- Email: [email protected]
Analysis of estrogenic activity of foodstuffs and cigarette smoke condensates using a yeast estrogen screening method
Food and Chemical Toxicology 41 (2003) 543–550 www.elsevier.com/locate/foodchemtox
Analysis of estrogenic activity of foodstuffs and cigarette smoke condensates using a yeast estrogen screening method T. Takamura-Enyaa,*, J. Ishiharaa, S. Taharab, S. Gotoc, Y. Totsukaa, T. Sugimuraa, K. Wakabayashia a Cancer Prevention Division, National Cancer Center Research Institute, 1-1 Tsukiji 5 chome, Chuo-ku, Tokyo 104-0045, Japan Division of Applied Bioscience, Graduate School of Agriculture, Hokkaido University, Kita-ku Sapporo 060-8589, Sapporo, Japan c Research Center for Material Cycles and Waste Management, National Institute for Environmental Studies, 16-2, Onogawa, Tsukuba, Ibaragi 305-0053, Japan
b
Accepted 14 October 2002
Abstract Hormone mimics present in our environment are of concern because such agents could potentially reduce fertility and increase sexual dysfunction in wildlife and increase the risk of breast and reproductive organ cancers in man. Therefore, monitoring of the levels of estrogenic compounds in environmental materials is essential in order to prevent their exposure to man and to discover potential harmful effects on human health. In the present study, we analyzed estrogenic activity in 23 foodstuffs and cigarette smoke condensate samples extracted with an organic solvent, using the yeast estrogen screening (YES) system. Three soybean-related foodstuffs (soy sauce, tofu, miso), beer, coffee and cigarette smoke condensates showed clear estrogenic activity in the YES system. HPLC fractionations followed by the YES of theseYES-positive samples revealed the presence of many estrogenic compounds in cigarette smoke condensates, whereas the other samples exerted estrogenic activities in only one or two fractions. Genistein was able to be isolated as the major active principle in soy sauce, tofu and miso, its concentration in these three foodstuffs ranging from 0.1 to 394 mg/g or ml. 8-Prenylnaringenin was also isolated from beer extracts as a major compound with estrogenic activity present at 0.22–4.0 ng/ml. Estrogenic activity of 8-prenylnaringenin with YES was 10-times as high as that of genistein, although it was 100-times less than that of 17b-estradiol. Based on our results in vitro, 10 mg miso and 10 ml beer can be calculated to have similar estrogenic activity to 1 pmole 17b-estradiol. It is very important that the effects of genistein and 8-prenylnaringenin on human health are elucidated. # 2003 Elsevier Science Ltd. All rights reserved. Keywords: Estrogenic activity; Foodstuffs; Yeast estrogen screening; Soybeans; Beer; Coffee; Cigarette smoke condensates; 8-Prenylnaringenin; Genistein
1. Introduction A wide variety of chemicals present in our environment have now been found to exhibit estrogenic properties or act like estrogen. These hormone mimics are hypothesized to be associated with reduced fertility and increased sexual dysfunction in wildlife animals as well as with an increased risk for breast and reproductive Abbreviations: DMSO, dimethyl sulfoxide; ONPG, O-nitrophenyl-b-d-galactopyranoside; YES, yeast estrogen screening. * Corresponding author. Tel.: +81-3-3542-2511x4551; fax: +81-33543-9305. E-mail address: [email protected] (T. Takamura-Enya).
tract cancers in humans (Colborn et al., 1993; Colborn, 1995; Daston et al., 1997; Safe, 2000). These compounds include natural substances found in plants (phytoestrogens), fungi (mycoestrogens) and those that are man-made (Waller et al., 1996; Denison and Helferich, 1998; IUPAC, 1998; Korach, 1998). Flavonoids such as genistein and coumestrol in some plant foods exhibit estrogenic activity (Breinholt and Larsen, 1998; IUPAC, 1998). b-Resorcylic acid lactones such as zearalenone are well-known fungus estrogens found in Fusarium, which infects corn (IUPAC, 1998). Certain pesticides and herbicides such as organochlorine compounds, combustion pollutants and phthalate plasticizers, have now been found to be
0278-6915/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved. PII: S0278-6915(02)00305-8
544
T. Takamura-Enya et al. / Food and Chemical Toxicology 41 (2003) 543–550
xenoestrogens with weak estrogenic activity (Soto et al., 1991; Routledge and Sumpter, 1996; Garrett et al., 1999; Graumann et al., 1999). Foodstuffs and other materials may contain plant- or fungal-derived estrogens and it is also conceivable that humans are exposed to known or unknown xenoestrogens through various routes in daily life (Clemetson et al., 1978; Ju et al., 2000; Lascombe et al., 2000). Dietary factors may affect in the production, metabolism and bioavailability of sex hormones and their impact on target tissues. Soy products have been suggested to be able to reduce serum estrogen levels (Nagata et al., 1998; Kumar et al., 2002; Maskarinec et al., 2002). Thus, it is essential to monitor the levels of these estrogenic compounds in the environment in order to undertake appropriate measures to prevent harmful effects. The purpose of the present study was to determine the estrogenicity of certain food items and cigarette smoke using an estrogen-inducible yeast screening assay (yeast estrogen screen; YES) (Nishikawa et al., 1999). This method employs a yeast strain, Saccharomyces cerevisiae, transfected with human estrogen receptor and an estrogen-responsible element linked to the reporter gene lac Z, encoding the enzyme b-galactosidase. Thus, in the presence of an estrogenic compound, b-galactosidase is synthesized to cause a colorimetric change of the substrate O-nitrophenyl-b-d-galactopyranoside (ONPG). Using this assay, we here found soy-derived materials, beer, coffee, and cigarette smoke condensates to contain estrogen-like substances. Moreover, the active principles in soy-derived materials and beer were able to be identified.
2. Materials and methods 2.1. Instrumentation Electron impact-mass spectra (EI-MS) were recorded on a Shimadzu GC–MS QP-5050A. 2.2. Sample preparation Food samples assayed in this study were as follows. Vegetables: adzuki (Phaselous angularis), carrots (Daucus sativus), Japanese horseradish (Eutrema wasabi), kidney beans (Phaselous vulgaris), sesame (Sesamum indicum) and spinach (Spinacia oleracea). Beverages: beer (brand A, lager, Tokyo, Japan), black tea (Ceylon), cocoa, instant coffee (blended, Hyogo, Japan), green tea (Shizuoka, Japan), pomegranite juice (containing 20% fresh pomegranite juice), mixed vegetable juice (containing 20 vegetables) and red wine (Bordeaux, France) and white wine (Cabernet Sauvignon, France). Processed foodstuffs: boiled corn, cheese (processed cheese, Hokkaido, Japan), miso (soybean paste, brand A, Ibaragi,
Japan), soy sauce (brand A, Ibaragi, Japan), tofu (soybean cake, brand A, Niigata, Japan), Worcester sauce (Worcester, UK). Cooked foods: broiled fish and grilled beef. All items were purchased from Japanese local grocery shops except for the broiled fish and grilled beef, which were prepared by grilling commercially available fish (cod fish, Gadus macrocephalus) and beef over a naked gas flame for 15 min in our laboratory. Cigarette smoke condensates were collected under standard conditions as described previously (Sato et al., 1977) with minor modifications. The cigarettes (84 mm length) were common Japanese brands with 15 mm length filter with charcoal filter (9 mm length). Cigarette smoking was performed in a small hood by an automatic smoking machine. Cigarettes were smoked to onethird butt length at a standard condition of a 35-ml puff volume per 2 s, once per min. The mainstream was collected on two superposed glass fiber papers (Pallflex 2500QAT 45 mm). From the above samples, solid samples (10–100 g) were extracted with 80% methanol under sonication for 30 min and evaporated to dryness. In the case of liquids, 1-l samples were pretreated with hexane (500 ml2), and then extracted with ethyl acetate containing 10% methanol. Hexane and ethyl acetate–methanol fractions were evaporated to dryness. All the extracted samples were dissolved in 1 ml of dimethyl sulfoxide (DMSO) and subjected to YES assay. Hexane fraction in each sample in this study did not show lac Z activity. 2.3. Yeast estrogen screen (YES) The yeast strain, which exerts two-hybrid based estrogen-inducible lac Z expression systems, is a kind gift of Dr. Nishikawa, Osaka University, Japan. (Nishikawa et al., 1999). Briefly, the yeast cells (strain Y190) were transformed with the two fusion plasmids, one of which is pGBT9 consisting of GAL4 DNA binding domain (GAL4DBD) and rat estrogen receptor a ligand binding domain (ER LBD), and the other was pGAD424 with GAL4 activation domain (GAL4AD) and TIF1 coactivator. The yeast strain Y190 originally harbors a GAL4 binding site upstream of lac Z reporter gene. If only GAL4DBD-ER interacts with GAL4AD-TIF1, GAL4AD recruits the basal transcriptional machinery to the promoter region of the lac Z. 2.4. -Galactosidase assay Yeast was grown overnight at 30 C in 10 ml of selective medium (Nishikawa et al., 1999). The next day, the 50 ml of overnight culture was diluted to 950 ml of fresh medium containing 10 ml of DMSO solution of sample and regrown overnight at 30 C. From this reaction culture, two sets of 300 ml of reaction culture in 1.5 ml sample tubes were prepared for b-galactosidase
T. Takamura-Enya et al. / Food and Chemical Toxicology 41 (2003) 543–550
assay and the other 100 ml of culture was taken for cell density measurements at a UV absorbance of 600 nm (A600). For the galactosidase assay, yeast cells in the 300 ml reaction culture were collected by centrifugation at 14,000 rpm at 4 C for 2 min and resuspended in 750 ml of Z-buffer (60 mm Na2HPO4, 40 mm NaH2PO4, 10 mm KCl, 1 mm MgSO4, 35 mm b-mercaptoethanol). The cells were pulverised by the addition of 25 ml of chloroform and 25 ml of 0.1% sodium dodecyl sulfate followed by vortexing. The reaction was incubated at 30 C for 10 min, then 250 ml of ONPG (4 mg/ml in Z buffer) was added. After incubation at 30 C for an appropriate time, the reactions were terminated by the addition of 400 ml of 1 m Na2CO3, the cell debris was removed by centrifugation at 14000 rpm for 2 min, and the absorbance at 420 nm (A420) of the sample was measured. Lac Z units were determined by the following formula: [A420/(A600 of 1/10 dilution of cellsvolume of culturelength of incubation)]1000. The assay was performed twice at each concentration and the data is the average of two independent assays. Standard deviations of this assay were within 10% at each concentration. 2.5. HPLC separation of YES-active samples Extracted samples from foodstuffs and cigarette smoke condensates containing estrogen-like substances, were first applied to a Sephadex LH-20 column (1.520 cm) with methanol as an eluant. Aliquots of 5 ml of eluate were collected and assayed with YES. b-Galactosidase positive fractions were then evaporated to dryness and redissolved in 60% methanol for further HPLC separation, performed by a Tosoh CCPM pump equipped with a Tosoh CM8020 UV detector at 254 nm at ambient temperature. A column of TSK-Gel ODS80Ts (4.6250 mm; Tosoh Co. Tokyo, Japan) was used with a gradient solvent system (A: 30% acetonitrile, B: 80% acetonitrile; linear gradient: 0 min, 0% B; 45 min, 60% B; 60 min, 100% B) at a flow rate of 1.0 ml/min. Each 1 ml of eluate was fractionated and aliquots were assayed with YES. Active fractions were then collected and further analyzed by other HPLC systems shown below. 2.6. Identification of the active principles in food samples by HPLC In the case of soybean products, the active HPLC fraction was evaporated and redissolved in 0.5 ml of 50% acetonitrile, then applied to a Capcell PAK UG80 column (4.6250 mm, Shiseido, Tokyo, Japan) and eluted with a linear gradient of 0–60% acetonitrile containing 0.1% trifluoroacetic acid over the course of 30 min at a flow rate of 1.5 ml/min. Monitoring was at a wavelength of 262 nm.
545
In the case of beer extracts, the active fraction was separated with a Shim-pack CLC-Ph column (4.6250 mm, Shimadzu, Kyoto, Japan) with a linear gradient system of 30–50% acetonitrile over the course of 45 min at a flow rate of 1.0 ml/min. The eluate was monitored at a wavelength of 254 nm.
3. Results 3.1. Screening for estrogenic activity in foodstuffs and cigarette smoke condensates We tested 23 food samples and cigarette smoke condensates by the YES assay, and found that beer, coffee, soy sauce, miso, tofu, cigarette smoke condensate extracts all showed estrogenic activity in a dose-dependent manner (Fig. 1). The other samples did not show clear lac Z activity within the doses tested in this study. The results showed that the activity of extracts from 10 ml of beer was almost the same as that of extracts from 10 mg of miso, both of which were almost equivalent to that of 17b-estradiol at 1 pmole. Extracts of 1 g of instant coffee, 0.1 g of tofu and 1 ml of soy sauce all showed similar b-galactosidase activity, around 20% of that of 1 pmole 17b-estradiol as a positive control. Cigarette smoke condensates also exhibited YES activity in a dose-dependent manner. 3.2. Identification of genistein as an estrogenic principle in soy products To obtain further insight into the chemical structure of estrogen-mimicking compounds present in YESpositive foods such as miso, tofu and soy sauce, we isolated the active principles using HPLC with a TSK-Gel ODS-80Ts column. A YES-active fraction was eluted at the same retention time as an authentic sample of genistein. Further purification of the fractions with HPLC with a Capcell PAK UG80 column confirmed genistein to be a major active compound, which contributed almost all the YES activity in the extracted materials from three soy-derived foods. As an example, results for a miso sample are shown in Fig. 2. The UV spectrum of the single peak compound with estrogenic activity was identical with that of the authentic sample of genistein. To confirm the general presence of genistein in soybean products, we further analyzed other brand samples as follows, miso (brand B and brand C), soy sauce (brand B and C), tofu (brand B and C). Extraction and pretreatment methods of these samples were the same as those prepared for YES samples. The levels of genistein in each sample are shown in Table 1. Every brand of soy products contained genistein, and concentrations in miso and tofu samples were 1.41–394 mg/g, and those in soy sauce were 0.1–2.62 mg/ml.
546
T. Takamura-Enya et al. / Food and Chemical Toxicology 41 (2003) 543–550
Fig. 1. Estrogenic activities of extract samples from several foodstuffs and cigarette smoke condensates in YES. The abscissa shows the original volume or weights of samples before extraction. Data for 17b-estradiol as a positive control are also given.
3.3. Identification of 8-prenylnaringenin as an active principle in beer extracts Beer extracts were separated using HPLC with a TSK-Gel ODS 80-Ts column and each fractionated sample was assayed with YES. Further purification of the active fraction by HPLC with a Shim-pack CLC-Ph column allowed the isolation of an active compound as a single peak fraction (Fig. 3). The UV/vis spectrum with absorption maxima at 294 and 336 nm of the active compound is shown in Fig. 4. Fig. 5 shows the EI-MS spectrum of the active compound, with a parent molecular ion peak of m/z=340. Based on these spectrum data, the estrogenic compound was speculated to be a prenylnaringenin derivative (Tahara et al., 1994). Direct comparison with various spectrum data and HPLC
analyses of authentic prenylnaringenin derivatives synthesized with reported methods (Tahara et al., 1994) indicated the estrogenic compound in beer to be 8-prenylnaringenin (Fig. 6). As shown in Fig. 6, 8-prenylnaringenin showed estrogenic activity in the YES assay in a dose-dependent manner, ranging from 108 m to 106 m. This activity was almost one-hundredth of that of 17b-estradiol. In contrast, 6-prenylnaringenin and 30 -prenylnaringenin and their parent flavanone, naringenin, exhibited no activities in the YES system at doses of 108–103 m. The concentration of 8-prenylnaringenin in a beer sample assayed in this study was calculated to be 4.0 ng/ml by HPLC analysis. Other beer samples, brand B and C, were also found to contain 8-prenylnaringenin, and their levels were 0.52 and 0.24 ng/ml, respectively (Table 2).
T. Takamura-Enya et al. / Food and Chemical Toxicology 41 (2003) 543–550
547
Fig. 2. Isolation of genistein as the active principle in miso. The HPLC separation was performed using a column of Capcell PAK UG80 eluted with a linear gradient of 0–60% acetonitrile containing 0.1% trifluoroacetic acid over the course of 30 min at a flow rate of 1.5 ml/min (upper panel). Each 1-ml eluate was fractionated and assayed with the YES (lower panel). The peak indicated with an arrow was confirmed to co-elute with authentic genistein, whose chemical structure is also shown. Other soy-derived foods, tofu and soy sauce showed similar chromatograms.
3.4. HPLC profiling of estrogenic substances in extracts of coffee and cigarette smoke condensates Extracts of coffee and cigarette smoke condensates were also subjected to LH-20 column and active fractions were further applied for HPLC. HPLC fractionation of coffee extracts showed one major YES active fraction at the retention times of 12–14 min (Fig. 7). Retention time
of this fraction differs from that of genistein, daizein, formononetin or biochanin A, that were known to be YES active compounds. Similar results were also obtained from other brands of instant coffee (data not shown), indicating coffee generally comprises estrogenic substances. This active fraction was further subjected to other HPLC systems (a Capcell PAK UG80 column
Table 1 Concentrations of genistein in soy products Soy productsa Miso Brand A Brand B Brand C
Amounts (mg/g or ml) 394 3.22 138
Soy sauce Brand A Brand B Brand C
0.10 2.41 2.62
Tofu Brand A Brand B Brand C
9.00 1.41 7.80
a
Each brand type of samples was as follows. Miso; brand A; Ibaragi, Japan, brand B; Nagano, Japan, brand C; Aichi, Japan. Soy sauce; brand A; Chiba, Japan, brand B; Chiba, Japan, brand C; Chiba, Japan. Tofu: brand A; Niigata, Japan, brand B; Niigata, Japan, brand C; Iwate, Japan.
Fig. 3. HPLC isolation of the active principle in beer extracts with estrogenic activity. HPLC separation, shown in upper part of the figure, was performed using a Shim-pack CLC-Ph column, with a linear gradient system of 30–50% acetonitrile over the course of 45 min at a flow rate of 1.0 ml/min. Each fraction was examined with the YES, and the resultant data are shown in the lower part of the figure.
548
T. Takamura-Enya et al. / Food and Chemical Toxicology 41 (2003) 543–550
Fig. 4. UV/vis spectrum of the active principle in beer extracts. The UV/vis spectrum was obtained in 60% acetonitrile using a photodiode array UV/vis spectrometer.
with a linear gradient of 0–80% acetonitrile containing 0.1% trifluoroacetic acid over the course of 30 min at a flow rate of 1.5 ml/min, or CLC-Ph column with a linear gradient system of 30–80% acetonitrile over the course of 45 min at a flow rate of 1.0 ml/min), but we could not gain any active fractions. This means that the major active fraction contains unstable YES active compound(s) and/or YES active samples in the further HPLC fractions is present but under the detection limit for YES. Extracts of cigarette smoke condensates were also separated by HPLC (Fig. 7). Almost of all fractions showed estrogenic activity in YES, suggesting cigarette smoke condensates is the crude mixture of YES active compounds. Because of the unstable property of sample and also the complexitiy of compounds in cigarette smoke condensates, we could not isolate the estrogenic compounds in these active fractions. We have also assayed extracts of original cigarettes with YES, but resulting in no YES activity. Thus, YES active compounds in cigarette smoke condensate were suggested to be pyro-synthesized from cigarettes. Table 2 Concentration of 8-prenylnaringenin in beer
Fig. 5. EI-MS spectrum of the active principle in beer extracts.
Beera
Amounts (ng/ml)
Brand A Brand B Brand C
4.0 0.52 0.22
a Brands and types of beer were as follows. Brand A: lager, Tokyo, Japan. Brand B: pilsner, Tokyo, Japan. Brand C: draft, Tokyo, Japan.
Fig. 6. Estrogenic activity of 8-prenylnaringenin in the YES. The abscissa shows the final concentration and the activity of 17b-estradiol as a positive control is given. The chemical structure of 8-prenylnaringenin is also shown.
Fig 7. HPLC separation of extracts of coffee and cigarette smoke condensates. Each 1-min fraction was assayed with YES and their activities are shown in upper panel for coffee, and lower panel for cigarette condensates. Details of HPLC conditions are shown in Material and Methods.
T. Takamura-Enya et al. / Food and Chemical Toxicology 41 (2003) 543–550
4. Discussion We document here estrogenic activity in miso, soysauce, tofu, beer, coffee and cigarette smoke condensates, with genistein found to be the major compound in the three soybean-related food materials. Concentrations of genistein in food samples assayed in this study were in the range from 0.1 to 394 mg/g or ml. In beer, 8-prenylnaringenin was identified as the estrogenic compound. This compound has been identified in hops (Humulus lupulus) (Milligan et al., 1999) and the roots of yellow lupin (Tahara et al., 1994) and proved to be estrogenic in an in vitro bioassay using human cancer cell lines (Milligan et al., 1999). The concentrations of 8-prenylnaringenin in beer established here were 0.2–4.0 ng/ml. No estrogenic activity was evident for naringenin, the parent flavanone, or the 6- and 30 -derivatives in the YES assay. It is interesting that only introduction of a prenyl alkenyl chain at the 8-position of naringenin results in estrogenic activity, at a level 10 times that of genistein and equivalent to that of coumestrol, although 100 times less than that of 17b-estradiol. The data obtained in the present study are of great interest in determining the daily intake of major estrogenic compounds. Based on the b-galactosidase activity in the YES assay, combined with the quantification of genistein and 8-prenylnaringenin, the estrogenic activities of 10–1000 mg miso and 10–50 ml of beer were calculated to be essentially similar to that of 1 pmole 17b-estradiol. In vivo levels of 17b-estradiol range from 0.2 to 1 nm, suggesting that estrogenic activity derived from food materials taken in our daily life may be less than that from in vivo 17b-estradiol. It is known that genistein can act both as an estrogen and an antiproliferative agent, depending on the dose and target tissues. Although a specific mechanism of action has not been identified, genistein inhibits tyrosine protein kinase (Akiyama et al., 1987), DNA topoisomerase II (Kaufuman, 1998), epidermal growth factorinduced phosphatidylinositol turnover (Imoto et al., 1988), S6 kinase activation (Linassier et al., 1990) and angiogenesis (Fotsis et al., 1993). High consumption of soybean-related products containing genistein has been suggested to contribute to a reduction in the risk of breast cancer in epidemiological studies (Messina et al., 1994). In animal studies, genistein was demonstrated to be effective in inhibiting development of mammary tumors in 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine-treated rats, although it was also reported to enhance azoxymethane-induced colon carcinogenesis in male rats (Rao et al., 1997; Ohta et al., 2000). Recently, genistein was also found to cause DNA breaks in MLL via topoisomerase II inhibition, resulting in chromosome translocations leading to leukemia (Strick et al., 2000). The other major active principle detected here, 8-prenylnaringenin, exerts a variety of actions including estrogenic
549
activity as identified by a reporter gene assay using human endometrial cell lines (Milligan et al., 1999). This compound is toxic to the fungus Cladosporium herbaum (Tahara et al., 1994), and is a specific CYP1A1 inhibitor (Henderson et al., 2000). 8-Prenylnaringenin had higher binding activity to estrogen receptor than genistein. Given the available information on multifunctionality of genistein and 8-prenylnaringenin, it is now important to determine their balance of adverse and beneficial effects in vivo, including estrogenicity. Our results showed that coffee also exhibits estrogenic activity, the identity of the most potent fraction differing from four authentic flavonoids, genistein, daizein, biochanin A and formononetin. Coffee has been known to contain estrogenic compounds and cafestol is the putative candidates for its estrogenicity (Chakravorty, 1943). However, in the YES system, cafestol did not stimulate any significant b-galactosidase activities. We are now attempting further purification of the most estrogenic fraction. Cigarette smoke condensates were also found to exhibit estrogen-like activity in the present YES assay. Chromatographic separation allowed demonstration of more potent estrogenicity but, surprisingly, many fractions were found to be YES-positive. The estrogenicity of cigarette smoke condensate extracts decreased within a relatively short period of time, suggesting unstable characteristics of the estrogenic compounds involved. Recent work also showed that some hydroxy-polyaromatic hydrocarbon had estrogenic activities (Machala et al., 2001). These compounds, present in cigarette smoke condensate, may contribute to YES activity. In summary, the present study showed that certain soy-derived foodstuffs, beer, coffee and cigarette smoke condensates exhibit estrogenic activity as determined by theYES assay. The active principles were identified as genistein and 8-prenylnaringenin in soy-derived materials and beer, respectively. Further studies are necessary to examine the in vivo estrogenic activity of each of these two compounds. It is also important to identify the active principles in coffee and cigarette smoke condensates. The results obtained should be very helpful for understanding the effects of environmental materials on human health.
Acknowledgements We are grateful to Dr. T. Watanabe (Kyoto Pharmaceutical University) for kindly performing the mass spectroscopy measurements and Dr. Nishikawa (Osaka University) for kindly providing of the yeast. This study was supported by a Grant-in-Aid for Cancer Research and a Grant-in-Aid for Research on Environmental Health from the Ministry of Health, Labour and Welfare, Japan.
550
T. Takamura-Enya et al. / Food and Chemical Toxicology 41 (2003) 543–550
References Akiyama, T., Ishida, J., Nakagawa, S., Ogawa, H., Watanabe, S., Itou, N., Shibuya, M., Fukami, Y., 1987. Genistein, a specific inhibitor of tyrosine-specific protein kinase. Journal of Biological Chemistry 262, 5592–5595. Breinholt, V., Larsen, J.C., 1998. Detection of weak estrogenic flavonoids using a recombinant yeast strain and a modified MCF7 cell proliferation assay. Chemical Research in Toxicology 11, 622–629. Chakravorty, P.N., Wesner, M.M., Levin, R.H., 1943. Cafesterol II. Journal of the Chemical Society 65, 929–932. Clemetson, C.A.B., de Carlo, S.J., Burney, G.A., Patel, T.J., Kozhiashvili, N., Taylor, R.A., 1978. Estrogens in food: the almond mystery. International Journal of Gynecology and Obstetrics 15, 515–521. Colborn, T., von Saal, F.S., Soto, A.M., 1993. Development effects of endocrine-disrupting chemicals in wildlife and humans. Environmental Health Perspectives 101, 378–384. Colborn, T., 1995. Environmental estrogens: health implications for humans and wildlife. Environmental Health Perspectives 103 (Suppl. 7), 135–136. Daston, G.P., Gooch, J.W., Breslin, W.J., Shuey, D.L., Nikiforov, A.I., Fico, T.A., Gorsuch, J.W., 1997. Environmental estrogens and reproductive health: a discussion of the human and environmental data. Reproductive Toxicology 11, 465–481. Denison, M.S., Helferich, W.G. (Eds.), 1998. Toxicant–Receptor Interactions. Taylor & Francis, Philadelphia. Fotsis, T., Pepper, M., Adlercreutz, H., Fleischmann, G., Hase, T., Montesano, R., Schweigerer, L., 1993. Genistein, a dietary-derived inhibitor of in vitro angiogenesis. Proceedings of the National Academy of Sciences of USA 90, 2690–2694. Garrett, S.D., Lee, H.A., Morgan, M.R.A., 1999. A nonisotopic estrogen receptor-based assay to detect estrogenic compounds. Nature Biotechnology 17, 1219–1222. Graumann, K., Breithofer, A., Jungbauer, A., 1999. Monitoring of estrogen mimics by a recombinant yeast assay: synergy between natural and synthetic compounds? Science of the Total Environment 225, 69–79. Henderson, M.C., Miranda, C.L., Stevens, J.F., Deinzer, M.L., Buhler, D.R., 2000. In vitro inhibition of human P450 enzymes by prenylated flavonoids from hops, Humulus lupulus. Xenobiotica 30, 235–251. Imoto, M., Yamashita, T., Sawa, T., Kurasawa, S., Naganawa, H., Takeuchi, T., Bao-quan, Z., Umezawa, K., 1988. Inhibition of cellular phoshatidylinositol turnover by psi-tectorigenin. FEBS Letters 230, 43–46. IUPAC, 1998. Natural and anthropogenic environmental oestrogens. Pure and Applied Chemistry 70. Ju, Y.H., Carlson, K.E., Sun, J., Pathak, D., Katzenellenbogen, B.S., Katzenellenbogen, J.A., Hefelich, W.G., 2000. Estrogenic effects of extracts from cabbage, fermented cabbage, and acidified Brussels sprouts on growth and gene expression of estrogen-dependent human breast cancer (MCF-7) cells. Journal of Agricultural and Food Chemistry 48, 4628–4634. Kaufuman, W.K., 1998. Human topoisomerase II function, tyrosin phosphorylation and cell cycle checkpoint. Proceedings of the Society for Experimental Biology and Medicine 217, 327–334. Korach, K.S. (Ed.), 1998. Reproductive and Developmental Toxicology. Marcel Dekker, New York. Kumar, N.B., Cantor, A., Allen, K., Riccardi, D., Cox, C.E., 2002. The specific role of isoflavones on estrogen metabolism in premenopausal women. Cancer 94, 1166–1174. Lascombe, I., Beffa, D., Ru¨egg, U., Tarradellas, J., Wahli, W., 2000. Estrogenic activity assessment of environmental chemicals using in
vitro assays: identification of two new estrogenic compounds. Environmental Health Perspectives 108, 621–629. Linassier, C., Pierre, M., Le Pecq, J., Pierre, J., 1990. Mechanisms of action in NIH-3T3 cells of genistein, an inhibitor of EFG receptor tyrosine kinase activity. Biochemistry and Pharmacology 39, 187– 193. Machala, M., Ciganek, M., Blaha, L., Minksova, K., Vondrack, J., 2001. Aryl hydrocarbon receptor-mediated and estrogenic activities of oxygenated polycyclic aromatic hydrocarbons and azaarenes originally identified in extracts of river sediments. Environmental Toxicology and Chemistry 20, 2736–2743. Maskarinec, G., Williams, A.E., Inouye, J.S., Stanczyk, F.Z., Franke, A.A., 2002. A randomized isoflavone intervention among premenopausal women. Cancer Epidemiology Biomarkers and Prevention 11, 195–201. Messina, M.J., Persky, V., Setchell, K.D.R., Barnes, S., 1994. Soy intake and cancer risk: a review of the in vitro and in vivo data. Nutrition and Cancer 21, 113–131. Milligan, S.R., Kalita, J.C., Heyerick, A., Rong, H., de Cooman, L., de Keukeleire, D., 1999. Identification of a potent phytoestrogen in hops (Humulus lipulus L.) and beer. Journal of Clinical Endocrinology and Metabolism 83, 2249–2252. Nagata, C., Takatsuka, N., Inaba, S., Kawakami, N., Shimizu, H., 1998. Effect of soymilk consumption on serum estrogen concentrations in premenopausal Japanese women. Jounal of the National Cancer Institute 90, 1830–1835. Nishikawa, J., Saito, K., Goto, J., Dakeyama, F., Matsuo, M., Nishihara, T., 1999. New screening methods for chemicals with hormonal activities using interaction of nuclear hormone receptor with coactivator. Toxicology and Applied Pharmacology 154, 76–83. Ohta, T., Nakatsugi, S., Watanabe, K., Kawamori, T., Ishikawa, F., Morotomi, M., Sugie, S., Toda, T., Sugimura, T., Wakabayashi, K., 2000. Inhibitory effects of Bifidobacterium-fermented soy milk on 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine-induced rat mammary carcinogenesis, with a partial contribution of its component isoflavones. Carcinogenesis 21, 937–941. Rao, C.V., Wang, C.X., Simi, B., Lubet, R., Kelloff, G., Steele, V., Reddy, B.S., 1997. Enhancement of experimental colon cancer by genistein. Cancer Research 57, 3717–3722. Routledge, E.J., Sumpter, J.P., 1996. Estrogenic activity of surfactants and some of their degradation products assessed using a recombinant yeast screen. Environmental Toxicology and Chemistry 15, 241–248. Safe, S.H., 2000. Endocrine disruptors and human health-is there a problem? An update. Environmental Health Perspectives 108, 487– 493. Sato, S., Seino, Y., Ohka, T., Yahagi, T., Nagao, M., Matsushima, T., Sugimura, T., 1977. Mutagenicity of smoke condensates from cigarettes, cigars and pipe tobacco. Cancer Letters 3, 1–8. Soto, A.M., Justicia, H., Wray, J.W., Sonnenschein, C., 1991. p-Nonyl-phenol: an estrogenic xenobiotic released from ‘‘modified’’ polystyrene. Environmental Health Perspectives 92, 167–173. Strick, R., Strissel, P.L., Borgers, S., Smith, S.L., Rowley, J.D., 2000. Dietary bioflavonoids induce cleavage in the MLL gene and may contribute to infant leukemia. Proceedings of the National Academy Sciences of USA 97, 4790–4795. Tahara, S., Katagiri, Y., Ingham, J.L., Mizutani, J., 1994. Prenylated flavonoids in the roots of yellow lupin. Phytochemistry 36, 1261– 1271. Waller, C.I., Oprea, T.I., Chae, K., Park, H.-K., Korach, K.S., Laws, S.C., Wiese, T.E., Kelce, W.R., Gray Jr., L.E., 1996. Ligand-based identification of environmental estrogens. Chemical Research in Toxicology 9, 1240–1248.
Analysis of estrogenic activity of foodstuffs and cigarette smoke condensates using a yeast estrogen screening method T. Takamura-Enyaa,*, J. Ishiharaa, S. Taharab, S. Gotoc, Y. Totsukaa, T. Sugimuraa, K. Wakabayashia a Cancer Prevention Division, National Cancer Center Research Institute, 1-1 Tsukiji 5 chome, Chuo-ku, Tokyo 104-0045, Japan Division of Applied Bioscience, Graduate School of Agriculture, Hokkaido University, Kita-ku Sapporo 060-8589, Sapporo, Japan c Research Center for Material Cycles and Waste Management, National Institute for Environmental Studies, 16-2, Onogawa, Tsukuba, Ibaragi 305-0053, Japan
b
Accepted 14 October 2002
Abstract Hormone mimics present in our environment are of concern because such agents could potentially reduce fertility and increase sexual dysfunction in wildlife and increase the risk of breast and reproductive organ cancers in man. Therefore, monitoring of the levels of estrogenic compounds in environmental materials is essential in order to prevent their exposure to man and to discover potential harmful effects on human health. In the present study, we analyzed estrogenic activity in 23 foodstuffs and cigarette smoke condensate samples extracted with an organic solvent, using the yeast estrogen screening (YES) system. Three soybean-related foodstuffs (soy sauce, tofu, miso), beer, coffee and cigarette smoke condensates showed clear estrogenic activity in the YES system. HPLC fractionations followed by the YES of theseYES-positive samples revealed the presence of many estrogenic compounds in cigarette smoke condensates, whereas the other samples exerted estrogenic activities in only one or two fractions. Genistein was able to be isolated as the major active principle in soy sauce, tofu and miso, its concentration in these three foodstuffs ranging from 0.1 to 394 mg/g or ml. 8-Prenylnaringenin was also isolated from beer extracts as a major compound with estrogenic activity present at 0.22–4.0 ng/ml. Estrogenic activity of 8-prenylnaringenin with YES was 10-times as high as that of genistein, although it was 100-times less than that of 17b-estradiol. Based on our results in vitro, 10 mg miso and 10 ml beer can be calculated to have similar estrogenic activity to 1 pmole 17b-estradiol. It is very important that the effects of genistein and 8-prenylnaringenin on human health are elucidated. # 2003 Elsevier Science Ltd. All rights reserved. Keywords: Estrogenic activity; Foodstuffs; Yeast estrogen screening; Soybeans; Beer; Coffee; Cigarette smoke condensates; 8-Prenylnaringenin; Genistein
1. Introduction A wide variety of chemicals present in our environment have now been found to exhibit estrogenic properties or act like estrogen. These hormone mimics are hypothesized to be associated with reduced fertility and increased sexual dysfunction in wildlife animals as well as with an increased risk for breast and reproductive Abbreviations: DMSO, dimethyl sulfoxide; ONPG, O-nitrophenyl-b-d-galactopyranoside; YES, yeast estrogen screening. * Corresponding author. Tel.: +81-3-3542-2511x4551; fax: +81-33543-9305. E-mail address: [email protected] (T. Takamura-Enya).
tract cancers in humans (Colborn et al., 1993; Colborn, 1995; Daston et al., 1997; Safe, 2000). These compounds include natural substances found in plants (phytoestrogens), fungi (mycoestrogens) and those that are man-made (Waller et al., 1996; Denison and Helferich, 1998; IUPAC, 1998; Korach, 1998). Flavonoids such as genistein and coumestrol in some plant foods exhibit estrogenic activity (Breinholt and Larsen, 1998; IUPAC, 1998). b-Resorcylic acid lactones such as zearalenone are well-known fungus estrogens found in Fusarium, which infects corn (IUPAC, 1998). Certain pesticides and herbicides such as organochlorine compounds, combustion pollutants and phthalate plasticizers, have now been found to be
0278-6915/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved. PII: S0278-6915(02)00305-8
544
T. Takamura-Enya et al. / Food and Chemical Toxicology 41 (2003) 543–550
xenoestrogens with weak estrogenic activity (Soto et al., 1991; Routledge and Sumpter, 1996; Garrett et al., 1999; Graumann et al., 1999). Foodstuffs and other materials may contain plant- or fungal-derived estrogens and it is also conceivable that humans are exposed to known or unknown xenoestrogens through various routes in daily life (Clemetson et al., 1978; Ju et al., 2000; Lascombe et al., 2000). Dietary factors may affect in the production, metabolism and bioavailability of sex hormones and their impact on target tissues. Soy products have been suggested to be able to reduce serum estrogen levels (Nagata et al., 1998; Kumar et al., 2002; Maskarinec et al., 2002). Thus, it is essential to monitor the levels of these estrogenic compounds in the environment in order to undertake appropriate measures to prevent harmful effects. The purpose of the present study was to determine the estrogenicity of certain food items and cigarette smoke using an estrogen-inducible yeast screening assay (yeast estrogen screen; YES) (Nishikawa et al., 1999). This method employs a yeast strain, Saccharomyces cerevisiae, transfected with human estrogen receptor and an estrogen-responsible element linked to the reporter gene lac Z, encoding the enzyme b-galactosidase. Thus, in the presence of an estrogenic compound, b-galactosidase is synthesized to cause a colorimetric change of the substrate O-nitrophenyl-b-d-galactopyranoside (ONPG). Using this assay, we here found soy-derived materials, beer, coffee, and cigarette smoke condensates to contain estrogen-like substances. Moreover, the active principles in soy-derived materials and beer were able to be identified.
2. Materials and methods 2.1. Instrumentation Electron impact-mass spectra (EI-MS) were recorded on a Shimadzu GC–MS QP-5050A. 2.2. Sample preparation Food samples assayed in this study were as follows. Vegetables: adzuki (Phaselous angularis), carrots (Daucus sativus), Japanese horseradish (Eutrema wasabi), kidney beans (Phaselous vulgaris), sesame (Sesamum indicum) and spinach (Spinacia oleracea). Beverages: beer (brand A, lager, Tokyo, Japan), black tea (Ceylon), cocoa, instant coffee (blended, Hyogo, Japan), green tea (Shizuoka, Japan), pomegranite juice (containing 20% fresh pomegranite juice), mixed vegetable juice (containing 20 vegetables) and red wine (Bordeaux, France) and white wine (Cabernet Sauvignon, France). Processed foodstuffs: boiled corn, cheese (processed cheese, Hokkaido, Japan), miso (soybean paste, brand A, Ibaragi,
Japan), soy sauce (brand A, Ibaragi, Japan), tofu (soybean cake, brand A, Niigata, Japan), Worcester sauce (Worcester, UK). Cooked foods: broiled fish and grilled beef. All items were purchased from Japanese local grocery shops except for the broiled fish and grilled beef, which were prepared by grilling commercially available fish (cod fish, Gadus macrocephalus) and beef over a naked gas flame for 15 min in our laboratory. Cigarette smoke condensates were collected under standard conditions as described previously (Sato et al., 1977) with minor modifications. The cigarettes (84 mm length) were common Japanese brands with 15 mm length filter with charcoal filter (9 mm length). Cigarette smoking was performed in a small hood by an automatic smoking machine. Cigarettes were smoked to onethird butt length at a standard condition of a 35-ml puff volume per 2 s, once per min. The mainstream was collected on two superposed glass fiber papers (Pallflex 2500QAT 45 mm). From the above samples, solid samples (10–100 g) were extracted with 80% methanol under sonication for 30 min and evaporated to dryness. In the case of liquids, 1-l samples were pretreated with hexane (500 ml2), and then extracted with ethyl acetate containing 10% methanol. Hexane and ethyl acetate–methanol fractions were evaporated to dryness. All the extracted samples were dissolved in 1 ml of dimethyl sulfoxide (DMSO) and subjected to YES assay. Hexane fraction in each sample in this study did not show lac Z activity. 2.3. Yeast estrogen screen (YES) The yeast strain, which exerts two-hybrid based estrogen-inducible lac Z expression systems, is a kind gift of Dr. Nishikawa, Osaka University, Japan. (Nishikawa et al., 1999). Briefly, the yeast cells (strain Y190) were transformed with the two fusion plasmids, one of which is pGBT9 consisting of GAL4 DNA binding domain (GAL4DBD) and rat estrogen receptor a ligand binding domain (ER LBD), and the other was pGAD424 with GAL4 activation domain (GAL4AD) and TIF1 coactivator. The yeast strain Y190 originally harbors a GAL4 binding site upstream of lac Z reporter gene. If only GAL4DBD-ER interacts with GAL4AD-TIF1, GAL4AD recruits the basal transcriptional machinery to the promoter region of the lac Z. 2.4. -Galactosidase assay Yeast was grown overnight at 30 C in 10 ml of selective medium (Nishikawa et al., 1999). The next day, the 50 ml of overnight culture was diluted to 950 ml of fresh medium containing 10 ml of DMSO solution of sample and regrown overnight at 30 C. From this reaction culture, two sets of 300 ml of reaction culture in 1.5 ml sample tubes were prepared for b-galactosidase
T. Takamura-Enya et al. / Food and Chemical Toxicology 41 (2003) 543–550
assay and the other 100 ml of culture was taken for cell density measurements at a UV absorbance of 600 nm (A600). For the galactosidase assay, yeast cells in the 300 ml reaction culture were collected by centrifugation at 14,000 rpm at 4 C for 2 min and resuspended in 750 ml of Z-buffer (60 mm Na2HPO4, 40 mm NaH2PO4, 10 mm KCl, 1 mm MgSO4, 35 mm b-mercaptoethanol). The cells were pulverised by the addition of 25 ml of chloroform and 25 ml of 0.1% sodium dodecyl sulfate followed by vortexing. The reaction was incubated at 30 C for 10 min, then 250 ml of ONPG (4 mg/ml in Z buffer) was added. After incubation at 30 C for an appropriate time, the reactions were terminated by the addition of 400 ml of 1 m Na2CO3, the cell debris was removed by centrifugation at 14000 rpm for 2 min, and the absorbance at 420 nm (A420) of the sample was measured. Lac Z units were determined by the following formula: [A420/(A600 of 1/10 dilution of cellsvolume of culturelength of incubation)]1000. The assay was performed twice at each concentration and the data is the average of two independent assays. Standard deviations of this assay were within 10% at each concentration. 2.5. HPLC separation of YES-active samples Extracted samples from foodstuffs and cigarette smoke condensates containing estrogen-like substances, were first applied to a Sephadex LH-20 column (1.520 cm) with methanol as an eluant. Aliquots of 5 ml of eluate were collected and assayed with YES. b-Galactosidase positive fractions were then evaporated to dryness and redissolved in 60% methanol for further HPLC separation, performed by a Tosoh CCPM pump equipped with a Tosoh CM8020 UV detector at 254 nm at ambient temperature. A column of TSK-Gel ODS80Ts (4.6250 mm; Tosoh Co. Tokyo, Japan) was used with a gradient solvent system (A: 30% acetonitrile, B: 80% acetonitrile; linear gradient: 0 min, 0% B; 45 min, 60% B; 60 min, 100% B) at a flow rate of 1.0 ml/min. Each 1 ml of eluate was fractionated and aliquots were assayed with YES. Active fractions were then collected and further analyzed by other HPLC systems shown below. 2.6. Identification of the active principles in food samples by HPLC In the case of soybean products, the active HPLC fraction was evaporated and redissolved in 0.5 ml of 50% acetonitrile, then applied to a Capcell PAK UG80 column (4.6250 mm, Shiseido, Tokyo, Japan) and eluted with a linear gradient of 0–60% acetonitrile containing 0.1% trifluoroacetic acid over the course of 30 min at a flow rate of 1.5 ml/min. Monitoring was at a wavelength of 262 nm.
545
In the case of beer extracts, the active fraction was separated with a Shim-pack CLC-Ph column (4.6250 mm, Shimadzu, Kyoto, Japan) with a linear gradient system of 30–50% acetonitrile over the course of 45 min at a flow rate of 1.0 ml/min. The eluate was monitored at a wavelength of 254 nm.
3. Results 3.1. Screening for estrogenic activity in foodstuffs and cigarette smoke condensates We tested 23 food samples and cigarette smoke condensates by the YES assay, and found that beer, coffee, soy sauce, miso, tofu, cigarette smoke condensate extracts all showed estrogenic activity in a dose-dependent manner (Fig. 1). The other samples did not show clear lac Z activity within the doses tested in this study. The results showed that the activity of extracts from 10 ml of beer was almost the same as that of extracts from 10 mg of miso, both of which were almost equivalent to that of 17b-estradiol at 1 pmole. Extracts of 1 g of instant coffee, 0.1 g of tofu and 1 ml of soy sauce all showed similar b-galactosidase activity, around 20% of that of 1 pmole 17b-estradiol as a positive control. Cigarette smoke condensates also exhibited YES activity in a dose-dependent manner. 3.2. Identification of genistein as an estrogenic principle in soy products To obtain further insight into the chemical structure of estrogen-mimicking compounds present in YESpositive foods such as miso, tofu and soy sauce, we isolated the active principles using HPLC with a TSK-Gel ODS-80Ts column. A YES-active fraction was eluted at the same retention time as an authentic sample of genistein. Further purification of the fractions with HPLC with a Capcell PAK UG80 column confirmed genistein to be a major active compound, which contributed almost all the YES activity in the extracted materials from three soy-derived foods. As an example, results for a miso sample are shown in Fig. 2. The UV spectrum of the single peak compound with estrogenic activity was identical with that of the authentic sample of genistein. To confirm the general presence of genistein in soybean products, we further analyzed other brand samples as follows, miso (brand B and brand C), soy sauce (brand B and C), tofu (brand B and C). Extraction and pretreatment methods of these samples were the same as those prepared for YES samples. The levels of genistein in each sample are shown in Table 1. Every brand of soy products contained genistein, and concentrations in miso and tofu samples were 1.41–394 mg/g, and those in soy sauce were 0.1–2.62 mg/ml.
546
T. Takamura-Enya et al. / Food and Chemical Toxicology 41 (2003) 543–550
Fig. 1. Estrogenic activities of extract samples from several foodstuffs and cigarette smoke condensates in YES. The abscissa shows the original volume or weights of samples before extraction. Data for 17b-estradiol as a positive control are also given.
3.3. Identification of 8-prenylnaringenin as an active principle in beer extracts Beer extracts were separated using HPLC with a TSK-Gel ODS 80-Ts column and each fractionated sample was assayed with YES. Further purification of the active fraction by HPLC with a Shim-pack CLC-Ph column allowed the isolation of an active compound as a single peak fraction (Fig. 3). The UV/vis spectrum with absorption maxima at 294 and 336 nm of the active compound is shown in Fig. 4. Fig. 5 shows the EI-MS spectrum of the active compound, with a parent molecular ion peak of m/z=340. Based on these spectrum data, the estrogenic compound was speculated to be a prenylnaringenin derivative (Tahara et al., 1994). Direct comparison with various spectrum data and HPLC
analyses of authentic prenylnaringenin derivatives synthesized with reported methods (Tahara et al., 1994) indicated the estrogenic compound in beer to be 8-prenylnaringenin (Fig. 6). As shown in Fig. 6, 8-prenylnaringenin showed estrogenic activity in the YES assay in a dose-dependent manner, ranging from 108 m to 106 m. This activity was almost one-hundredth of that of 17b-estradiol. In contrast, 6-prenylnaringenin and 30 -prenylnaringenin and their parent flavanone, naringenin, exhibited no activities in the YES system at doses of 108–103 m. The concentration of 8-prenylnaringenin in a beer sample assayed in this study was calculated to be 4.0 ng/ml by HPLC analysis. Other beer samples, brand B and C, were also found to contain 8-prenylnaringenin, and their levels were 0.52 and 0.24 ng/ml, respectively (Table 2).
T. Takamura-Enya et al. / Food and Chemical Toxicology 41 (2003) 543–550
547
Fig. 2. Isolation of genistein as the active principle in miso. The HPLC separation was performed using a column of Capcell PAK UG80 eluted with a linear gradient of 0–60% acetonitrile containing 0.1% trifluoroacetic acid over the course of 30 min at a flow rate of 1.5 ml/min (upper panel). Each 1-ml eluate was fractionated and assayed with the YES (lower panel). The peak indicated with an arrow was confirmed to co-elute with authentic genistein, whose chemical structure is also shown. Other soy-derived foods, tofu and soy sauce showed similar chromatograms.
3.4. HPLC profiling of estrogenic substances in extracts of coffee and cigarette smoke condensates Extracts of coffee and cigarette smoke condensates were also subjected to LH-20 column and active fractions were further applied for HPLC. HPLC fractionation of coffee extracts showed one major YES active fraction at the retention times of 12–14 min (Fig. 7). Retention time
of this fraction differs from that of genistein, daizein, formononetin or biochanin A, that were known to be YES active compounds. Similar results were also obtained from other brands of instant coffee (data not shown), indicating coffee generally comprises estrogenic substances. This active fraction was further subjected to other HPLC systems (a Capcell PAK UG80 column
Table 1 Concentrations of genistein in soy products Soy productsa Miso Brand A Brand B Brand C
Amounts (mg/g or ml) 394 3.22 138
Soy sauce Brand A Brand B Brand C
0.10 2.41 2.62
Tofu Brand A Brand B Brand C
9.00 1.41 7.80
a
Each brand type of samples was as follows. Miso; brand A; Ibaragi, Japan, brand B; Nagano, Japan, brand C; Aichi, Japan. Soy sauce; brand A; Chiba, Japan, brand B; Chiba, Japan, brand C; Chiba, Japan. Tofu: brand A; Niigata, Japan, brand B; Niigata, Japan, brand C; Iwate, Japan.
Fig. 3. HPLC isolation of the active principle in beer extracts with estrogenic activity. HPLC separation, shown in upper part of the figure, was performed using a Shim-pack CLC-Ph column, with a linear gradient system of 30–50% acetonitrile over the course of 45 min at a flow rate of 1.0 ml/min. Each fraction was examined with the YES, and the resultant data are shown in the lower part of the figure.
548
T. Takamura-Enya et al. / Food and Chemical Toxicology 41 (2003) 543–550
Fig. 4. UV/vis spectrum of the active principle in beer extracts. The UV/vis spectrum was obtained in 60% acetonitrile using a photodiode array UV/vis spectrometer.
with a linear gradient of 0–80% acetonitrile containing 0.1% trifluoroacetic acid over the course of 30 min at a flow rate of 1.5 ml/min, or CLC-Ph column with a linear gradient system of 30–80% acetonitrile over the course of 45 min at a flow rate of 1.0 ml/min), but we could not gain any active fractions. This means that the major active fraction contains unstable YES active compound(s) and/or YES active samples in the further HPLC fractions is present but under the detection limit for YES. Extracts of cigarette smoke condensates were also separated by HPLC (Fig. 7). Almost of all fractions showed estrogenic activity in YES, suggesting cigarette smoke condensates is the crude mixture of YES active compounds. Because of the unstable property of sample and also the complexitiy of compounds in cigarette smoke condensates, we could not isolate the estrogenic compounds in these active fractions. We have also assayed extracts of original cigarettes with YES, but resulting in no YES activity. Thus, YES active compounds in cigarette smoke condensate were suggested to be pyro-synthesized from cigarettes. Table 2 Concentration of 8-prenylnaringenin in beer
Fig. 5. EI-MS spectrum of the active principle in beer extracts.
Beera
Amounts (ng/ml)
Brand A Brand B Brand C
4.0 0.52 0.22
a Brands and types of beer were as follows. Brand A: lager, Tokyo, Japan. Brand B: pilsner, Tokyo, Japan. Brand C: draft, Tokyo, Japan.
Fig. 6. Estrogenic activity of 8-prenylnaringenin in the YES. The abscissa shows the final concentration and the activity of 17b-estradiol as a positive control is given. The chemical structure of 8-prenylnaringenin is also shown.
Fig 7. HPLC separation of extracts of coffee and cigarette smoke condensates. Each 1-min fraction was assayed with YES and their activities are shown in upper panel for coffee, and lower panel for cigarette condensates. Details of HPLC conditions are shown in Material and Methods.
T. Takamura-Enya et al. / Food and Chemical Toxicology 41 (2003) 543–550
4. Discussion We document here estrogenic activity in miso, soysauce, tofu, beer, coffee and cigarette smoke condensates, with genistein found to be the major compound in the three soybean-related food materials. Concentrations of genistein in food samples assayed in this study were in the range from 0.1 to 394 mg/g or ml. In beer, 8-prenylnaringenin was identified as the estrogenic compound. This compound has been identified in hops (Humulus lupulus) (Milligan et al., 1999) and the roots of yellow lupin (Tahara et al., 1994) and proved to be estrogenic in an in vitro bioassay using human cancer cell lines (Milligan et al., 1999). The concentrations of 8-prenylnaringenin in beer established here were 0.2–4.0 ng/ml. No estrogenic activity was evident for naringenin, the parent flavanone, or the 6- and 30 -derivatives in the YES assay. It is interesting that only introduction of a prenyl alkenyl chain at the 8-position of naringenin results in estrogenic activity, at a level 10 times that of genistein and equivalent to that of coumestrol, although 100 times less than that of 17b-estradiol. The data obtained in the present study are of great interest in determining the daily intake of major estrogenic compounds. Based on the b-galactosidase activity in the YES assay, combined with the quantification of genistein and 8-prenylnaringenin, the estrogenic activities of 10–1000 mg miso and 10–50 ml of beer were calculated to be essentially similar to that of 1 pmole 17b-estradiol. In vivo levels of 17b-estradiol range from 0.2 to 1 nm, suggesting that estrogenic activity derived from food materials taken in our daily life may be less than that from in vivo 17b-estradiol. It is known that genistein can act both as an estrogen and an antiproliferative agent, depending on the dose and target tissues. Although a specific mechanism of action has not been identified, genistein inhibits tyrosine protein kinase (Akiyama et al., 1987), DNA topoisomerase II (Kaufuman, 1998), epidermal growth factorinduced phosphatidylinositol turnover (Imoto et al., 1988), S6 kinase activation (Linassier et al., 1990) and angiogenesis (Fotsis et al., 1993). High consumption of soybean-related products containing genistein has been suggested to contribute to a reduction in the risk of breast cancer in epidemiological studies (Messina et al., 1994). In animal studies, genistein was demonstrated to be effective in inhibiting development of mammary tumors in 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine-treated rats, although it was also reported to enhance azoxymethane-induced colon carcinogenesis in male rats (Rao et al., 1997; Ohta et al., 2000). Recently, genistein was also found to cause DNA breaks in MLL via topoisomerase II inhibition, resulting in chromosome translocations leading to leukemia (Strick et al., 2000). The other major active principle detected here, 8-prenylnaringenin, exerts a variety of actions including estrogenic
549
activity as identified by a reporter gene assay using human endometrial cell lines (Milligan et al., 1999). This compound is toxic to the fungus Cladosporium herbaum (Tahara et al., 1994), and is a specific CYP1A1 inhibitor (Henderson et al., 2000). 8-Prenylnaringenin had higher binding activity to estrogen receptor than genistein. Given the available information on multifunctionality of genistein and 8-prenylnaringenin, it is now important to determine their balance of adverse and beneficial effects in vivo, including estrogenicity. Our results showed that coffee also exhibits estrogenic activity, the identity of the most potent fraction differing from four authentic flavonoids, genistein, daizein, biochanin A and formononetin. Coffee has been known to contain estrogenic compounds and cafestol is the putative candidates for its estrogenicity (Chakravorty, 1943). However, in the YES system, cafestol did not stimulate any significant b-galactosidase activities. We are now attempting further purification of the most estrogenic fraction. Cigarette smoke condensates were also found to exhibit estrogen-like activity in the present YES assay. Chromatographic separation allowed demonstration of more potent estrogenicity but, surprisingly, many fractions were found to be YES-positive. The estrogenicity of cigarette smoke condensate extracts decreased within a relatively short period of time, suggesting unstable characteristics of the estrogenic compounds involved. Recent work also showed that some hydroxy-polyaromatic hydrocarbon had estrogenic activities (Machala et al., 2001). These compounds, present in cigarette smoke condensate, may contribute to YES activity. In summary, the present study showed that certain soy-derived foodstuffs, beer, coffee and cigarette smoke condensates exhibit estrogenic activity as determined by theYES assay. The active principles were identified as genistein and 8-prenylnaringenin in soy-derived materials and beer, respectively. Further studies are necessary to examine the in vivo estrogenic activity of each of these two compounds. It is also important to identify the active principles in coffee and cigarette smoke condensates. The results obtained should be very helpful for understanding the effects of environmental materials on human health.
Acknowledgements We are grateful to Dr. T. Watanabe (Kyoto Pharmaceutical University) for kindly performing the mass spectroscopy measurements and Dr. Nishikawa (Osaka University) for kindly providing of the yeast. This study was supported by a Grant-in-Aid for Cancer Research and a Grant-in-Aid for Research on Environmental Health from the Ministry of Health, Labour and Welfare, Japan.
550
T. Takamura-Enya et al. / Food and Chemical Toxicology 41 (2003) 543–550
References Akiyama, T., Ishida, J., Nakagawa, S., Ogawa, H., Watanabe, S., Itou, N., Shibuya, M., Fukami, Y., 1987. Genistein, a specific inhibitor of tyrosine-specific protein kinase. Journal of Biological Chemistry 262, 5592–5595. Breinholt, V., Larsen, J.C., 1998. Detection of weak estrogenic flavonoids using a recombinant yeast strain and a modified MCF7 cell proliferation assay. Chemical Research in Toxicology 11, 622–629. Chakravorty, P.N., Wesner, M.M., Levin, R.H., 1943. Cafesterol II. Journal of the Chemical Society 65, 929–932. Clemetson, C.A.B., de Carlo, S.J., Burney, G.A., Patel, T.J., Kozhiashvili, N., Taylor, R.A., 1978. Estrogens in food: the almond mystery. International Journal of Gynecology and Obstetrics 15, 515–521. Colborn, T., von Saal, F.S., Soto, A.M., 1993. Development effects of endocrine-disrupting chemicals in wildlife and humans. Environmental Health Perspectives 101, 378–384. Colborn, T., 1995. Environmental estrogens: health implications for humans and wildlife. Environmental Health Perspectives 103 (Suppl. 7), 135–136. Daston, G.P., Gooch, J.W., Breslin, W.J., Shuey, D.L., Nikiforov, A.I., Fico, T.A., Gorsuch, J.W., 1997. Environmental estrogens and reproductive health: a discussion of the human and environmental data. Reproductive Toxicology 11, 465–481. Denison, M.S., Helferich, W.G. (Eds.), 1998. Toxicant–Receptor Interactions. Taylor & Francis, Philadelphia. Fotsis, T., Pepper, M., Adlercreutz, H., Fleischmann, G., Hase, T., Montesano, R., Schweigerer, L., 1993. Genistein, a dietary-derived inhibitor of in vitro angiogenesis. Proceedings of the National Academy of Sciences of USA 90, 2690–2694. Garrett, S.D., Lee, H.A., Morgan, M.R.A., 1999. A nonisotopic estrogen receptor-based assay to detect estrogenic compounds. Nature Biotechnology 17, 1219–1222. Graumann, K., Breithofer, A., Jungbauer, A., 1999. Monitoring of estrogen mimics by a recombinant yeast assay: synergy between natural and synthetic compounds? Science of the Total Environment 225, 69–79. Henderson, M.C., Miranda, C.L., Stevens, J.F., Deinzer, M.L., Buhler, D.R., 2000. In vitro inhibition of human P450 enzymes by prenylated flavonoids from hops, Humulus lupulus. Xenobiotica 30, 235–251. Imoto, M., Yamashita, T., Sawa, T., Kurasawa, S., Naganawa, H., Takeuchi, T., Bao-quan, Z., Umezawa, K., 1988. Inhibition of cellular phoshatidylinositol turnover by psi-tectorigenin. FEBS Letters 230, 43–46. IUPAC, 1998. Natural and anthropogenic environmental oestrogens. Pure and Applied Chemistry 70. Ju, Y.H., Carlson, K.E., Sun, J., Pathak, D., Katzenellenbogen, B.S., Katzenellenbogen, J.A., Hefelich, W.G., 2000. Estrogenic effects of extracts from cabbage, fermented cabbage, and acidified Brussels sprouts on growth and gene expression of estrogen-dependent human breast cancer (MCF-7) cells. Journal of Agricultural and Food Chemistry 48, 4628–4634. Kaufuman, W.K., 1998. Human topoisomerase II function, tyrosin phosphorylation and cell cycle checkpoint. Proceedings of the Society for Experimental Biology and Medicine 217, 327–334. Korach, K.S. (Ed.), 1998. Reproductive and Developmental Toxicology. Marcel Dekker, New York. Kumar, N.B., Cantor, A., Allen, K., Riccardi, D., Cox, C.E., 2002. The specific role of isoflavones on estrogen metabolism in premenopausal women. Cancer 94, 1166–1174. Lascombe, I., Beffa, D., Ru¨egg, U., Tarradellas, J., Wahli, W., 2000. Estrogenic activity assessment of environmental chemicals using in
vitro assays: identification of two new estrogenic compounds. Environmental Health Perspectives 108, 621–629. Linassier, C., Pierre, M., Le Pecq, J., Pierre, J., 1990. Mechanisms of action in NIH-3T3 cells of genistein, an inhibitor of EFG receptor tyrosine kinase activity. Biochemistry and Pharmacology 39, 187– 193. Machala, M., Ciganek, M., Blaha, L., Minksova, K., Vondrack, J., 2001. Aryl hydrocarbon receptor-mediated and estrogenic activities of oxygenated polycyclic aromatic hydrocarbons and azaarenes originally identified in extracts of river sediments. Environmental Toxicology and Chemistry 20, 2736–2743. Maskarinec, G., Williams, A.E., Inouye, J.S., Stanczyk, F.Z., Franke, A.A., 2002. A randomized isoflavone intervention among premenopausal women. Cancer Epidemiology Biomarkers and Prevention 11, 195–201. Messina, M.J., Persky, V., Setchell, K.D.R., Barnes, S., 1994. Soy intake and cancer risk: a review of the in vitro and in vivo data. Nutrition and Cancer 21, 113–131. Milligan, S.R., Kalita, J.C., Heyerick, A., Rong, H., de Cooman, L., de Keukeleire, D., 1999. Identification of a potent phytoestrogen in hops (Humulus lipulus L.) and beer. Journal of Clinical Endocrinology and Metabolism 83, 2249–2252. Nagata, C., Takatsuka, N., Inaba, S., Kawakami, N., Shimizu, H., 1998. Effect of soymilk consumption on serum estrogen concentrations in premenopausal Japanese women. Jounal of the National Cancer Institute 90, 1830–1835. Nishikawa, J., Saito, K., Goto, J., Dakeyama, F., Matsuo, M., Nishihara, T., 1999. New screening methods for chemicals with hormonal activities using interaction of nuclear hormone receptor with coactivator. Toxicology and Applied Pharmacology 154, 76–83. Ohta, T., Nakatsugi, S., Watanabe, K., Kawamori, T., Ishikawa, F., Morotomi, M., Sugie, S., Toda, T., Sugimura, T., Wakabayashi, K., 2000. Inhibitory effects of Bifidobacterium-fermented soy milk on 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine-induced rat mammary carcinogenesis, with a partial contribution of its component isoflavones. Carcinogenesis 21, 937–941. Rao, C.V., Wang, C.X., Simi, B., Lubet, R., Kelloff, G., Steele, V., Reddy, B.S., 1997. Enhancement of experimental colon cancer by genistein. Cancer Research 57, 3717–3722. Routledge, E.J., Sumpter, J.P., 1996. Estrogenic activity of surfactants and some of their degradation products assessed using a recombinant yeast screen. Environmental Toxicology and Chemistry 15, 241–248. Safe, S.H., 2000. Endocrine disruptors and human health-is there a problem? An update. Environmental Health Perspectives 108, 487– 493. Sato, S., Seino, Y., Ohka, T., Yahagi, T., Nagao, M., Matsushima, T., Sugimura, T., 1977. Mutagenicity of smoke condensates from cigarettes, cigars and pipe tobacco. Cancer Letters 3, 1–8. Soto, A.M., Justicia, H., Wray, J.W., Sonnenschein, C., 1991. p-Nonyl-phenol: an estrogenic xenobiotic released from ‘‘modified’’ polystyrene. Environmental Health Perspectives 92, 167–173. Strick, R., Strissel, P.L., Borgers, S., Smith, S.L., Rowley, J.D., 2000. Dietary bioflavonoids induce cleavage in the MLL gene and may contribute to infant leukemia. Proceedings of the National Academy Sciences of USA 97, 4790–4795. Tahara, S., Katagiri, Y., Ingham, J.L., Mizutani, J., 1994. Prenylated flavonoids in the roots of yellow lupin. Phytochemistry 36, 1261– 1271. Waller, C.I., Oprea, T.I., Chae, K., Park, H.-K., Korach, K.S., Laws, S.C., Wiese, T.E., Kelce, W.R., Gray Jr., L.E., 1996. Ligand-based identification of environmental estrogens. Chemical Research in Toxicology 9, 1240–1248.