"Disposition of soy isoflavones in normal human breast tissue1–4"
Disposition of soy isoﬂavones in normal human breast tissue1–4 Selin Bolca, Mireia Urpi-Sarda, Phillip Blondeel, Nathalie Roche, Lynn Vanhaecke, Sam Possemiers, Nawaf Al-Maharik, Nigel Botting, Denis De Keukeleire, Marc Bracke, Arne Heyerick, Claudine Manach, and Herman Depypere ABSTRACT The isoﬂavones genistein and daidzein, which are present in Background: Despite decades of research on the relation between soy-derived foods and dietary supplements, as well as equol, soy and breast cancer, questions regarding the absorption, metabo- a microbial metabolite of daidzein (11), are partial agonists of lism, and distribution of isoﬂavones in breast tissue largely remain estrogen receptors (ERs) a and b, with a higher in vitro afﬁnity unanswered. to ERb than ERa (12–14). Consequently, depending on the Objective: We evaluated the potential health effects of isoﬂavone endogenous estrogen concentrations and the target tissue, these consumption on normal breast tissue; isoﬂavone concentrations, compounds may interfere with ER signaling. Furthermore, metabolites, and biodistribution were investigated and compared several other ER-independent mechanisms of action, such as the with 17b-estradiol exposure. inhibition of steroidogenic enzyme activities, modulation of Downloaded from www.ajcn.org by guest on July 12, 2011 Design: In this dietary intervention study, healthy women were growth-factor action, down-regulation of tyrosine and other randomly allocated to a soy milk (n = 11; 16.98-mg genistein and protein kinases, and the inhibition of angiogenesis, have been 5.40-mg daidzein aglycone equivalents per dose), soy supplement suggested for isoﬂavones on the basis of experiments with ani- (n = 10; 5.27-mg genistein and 17.56-mg daidzein aglycone equiv- mal and cellular models (15, 16). However, the physiologic alents per dose), or control (n = 10) group. After a run-in period !4 d, relevance of these effects in humans is unclear. 3 doses of soy milk or soy supplements were taken daily for 5 d As recently reviewed by Messina and Wu (1) and Stubert and before an esthetic breast reduction. Blood and breast biopsies were Gerber (2), the relation between soy food, isoﬂavone intake, and collected during surgery and analyzed with liquid chromatography– breast cancer has been investigated for a few decades. Never- tandem mass spectrometry. theless, the potential health effects of soy and its isoﬂavones on Results: After soy administration, genistein and total daidzein con- breast tissue cannot be completely understood until their bio- centrations, which were expressed as aglycone equivalents, ranged availability has been fully established (17). In particular, ques- from 135.1 to 2831 nmol/L and 105.1 to 1397 nmol/L, respectively, in hydrolyzed serum and from 92.33 to 493.8 pmol/g and 22.15 to 1 770.8 pmol/g, respectively, in hydrolyzed breast tissue. The major From the Laboratory of Microbial Ecology and Technology, Faculty of Bioscience Engineering (SB and SP), the Laboratory of Chemical Analysis, metabolites identiﬁed in nonhydrolyzed samples were genistein- Faculty of Veterinary Medicine (LV), and the Laboratory of Pharmacognosy 7-O-glucuronide and daidzein-7-O-glucuronide, with an overall glu- and Phytochemistry, Faculty of Pharmaceutical Sciences (SB, DDK, and curonidation of 98%. Total isoﬂavones showed a breast adipose/ AH), Ghent University, Ghent, Belgium; the Nutrition and Food Science glandular tissue distribution of 40:60, and their mean (6SEM) de- Department, XaRTA, INSA. Pharmacy Faculty, University of Barcelona, rived 17b-estradiol equivalents toward estrogen receptor b were Barcelona, Spain (MU-S); the Departments of Plastic and Reconstructive 21 6 4-fold and 40 6 10-fold higher than the 17b-estradiol con- Surgery (PB and NR) and Uro-Gynaecology (HD) and the Laboratory of centrations in adipose (0.283 6 0.089 pmol/g, P , 0.001) and Experimental Cancer Research, Department of Experimental Cancer Re- glandular (0.246 6 0.091 pmol/g, P = 0.001) fractions, respectively. search, Radiotherapy and Nuclear Medicine (MB), Ghent University Hos- Conclusion: After intake of soy milk and soy supplements, isoﬂa- pital, Ghent, Belgium; the School of Chemistry, University of St Andrews, vones reach exposure levels in breast tissue at which potential St Andrews, United Kingdom (NA-M and NB); and Institut National de la ´ Recherche Agronomique, UMR 1019, Unite Nutrition Humaine, Centre health effects may occur. Am J Clin Nutr 2010;91:976–84. ` Clermont-Theix, St Genes Champanelle, France (CM). 2 The study and analyses were conducted independently from the support- ing agencies. INTRODUCTION 3 Supported by Alpro NV and Frutarom Netherlands BV, a PhD grant Estrogens regulate the development and function of many from the Institute for the Promotion of Innovation through Science and tissues in men and women but are also implicated in the etiology Technology in Flanders (IWT-Vlaanderen; to SB), and an FPI fellowship of breast cancer. Therefore, estrogen-induced cell proliferation from the Spanish Ministry of Science and Innovation (to MU-S). Soy milk has been a major focus in breast cancer research. Despite the (Alpro Soya Drink Nature, Alpro NV, Wevelgem, Belgium) and soy supple- substantially lower breast-cancer prevalence in populations with ments (Frutarom Netherlands BV, Veenendaal, Netherlands) were kindly provided by their manufacturers. a high soy consumption (1, 2), the ﬁndings of Allred et al (3–7) 4 Address correspondence to A Heyerick, Laboratory of Pharmacognosy provoked a heated debate about the safety of dietary phytoes- and Phytochemistry, Faculty of Pharmaceutical Sciences, Ghent University, trogens, such as isoﬂavones, especially for patients with existing Harelbekestraat 72, B-9000 Ghent, Belgium. E-mail: email@example.com. estrogen-sensitive tumors and women at a high risk of developing Received October 23, 2009. Accepted for publication January 6, 2010. breast cancer (1, 8–10). First published online February 17, 2010; doi: 10.3945/ajcn.2009.28854. 976 Am J Clin Nutr 2010;91:976–84. Printed in USA. Ó 2010 American Society for Nutrition ISOFLAVONES IN HUMAN BREAST TISSUE 977 tions regarding the absorption, metabolism, and distribution of and magnesium stearate, and had a titanium-dioxide and iron- these compounds in the target tissue need to be clariﬁed because oxide coating. these processes highly determine whether isoﬂavones may elicit protective or rather adverse responses with respect to breast Subjects carcinogenesis. Two studies (18, 19) quantiﬁed isoﬂavones in human nipple aspirate ﬂuid and/or breast-tissue homogenates A total of 31 generally healthy Belgian or Dutch women who after soy supplementation. However, only total concentrations were scheduled for an esthetic breast reduction were recruited for after enzymatic hydrolysis were reported, even though the this study. The exclusion criteria were breast cancer, an antibiotic conjugation to glucuronic acid and/or sulfate moieties is known treatment within the previous month, and a soy allergy. to alter the pharmacologic proﬁles of genistein and daidzein (20). Ethical approval was granted by the Ethics Committee of the In addition, information is missing concerning the biodistribution Ghent University Hospital (EC UZG 2005/022; initial re- of isoﬂavones within breast tissue, although cell-type-speciﬁc cruitment date: 20 February 2007). The volunteers were fully responses to estrogen exposure have been reported (21), and informed about the aims of the study and gave written consent to aromatase, unlike steroid sulfatase and sulfotransferase, is almost participate in the study. exclusively expressed in the adipose part of normal breast tissue (22). Therefore, we assessed the concentrations, nature of Study design metabolites (free or conjugated), and biodistribution (adipose or glandular tissue) of orally administered isoﬂavones in breast This study was a randomized dietary intervention trial with tissue with a randomized dietary intervention trial in generally a run-in period !4 d and a supplementation phase of 5 d before healthy women undergoing an esthetic breast reduction. In ad- breast reduction. After eligibility assessment, volunteers were dition, the estrogenic potential of the isoﬂavone exposure was randomly allocated to the soy milk (n = 11), soy supplement (n = Downloaded from www.ajcn.org by guest on July 12, 2011 compared with the endogenous 17b-estradiol (E2) concentrations 10), or control (n = 10) groups. All participants were counseled measured in the breast biopsies. not to change their habitual, Western-type dietary patterns but were asked to abstain from soy-based products during the ex- perimental period. A detailed list of isoﬂavone-containing foods SUBJECTS AND METHODS and dietary supplements was distributed to guide the volunteers in this respect. In addition, subjects were instructed to report Chemicals every case of doubt or fortuitous consumption and to provide Genistein, daidzein, and equol were purchased from detailed information on that eating occasion including the type ` Extrasynthese (Genay, France), and dihydrodaidzein and O- and portion size of the food and supplements. desmethylangolensin were purchased from Plantech UK During the supplementation phase, either 250 mL soy milk or (Reading, United Kingdom). Standards of genistein-7- one soy supplement was taken daily with breakfast, lunch, and O-glucuronide and daidzein-7-O-glucuronide were chemically dinner. The control group did not receive any supplementation synthesized according to the method of Al-Maharik and Bot- before surgery. Compliance was evaluated by a subject inquiry ting (23). For the hydrolysis of conjugated isoﬂavones, a 33-g/L and urinary phytoestrogen excretion. solution of type H-1 Helix pomatia extract (minimum of 300 U Subjects delivered 2 spot urine samples: a ﬁrst one after the b-glucuronidase/mg and 15.3 U sulfatase/mg; Sigma-Aldrich, run-in period and a second before anesthesia. During surgery Bornem, Belgium) in sodium acetate buffer (0.1 mol/L, pH = 5) (12–18 h after the last soy supplementation), blood and breast was prepared. 4-Hydroxybenzophenone in methanol was used as biopsies were collected. Serum was obtained by centrifugation an internal standard in the quantitative analyses of urine and [10 min at 600 · g, room temperature (RT)] after coagulation. serum (400 lmol/L) and breast tissue (40 lmol/L). Aliquots of urine and serum samples were stored at 220°C until analysis. The tissue samples were immediately frozen in liquid nitrogen and stored at 280°C until analysis. The timings of the Isoﬂavone preparations last soy supplementation and sampling were noted. In addition, The isoﬂavone preparations used in this study were com- a general questionnaire was used to obtain information on each mercially available soy milk derived from whole soybeans (Alpro subject’s history of antibiotic treatments, hormonal therapies, Soya Drink Nature; Alpro NV, Wevelgem, Belgium) in 250-mL use of any other medication, food-supplement intakes, and an- cartons and a soy germ powder (SoyLife EXTRA; Frutarom thropometric measures, whereas dietary habits were assessed Netherlands BV, Veenendaal, Netherlands) formulated as tablets. with a validated food-frequency questionnaire (24). The inves- One batch of each preparation was analyzed in triplicate at the tigators were blinded to the treatments when working with the study onset and closure. samples. One portion of soy milk (250 mL) contained a mean (6SEM) of 16.98 6 0.76-mg genistein and 5.40 6 0.22-mg daidzein Analytic methods aglycone equivalents (ratio: 3.14) and 8.25 g protein, 7 g car- bohydrates, 4.75 g lipids, 1.5 g ﬁber, 0.375 lg vitamin B-12, Isoﬂavones in soy milk and soy supplements 0.6 lg riboﬂavin, and minerals. Before solid-phase extraction, the soy milk was diluted 10-fold Each soy supplement contained 5.27 6 0.04-mg genistein and in 0.1 mol/L hydrochloric acid. The C18 silica columns (5 mL, 17.56 6 0.28-mg daidzein aglycone equivalents (ratio: 0.30) and 500 mg; Bond Elut, Varian, St-Katelijne-Waver, Belgium) were was ﬁlled with cellulose along with the common processing preconditioned with 5 mL methanol, 5 mL water, and 5 mL aids, such as sodium carboxymethylcellulose, silicon dioxide, of 10 mmol/L hydrochloric acid, consecutively. After sample 978 BOLCA ET AL application (10 mL), the cartridges were rinsed with 5 mL of 20 lL internal standard, were mixed with sodium acetate buffer 10 mmol/L hydrochloric acid, and the compounds of interest (0.1 mol/L, pH = 5; 50:50, vol:vol), incubated with 100 lL were eluted with 5 mL methanol by using a VacMaster 20 sample b-glucuronidase/sulfatase for 24 h at 37°C, and treated with 800 processing unit (IST, Mid Glamorgan, United Kingdom). Finally, lL of 200 mmol/L hydrochloric acid in methanol. After cen- the solvent was evaporated at RT under a gentle stream of ni- trifugation (5 min at 12,500 · g, 4°C), the supernatant ﬂuid was trogen gas, and the samples were reconstituted in 1 mL methanol stored at 220°C until analysis. per 0.1 mol/L hydrochloric acid (80:20, vol:vol) and analyzed by To study the biodistribution of isoﬂavones within the breast, HPLC-ultraviolet detection. The soy supplements were ground in tissue samples were ﬁrst dissected into fractions containing al- a mortar, and 500 mg powder was extracted with 10 mL ace- most exclusively either pure fat or glandular tissue on the basis of tonitrile per 0.2 mol/L hydrochloric acid (50:50, vol:vol) and gross inspection. Areas of adipose tissue intimately intermixed diluted 5-fold in methanol before HPLC-ultraviolet analysis. with ﬁbroglandular tissue were avoided, and connective tissue Genistin, acetyl genistin, malonyl genistin, genistein, daidzin, was removed. Next, ’250 mg adipose or glandular breast tissue, acetyl daidzin, malonyl daidzin, and daidzein were quantiﬁed with 20 lL internal standard, were homogenized in 2.25 mL ice- with a Waters 2695 Alliance separations module and 996 pho- cold 200 mmol/L hydrochloric acid in methanol/water (70:30, todiode array detector (Waters, Milford, MA) combined with an vol:vol) and 0.25 mL hexane with a T10 ULTRA-TURRAX XTerra MS C18 reversed-phase column (5 lm; 250 · 4.6 mm; homogenizer (Ika, Werk Staufen, Germany). After centrifuga- Waters) at 35°C and by using a gradient of solvent A (ie, tion (10 min at 1200 · g, RT), the supernatant ﬂuid was col- 6.6 mmol aqueous formic acid/L) and solvent B (ie, 6.6 mmol lected, and the pellet was extracted with 1 mL ice-cold 200 formic acid in acetonitrile/L) with the following elution proﬁle mmol/L hydrochloric acid in methanol/water (70:30, vol:vol). and a ﬂow rate of 1.5 mL/min: 0–20 min, from 10% to 30% Pooled supernatant ﬂuids were evaporated to dryness at 37°C solvent B in solvent A; 20.1–27 min, 30% solvent B in solvent under a gentle stream of nitrogen gas, reconstituted in 250 lL Downloaded from www.ajcn.org by guest on July 12, 2011 A; and 27.1–30 min, from 30% to 10% solvent B in solvent A. sodium acetate buffer (0.1 mol/L, pH = 5), and incubated with The injection volume was 20 lL. Ultraviolet detection was 100 lL b-glucuronidase/sulfatase for 24 h at 37°C. Finally, conducted at 254 nm for all compounds. hydrolyzed samples were treated with 1 mL of 200 mmol/L hydrochloric acid in methanol and centrifuged (5 min at Isoﬂavones in hydrolyzed urine 12,500 · g, 4°C), and supernatant ﬂuids were stored at 220°C Total isoﬂavones in urine were quantiﬁed by enzymatic hy- until analysis. drolysis and liquid-liquid extraction by using a method validated Quantiﬁcation of genistein, daidzein, dihydrodaidzein, equol, by Wyns et al (25). Brieﬂy, 2 mL urine, with 20 lL internal O-desmethylangolensin, and the internal standard in hydrolyzed standard, was mixed with sodium-acetate buffer (0.1 mol/L, serum and tissue samples was performed by HPLC-MS/MS with pH = 5) (50:50, vol:vol), incubated with 30 lL b-glucuronidase/ a Thermo Scientiﬁc Accella autosampler and pump system and sulfatase for 1 h at 37°C, and extracted twice with 5 mL diethyl a triple quadrupole tandem MS operating in positive and nega- ether. Pooled extracts were evaporated to dryness at 37°C under tive electrospray ionization modes (TSQ Vantage; Thermo Sci- a gentle stream of nitrogen gas, reconstituted in 100 lL of the entiﬁc, San Jose, CA), combined with a Hypersil GOLD C18 mobile phase, and stored at 220°C before quantitative HPLC- reversed-phase column (3 lm, 2.1 · 50 mm; Thermo Scientiﬁc) ultraviolet/mass spectrometer (MS) analysis. at 35°C, and by using a gradient of solvent A (ie, 26.5 mmol Genistein, daidzein, dihydrodaidzein, equol, O-desmethyl- aqueous formic acid/L) and solvent B (ie, 26.5 mmol/L formic angolensin, and the internal standard in hydrolyzed urine were acid in acetonitrile) with the following elution proﬁle and a ﬂow quantiﬁed with a reversed-phase HPLC (XBridge C18 column, rate of 300 lL/min: 0–0.5 min, 10% solvent B in solvent A; 0.5– 3.5 lm, 150 · 3.0 mm; Waters) coupled to a photodiode array 2.5 min, from 10% to 100% solvent B in solvent A; 2.5–4.5 min, detector and single quadrupole MS operating in the positive 100% solvent B; 4.5–4.6 min, from 100% to 10% solvent B in atmospheric pressure chemical ionization mode (Hewlett- solvent A; and 4.6–5.1 min, 10% solvent B in solvent A. The in- Packard 1200 series; Agilent Technologies, Santa Cara, CA) jection volume was 10 lL. Ionization and detection variables were according to Wyns et al (25). The limits of detection (LOD, optimized during infusion experiments with standards of genis- signal/noise = 3) and limits of quantiﬁcation (LOQ, signal/noise = tein, daidzein, dihydrodaidzein, equol, O-desmethylangolensin, 10) were in the lower nanomoles-per-liter range for all analytes and 4-hydroxybenzophenone. MS/MS data were collected in a se- (LOD: 3.00–29.74 nmol/L; LOQ: 9.98–99.11 nmol/L). lected reaction monitoring mode by monitoring speciﬁc transitions of parent and product ions for each analyte: genistein (m/z 271/91/ Creatinine in nonhydrolyzed urine 215), daidzein (m/z 253/208/223), dihydrodaidzein (m/z 257/123/ 163), equol (m/z 244/123/133), O-desmethylangolensin (m/z 257/ To allow the standardization of diuresis, the urinary excretion 108/136), and 4-hydroxybenzophenone (m/z 199/77/121). Linear of creatinine was measured according to the conventional kinetic (r2 . 0.97) calibration curves were obtained over a range of ´ Jaffe method as described by Bolca et al (26). On the basis of 0.05–10.00 lmol/L in serum and 0.04–8.00 nmol/g in tissue. The a creatinine clearance rate of 0.163 mmol/(d Á kg) (27), daily LOD and LOQ were in the lower nanomoles-per-liter range (LOD: urinary isoﬂavone excretions were calculated. 0.26–44.22 nmol/L; LOQ: 0.66–147.18 nmol/L) and picomoles- per-gram range (LOD: 0.38–52.92 pmol/g; LOQ: 1.08–175.50 Isoﬂavones in hydrolyzed serum and breast tissue pmol/g) for all analytes in hydrolyzed serum and breast tissue, re- A sample treatment preceding the quantitative HPLC-MS/MS spectively, except for equol (LOD: 328.68 nmol/L and 445.50 analysis of isoﬂavone aglycones in hydrolyzed serum was based pmol/g; LOQ: 1.10 lmol/L and 1.49 nmol/g). The intra- and in- on the protocol of Guy et al (28). Serum samples (200 lL), with terassay CVs were 5% and 15%, respectively. ISOFLAVONES IN HUMAN BREAST TISSUE 979 Isoﬂavones in nonhydrolyzed serum and breast tissue glandular breast tissue was homogenized in 5 mL ethanol/water Semiquantitative analyses of isoﬂavone metabolites in non- (70:30, vol:vol) with a T10 ULTRA-TURRAX homogenizer hydrolyzed serum and tissue samples were based on the protocol (Ika) and, after precipitation (2 · 24 h at 220°C), extracted with of Guy et al (28). Brieﬂy, 200 lL serum were mixed with 500 lL 5 mL ethyl acetate/hexane (60:40, vol:vol). The organic phase 200 mmol/L hydrochloric acid in methanol, centrifuged (4 min was evaporated to dryness at 37°C under a gentle stream of at 12,500 · g, RT), and analyzed by HPLC-MS/MS. Approxi- nitrogen gas and reconstituted in 500 lL steroid-free serum mately 250 mg breast tissue were homogenized in 1 mL ice-cold (Std0-DRG; DRG Instruments GmbH, Marburg, Germany). Fi- 200 mmol/L hydrochloric acid in methanol/water (70:30, vol: nally, samples were analyzed for E2 with a commercial quan- vol) with a System POLYTRON PT2100 homogenizer (Kine- titative immunoassay (EIA-4499, DRG Instruments GmbH). matica AG, Luzern, Switzerland). After centrifugation (10 min According to the manufacturer, this kit had a sensitivity ,5.13 at 1200 · g, 4°C), the supernatant ﬂuid was defatted with 1 mL pmol/L serum; an intra- and interassay CV of 6.4% and 7.6%, hexane (centrifugation: 10 min at 12,500 · g, RT), evaporated to respectively; and a cross-reactivity of 0.2% with estrone, 0.05% dryness at RT under a gentle stream of nitrogen gas, and recon- with estriol, and ,0.05% with E2. The cross-reactivity with stituted in 250 lL methanol. Finally, samples were centrifuged isoﬂavones was estimated as 0.01%. (4 min at 12,500 · g, RT), and analyzed by HPLC-MS/MS. As described by Guy et al (28), semiquantitative analyses of Statistical analyses isoﬂavone metabolites in nonhydrolyzed serum and tissue samples were performed by HPLC-MS/MS with a Hewlett- A statistical program (SPSS for Windows version 15.0; SPSS, Packard 1100 HPLC system (Agilent Technologies, Waldbronn, Chicago, IL) was used for all statistical analyses. Results were Germany) and an Applied Biosystems API 2000 triple quadru- considered statistically signiﬁcant at an a 2-tailed level of 0.05. pole tandem MS (PE Sciex, Concord, Ontario, Canada) equipped Means and SEMs of urine, serum, and tissue concentrations Downloaded from www.ajcn.org by guest on July 12, 2011 with a Turbo IonSpray source operating in the positive-ion were calculated. Tests for normality and equality of the var- mode. A SymmetryShield C18 reversed-phase column (5 lm, 2.1 iances were performed by using the Kolmogorov-Smirnov and · 150 mm; Waters) with a linear gradient of solvent A (ie, Levene tests, respectively. Intrasubject comparisons were eval- 26.5 mmol aqueous formic acid/L) and solvent B (ie, acetoni- uated with the paired Student’s t test or Wilcoxon’s matched- trile) was used with the following elution proﬁle at 400 lL/min: pairs signed-rank test, whereas the Student’s t test, analysis of 0–15 min, from 0% to 100% solvent B; 15–19 min, 100% sol- variance with Bonferroni correction, or Mann-Whitney U test vent B, and a 6-min reequilibration. The injection volume was were used to compare means between groups. Associations were 20 lL. Ionization and detection variables were optimized during described by using paired Pearson’s correlation coefﬁcients or infusion and ﬂow injection analysis experiments with standards nonparametric Spearman’s correlations. With the use of the of genistein, daidzein, equol, genistein-7-O-glucuronide, and TwoStep cluster-analysis protocol (SPSS for Windows version daidzein-7-O-glucuronide (28). 15.0; SPSS), subjects were phenotyped as weak, moderate, or To determine the nature of isoﬂavone metabolites in serum and strong equol producers on the basis of the urinary excretion of tissue, MS/MS data were collected in a multiple-reaction mon- daidzein aglycone equivalents as equol (equol/total daidzein; itoring mode by monitoring speciﬁc transitions of parent and total daidzein = daidzein + dihydrodaidzein + equol + O- product ions for each analyte as follows: genistein (m/z 271/91), desmethylangolensin) (30). dihydrogenistein (m/z 273/255), daidzein (m/z 255/91), dihy- drodaidzein (m/z 257/95), equol (m/z 244/133), genistein-7-O- glucuronide (m/z 447/271), dihydrogenistein (m/z 449/273), RESULTS daidzein-7-O-glucuronide (m/z 441/255), dihydrodaidzein- Study population O-glucuronide (m/z 443/257), equol-O-glucuronide (m/z 419/ 244), O-desmethylangolensin-O-glucuronide (m/z 445/259), gen- A total of 31 generally healthy women undergoing an esthetic istein-sulfoglucuronide (m/z 527/271), daidzein-sulfoglucuronide breast reduction, all complying with the study protocol, partic- (m/z 511/255), genistein-sulfate (m/z 351/271), daidzein-sulfate ipated in this study. On the basis of self-reported weight and (m/z 335/255), and equol-sulfate (m/z 323/244). For the quan- height measurements, their age and BMI (in kg/m2) ranged from titative analyses, linear (r2 . 0.99) calibration curves were ob- 18 to 62 y and 19 to 36, respectively, and did not differ sig- tained for genistein, daidzein, genistein-7-O-glucuronide, and niﬁcantly (P = 0.898 and P = 0.103, respectively) between daidzein-7-O-glucuronide over a range of 0.01–1.00 lmol/L in groups. None of the participants reported a change in body blank serum and tissue homogenate, and the LOQ were esti- weight after soy supplementation. Eight women (27%; 2 women mated between 8.3 and 30 nmol/L. in the soy milk group, 3 women in the soy supplement group, 3 women in the control group) were in the follicular phase of their menstrual cycle, and 8 women (27%; 3 women in the soy milk E2 in nonhydrolyzed breast tissue group, 4 women in the soy supplement group, 1 woman in the To study the biodistribution of E2 within the breast, tissue control group) were in the luteal phase, whereas 14 women samples were dissected into fractions containing, almost ex- (46%; 6 women in the soy milk group, 3 women in the soy clusively, either pure fat or glandular tissue on the basis of supplement group, 5 women in the control group) were meno- a macroscopic inspection. Areas of adipose tissue intimately pausal. Seven women followed a therapy with exogenous es- intermixed with ﬁbroglandular tissue were avoided, and con- trogens [ie, oral contraceptives (19%; 1 woman in the soy milk nective tissue was removed. Estrogens were extracted as de- group, 4 women in the soy supplement group, 1 woman in the scribed by Chetrite et al (29) as follows: ’200 mg adipose or control group) or intrauterine contraceptives (3%; 1 woman in 980 BOLCA ET AL TABLE 1 Urinary excretion and serum concentrations of genistein, daidzein, dihydrodaidzein, equol, and O-desmethylangolensin algycone equivalents 12–18 h after isoﬂavone supplementation1 Soy milk Soy supplements n Mean 6 SEM n Mean 6 SEM P2 Urine (lmol/d) 11 10 Genistein 11 2077.7 6 1165.4 10 267.50 6 88.49 0.138 Daidzein 11 148.35 6 72.01 10 214.22 6 73.36 0.533 Dihydrodaidzein 11 12.58 6 6.03 10 18.21 6 11.44 0.470 Equol 11 154.96 6 79.65 10 287.60 6 148.93 0.419 O-Desmethylangolensin 11 220.33 6 144.98 10 215.09 6 107.98 0.057 Serum (nmol/L) 11 10 Genistein 11 797.04 6 237.27 10 217.89 6 27.61 0.002 Daidzein 11 196.10 6 53.53 10 315.79 6 99.61 0.291 Dihydrodaidzein 11 90.08 6 17.60 10 146.34 6 26.56 0.085 Equol 4 590.93 6 118.34 3 900.52 6 81.40 0.103 O-Desmethylangolensin 11 130.25 6 44.35 10 137.68 6 35.76 0.899 1 Isoﬂavone supplementation: 3 doses of soy milk (ie, 50.94-mg genistein and 16.20-mg daidzein aglycone equiv- alents/d) or soy supplements (15.81-mg genistein and 52.68-mg daidzein aglycone equivalents/d) daily for 5 d. 2 Student’s t test was used for all analytes under investigation except genistein, dihydrodaidzein, and O-desmethy- langolensin in urine and genistein in serum, for which the Mann-Whitney U test was used. Downloaded from www.ajcn.org by guest on July 12, 2011 the control group)], whereas 2 women were treated with anti- conjugated aglycones and deconjugated (sulfo)glucuronides and estrogens (6%; 1 woman in the soy supplement group, 1 woman sulfates, measured in hydrolyzed urine and serum (Table 1) and in the control group). All individuals had an average fat and ﬁber hydrolyzed adipose and glandular tissue (Table 2), and com- intake (24) and Western-type dietary patterns. pared between treatment groups. None of the control samples, On the basis of their urinary excretion proﬁles after soy ie, all urine samples collected at the end of the run-in phase and supplementation, 3 participants (14%; 2 subjects in the soy milk those of the control group after the intervention phase, contained group, 1 subject in the soy supplement group) were phenotyped as detectable amounts of soy-derived phytoestrogens. After iso- moderate equol producers (45.2 6 5.1% equol), and 6 partic- ﬂavone supplementation, estimated daily urinary isoﬂavone ipants (29%; 3 subjects in the soy milk group, 3 subjects in the excretion varied considerably between individuals and was in soy supplement group) were phenotyped as strong equol pro- the micromoles-per-day range (6.56–13,138 lmol genistein/ ducers (76.2 6 4.1% equol). d and 2.38–2234 lmol total daidzein/d). Genistein and total daidzein serum concentrations ranged from 135.1 to 2831 nmol/ L and 105.1 to 1397 nmol/L, respectively, with genistein/total Total exposure daidzein ratios (G/D) of 1.56 6 0.44 and 0.28 6 0.04 (P = Exposure to genistein, daidzein, and its microbial metabolites 0.015) after soy milk and soy supplement administration, re- after soy supplementation was assessed as the sum of un- spectively. Seven women (33%; 4 women in the soy milk group, TABLE 2 Breast adipose and glandular tissue concentrations of genistein, daidzein, dihydrodaidzein, equol, and O- desmethylangolensin algycone equivalents 12–18 h after isoﬂavone supplementation1 Soy milk Soy supplements n Mean 6 SEM n Mean 6 SEM P2 Adipose tissue (pmol/g) 11 10 Genistein 11 183.59 6 14.84 10 146.79 6 31.27 0.287 Daidzein 11 45.35 6 6.58 10 61.96 6 13.99 0.282 Dihydrodaidzein 2 116.38 6 44.58 1 234.26 — Equol 11 ,LOD 10 ,LOD — O-Desmethylangolensin 11 15.10 6 3.62 16.82 6 6.19 0.808 Glandular tissue (pmol/g) 11 10 Genistein 11 283.71 6 35.88 10 148.85 6 10.66 0.004 Daidzein 11 57.26 6 10.65 10 89.16 6 15.54 0.101 Dihydrodaidzein 3 208.40 6 67.82 2 368.52 6 171.71 — Equol 1 559.39 1 446.25 — O-Desmethylangolensin 11 30.80 6 13.86 10 22.33 6 5.11 0.588 1 Isoﬂavone supplementation: 3 doses of soy milk (ie, 50.94-mg genistein and 16.20-mg daidzein aglycone equiv- alents/d) or soy supplements (15.81-mg genistein and 52.68-mg daidzein aglycone equivalents/d) daily for 5 d. LOD, limit of detection. 2 Student’s t test was used for all analytes under investigation. ISOFLAVONES IN HUMAN BREAST TISSUE 981 3 women in the soy supplement group) had circulating equol Breast tissue distribution concentrations .LOD. In breast adipose and glandular tissue, Signiﬁcantly higher concentrations of total isoﬂavones were exposure levels of 92.33–493.8 pmol genistein/g and 22.15– measured in glandular breast fractions compared with adipose 770.8 pmol total daidzein/g were measured. The G/D in both tissues, with an adipose/glandular tissue distribution of 40:60 tissue fractions (adipose soy milk compared with the soy sup- (P = 0.025). Genistein showed an adipose/glandular tissue dis- plement: 3.00 6 0.44 compared with 1.96 6 0.20, P = 0.046; tribution of 41:59 and 47:53 after soy milk consumption (P = glandular soy milk compared with the soy supplement: 3.64 6 0.002) and soy supplement intake (P = 0.453), respectively. 0.77 compared with 1.16 6 0.20, P = 0.009) correlated well Analogously, total daidzein equivalents were more abundant in with those shown in serum (adipose-serum r = 0.490, P = 0.024; the glandular samples of both treatment groups (soy milk: 42:58, glandular-serum r = 0.574, P = 0.007). However, no signiﬁcant P = 0.286; soy supplement: 34:66, P = 0.047). However, the correlations were observed between the serum and tissue con- G/D were similar in both tissue fractions of the soy milk group centrations of genistein, total daidzein, and total isoﬂavones (ie, (adipose compared with glandular: 3.00 6 0.44 compared with genistein + total daidzein). Unlike O-desmethylangolensin, equol 3.64 6 0.77, P = 0.351) and soy supplement group (adipose was shown in only 2 glandular tissue samples (10%; 1 sample in compared with glandular: 1.96 6 0.20 compared with 1.16 6 the soy milk group; 1 sample in the soy supplement group). 0.20, P = 0.051). Phase II metabolism On isoﬂavone supplementation, the major metabolites iden- Physiologic relevance tiﬁed in nonhydrolyzed serum and breast tissue were genistein- The exposure to isoﬂavones and their microbial metabolites 7-O-glucuronide and daidzein-7-O-glucuronide, whereas neither was translated to an overall exposure to E2 equivalents toward ERa and ERb (E2a and E2b equivalents), on the basis of the Downloaded from www.ajcn.org by guest on July 12, 2011 glucuronides nor aglycones of microbial daidzein metabolites were detected (Table 3). We assume that the other metabolites generally accepted dose-addition concept (32) and thereby as- detected with MS/MS transitions 447/271 and 441/255 were 4#- suming relative estrogenic potencies toward ERa of 1/100, 1/ O-glucuronides of genistein and daidzein, respectively (31). In 5000, and 1/200 and toward ERb of 1/2, 1/100, and 1/35 for serum, similar concentrations of both 7-O-glucuronides were genistein, daidzein, and equol, respectively (12–14), and an shown in nearly all (.90%) samples of the soy milk group, overall 98% attenuation because of glucuronidation (33–35) whereas, after soy supplement intake, daidzein-7-O-glucuronide (Table 4). These isoﬂavone-derived E2a and E2b equivalents was the most abundant metabolite in all cases with only traces of showed adipose/glandular tissue distributions of 42:58 (P = genistein-7-O-glucuronide in 6 (67%) samples. Like serum, 0.029) and 44:57 (P = 0.037), respectively, whereas no signiﬁ- breast tissue homogenates collected after soy supplementation cant differences were observed between adipose and glandular contained mostly 7-O-glucuronidated isoﬂavones, but these E2 concentrations (56:44, P = 0.102). On both soy milk and soy were not detectable in all samples. Only traces of genistein and supplement administration, breast tissue was exposed to esti- daidzein aglycones were observed, and an overall total glucur- mated concentrations of isoﬂavone-derived E2b equivalents, which onidation of 98% was estimated. signiﬁcantly exceeded the endogenous E2 tissue concentrations TABLE 3 Serum and breast tissue homogenate concentrations of genistein, daidzein, and their phase II metabolites 12–18 h after isoﬂavone supplementation1 Soy milk Soy supplements n Mean 6 SEM n Mean 6 SEM P2 Serum (nmol/L) 11 9 Genistein-7-O-glucuronide 10 122 6 68 6 11 6 4 0.030 Genistein-4#-O-glucuronide 6 14 6 5 2 4 6 0.13 0.101 Genistein 2 67 6 19 9 ,LOD — Daidzein-7-O-glucuronide 11 151 6 50 9 294 6 119 0.249 Daidzein-4#-O-glucuronide 9 16 6 6 7 39 6 19 0.238 Daidzein 3 33 6 7 4 49 6 15 0.445 Breast tissue (pmol/g) 11 10 Genistein-7-O-glucuronide 8 268 6 179 4 191 6 65 0.776 Genistein-4#-O-glucuronide 5 86 6 46 4 45 6 7 0.463 Genistein 4 12 6 23 3 8 6 23 0.195 Daidzein-7-O-glucuronide 5 145 6 72 7 498 6 226 0.178 Daidzein-4#-O-glucuronide 2 97 6 18 3 128 6 27 — Daidzein 6 8 6 13 6 22 6 43 0.010 1 Isoﬂavone supplementation: 3 doses of either soy milk (ie, 50.94-mg genistein and 16.20-mg daidzein aglycone equivalents/d) or soy supplements (15.81-mg genistein and 52.68-mg daidzein aglycone equivalents/d) daily for 5 d.; LOD, limit of detection. 2 Student’s t test was used for all analytes under investigation. 3 Values less than the limits of quantiﬁcation. 982 BOLCA ET AL (adipose: P , 0.001 and P = 0.003, respectively; glandular: P = to genistein and daidzein as occur through a rather high soy food 0.004 and P = 0.005, respectively). or supplement intake. In agreement with previous reports (28, 31, 39), isoﬂavone glucuronides were the predominant circulating metabolites, with DISCUSSION preferential conjugation at the C7-position for genistein and There is a lot of controversy about the soy–breast cancer daidzein. No monosulfates or sulfoglucuronides of genistein and hypothesis (1, 2). Although soy products contain several bio- daidzein were observed, whereas diglucuronides and disulfates active phytochemicals, most cancer research has focused on were not monitored. Like serum, breast tissue contained mostly 7-O-glucuronidated isoﬂavones, but these were not detected in isoﬂavones and genistein in particular. However, to properly test all samples. Other than the study of Guy et al (28), we are not and evaluate the suggested mechanisms of action on breast tissue, aware of additional reports on the disposition of phase II me- information on the concentrations of orally administered iso- tabolites in human tissue. ﬂavones that actually reach their target site in a bioactive form is The question is whether the exposure to isoﬂavones, as ob- needed. Our results indicate that, 12–18 h after the last soy milk served in the current study, could result in any protective or or soy supplement intake, breast adipocytes and mammary gland adverse response related to breast carcinogenesis. Given the epithelial cells were exposed to up to 20–25 pmol/g total iso- complexity of this interaction, our information on in situ con- ﬂavone aglycones and 900–1150 pmol/g total isoﬂavone glu- centrations, in addition to the current state of knowledge, only curonides. Because total isoﬂavone-derived E2b equivalents allows for speculation on the potential activities of orally ad- were, on average, 21 and 40 times more abundant than the en- ministered isoﬂavones on breast tissue. The prevailing opinion is dogenous E2 in adipose and glandular breast tissue, respectively, that treatments that trigger ER antagonistic effects, such as ta- the clinical implications of these ﬁndings require further moxifen, and/or reduce E2 and estrone concentrations, such as Downloaded from www.ajcn.org by guest on July 12, 2011 investigation. aromatase inhibitors, are protective against breast cancer and Although differences in the time of sampling, dosing regimen, favorably affect the course of breast cancer once diagnosed. and formulation (36) often hamper sound comparisons between Therefore, we discuss the soy–breast cancer hypothesis in the dietary intervention trials, the hydrolyzed serum and tissue context of the potential of isoﬂavones to interfere with ER- isoﬂavone concentrations observed in this study were in agree- mediated signaling and estrogen synthesis and deprivation. ment with previous reports on human breast exposure (18, 19, 37) Both ER isoforms, ie, ERa and ERb, are expressed in breast and prostate exposure (28). Isoﬂavone tissue concentrations were tissue. Estrogen-induced cell proliferation and breast carcino- lower than the corresponding serum concentrations (19, 28). genesis have been mainly linked to ERa signaling, whereas ERb Moreover, the consistent lack of a correlation between serum and can antagonize ERa-dependent transcription (40). This alleged tissue concentrations (18, 19) suggests that isoﬂavone serum protective ERb-mediated inhibition of ERa signaling, combined concentrations do not predict tissue disposition. Nevertheless, the with the greater afﬁnity of isoﬂavones for ERb than for ERa G/D ratios correlated well between the different samples and (12–14), fueled the enthusiasm regarding soy consumption as reﬂected the composition of the isoﬂavone preparations (28). a possible environmental factor responsible for the striking Because of the well-known large interindividual variation in geographic differences in breast cancer occurrence (1, 2). phytoestrogen bioavailability (38), the current study was not However, ligand-speciﬁc conformational changes in ER on sufﬁciently powered to detect differences related to the isoﬂavone binding result in unique coregulator recruitment and transcrip- formulation (soy milk compared with soy supplements) (30). tional and ﬁnal biological responses, as illustrated in animal However, our main objective was to measure the exposure ranges studies (41, 42). Thus, all ligands and metabolites, and even TABLE 4 Breast adipose and glandular tissue exposure to calculated total isoﬂavone-derived 17b-estradiol a (E2a) and E2b equivalents and unconjugated endogenous E2 12–18 h after isoﬂavone supplementation1 Soy milk (n = 11) Soy supplements (n = 10) P2 Adipose tissue (pmol/g) Isoﬂavone-derived E2a 0.04 6 0.01 0.03 6 0.01 0.295 Isoﬂavone-derived E2b 1.85 6 0.15 1.48 6 0.32 0.295 Endogenous E2 0.41 6 0.16 0.14 6 0.05 0.133 Glandular tissue (pmol/g) Isoﬂavone-derived E2a 0.06 6 0.01 0.04 6 0.01 0.044 Isoﬂavone-derived E2b 2.88 6 0.38 1.53 6 0.11 0.004 Endogenous E2 0.35 6 0.17 0.13 6 0.15 0.944 1 All values are means 6 SEMs. Isoﬂavone-derived E2a and E2b equivalents were estimated from isoﬂavone tissue concentrations (in pmol/g) by using the dose-addition concept and taking the relative estrogenic potencies and glucuroni- dation into consideration: isoﬂavone-derived E2a = 0.02 · (0.01 · genistein + 0.0002 · daidzein + 0.005 · equol) and isoﬂavone-derived E2b = 0.02 · (0.5 · genistein + 0.01 · daidzein + 0.029 · equol). Isoﬂavone supplementation: 3 doses of soy milk (ie, 50.94-mg genistein and 16.20-mg daidzein aglycone equivalents/d) or soy supplements (15.81-mg genistein and 52.68-mg daidzein-equivalents/d) daily for 5 d. 2 Student’s t test was used for all values except endogenous E2 in glandular breast tissue, for which the Mann-Whitney U test was used. ISOFLAVONES IN HUMAN BREAST TISSUE 983 mixtures of ligands and metabolites, should be evaluated. Glu- several phytochemicals with multiple and perhaps additive or curonidation may increase [eg, morphine (43)] or decrease [eg, interfering activities. Finally, the LOD and LOQ hindered the tamoxifen (44)] the bioactivity of its substrate. The latter seems accurate quantiﬁcation of isoﬂavone aglycones, in particular of to be the case for the estrogenicity of genistein and daidzein equol, and an underestimation of isoﬂavone conjugates in en- (33–35). The in vitro results of Zhang et al (35) suggest that zymatically hydrolyzed tissues was reported (54). Moreover, isoﬂavone glucuronides need to reach intracellular concen- theoretical concepts and in vitro data were used to translate trations exceeding 105-106 times those of endogenous E2 to aglycones concentrations into an overall exposure to E2 equiv- compete for ER binding. Therefore, we conclude that the con- alents. centrations observed in breast tissue are 100–1000 times too low Taking these limitations and assumptions into consideration, to result in direct ER-mediated effects. Although genistein and we conclude that, after soy milk and soy supplement intakes, daidzein glucuronides may act as a source of tissue aglycones by isoﬂavones can reach exposure levels in breast tissue with po- means of an in situ glucuronidase activity, intracellular UDP- tential health effects. This study provides data for a more glucuronosyltransferases may catalyze the opposite reaction. comprehensive evaluation of the soy–breast cancer relation on Therefore, only aglycones were taken into consideration to es- the basis of physiologically relevant exposure levels and timate the overall E2a- and E2b-equivalent exposure. The total metabolites. isoﬂavone-derived E2b equivalents signiﬁcantly exceeded the E2 The authors’ responsibilities were as follows—SB: designed the study, tissue concentrations, which were in agreement with literature recruited and followed the participants, analyzed all samples, performed reports (29, 45), suggesting that, in this case, soy consumption the statistical analyses, interpreted the data, and wrote the initial draft of could elicit partial ERb agonistic (46) effects in human breast the manuscript; MU-S, LV, and CM: assisted in the analyses and interpretation tissue. of the LC-MS/MS data; PB and NR: recruited participants and collected sam- Breast tissue E2 concentrations are maintained by the active ples; NA-M and NB: provided glucuronide standards; SP, DDK, MB, AH, and Downloaded from www.ajcn.org by guest on July 12, 2011 uptake of circulating estrogens and/or local synthesis (the in- HD: assisted in the study design and data interpretation and provided signif- tracrine organ concept) (47). The putative attenuation of in situ icant advice; and all authors: participated in critically revising the manuscript. None of the authors had a personal or ﬁnancial conﬂict of interest. steroidogenesis through the inhibition of aromatase, sulfo- transferase, and 17b-hydroxysteroid dehydrogenase by iso- ﬂavones is difﬁcult to evaluate because most experiments were REFERENCES performed at supraphysiologic concentrations (.1 lmol/L), 1. Messina M, Wu AH. Perspectives on the soy–breast cancer relation. Am and, to our knowledge, little or no information is available on J Clin Nutr 2009;89:1673S–9S. the ability of microbial and phase II metabolites and mixtures of 2. Stubert J, Gerber B. Isoﬂavones – mechanism of action and impact on breast cancer risk. Breast Care 2009;4:22–9. isoﬂavones to modulate E2 synthesis and metabolism. However, 3. Allred CD, Allred KF, Ju YH, Virant SM, Helferich WG. Soy diets con- at 100 nmol/L, genistein was shown to decrease in vitro 17b- taining varying amounts of genistein stimulate growth of estrogen- hydroxysteroid dehydrogenase type I activity and andro- dependent (MCF-7) tumors in a dose-dependent manner. Cancer Res stenedione and estrone-stimulated MCF-7 cell proliferation with 2001;61:5045–50. 4. Allred CD, Ju YH, Allred KF, Chang J, Helferich WG. Dietary genistin a concomitant lowered E2 production (48). In addition, a com- stimulates growth of estrogen-dependent breast cancer tumors similar to bination of genistein, biochanin A, and daidzein, all at 100 that observed with genistein. Carcinogenesis 2001;22:1667–73. nmol/L, showed aromatase mRNA down-regulation in human 5. Allred CD, Allred KF, Ju YH, Goeppinger TS, Doerge DR, Helferich granulose-luteal cells (49). Conversely, in pre- and post- WG. Soy processing inﬂuences growth of estrogen-dependent breast cancer tumors. Carcinogenesis 2004;25:1649–57. menopausal women, the consumption of soy isoﬂavones showed 6. Hsieh CY, Santell RC, Haslam SZ, Helferich WG. Estrogenic effects of no signiﬁcant effect on circulating total E2 or estrone concen- genistein on the growth of estrogen receptor-positive human breast trations, as reviewed by Hooper et al (50). cancer (MCF-7) cells in vitro and in vivo. Cancer Res 1998;58:3833–8. The current study has some limitations. First, sampling was 7. Ju YH, Allred CD, Allred KF, Karko KL, Doerge DR, Helferich WG. Physiological concentrations of dietary genistein dose-dependently done at a single time point, 12–18 h after the last soy admin- stimulate growth of estrogen-dependent human breast cancer (MCF-7) istration. Although steady state levels were reached after 5 d of tumors implanted in athymic nude mice. J Nutr 2001;131:2957–62. regular intake throughout the day (51), diurnal ﬂuctuations were 8. Eisenbrand G. Isoﬂavones as phytoestrogens in food supplements and expected because of the discontinued dosing during the night, as dietary foods for special medical purposes. Mol Nutr Food Res 2007;51: 1305–12. shown for daidzein and genistein serum concentrations (36). 9. Eisenbrand G. Untitled – Answer to Dr. Messina’s letter to the editor. Second, tissue was obtained from a small, heterogeneous group of Mol Nutr Food Res 2008;52:737–8. generally healthy women with mammary hypertrophy, and it 10. Messina M. Conclusion that isoﬂavones exert estrogenic effects on is not known to what extent our ﬁndings can be extrapolated breast tissue and may raise breast cancer risk unfounded. Mol Nutr Food Res 2008;52:299–300. to the general population, despite their normal urine and 11. Setchell KDR, Brown NM, Lydeking-Olsen E. The clinical importance serum concentrations and E2 tissue concentrations. For instance, of the metabolite equol – a clue to the effectiveness of soy and its breast tumors have lowered ERb expression (40), enhanced isoﬂavones. J Nutr 2002;132:3577–84. b-glucuronidase, and decreased UDP-glucuronosyltransferase 12. Kostelac D, Rechkemmer G, Briviba K. Phytoestrogens modulate binding response of estrogen receptors a and b to the estrogen response activities (52) and, overall, altered estrogen metabolism result- element. J Agric Food Chem 2003;51:7632–5. ing in higher E2 concentrations (29, 53) and therefore a mark- 13. Kuiper GGJM, Lemmen JG, Carlsson B, et al. Interaction of estrogenic edly different hormonal environment. Similarly, isoﬂavone chemicals and phytoestrogens with estrogen receptor b. Endocrinology disposition may be different in utero; during childhood, puberty, 1998;139:4252–63. 14. Mueller SO, Simon S, Chae K, Metzler M, Korach KS. Phytoestrogens or pregnancy; and in men. Third, the very-short-term isoﬂavone and their human metabolites show distinct agonistic and antagonistic supplementations applied in this study do not reﬂect a typical properties on estrogen receptor a (ERa) and ERb in human cells. Tox- dietary phytoestrogen exposure, which is often a combination of icol Sci 2004;80:14–25. 984 BOLCA ET AL 15. Adlercreutz H. Phytoestrogens and breast cancer. J Steroid Biochem in the MCF-7 human breast cancer cell line. J Nutr Biochem 2008;19: Mol Biol 2002;83:113–8. 739–45. 16. Mense SM, Hei TK, Ganju RK, Bhat HK. Phytoestrogens and breast 35. Zhang Y, Song TT, Cunnick JE, Murphy PA, Hendrich S. Daidzein and cancer prevention: possible mechanisms of action. Environ Health genistein glucuronides in vitro are weakly estrogenic and activate human Perspect 2008;116:426–33. natural killer cells at nutritionally relevant concentrations. J Nutr 1999; 17. Steiner C, Arnould S, Scalbert A, Manach C. Isoﬂavones and the pre- 129:399–405. vention of breast and prostate cancer: new perspectives opened by nu- 36. Gardner CD, Chatterjee LM, Franke AA. Effects of isoﬂavone supple- trigenomics. Br J Nutr 2008;99:ES78–108. ments vs. soy foods on blood concentrations of genistein and daidzein in 18. Hargreaves DF, Potten CS, Harding C, et al. Two-week dietary soy adults. J Nutr Biochem 2009;20:227–34. supplementation has an estrogenic effect on normal premenopausal 37. Pumford SL, Morton MM, Turkes A, Grifﬁths K. Determination of the breast. J Clin Endocrinol Metab 1999;84:4017–24. isoﬂavonoids genistein and daidzein in biological samples by gas 19. Maubach J, Depypere HT, Goeman J, et al. Distribution of soy-derived chromatography-mass spectrometry. Ann Clin Biochem 2002;39: phytoestrogens in human breast tissue and biological ﬂuids. Obstet 281–92. Gynecol 2004;103:892–8. 38. Lampe JW, Chang J-L. Interindividual differences in phytochemical 20. Williamson G, Barron D, Shimoi K, Terao J. In vitro biological prop- metabolism and disposition. Semin Cancer Biol 2007;17:347–53. erties of ﬂavonoid conjugates found in vivo. Free Radic Res 2005;39: 39. Zhang Y, Hendrich S, Murphy PA. Glucuronides are the main isoﬂavone 457–69. metabolites in women. J Nutr 2003;133:399–404. 21. Katzenellenbogen JA, OMalley BW, Katzenellenbogen BS. Tripartite 40. Gustafsson JA. Estrogen receptor b – a new dimension in estrogen steroid hormone receptor pharmacology: Interaction with multiple ef- mechanism of action. J Endocrinol 1999;163:379–83. fector sites as a basis for the cell- and promoter-speciﬁc action of these 41. Gallo D, Zannoni GF, Martinelli E, et al. Estradiol and phytoestrogens hormones. Mol Endocrinol 1996;10:119–31. differently inﬂuence the rodent postmenopausal mammary gland. 22. Geisler J. Breast cancer tissue estrogens and their manipulation with Menopause 2006;13:72–9. aromatase inhibitors and inactivators. J Steroid Biochem Mol Biol 2003; 42. Wood CE, Hester JM, Appt SE, Geisinger KR, Cline JM. Estrogen 86:245–53. effects on epithelial proliferation and benign proliferative lesions in 23. Al-Maharik N, Botting NP. A facile synthesis of isoﬂavone 7-O- the postmenopausal primate mammary gland. Lab Invest 2008;88: glucuronides. Tetrahedron Lett 2006;47:8703–6. 938–48. Downloaded from www.ajcn.org by guest on July 12, 2011 24. Bolca S, Huybrechts I, Verschraegen M, De Henauw S, Van de Wiele T. 43. Frances B, Gout R, Monsarrat B, Cros J, Zajac JM. Further evidence that Validity and reproducibility of a self-administered semi-quantitative morphine-6-b-glucuronide is a more potent opioid agonist than mor- food-frequency questionnaire for estimating usual daily fat, ﬁbre, al- phine. J Pharmacol Exp Ther 1992;262:25–31. cohol, caffeine and theobromine intakes among Belgian post- 44. Zheng Y, Sun D, Sharma AK, Chen G, Amin S, Lazarus P. Elimination menopausal women. Int J Environ Res Public Health 2009;6:121–50. of antiestrogenic effects of active tamoxifen metabolites by glucur- 25. Wyns C, Bolca S, De Keukeleire D, Heyerick A. Development of onidation. Drug Metab Dispos 2007;35:1942–8. a high-throughput LC-APCI-MS method for the determination of 13 45. Falk RT, Gentzschein E, Stanczyk FZ, et al. Measurement of sex steroid phytoestrogens (including gut microbial metabolites) in human urine hormones in breast adipocytes: Methods and implications. Cancer Epi- and serum. Planta Med 2008;74:1100 (abstr). demiol Biomarkers Prev 2008;17:1891–5. 26. Bolca S, Wyns C, Possemiers S, et al. Cosupplementation of isoﬂavones, 46. Pike ACW, Brzozowski AJ, Hubbard RE, et al. Structure of the prenylﬂavonoids, and lignans alters human exposure to phytoestrogen- ligand-binding domain of oestrogen receptor b in the presence of derived 17b-estradiol equivalents. J Nutr 2009;139:2293–300. 27. Junge W, Wilke B, Halabi A, Klein G. Determination of reference in- a partial agonist and full antagonist. EMBO J 1999;18:4608–18. 47. Pasqualini JR, Chetrite GS. Recent insight on the control of enzymes tervals for serum creatinine, creatinine excretion and creatinine clear- ´ ance with an enzymatic and a modiﬁed Jaffe method. Clin Chim Acta involved in estrogen formation and transformation in human breast 2004;344:137–48. cancer. J Steroid Biochem Mol Biol 2005;93:221–36. 28. Guy L, Vedrine N, Urpi-Sarda M, et al. Orally administered isoﬂavones 48. Brooks JD, Thompson LU. Mammalian lignans and genistein decrease are present as glucuronides in the human prostate. Nutr Cancer 2008;60: the activities of aromatase and 17b-hydroxy steroid dehydrogenase in 461–8. MCF-7 cells. J Steroid Biochem Mol Biol 2005;94:461–7. 29. Chetrite GS, Cortes-Prieto J, Philippe JC, Wright F, Pasqualini JR. 49. Rice S, Mason HD, Whitehead SA. Phytoestrogens and their low dose Comparison of estrogen concentrations, estrone sulfatase and aromatase combinations inhibit mRNA expression and activity of aromatase in activities in normal, and in cancerous, human breast tissues. J Steroid human granulosa-luteal cells. J Steroid Biochem Mol Biol 2006;101: Biochem Mol Biol 2000;72:23–7. 216–25. 30. Bolca S, Possemiers S, Herregat A, et al. Microbial and dietary factors 50. Hooper L, Ryder JJ, Kurzer MS, et al. Effects of soy protein and iso- are associated with the equol producer phenotype in healthy post- ﬂavones on circulating hormone concentrations in pre- and post- menopausal women. J Nutr 2007;137:2242–6. menopausal women: a systematic review and meta-analysis. Hum 31. Hosoda K, Furuta T, Yokokawa A, Ogura K, Hiratsuka A, Ishii K. Reprod Update 2009;15:423–40. Plasma proﬁling of intact isoﬂavone metabolites by high-performance 51. Mathey J, Lamothe V, Coxam V, Potier M, Sauvant P, Bennetau- liquid chromatography and mass spectrometric identiﬁcation of ﬂavone Pelissero C. Concentrations of isoﬂavones in plasma and urine of post- glycosides daidzin and genistin in human plasma after administration of menopausal women chronically ingesting high quantities of soy kinako. Drug Metab Dispos 2008;36:1485–95. isoﬂavones. J Pharm Biomed Anal 2006;41:957–65. 32. Kortenkamp A. Ten years of mixing cocktails: a review of combination 52. Albin N, Massaad L, Toussaint C, et al. Main drug-metabolizing enzyme effects of endocrine-disrupting chemicals. Environ Health Perspect systems in human breast tumors and peritumoral tissues. Cancer Res 2007;115(suppl 1):98–105. 1993;53:3541–6. 33. Kinjo J, Tsuchihashi R, Morito K, et al. Interactions of phytoestrogens 53. Pasqualini JR, Chetrite G, Blacker C, et al. Concentrations of estrone, with estrogen receptors a and b (III). Estrogenic activities of soy iso- estradiol, and estrone sulfate and evaluation of sulfatase and aromatase ﬂavone aglycones and their metabolites isolated from human urine. Biol activities in pre- and postmenopausal breast cancer patients. J Clin Pharm Bull 2004;27:185–8. Endocrinol Metab 1996;81:1460–4. 34. Pritchett LE, Atherton KM, Mutch E, Ford D. Glucuronidation of the 54. Gu L, Laly M, Chang HC, et al. Isoﬂavone conjugates are under- soyabean isoﬂavones genistein and daidzein by human liver is related to estimated in tissues using enzymatic hydrolysis. J Agric Food Chem levels of UGT1A1 and UGT1A9 activity and alters isoﬂavone response 2005;53:6858–63.