Weber, K.D.R. Setchell, D.M. Stocco, E.D. Lephart, Dietary exclusive 3␤-hydroxysteroid dehydrogenase/5–4-isomerase in soy-phytoestrogens decrease testosterone levels and prostate weight human adrenals and gonads, Mol. Endocrinol. 5 (1991) 1147–1157. without altering LH, prostate 5␣-reductase or testicular steroidogenic [29] F. Labrie, J. Simard, V. Luu-The, G. Pelletier, A. Bélanger, acute regulatory peptide levels in adult male Sprague–Dawley rats, Y. Lachance, H.F. Zhao, C. Labrie, N. Breton, Y. de Launoit, J. Endocrinol. 170 (2001) 591–599.