{"paper_id":"0e414c77-55f5-45be-a0f7-dd48e78bab25","body_text":"709Reprod. Nutr. Dev. 45 (2005) 709–720\n© INRA, EDP Sciences, 2005\nDOI: 10.1051/rnd:2005055\nOriginal article\nThe effects of dietary phytoestrogens on aromatase activity \nin human endometrial stromal cells\nKatie M. EDMUNDSa, Alison C. HOLLOWAYa, Denis J. CRANKSHAWa, \nSanjay K. AGARWALb, Warren G. FOSTERa*\na Reproductive Biology Division, Department of Obstetrics and Gynecology, McMaster University, \nHamilton, Ontario, USA\nb Department of Reproductive Medicine, University of California, San Diego, California, USA\n(Received 14 January 2005; accepted 26 May 2005)\nAbstract – Dietary phytoestrogens have been reported to inhibit arom atase activity in placental\nmicrosomes, but the effects in the human endometrium are unknown. Aromatase, the rate-limiting\nenzyme in the conversion of androgens to estrogens, has recently been shown to be expressed in the\nendometrium of women with endometriosis and is thought to play a role in the pathophysiology of\nthis disease. Therefore,  the objective of this study was to sc reen dietary phytoestrogens for their\nability to inhibit aromatase activity in human endometrial stromal cells (ESC) and identify potential\nnovel therapeutic agents for the treatment of endometriosis. The inhibition of aromatase activity by\ndirect interaction with the dietary phytoestrogens genistein, daidzein, chrysin, and naringenin was\ntested in a cell free assay. Furthermore, test compound effects on aromatase activity in ESC cultures\nwere also examined. Genistein and daidzein were inactive in the human recombinant aromatase assay\nwhereas naringenin and chrysin inhibited aromatase activity. However, genistein (1 nM to 1 mM)\nstimulated aromatase activity in ESC whereas other phytoestrogens had no effect. Immunopositive\naromatase cells were demonstrated  in genistein-treated ESC but not  in untreated control cultures.\nTaken together, our data suggest th at genistein can increase aromat ase activity in ESC likely via\nincreased enzyme expression.\nphytoestrogens / endometriosis / aromatase / genistein / endometrium\n1. INTRODUCTION\nCytochrome P450 aromatase (P450AROM)\nis the rate limiting enzyme that catalyzes the\nconversion of androstenedione and testo-\nsterone to estrone and 17 β-estradiol,\nrespectively. While the ovaries are the pri-\nmary source of estrogen in the body, local\nproduction of estrogen by other tissues has\nalso been demonstrated in estrogen depend-\nent diseases such as breast cancer [1, 2] and\nendometriosis [3–5]. Endometriosis is a\ncommon gynecologic disorder that is char-\nacterized by the pres ence of endometrial\nglands and stroma outside of the uterine\ncavity. Endometriosis is an estrogen depend-\nent disease [7, 8] that affects approximately\n14% of all women of reproductive age, and\n30–50% of infertile women [9]. Local pro-\nduction of estrogens by ectopic endometrial\n* Corresponding author: fosterw@mcmaster.ca\nArticle published by EDP Sciences and available at \nhttp://www.edpsciences.org/rnd\nor \nhttp://dx.doi.org/10.1051/rnd:2005055\n\n\n710 K.M. Edmunds et al.\nimplants in women with endometriosis may\nexplain treatment failures and the persist-\nence of recalcitrant en dometriosis in post-\nmenopausal women [6]. Therefore targeted\ninhibition of local estrogen production in\nendometriotic lesions by inhibition of aro-\nmatase activity may have a place in the\nmanagement of this disease. As endometri-\nosis is the leading cause of hospitalization\nfor gynecologic surgery [10], thus novel,\nsafe and effective treatment options are\nurgently needed.  \nPhytoestrogens are a class of plant estro-\ngens that include isoflavones, flavones, fla-\nvonones and several mycotoxins such as\ncoumestrol and zeralanone. Phytoestrogens\nare thought to have health benefits such as\nproviding protection against breast cancer\ndevelopment [11, 12] and are potentially\nuseful in the management of menopausal\nsymptoms [13, 14]. Dietary factors such as\nphytoestrogens have been shown to inhibit\naromatase activity [15, 16] without altering\nplasma estrogen concentrations [17]. There-\nfore, therapeutic use of phytoestrogens may\nbe of benefit to women with endometriosis.  \nSoy-based foods have high phytoestro-\ngen content of which genistein is the dom-\ninant isoflavone [18]. Chrysin, a flavone\nfound in the plant Passiflora coerula and\nnaringenin, a flavo none found in citrus\nfruits, have been shown to inhibit aromatase\nactivity in hepatocytes and placental micro-\nsomes in vitro  [19, 20]. However, the\neffects of these comp ounds on aromatase\nexpression are unknown. Furthermore, aro-\nmatase expression is regulated via different\npromoter regions in a tissue specific man-\nner [21] and thus the effects of phytoestro-\ngens on endometrial aromatase expression\nand activity are also unknown. Therefore,\nthe objective of this study was to screen sev-\neral phytoestrogens for their ability to\ndirectly inhibit arom atase activity and to\ndetermine the effect of dietary phytoestro-\ngens on aromatase expression and activity\nin human endometrial stromal cells. Genis-\ntein and daidzein, the dominant phytoestro-\ngens in the diet, together with chrysin and\nnaringenin, two phytoestrogens previously\nshown to inhibit aromatase activity were\nselected as the test compounds for this\nstudy. We hypothesized that phytoestro-\ngens will inhibit aromatase activity in\nendometrial stromal cell cultures and thus\npotentially provide a novel therapeutic\noption that is both natural and effective in\nthe management of endometriosis. \n2. MATERIALS AND METHODS\n2.1. Cell-free assay\nThe ability of the test compounds to\ninteract directly with the enzyme to alter\naromatase activity was investigated in a cell\nfree assay by modification of the fluores-\ncence assay described previously [22],\nusing human recombinant aromatase\nexpressed in insect cell microsomes (CYP19\nsuprasomes BD Gentest Biosciences,\nWoburn, USA) and 0.25 µM dibenzylfuo-\nrescein (BD Gentest Biosciences, Woburn,\nUSA) as the substrate. The ability of test\ncompounds (1 pM–100 µM in 0.1 M potas-\nsium phosphate buffer pH 7.4) to inhibit\naromatase enzyme activ ity (0.4 pmol aro-\nmatase/well was examined by incubation in\nthe presence of cofactors (40 µM NADP,\n100 µM Glucose-6-phosphate, 100 µM\nMgCl\n2) and DMSO (1%). Assays were per-\nformed in a 96-well black walled culture\nplate (Becton Dickinso n, Franklin Lakes\nUSA) in a total volume of 202 µL. Reac-\ntions were started by addition of 50 µL of\nprewarmed (37 °C) enzyme to the pre-\nwarmed plates. Blank wells contained\n50 µL of buffer in place of the enzyme. The\nplate was incubated at 37 ºC for 1 h and pre-\nliminary experiments showed that enzyme\nactivity was linear up to 90 min. The reac-\ntion was stopped by the addition of 75 µL\nof 2 M NaOH to each well. Fluorescence\nwas measured using a PerkinElmer HTS\n7000 Bio Assay Reader  at an excitation\nwavelength = 485 nm and emission = 535 nm.\nFluorescein (Sigma Aldrich, Oakville,\nCanada) was used as the standard.\n\nGenistein induced increased aromatase activity 711\nNon-linear least-squares regression analy-\nsis was used to fit inhibition curves to the\nequation:\n \nwhere E\nmax and Emin are the maximum and\nminimum effects of the test compound,\nrespectively. pIC\n50 is the negative log of the\nmolar concentration of the compound that\nproduces 50% inhibition of enzyme and log\nC is the molar concentration of the com-\npound that produces the effect E. IC\n50 val-\nues were converted to Kis using the Cheng\nPrusoff equation [23]: \nKi = IC50/(1 + S/Km))\nwhere S is the substrate concentration and\nKm is the Michaelis constant for the\nenzyme. The K m and maximum velocity\n(Vmax) of the enzyme reactions were deter-\nmined under the same conditions as\ndescribed for inhibition experiments except\nthat the substrate concentration varied\nbetween 0 and 0.4 µM. Data for these exper-\niments were fit by non-linear least-squares\nregression to:\nV = (Vmax × S)/(Km + S)\nwhere V is the reaction velocity at substrate\nconcentration S.\n2.2. Endometrial stromal cell culture\nEndometrial biopsies were obtained\nfrom eighteen women aged 27–44 (mean\n(± SD) of 38.3 ± 6.0 years) undergoing\nbenign gynecologic surgery at McMaster\nUniversity Medical Centre. Informed con-\nsent was obtained from each patient by a\nresearch nurse and all procedures were car-\nried out in accordance with approval of the\nMcMaster University  Research Ethics\nBoard. Among the eighteen patients included\nin this study, eleven had a laparoscopic\ndiagnosis of endometriosis and seven did\nnot have any evidence of pelvic endometri-\nosis. None of the study subjects had\nreceived endocrine therapy in the previous\nsix months before surgery. Endometrial tis-\nsue (1–2 g) obtained at hysterectomy was\nrinsed in Hanks’ balanced salt solution (HBSS)\ncontaining 200 units·mL\n–1 penicillin,\n0.2 mg·mL–1 streptomycin and 0.5µg·mL–1\namphotericin B (Sigma Aldrich, Oakville\nCanada) to remove blood and debris.  Sep-\naration of the endometrial stromal cells was\nperformed as previously described [24].\nBriefly, the tissue was minced into 1 mm\n3\nfragments and digested for 2.5 h at 37 °C\nin medium containing collagenase type IA\n(2 mg·mL–1, Sigma-Aldrich, Oakville,\nCanada). After digestion, the remaining tis-\nsue fragments were mechanically dispersed\nand the dispersed cells were filtered\nthrough a 100 µm and subsequently a\n40 µm cell strainer (Becton Dickson, Franklin\nLakes, USA). Centrifugation (10 min, 725 ×\ng)  was used to pellet the cells after which\ntime they were resuspended in 3 mL of plat-\ning media [DMEM:F12, 4% FBS, 1% ITS+\nand 1% antibiotic antimycotic solution\n(100 units·mL\n–1 penicillin, 0.1 mg·mL –1\nstreptomycin and 0.25 g·mL –1  amphoter-\nicin B (Sigma Aldrich, Oakville, Canada)].\nRed blood cells were removed by layering\nthe cell suspension over 3 mL of Ficoll-\nPaque PLUS (Amersham Biosciences,\nUppsala, Sweden) in a sterile 15 mL poly-\npropylene tube. The solution was centri-\nfuged for 10 min at 400 × g. The media/\nFicoll interface layer containing the stromal\ncells was plated into 48 well Falcon tissue\nculture plates (Becton  Dickson, Franklin\nLakes, USA) at a density of 200 000 cells/\nwell/0.5 mL. Media was changed after 48 h\nand the cells were treated after 96 h in cul-\nture, when the cells were near confluence.\nPurity of the cell preparation was confirmed\nby immunostaining for vimentin (mesen-\nchymal cell marker) and cytokeratin (epi-\nthelial cell marker) as described below.  \n2.3. Cell treatment and aromatase \nactivity assay\nCells were washed twice in HBSS and\nincubated for a minimum of 1 h in serum-free\nDMEM-F12 containing 100 units·mL\n–1\npenicillin, 0.1 mg·mL –1 streptomycin and\nEE min  +=\nEmax Emin–() /1 1 0\npIC 50 Clog––\n+()\n\n712 K.M. Edmunds et al.\n0.25µg·mL–1 amphotericin B (Sigma Aldrich,\nOakville, Canada) prior to treatment for 24 h\nwith increasing log concentrations (10 –9–\n10–4 M) of genistein, daidzein, naringenin\nor chrysin (Sigma Aldrich, Oakville, Can-\nada) diluted in serum free media. To exam-\nine the role of estrogen receptor mediated\neffects, the cells were also treated with gen-\nistein in the presence of a non-selective\nestrogen receptor antagonist (ICI 182,780;\nTocris, Ellisville, USA). After 24 h, the\ntreatment media was removed and replaced\nwith 500 µL of [1 β-\n3H]-androstenedione\n[2.5 µCi·mL–1] (Perkin Elmer, Boston, USA)\nin DMEM-F12 (containing 100 units·mL–1\npenicillin, 0.1 mg·mL –1 streptomycin and\n0.25 µg·mL–1 amphotericin B) for 4 h at\n37 °C.\nAromatase activity was assayed using a\nradiometric technique that quantifies the\nincorporation of tritium from [1 β-3H]-\nandrostenedione into 3H-labeled water as\npreviously described [25]. Briefly, aro-\nmatase activity was determined by transfer-\nring 300 µL of the incubation medium to\nglass tubes, adding 300 µL of dextran\ncoated activated charcoal (250 mg·mL –1,\nBD Biosciences, Oakville, Canada) to each\ntube and incubating for 2 h at 4° C. The\nsamples were then centrifuged (15 min,\n2500 × g) and the tritiated water content was\ndetermined by counting the supernatant in\n5 mL of scintillation fluid (Aqueous Count-\ning Scintillant, Amersham, England) in a\nliquid scintillation counter. To control for\nvariation in the number of cells in each well,\nthe aromatase activity was normalized to\nthe cell protein content in each well as\ndetermined by the Bradford method. Due to\nvariation in basal aromatase activity between\npatients, normalized aromatase activity was\nconverted to a percentage of the control\nlevel for each culture. The aromatase assay\nis based on the release of tritiated water and\nthe specificity of the assay was determined\nby co-incubation with 4-hydroxyandros-\ntendione an irreversible inhibitor of the cat-\nalytic activity of aromatase [26] to block the\nformation of tritiated water.\n2.4. Immunocytochemistry\nCells were seeded into 8 well Lab-Tek\nchamber slides (BD Biosciences, Oakville,\nCanada) at a density of 200 000 cells/well/\n0.5 mL. Media was changed after 48 h and\nthe cells were treated with genistein (10\n–6)\nafter 96 h in culture. After 24 h of treatment,\nthe cells were fixed in 10% neutral buffered\nformalin, washed in PBS, and endogenous\nperoxidase activity was quenched by incu-\nbating the cells in 3% hydrogen peroxide (in\nmethanol) for 5 min. The cells were washed\nin PBS, incubated with the primary anti-\nbodies (Dako Diagnostics, Mississauga,\nCanada) for cytokeratin (1:50), and vimen-\ntin (1:50) for 1 h at room temperature and\nimmunostaining was identified using\nEnVision (Dako Diagnostics, Mississauga,\nCanada) with diaminobenzidine (Sigma-\nAldrich, Oakville, Canada) as the chro-\nmogen. The cells were counterstained with\nCarazzi hematoxylin. For negative controls,\nthe cells were incubated with non-immune\nserum in place of the primary antibodies. To\nstain for the presence of aromatase in the\ngenistein treated cultures and untreated\ncontrols, immunohistochemistry was per-\nformed on the chamberslides using a pri-\nmary monoclonal mouse antibody against\nhuman aromatase (1:50 Serotec, Raleigh,\nUSA). Immunostaining was identified with\nthe avidin-biotin-peroxidase technique using\nthe Vectastain kit (Vector Laboratories,\nBurlington, Canada) with diaminobenza-\ndine as the chromogen and Carazzi hema-\ntoxylin as a counter stain.\n2.5. Statistical analyses\nData were analyzed for equal variance\nand normal distribution. An effect of treat-\nment on ESC aromatase activity was tested\nusing a one-way analysis of variance\n(ANOVA) and differences between doses\nwere determined using the Tukey multiple\ncomparison method. A p value < 0.05 was\nconsidered to be statistically significant for\nall procedures used.\n\nGenistein induced increased aromatase activity 713\n3. RESULTS\n3.1. Recombinant human aromatase \nactivity\nAromatase activity in the presence of\nincreasing substrate yielded a Km of 0.26µM\n(pKm = 6.6 ± 0.2) and a Vmax of 2.2 ± 1 pmol\nfluorescein released per mol enzyme per\nminute (Fig. 1). Naringenin (Ki = 0.3 µM)\nand chrysin (Ki = 1 µM) were potent inhib-\nitors of recombinan t human aromatase\nwhereas genistein and daidzein were weak\n(Ki > 50 µM) inhibitors (Fig. 2).\n3.2. Aromatase activity after \nphytoestrogen treatment \nof endometrial stromal cells\nImmunocytochemical staining for cells\nof mesenchymal origin and epithelial cells\nillustrated that our cultures consisted of\nFigure 1. Michaelis-Menten plot of the diben-\nzyfluorescein deakylase activity of recombinant\nhuman aromatase deter mined as described in\nmaterials and methods. Points represent means\nand standard errors of triplicates within a single\nexperiment. K\nm and Vmax values from this expe-\nriment were 0.22 µM and 1.4 pmol/pmol/min,\nrespectively.\nFigure 2. Human recombinant aromatase activity as indicated by fluorimetrically quantified DBF\ndealkylase after treatment with naringenin, chrysin, genistein and dadizein. Each data point is the\nmean (± SEM) from three separated experiments. Naringenin and chrysin were effective inhibitors\nof the enzyme with a Ki = 0.3 and 1.0 µM, respectively, while genistein and daidzein were ineffective\nas shown by a Ki > 50 µM.\n\n714 K.M. Edmunds et al.\ngreater than 99% endometrial stromal cells\n(data not shown).\nPhytoestrogen treatment did not attenu-\nate aromatase activity in ESC from women\nwith endometriosis (n = 11) at any concen-\ntration tested (Fig. 3). However, genistein\n(10\n–9–10–6 M) treatment of ESC from\nwomen without endometriosis (n = 7) resulted\nin a significant increase in aromatase activ-\nity (P < 0.05) to approximately 150% above\nthe activity observed in untreated ESC from\nthe same patient (Fig. 4), whereas daidzein,\nnaringenin and chrysin treatment had no\neffect. Furthermore, the genistein induced\nincrease in aromatase activity was not atten-\nuated by co-treatment with the estrogen\nreceptor antagonist ICI 182,780 ( P > 0.1,\nFig. 5). \n3.3. Immunocytochemistry\nImmunopositive aromatase staining was\nevident as a diffuse brown cytoplasmic pre-\ncipitate (Fig. 6) that was absent in control\ncultures where the primary antibody was\nsubstituted with non-immune serum. Immu-\nnopositive staining was focally present in\nsome but not all genistein (10\n–6 M) treated\nESC from eutopic endometrium of women\nwithout endometriosis. Moreover, no immu-\nnoreactive aromatase staining was visible in\nthe untreated ESC taken from the same\npatient. \n4. DISCUSSION\nThe objective of the current study was to\nscreen dietary phytoestrogens for their abil-\nity to inhibit human recombinant aromatase\nactivity and to determine the effect of die-\ntary phytoestrogens on endometrial stromal\ncell aromatase activity in culture. Although\nnaringenin and chrysin inhibited aromatase\nin our cell-free assay, they were ineffective\nin endometrial stromal cell cultures from\nFigure 3. Aromatase activity was unchanged in genistein treated endometrial stromal cell cultures\nfrom women with endometriosis ( n = 11). The control bar represen ts the aromatase activity from\nthe vehicle treated cells from each of the patients and the data bars represent the aromatase activity\nof the cells following treatment with genistein represented as percent of control. The control value\nhas arbitrarily been set to 100% and data are presented as the mean ± SEM. \n\nGenistein induced increased aromatase activity 715\nFigure 4. The effects of genistein treatment for 24 h on aromatase activity in endometrial stromal\ncells obtained from eutopic endometrium of women without endometriosis (n = 7). The control bar\nrepresents the aromatase activity from the untreated cells from each of the patients and the data bars\nrepresent the aromatase activity of the cells following treatment with genistein represented as percent\nof control. The results are the mean (± SEM) fro m seven different cultures. Values with different\nsuperscripts are significantly (P < 0.05) different.\nFigure 5. The effects of 10\n–6 M genistein (GEN) alone and in combination with 10–6 M ICI 182 780\n(ICI), and non-selective estrogen receptor antagonist, on aromatase activity obtained from the euto-\npic endometrium of women without endometriosis (n = 3). The results are the mean (± SEM) from\nthree different cultures. Means identified with a different letter were si gnificantly different ( P =\n0.008).\n\n716 K.M. Edmunds et al.\nwomen with and without endometriosis and\nthus are unlikely to have any potential ther-\napeutic benefit in the management of\nendometriosis. In contrast, genistein, the\ndominant isoflavone found in soy-based\nfoods, was inactive in the cell-free assay but\nto our surprise increased aromatase activity\nin endometrial stromal cells of women\nwithout endometriosis. These data suggest\nthat the observed effects of genistein are not\nmediated through direct effects of genistein\non enzyme activity but indirectly via\nenhanced aromatase expression in endome-\ntrial stromal cells or via intermediates on\naromatase activity. This point is supported\nby evidence of immunocytochemical stain-\ning for aromatase in genistein treated but\nnot untreated cells. In our study, the con-\ncentrations of genistein that were used to\ntreat the human endometrial stromal cells\n(1 nM to 10 mM) correspond to the serum\nconcentrations of both Asian and Cauca-\nsian women who are consuming soy-based\nfoods [27–29] and thus are considered to be\nphysiologically relevant. Taken together,\nour results suggest that while phytoestro-\ngens may have health benefits such as the\nproposed protection against breast cancer\ndevelopment [11, 12], genistein is unlikely\nto have any therapeutic value in the man-\nagement of endometriosis and more impor-\ntantly may increase aromatase activity in\nthe endometrium and thus could be an\nimportant factor in the pathobiology of this\nenigmatic disease.\nIn the present study, aromatase was not\ndetected by immunohistochemistry in con-\ntrol cultures of endometrial stromal cells\nfrom women without endometriosis. In\naddition, aromatase activity of vehicle\ntreated endometrial cells was at background\nlevels for the assay and thus supports the\nview that aromatase is either absent or\ninhibited in the endometrium from women\nwithout endometriosis. Our findings are in\nagreement with prior studies in which aro-\nmatase cytochrome P450 has been reported\nto be expressed in the endometrium of\nwomen with endometriosis but is either\nabsent [4], or expressed at low levels in the\nendometrium of women without endome-\ntriosis [30]. Therefore, the patients in the\ncurrent study were grouped into two cate-\ngories: endometriotic and non-endometri-\notic. None of the phytoestrogens tested\ninhibited aromatase activity of the ESC\nfrom women with endometriosis. However,\nin the current study, genistein-treatment of\nFigure 6. Immunocytochemical staining for aromatase in untreated cells (A) and cells treated with\n10–6 M genistein (B) reveals positive staining in the treated cells (arrows).\n\nGenistein induced increased aromatase activity 717\nESC from women without endometriosis\ninduced an increase in aromatase activity to\n150% of the untreated controls similar to\nthe findings using adrenocortical carci-\nnoma cell lines treated with herbicides [31,\n32]. Furthermore, our results are harmoni-\nous with the previous finding that genistein\n(30 µM) increased aromatase activity 3 fold\nin the H295R human adrenocortical carci-\nnoma cell line [33]. Hence, genistein treat-\nment-induced changes in aromatase activity\ncould lead to increased local levels of estro-\ngens in the endometrium. However, the\nfunctional significance of genistein induced\nchanges in aromatase activity is unknown.\nA previous study has demonstrated that\ngenistein treatment increased cell prolifer-\nation and was weak ly estrogenic in\nendometrial stromal cell and Ishikawa cell\ncultures [34]. However, genistein treatment\nantagonized the effects of estradiol in these\ncultures suggesting that genistein is a com-\npetitive antagonist of estradiol. Therefore,\na genistein induced increase in aromatase\nactivity and local estrogen production in the\nendometrium could be relevant in hypoes-\ntrogenic states such as menopause. While\ngenistein treatment was without effect on\nthe endometrium of macaque monkeys\nwith surgically induced menopause [35],\nour proposal is supported by the observa-\ntion that endometrial hyperplasia was sig-\nnificantly more prevalent in postmenopausal\nwomen receiving soy tablets vs. a reference\ngroup that received a placebo [36]. Moreo-\nver, a recent study [37] has also shown that\nhigh dose phytoestrogens can reverse the\nantiestrogenic effects of clomiphene citrate\non the endometrium. Hence, we propose\nthat the effects of soy isoflavones, including\ngenistein on the endometrium is complex\nand requires further study.\nThe mechanism through which genistein\ntreatment increased aromatase activity in\nthe endometrium remains unknown. Although\nestradiol has been shown to increase aro-\nmatase activity in ESC cultures [38], sev-\neral distinct lines of  evidence lead us to\nsuggest that genistein is not acting through\nan estrogen receptor mediated pathway in\nour cultures to increase aromatase activity.\nGenistein is a preferential estrogen receptor\n(ER)-β agonist and has been shown to have\nestrogenic actions in a variety of tissues in\nthe rat [39, 40]. However, ER-α , not ER-β\nis the dominant ER sub-type expressed in\nthe endometrium [41]. Furthermore, it is\nunlikely that genistein causes stimulation\nof aromatase in the endometrium by acting\nthrough a functional estrogen receptor path-\nway because we have shown that the stim-\nulation of aromatase in ESC by genistein is\nnot attenuated by co-t reatment of the cells\nwith ICI 182,780 which is a non-selective\nestrogen receptor antagonist. We therefore\npropose that it is unlikely that genistein\nstimulates aromatase activity in endome-\ntrial stromal cell cultures by acting directly\nvia the ER. Alternatively, we propose that\ngenistein can stimul ate aromatase activity\nin the endometrium through inhibition of\nphosphodiesterase activity and result in\nincreased levels of cAMP. Support for this\nproposal comes from evidence that genis-\ntein inhibits cAMP-phosphodiesterase\nactivity in a variety of cell types [42–44]. In\naddition, aromatase expression in the\nendometrium is regulated through cAMP-\ninduced promoter II [45] and cAMP-treat-\nment has previously been shown to result in\na 26–60 fold increase in endometrial aro-\nmatase activity [46]. \nSoy products are widely believed by the\npublic to provide health benefits. The Food\nand Drug Administration has released an\napproval for foods that contain at least\n6.25 g of soy protein/serving to contain a\ncardiovascular health claim (November 10,\n1999; No. 279) and this has led to a plethora\nof soy-based and fortified foods as well as\nsoy supplements to emerge on to the market\n[47]. Phytoestrogens are efficiently absorbed\nafter ingestion and their bioavailability is\nhigh enough to have biological effects [48].\nFurthermore, contemporary studies reveal\nthat non-Asian women are ingesting increas-\ning amounts of phytoestrogens in their diet\nas part of a trend towards a healthier lifestyle\n[49, 50]. Despite potential health benefits\nfor women of some age groups, we speculate\n\n718 K.M. Edmunds et al.\nthat genistein consumption by women of\nreproductive age may have associated health\nrisks. Moreover, epidemiological evidence\ndemonstrates that Oriental women have a\nhigher incidence of  endometriosis than\nCaucasian women suggesting a link between\nendometriosis and dietary phytoestrogens,\nas Asian diets are high in soy isoflavones\n[51, 52]. Hence, consumption of soy prod-\nucts by women of reproductive age may not\nbe without conseque nce for endometrial\naromatase activity and potentially endome-\ntriosis.\nIn summary, the resu lts of this study\ndemonstrate that dietary compounds, such\nas genistein which is present in foods\nincluding soy milk and tofu that the general\npublic views as healthy alternatives to tra-\nditional foods in the North American diet,\ncan increase the local production of estro-\ngen in the ESC. Genistein-induced changes\nin endometrial aromatase activity may have\ndetrimental effects which could lead to\nincreased risk for estrogen-dependent dis-\neases which involve the dysregulation of\naromatase such as endometriosis, adenom-\nyosis and uterine leiomyomas.  \nACKNOWLEDGEMENTS\nThe authors gratefully acknowledge the tech-\nnical assistance of Ma ry Louise Beecroft, RN\nScott Phillips and Vas ko (Bill) Georgievski.\nThis study was supported by funding from the\nCanadian Network of  Toxicology Centres\n(WGF) and the Canadian Institutes of Health\nResearch (MOP 62681; WGF and ACH).\nREFERENCES\n[1] Maggiolini M, Carpino A, Bonofiglio D,\nPezzi V, Rago V, Marsico S, Picard D, Ando\nS. The direct proliferative stimulus of dehy-\ndroepiandrosterone on MCF7 breast cancer\ncells is potentiated by overexpression of aro-\nmatase. Mol Cell Endocrinol 2001, 184: 163–\n171.\n[2] Sasano H, Harada N. Intratumoral aromatase\nin human breast, endometrial, and ovarian\nmalignancies. Endocr Rev 1998, 19: 593–607.\n[3] Noble LS, Simpson ER, Johns A, Bulun SE.\nAromatase expression in endometriosis. J\nClin Endocrinol Metab 1996, 81: 174–179.\n[4] Kitawaki J, Noguchi T, Amatsu T, Maeda K,\nTsukamoto K, Yamamoto T, Fushiki S,\nOsawa Y, Honjo H. Expression of aromatase\ncytochrome P450 protein and messenger\nribonucleic acid in human endometriotic and\nadenomyotic tissues but not in normal\nendometrium. Biol Reprod 1997, 57: 514–519.\n[5] Kitawaki J, Kusuki I, Koshiba H, Tsukamoto\nK, Fushiki S, Honjo H. Detection of aromatase\ncytochrome P-450 in endometrial biopsy\nspecimens as a diagnostic test for endometri-\nosis. Fertil Steril 1999, 72: 1100–1106.\n[6] Takayama K, Zeitoun K, Gunby RT, Sasano\nH, Carr BR, Bulun SE. Treatment of severe\npost menopausal endometriosis with an aro-\nmatase inhibitor. Fertil Steril 1998, 69: 709–713.\n[7] Dizerega GS, Barber DL, Hodgen GD.\nEndometriosis: role of ovarian steroids in ini-\ntiation, maintenance, and suppression. Fertil\nSteril 1980, 33: 649–653.\n[8] Kitawaki J, Kado N, Ishihara H, Koshiba H,\nKitaoka Y, Honjo H. Endometriosis: the\npathophysiology as an estrogen-dependent\ndisease. J Steroid Biochem Mol Biol 2003, 83:\n149–155.\n[9] Rice VA. Conventional medical therapies for\nendometriosis. Ann N Y Acad Sci 2002, 955:\n343–352.\n[10] Ajossa S, Mais V, Guerrierero S, Paoletti AM,\nCaffiero A, Murgia C, Melis GB. The preva-\nlence of endometriosis in premenopausal\nwomen undergoing gynecological surgery.\nClin Exp Obstet Gynecol 1994, 21: 195–197. \n[11] Hilakivi-Clarke L, Ch o E, De Assis S, Olivo\nS, Ealley E, Bouker KB, Welch JN, Kan G,\nClarke R, Cabanes A. Maternal and prepuber-\ntal diet, mammary development and breast\ncancer risk. J Nutr 2001, 131: 154S–157S.\n[12] Katadre M, Osborne M, Telang NT. Soy iso-\nflavone genistein modulates cell cycle pro-\ngression and induces apoptosis in HER-2/neu\noncogene expressing human breast epithelial\ncells. Int J Oncol 2002, 21: 809–815. \n[13] Hale G, Bievre M, Hughes C. Exploring the\nrole of progestins and phytoestrogens in men-\nopause. Integr Med. 1999, 2: 133–141.\n[14] Vincent A, Fitzpatrick LA. Soy isoflavones:\nare they useful in menopause? Mayo Clin Proc\n2000, 75: 1174–1184. \n[15] Campbell DR, Kurzer  MS. Flavonoid inhibi-\ntion of aromatase enzyme activity in human\npreadipocytes. J Steroid Biochem Mol Biol\n1993, 46: 381–388.\n\nGenistein induced increased aromatase activity 719\n[16] Adlercreutz H, Bannwart C, Wahala K,\nMakela T, Brunow G, Hase T, Arosemena PJ,\nKellis Jr. JT, Vickery LE. Inhibition of human\naromatase by mammalian lignans and isofla-\nvonoid phytoestrogens. J Steroid Biochem\nMol Biol 1993, 44: 147–153.\n[17] Celec P, Ostatnikova D, Caganova M,\nZuchova S, Hodosy J, Putz Z, Bernadic M,\nKudela M. Endocrine and cognitive effects of\nshort-time soybean consumption in women.\nGynecol Obstet Invest 2004, 59: 62–66.\n[18] Tham DM, Gardner CD, Haskell WL. Poten-\ntial health benefits of dietary phytoestrogens:\na review of the clinical, epidemiological, and\nmechanistic evidence. J Clin Endocrinol\nMetab 1998, 83: 2223–2235.\n[19] Kao Y, Zhou C, Sherman M, Laughton CA,\nChen S. Molecular basis of the inhibition of\nhuman aromatase (estrogen synthetase) by\nflavone and isoflavones phytoestrogens: a\nsite-directed mutagenesis study. Environ\nHealth Perspect 1998, 106: 85–92.\n[20] Jeong HJ, Shin YG, Kim IH, Pezzuto JM.\nInhibition of aromatase activity by flavonoids.\nArch Pharm Res 1999, 22: 309–312.\n[21] Simpson ER, Zhao Y, Agarwal VR, Michael\nMD, Bulun SE, Hinshelwood MM, Graham-\nLorence S, Sun T, Fisher CR, Qin K, Mendel-\nson CR. Aromatase Expression in Health and\nDisease. Recent Prog Horm Res 1997, 52:\n185–213.\n[22] Stresser DM, Turner SD, McNamara J,\nStocker P, Miller VP, Crespi CL, Patten CJ.\nA high-throughput screen to identify inhibi-\ntors of aromatase (C YP19). Anal Biochem\n2000, 284: 427–430.\n[23] Cheng Y, Prusoff WH. Relationship between\nthe inhibition constant (K\nI) and the concentra-\ntion of inhibitor which causes 50 per cent inhi-\nbition (I50) of an enzymatic reaction. Biochem\nPharmacol 1973, 22: 3099–3108.\n[24] Arnold JT, Kaufman DG, Seppälä M, Lessey\nBA. Endometrial stromal cells regulate epi-\nthelial cell growth in vitro: a new co-culture\nmodel. Hum Reprod 2001, 16: 836–845.\n[25] Garzo VG, Dorrington JH. Aromatase activity\nin human granulosa cells during follicular\ndevelopment and the modulation by follicle-\nstimulating hormone and insulin. Am J Obstet\nGynecol 1984, 148: 657–662.\n[26] Brodie AM, Shwarzel WC, Shaikh AA,\nBrodie HJ. The effect of an aromatase inhib-\nitor, 4-hydroxy-4-androstene-3,17-dione, on\nestrogen-dependent processes in reproduction\nand breast cancer. Endocrinology 1977, 100:\n1684–1695.\n[27] Valentin-Blasini L, Blount BC, Caudill SP,\nNeedham LL. Urinary and serum concentra-\ntions of seven phytoestrogens in a human ref-\nerence population subset. J Exp Anal Environ\nEpidemiol 2003, 13: 276–282.\n[28] Fanti P, Stephenson TJ, Kaariainen IM,\nRezkalla B, Tsukam oto Y, Morishita T,\nNomura M, Kitiyakara C, Custer LJ, Franke\nAA. Serum isoflavones and soya food intake\nin Japanese, Thai and American end-stage\nrenal disease patients on chronic haemodialy-\nsis. Nephrol Dial Transplant 2003, 18: 1862–\n1868.\n[29] Grace PB, Taylor JI, Botting NP, Fryatt T,\nOldfield MF, Al-Mah arik N, Bingham SA.\nQuantification of isoflavones and lignans in\nserum using isotope di lution liquid chroma-\ntography/tandem mass spectrometry. Rapid\nCommun Mass Spectrom 2003, 17: 1350–1357.\n[30] Taga S, Yoshida N, Sekiba K. Distribution\nand cyclic change of aromatase cytochrome P-\n450 activity in human uteri. J Steroid Biochem\nMol Biol 1990, 37: 741–745.\n[31] Sanderson JT, Seinen W, Giesy JP, van den\nBerg M. 2-chloro-s-triazine herbicides induce\naromatase (CYP19) activity in H295R human\nadrenocortical carcinoma cells: A novel\nmechanism for estrogenicity. Toxicol Sci\n2000, 54: 121–127.\n[32] Sanderson JT, Boerma J, Lansbergen GW,\nden Berg M. Induction and inhibition of aro-\nmatase (CYP19) activity by various classes of\npesticides in H295R human adrenocortical\ncarcinoma cells. Toxicol Appl Pharmacol\n2002, 182: 44–54.\n[33] Sanderson JT, Hordijk J, Denison MS,\nSpringsteel MF, Nantz MH, van den Berg M.\nInduction and inhibition of aromatase\n(CYP19) activity by natural and synthetic fla-\nvonoid compounds in H295R human adreno-\ncortical carcinoma cells. Toxicol Sci 2004, 82:\n70–79.\n[34] Kayisli UA, Aksu CAH, Berkkanoglu M,\nArici A. Estrogenici ty of isoflavones on\nhuman endometrial stromal and glandular\ncells. J Clin Endocrinol Metab 2002, 87:\n5539–5544.\n[35] Foth D, Cline JM. Effects of mammalian and\nplant estrogens on mammary glands and uteri\nof macaques. Am J Clin Nutr 1998, 68:\n1413S–1417S.\n[36] Unfer V, Casini ML, Costabile L, Mignosa M,\nGerli S, Di Renzo GC. Endometrial effects of\nlong-term treatment with phytoestrogens a\nrandomized, double-blind,  placebo-control-\nled study. Fertil Steril 2004, 82: 145–148.\n[37] Unfer V, Casini ML, Costabile L, Mignosa M,\nGerli S, Di Renzo GC. High dose of phytoes-\ntrogens can reverse the antiestrogenic effects\nof clomiphene citrate on the endometrium in\n\n720 K.M. Edmunds et al.\nTo access this journal online: \nwww.edpsciences.org\npatients undergoing intrauterine insemina-\ntion: A randomized trial. J Soc Gynecol Inves-\ntig 2004, 11: 322–328.\n[38] Bulun SE, Yang S, Fang Z, Gurates B, Tamura\nM, Sebastian S. Estrogen production and\nmetabolism in endometriosis. Ann N Y Acad\nSci 2002, 955: 75–85.\n[39] Santell RC, Chang YC, Nair MG, Helferich\nWG. Dietary genistein exerts estrogenic\neffects upon the uterus, mammary gland and\nthe hypothalamic/pituitary axis in rats. J Nutr\n1997, 127: 263–269.\n[40] Diel P, Smolnikar K, Schulz T, Laudenbach-\nLeschowski U, Michna H, Vollmer G. Phy-\ntoestrogens and carcinogenesis-differential\neffects of genistein in experimental models of\nnormal and malignant rat endometrium. Hum\nReprod 2001, 16: 997–1006.\n[41] Weihua Z, Saji S,  Makinen S, Cheng G,\nJenson EV, Warner M, Gustafsson J. Estrogen\nreceptor (ER) β, a modulator of ER α  in the\nuterus. PNAS 2000, 97: 5936–5941.\n[42] Kuppusamy UR, Das NP. Effects of flavo-\nnoids on cyclic AMP phosphodiesterase and\nlipid mobilization in rat adipocytes. Biochem\nPharmacol 1992, 44: 1307–1315.\n[43] Nichols MR, Morimoto BH. Tyrosine kinase-\nindependent inhibition of cyclic-AMP phos-\nphodiesterase by genistein and tyrphostin 51.\nArch Biochem Biophys 1999, 366: 224–230.\n[44] Burvall KM, Palmberg L, Larsson K. The\ntyrosine kinase inhibitor genistein increases\nbasal cAMP and poten tiates forskolin-\ninduced cAMP accumulation in A549 human\nairway epithelial cells. Mol Cell Biochem\n2002, 240: 131–133.\n[45] Sofi M, Young MJ, Papamakarios T, Simpson\nER, Clyne CD. Role of CRE-binding protein\n(CREB) in aromatase expression in breast adi-\npose. Breast Cancer Res Treat 2003, 79: 399–\n407.\n[46] Noble LS, Takayama K, Zeitoun KM, Putman\nJM, Johns DA, Hinshelwood MM, Agarwal\nVR, Zhao Y, Carr BR, Bulun SE. Prostaglan-\ndin E\n2 stimulates aromatase expression in\nendometriosis-derived stromal cells. J Clin\nEndocrinol Metab 1997, 82: 600–606.\n[47] Setchell KD, Brown NM, Desai P, Zimmer-\nNechemias L, Wolfe BE, Brashear WT, Kir-\nschner AS, Cassidy A, Heubi JE. Bioavaila-\nbility of pure isoflavones in healthy humans\nand analysis of commercial soy isoflavone\nsupplements. J Nutr 2001, 131 (Suppl):\n1362S–1375S.\n[48] Ross JA, Kasum CM. Dietary flavonoids: bio-\navailability, metabolic effects, and safety.\nAnnu Rev Nutr 2002, 22: 19–34. \n[49] Goodman MT, Wilkens LR, Hankin JH, Lyu\nLC, Wu AH, Kolonel LN. Association of soy\nand fiber consumption with the risk of\nendometrial cancer. Am J Epidemiol 1997,\n146: 294–306.\n[50] Guthrie JR, Ball M, Murkies A, Dennerstein\nL. Dietary phytoestrogen intake in mid-life\nAustralian-born women: relationship to health\nvariables. Climacteric 2000, 3: 254–261.\n[51] Miyazawa K. Incidence of endometriosis\namong Japanese women. Obstet Gynecol\n1976, 48: 407–409.\n[52] Arumugam K, Templeton AA. Endometriosis\nand race. Aust N Z J Obstet Gynaecol 1992,\n32: 164–165.","source_license":"public-domain-us","license_restricted":false}