Experimental Biology and Medicine

On Dec. 3-6, 1995, the National Center for Toxicological Research sponsored an international meeting to provide a forum for presentation and discussion of recent findings in phytoestrogen research. There is widespread exposure of the human population as well as other animal species to plant chemicals possessing estrogenic activity. This fact is the primary driving force for research in the field. The pace of investigation of the numerous roles that these chemicals play in cellular processes and their impact on human health has clearly quickened since the publication of papers from the Second International Conference on Phytoestrogens (see Proc SOC Exp Biol Med 208: 1-1 38, 1995). This is reflected in the increasing diversity of investigations, as demonstrated by the papers in this issue. In addition to the papers, I call attention to the announcement which provides instructions for access to NCTR's Phytoestrogen Bibliography located on the Internet.

Our diet and environment have changed rapidly since the start of industrialization. Our partially genetically determined behavior leads, in the affluent society, to the consumption of a diet rich in cholesterol, fat, energy, and salt and to a rapid increase in the incidence of so-called Western diseases. These diseases include all the major causes of death like various cancers and coronary heart disease. The industrially produced highly palatable diet contains too little essentially protective compounds, and our sedentary lifestyle, smoking, high alcohol consumption, and other behaviors further contribute to the development of disease. In fact cancer may be a relatively young disease going back to Hippocrates and ancient Greece about 2500 years ago, suggesting that evolution plays an important role.

In this presentation I will discuss the importance of the relations between phytoestrogens, mainly flavonoids and lignans, and steroids during evolution and the possible role of these evolutionary events and our intestinal microflora for adequate nutrition and availability of cancer protective compounds such as phytoestrogens.

Steroid Hormone-Dependent Cancer and the Structural Similarities Between Steroids and Phytoestrogens

Breast, endometrial, and prostate cancers are steroid hormone-dependent, and steroid hormone receptors have also been found in colon and other cancers. Estrogens play a definite role in the development of breast and endometrial cancer, but, in addition, they may have important roles with regard to prostate (1) and colon (2) cancer. After identification of the unknown phenolic compounds in urine (3–7), the structural similarities between the diphenolic lignans and isoflavonoids on one hand, and estrogens on the other hand and their very close molecular weights immediately suggested that they could interfere with steroid hormone metabolism or action at the cellular level or that they by themselves could have biological effects in the body. Then we found that postmenopausal breast cancer patients, living in Boston, Massachusetts, after one year's follow-up, had a very low excretion of lignans and somewhat lower equol in urine compared to vegetarians (8).

Exposure to estrogenic compounds may pose a developmental hazard to infants. Soy products, which contain the phytoestrogens, genistein and daidzein, are becoming increasingly popular as infant foods. To begin to evaluate the potential of the phytoestrogens in these products to affect infants, we measured total genistein and daidzein contents of commercially available soy-based infant formulas, infant cereals, dinners, and rusks. We also assayed phytoestrogens in dairy-based formulas and in breast milk from omnivorous or vegetarian mothers. In most cases, the gluco-side forms of the phytoestrogens were hydrolyzed before separation by HPLC.

Mean (±SEM) total genistein and daidzein contents in four soy infant formulas were 87 ± 3 and 49 ± 2 μg/g, respectively. The phytoestrogen content of cereals varied with brand, with genistein ranging from 3–287 μg/g and daidzein from 2–276 μg/g. By contrast, no phytoestrogens were detected in dairy-based infant formulas or in human breast milk, irrespective of the mother's diet (detection limit = 0.05 μg/ml). When fed according to the manufacturer's instruction, soy formulas provide the infant with a daily dose rate of total isoflavones (i.e., genistein + daidzein) of approximately 3 mg/kg body weight, which is maintained at a fairly constant level between 0–4 months of age. Supplementing the diet of 4-month-old infants with a single daily serving of cereal can increase their isoflavone intake by over 25%, depending on the brand chosen. This rate of isoflavone intake is much greater than that shown in adult humans to alter reproductive hormones. Since the available evidence suggests that infants can digest and absorb dietary phytoestrogens in active forms and since neonates are generally more susceptible than adults to perturbations of the sex steroid milieu, we suggest that it would be highly desirable to study the effects of soy isoflavones on steroid-dependent developmental processes in human babies.

The current interest in the role of dietary isoflavonoids, particularly the soy isoflavone genistein, in lowering the risk of several chronic diseases, has led to the need for rapid, sensitive and precise assays for isoflavones and their metabolites in food matrices and in various physiological fluids and tissues. HPLC has the advantage over GC-based methods in that all the conjugated and unconjugated isoflavonoids and their metabolites can be separated and analyzed without the need for derivatization. An important advance in mass spectrometry has been the introduction of effective interfaces between the HPLC and the mass spectrometer, namely the electrospray ionization (ESI) and the heated nebulizer-atmospheric pressure chemical ionization (HN-APCI) interfaces. Because of the isoflavonoid concentrations in fluids such as bile or urine, preliminary extraction, so essential for GC-MS and many other analytical methods, is not necessary. This immediately overcomes the thorny issue of finding an effective solid-phase extraction procedure. Using reversed-phase HPLC-ESI-MS, it is possible to obtain a mass/intensity map of all isoflavonoid metabolites in a single 20 min analysis. Analysis of isoflavonoid conjugates in serum/plasma samples requires initial extraction, but the conjugates can be measured intact either by capillary reversed-phase HPLC-ESI-MS or on regular reversed-phase columns by HPLC-HN-APCI-MS. In both cases, specificity is obtained by causing the parent isoflavonoid molecular ions to undergo collision-induced dissociation to form specific daughter ions in a triple quadrupole MS instrument. When it is only necessary to measure the total isoflavonoids and their metabolites in blood, hydrolysis can be performed directly in serum/plasma samples and isoflavonoids recovered by ether or ethyl acetate solvent extraction. The isoflavone aglucones can be analyzed by HPLC-MS under isocratic solvent conditions, thereby drastically shortening analysis time and opening up prospects for automation. Therefore, HPLC-MS is a technique that is broadly applicable to the major issues in phytoestrogen research.

A fast, precise and selective diode array HPLC method is presented for the extraction and analysis of soy isoflavonoids from foods and from human urine, plasma, and breast milk in support of mechanistic and epidemiologic studies assessing the potential cancer protective role of soya or isoflavones. Solid phase or solvent extraction was chosen for isolation, and enzymatic or acid hydrolysis procedures were used for aglycone production depending on the matrix to be analyzed. C-18 reversed-phase HPLC was applied to selectively separate and quantitate daidzein (1), 2 glycitein (3), 2 and genistein (4), 2 including their malonyl (a) 2 and acetyl (b) 2 esters, and their mammalian metabolites equol (6) 2 and Odesmethylangolensin (7), 2 as well as formononetin (2), 2 biochanin-A (5), 2 and coumestrol (8) 2 using a gradient elution system. UV absorbance scans and authentic standards were applied for identification purposes, additional to fluorometric monitoring, electrochemical detection, and GC/MS analysis after trimethyl silylation. Detection limits of 20-μl injections were found to be 1.09, 0.53, 3.28, and 1.00 pmoles for daidzein, genistein, equol, and Odesmethylangolensin (DMA), respectively, by monitoring at the individual compound's absorption maximum. The proposed method was applied to monitor isoflavone levels in soy foods and in human plasma, urine and breast milk after challenge with roasted soybeans. Implications of the presented results on the potential activity of isoflavones to prevent cancer by exposing newborn infants to these agents are discussed.

The study of phytoestrogens in food sources and their metabolism, effects, and mechanism of action in animals requires very selective and often sensitive analytical techniques. We have applied coulometric array detection, which uses a series of flow-through electrochemical sensors each providing 100% electrolytic efficiency, for measurement of a variety of phytochemicals in complex matrices. Recent work has involved the resolution of coumestrol (COM), daidzein (DE), daidzin (DI), diethylstil-bestrol (DES), enterodiol (ED), enterolactone (EL), equol (EQ), estradiol (E2), estriol (E3), estrone (E), genistein (GE), and quercetin (QE). Binary gradient reversed-phase (C18) chromatography was used with a sodium acetate buffer (pH 4.8)-methanol-acetonitrile solvent system. Eight coulometric sensors were set at 260, 320, 380, 440, 500, 560, 620, and 680 mV (vs Pd reference). Compounds were resolved in 30 min via both their oxidation/reduction characteristics and chromatographic behavior. Respective maximal oxidation potentials (mV) were: COM = 380; DE = 500; DI = 620; DES = 440; ED = 620; EL = 620; EQ = 560; E2 = 560; E3 = 560; E1 = 560; GE = 500; and QE = 260 with limits of detection of 5–50 pg. Uterine tissue homogenates (30 mg/ml in Tris-EDTA) and plasma from Sprague-Dawley rats sacrificed 1 hr after sc injection with either vehicle, dimethylsulfoxide, 10 μg DES, or 1.0 mg EQ were analyzed before and after enzymatic hydrolysis with β-glucuronidase/sulfatase. Urine samples from humans receiving a Boston-area diet with or without soy protein isolate supplements were also analyzed. Ethanol extracts were evaporated and reconstituted in 20% methanol before HPLC analysis. DE, ED, EL, EQ, and GE were determined in urine with less than 5% (R.S.D.) intraassay imprecision and 85%-102% recovery. Levels (ng/ml) of GE (1.8), QE (11.2), and EQ (1.7) were found in control plasma before hydrolysis and GE (293), QE (183), and EQ (22) after hydrolysis. Higher concentrations, corresponding to sc injection, in free and total EQ were found in both tissue and plasma.

Formononetin is an isoflavonoid phytoestrogen found in animal fodder and in certain human foodstuffs. Since high consumption of formononetin has been implicated in infertility among ruminants and may relate to human breast cancer, we developed a radioimmunoassay method and applied it to quantitate formononetin in murine plasma and mammary glandular tissue for animal model studies.

The radioimmunoassay utilized an antiserum raised in rabbits following immunization with formononetin-7-O(carboxymethyl)ether coupled to bovine serum albumin. The tracer was an [³H]leucine derivative of formononetin synthesized by mixed anhydride reaction between formononetin-7-O-(carboxymethyl)ether and [³H]leucine. The bound and free forms of formononetin were separated by adding dextran-coated charcoal. This radioimmunoassay procedure enabled the quantification of 4 ng/ml of plasma or 50 pg/mg of mammary tissue, and the antiserum showed no marked cross-reaction with the reactants tested. The reliability and reproducibility of the assay were demonstrated by intra- and inter-assay variation that was 6.5% and 11.9%, respectively. This radioimmunoassay was compared with a high-performance liquid chromatographic method by determining concentrations of formononetin in ethanol extracts of red clover. Good correlation existed between the radioimmunoassay and high-performance liquid chromatographic method (r = 0.980), but the radioimmunoassay values were on the average 5% higher than high-performance liquid chromatographic values. Furthermore, we used the radioimmunoassay to assess the pharmacokinetics of formononetin in castrated female BALB/c mice after a single subcutaneous injection at 40 mg/kg dose. The sensitivity of the radioimmunoassay permitted the detection of a prolonged elimination phase for formononetin in both plasma and mammary glandular tissues, with mean elimination half-lives of 2 and 2.5 hr, respectively.

A specific, sensitive, and fast radioimmunoassay procedure has been developed for the determination of formononetin in murine plasma and mammary glandular tissue. We conclude that the presented radioimmunoassay is a valid alternative method to the quantification of formononetin in biological fluids and/or plants.

Analysis of a group of phytoestrogens indicates that the majority of these chemicals are devoid of a 6Å distance descriptor biophore. This 6Å biophore is associated with the carcinogenicity of a set of chemicals in the mouse subset of the Carcinogenic Potency Database assembled by Gold et al. The prevalence of non-DNA-reactive carcinogens and chemicals endowed with estrogenic activity included in the group of chemicals possessing 6Å descriptor suggests that it describes a ligand binding site on an estrogen receptor. Evidence is presented that estrogens with and without the 6Å biophore bind to alternate receptors or to similar receptors but with different affinities. Since the 6Å biophore was identified based upon a carcinogenicity database, it is conceivable that binding to the putative receptor that recognizes this biophore is associated with carcinogenicity. Alternatively, estrogens devoid of this 6Å biophore may be noncarcinogenic, suggesting that carcinogenicity and estrogenicity may be separate phenomena.

The synthesis of novel mammalian metabolites of dietary isoflavones, dihydrodaidzein (4′,7-dihydroxyisoflavanone) 7, 2 dihydrogenistein (4′,5,7-trihydroxy-isoflavanone) 9, 6′-hydroxy-O-demethylangolensin [1-(2,4,6-trihydroxyphenyl)-2-(4-hydroxyphenyl)propan-1-one] 10, and cis-and trans-4′,7-dihydroxyisoflavan-4-ols 11, 12 is described, and their characteristics by physical and chemical constants given for the first time.

The biological activity of natural estrogens is influenced by the degree to which they bind to serum proteins. To determine directly how serum affected the uptake of estradiol, we compared the whole cell uptake of [³H]estradiol in intact MCF-7 human breast cancer cells from serum-free medium with the uptake from 100% serum from adult men. In estrogen receptor saturation assays, 28.9 times more estradiol was required in serum to occupy the same number of estrogen receptors as was required in serum-free medium (SFM), suggesting that the effective free fraction of estradiol in adult male serum was 3.46% (1 + 28.9). Since most xenoestrogens are not available in tritium-labeled form, the cell uptake of unlabeled xenoestrogens could not be measured directly with saturation analysis. Therefore, we developed the relative binding affinity-serum modified access (RBA-SMA) assay to determine the effect of serum on the access of nonradioactive xenoestrogens to estrogen receptors within intact MCF-7 cells. Serum modified access (SMA) was calculated by dividing the relative binding affinity (RBA, relative to estradiol) measured in 100% serum, by the RBA measured in serum-free medium. An SMA > 1 indicated that the xenoestrogen had greater access to estrogen receptors than estradiol from serum. In contrast, an SMA < 1 indicated that the xenoestrogen had less access to estrogen receptors from serum than did estradiol. The synthetic estrogen diethylstilbestrol (DES) binds poorly to sex hormone binding globulin (SHBG), and DES showed enhanced access in serum, SMA = 6.2. Additional calculations through the K_i (inhibition constant) indicated that this corresponded to an effective free fraction of 26.9% for DES in serum. The phytoestrogens, coumestrol, genistein, and equol, showed substantial enhanced access in serum, over 10-fold relative to estradiol (SMA = 12.1, 10.3, and 11.3, respectively), and effective free fractions in serum of 47.8, 45.8, and 49.7%, respectively. Since most in vitro assays of xenoestrogens do not address how serum influences their bioactivity, the estrogenic activity of these phytoestrogens would be underestimated. Conversely, biochanin A showed decreased access from serum (SMA = 0.44) and had an effective free fraction of 2.4%; its estrogenic activity would be overestimated in serum-free assays.

Several flavonoids and isoflavonoids were found to inhibit 17β-oxidoreduction of estrogens by the purified 17β-HSOR type 1, or in cell lines expressing 17β-HSOR type 1 enzyme (T-47D breast cancer cells) or type 2 (PC-3 prostate cancer cells). The structural demands for the inhibition of estrone (E₁) reduction and estradiol (E₂) oxidation catalyzed by 17β-HSOR types 1 and 2, respectively, were not identical. Flavones, flavanones, and isoflavones hydroxylated at both the double ring (positions 5 and 7) and ring B (position 4′) were the most potent inhibitors of E₁ reduction in T-47D cells, and by the purified type 1 enzyme whereas flavones hydroxylated at positions 3,5, and 7 of rings A and C, with or without a hydroxyl group in ring B, were capable of inhibiting E₂ oxidation in PC-3 cells. Change to flavanone structure, or hydroxylation at position 3 of ring C of flavones, or methylation of the hydroxyl group at position 4′ of ring B of flavones and isoflavones reduced or abolished their inhibitory activity on E₁ reduction in T-47D cells. On the contrary, hydroxyl group at position 3 of flavones (flavonol structure) markedly increased the inhibition of E₂ oxidation in PC-3 cells. Thus, changes in the number and location of hydroxyl groups may discriminate inhibition of E₁, reduction and E₂ oxidation. Some of the differences may be due to differences in pharmacokinetics of these compounds in T-47D and PC-3 cells. Inhibition of 17β-HSORs could lead to an alteration in the availability of the highly active endogenous estrogen, but the effects of these compounds in vivo cannot be predicted on the basis of these results alone. Some of these compounds (isoflavones) are estrogenic per se, and they may replace endogenous estrogens, whereas flavones are only very weakly estrogenic or nonestrogenic. Regarding prevention or treatment of estrogen-related diseases, apigenin, coumestrol, and genistein raise special interest.

There is considerable interest in the role(s) of plant-derived compounds such as bioflavonoids in regulating steroid hormone action in mammals, and in particular, the possible effects of the bioflavonoids on the growth of steroid-dependent breast and prostate tumors and on possible abnormal development of steroid-sensitive tissues. Studies of the hormone-like actions of bioflavonoids often use fetal or neonatal rats, which contain high levels of serum alpha-fetoprotein (AFP), a protein that binds estradiol with a K _d ~ 5 × 10⁻⁹ M. Interaction of bioflavonoids with AFP could affect the availability of estrogens to estrogen-responsive cells, as well as the actions of bioflavonoids. These considerations motivated us to study the effect of several flavonoids (quercetin, rutin, naringenin, chrysin, apigenin, kaempferol, myricetin, morin, fisetin) and isoflavonoids (daidzein, genistein) on estrogen binding to rat AFP. We found that naringenin, a flavanone, and quercetin and kaempferol, flavonols, inhibit estrogen binding to AFP with apparent K _ds of about 5 × 10⁻⁷ M. To our surprise, the two isoflavonoids, daidzein and genistein, have K _ds of about 5 × 10⁻⁶ M for AFP. This 1Q-fold difference in affinity for AFP between flavonoids and isoflavonoids suggests that AFP has a specificity for the flavonoid structure. Moreover, the affinities of bioflavonoids for rat AFP are sufficiently high to suggest that flavonoids and isoflavonoids could modulate estradiol and estrone binding to rat AFP in vivo, when present at dietary levels. Additionally, the potency of the plant estrogens may be altered by binding to AFP. The flavonoids that we tested have different hydroxyl and glucoside substituents on the A, B, and C rings, which allows us to define some of the spatial requirements for binding to AFP. We find that 5,7-hydroxyl groups in ring A and a 4′-hydroxyl group in ring B are important for binding to AFP. This information, combined with molecular modeling studies, may elucidate the molecular basis for recognition of flavonoids and estrogens by AFP. Also, these findings indicate that the flavonoid levels in the diet need to be considered in studies of the effects of various xenobiotics and endocrine manipulations on experimental animals, particularly during development when serum estrogen binding protein concentrations are often elevated. Finally, bioflavonoids should be useful tools for understanding the variety of estrogen actions initiated by different structural classes of estrogens.

Increased 16β-hydroxylation of estradiol has been shown to be associated with heightened cancer risk in estrogen responsive tissue. Certain types of human papillomavirus (HPV) are cofactors for cancer in the cervix, an estrogen sensitive tissue. We have demonstrated that estradiol and 16β-hydroxyestrone increased the number of cells positive for proliferating cell nuclear antigen in HPV immortalized keratinocytes, the in vitro correlate of the premalignant keratinocyte. These estrogens caused the abnormal proliferation and anchorage independent growth, which correlates with malignant conversion. Indole-3-carbinol, a phytochemical in cruciferous vegetables known to preferentially induce 2-hydroxylation with minimal effect on 16β-hydroxylation, markedly blocked the ability of estradiol to increase anchorage independent growth. The results indicate that 16β-hydroxyestrone increases the malignant phenotype of HPV immortalized keratinocytes. However, indole-3-carbinol will block this response.

Three DNA damage-responsive cell cycle checkpoints can be shown to operate in diploid human fibroblasts. One checkpoint arrests growth in G1, another inhibits replicon initiation in S phase cells, and the third delays progression from G2 into mitosis. Progression from G2 into M is controlled in part by a cyclin-dependent kinase (cyclin B/Cdk1) that is regulated by tyrosine phosphorylation. Phosphorylation of Tyr15 on Cdk1 is inhibitory for kinase activity. Activation of cyclin B/Cdk1 at the onset of mitosis is accomplished by a phosphatase, Cdc25C, that interacts with cyclin B/Cdk1 in an autocatalytic feedback loop to remove the inhibitory phosphate at Tyr15 and activate kinase activity. DNA damage triggers G2 delay by inhibiting formation of the autocatalytic feedback loop so that dephosphorylation of Tyr15 does not occur. This suppression of activation of cyclin B/Cdk1 appears to account for the failure of damaged G2 cells to progress into mitosis. Once the damage to DNA is repaired, cells resume progression into mitosis as the cycle is re-engaged. The isoflavone genistein inhibits tyrosine kinases, including one that phosphorylates Cdk1 on Tyr15. This kinase, p56/p53^lyn is rapidly induced by treatments that trigger cell cycle checkpoints (ionizing radiation, cytosine arabinoside), suggesting that this kinase may actively delay the onset of mitosis by phosphorylating Tyr15 on Cdk1.

Genistein also inhibits type II DNA topoisomerase to produce a form of DNA damage that triggers all of the DNA damage-responsive cell cycle checkpoints. A brief 10 min incubation with the topoisomerase poison amsacrine was sufficient to trigger the S phase checkpoint response and inhibit replicon initiation. Inhibition of replicon initiation by 1 μM amsacrine was maximal 20–30 min after drug treatment and by 120 min, the checkpoint response had decayed to allow near control rates of replicon initiation. Topoisomerase II poisons also are powerful clastogens inducing lethal and carcinogenic chromosomal aberrations. Type II topoisomerase can break DNA in a region of chromosome 11q23 that contains the ataxia telangiectasia gene (ATM). The ATM gene controls all of the DNA damage-responsive cell cycle checkpoints. Chromosomal aberrations in 11q23 are frequently seen in acute myeloid leukemia that develops as a consequence of etoposide chemotherapy. Thus, topoisomerase poisons such as genistein may trigger chromatid breakage to inactivate AT gene function, disable cell cycle control, and induce genetic instability.

Equol is an isoflavonoid phytoestrogen produced from the soy isoflavone daidzein by gut microflora. Not all humans produce equol from daidzein, presumably due to differences in colonic bacterial populations among individuals. Previously, smaller studies reported that approximately 30% of participants excreted equol when consuming soy. The purpose of our study was to determine the prevalence of equol excreters in a larger sample and to examine what dietary components might influence the tendency to be an equol excreter. Thirty men and thirty women consumed a soy protein beverage containing 22 mg genistein and 8 mg daidzein for 4 days as a supplement to their habitual diets. The mean daily nutrient content of their habitual intakes was determined from 4-day food records. On Day 4, participants provided a 24-hour urine collection. Urinary isoflavonoid (genistein, daidzein, equol, and O-desmethy-langolensin) excretion was measured by gas chromatography-mass spectrometry. Twenty-one of the 60 participants (35%) excreted equol (> 2000 nmol/day) after 3 days of consuming the soy supplement. Daily equol excretion ranged from 2,134–20,301 nmol/day in the excreters and 21–233 nmol/day in the nonexcreters. There was no difference in equol excreter prevalence between men (43%) and women (27%). Daily excretion of daidzein, genistein, and O-desmethylangolensin was similar between equol excreters and nonexcreters and between men and women. Among the women, equol excreters consumed a significantly higher percentage of energy as carbohydrate and greater amounts of plant protein and dietary fiber, both as soluble and insoluble fiber compared to nonexcreters. Such differences were not observed in the men, who overall had significantly higher fiber intakes than the women. These data suggest that, among women, dietary fiber or other components of a high-fiber diet may promote the growth and/or the activity of bacterial populations responsible for equol production in the colon.

The estrogenic actions of dietary phytoestrogens have raised concerns regarding the potential DES-like developmental effects on the female genital tract, but the growing evidence of cardioprotective benefits of dietary soybean estrogens provides the impetus to assess the effects of these compounds in adult female models of the menopause. We conducted an experiment in ovariectomized rats to determine the independent effects of dietary soybean estrogens (SBE) and the interactions of these agents with the commonly used pharmaceutical estrogen preparation (conjugated equine estrogens, CEE) in the vagina and uterus. We looked at the effects of SBE and CEE, alone and in combination, on uterine weight, body weight, vaginal cytology, uterine luminal epithelial height, and immunohistochemical staining for proliferating cell nuclear antigen (PCNA), lactoferrin (Ltf), and apoptosis. Ovariectomized rats were fed diets containing casein or soybean protein (SBE, low dose = 11.6 mg isoflavones/1800 cal; high dose = 117.8 mg/1800 cal), with no CEE, low dose CEE (0.313 mg/1800 cal), or high dose CEE (0.625 mg/1800 cal) added. In this study, SBE did not demonstrate estrogenic activity for uterine weight or vaginal cytology. We also found no estrogenic effects of these doses of SBE for PCNA, apoptosis, Ltf staining, or for LEH measurements. In addition, our results regarding the interactions of SBE and CEE do not show any evidence that the combination is additive in effect. On the contrary, the LEH response induced by low levels of CEE, was reduced by high levels of SBE. Furthermore, the Ltf response induced by CEE also was reduced by high levels of SBE. This suggests that high doses of SBE may antagonize the estrogen-agonist actions of low doses of CEE in the rat uterus. Our results in the ovariectomized rat model of menopause suggest that dietary soybean estrogens will not elicit a pattern of effects that simply recapitulates those of steroidal estrogens.

The ovariectomized (OVX), lactating rat model has been used to investigate the skeletal effects of the plant estrogen, genistein, over a 14-day period. The OVX, lactating rat on a low-calcium diet loses slightly more than 50% of its bone mineral mass during the first 2 weeks of lactation, and we have demonstrated that estrogen treatment can significantly reduce the loss of femoral mass (ash weight). Following OVX, the rats were assigned to treatment or control groups (both placebo and positive control with estrogen replacement). The treatment groups received one of three doses of a genistein-rich preparation each day via the feed for 2 weeks, after which time the pups began to have an interest in solid feed. A positive control group received conjugated estrogen in the feed. The genistein doses were: low (0.5 mg/d); intermediate (1.6 mg/d); and high (5.0 mg/d). Measurements included ash weights of the femur, scanning electron microscopy (SEM) of the proximal tibia, and uterine weights. SEM results were as follows: (1) at the low dose genistein was approximately equally effective to estrogen in the retention of cancellous bone tissue, as reflected in the number and density of trabeculae in hemisections of the tibial subepiphyseal region, but at high doses genistein was less effective; and (2) rats treated with low-dose genistein, like estradiol, had rougher endosteal surfaces and smaller pores on these surfaces than untreated control rats. Mean ash weights of the entire femur were highest in the rats treated with the low dose compared to control rats (P < 0.05), and they were higher than ash weights of rats administered the intermediate or high doses of genistein. The mean ash weights of the femurs were consistent with the genistein effects on the tibias observed by SEM. In summary, a biphasic response to the genistein preparation was found in this OVX rat model. Interpretation of the results suggests that, at the low dose, genistein appears to be an agonist at the estrogen receptor locus, whereas at higher doses the genistein is less effective and may even have adverse effects on bone cells. These findings of a biphasic effect of genistein (i.e., an inverted U effect) are consistent with those of other recent reports in the literature on isolated bone cells and on reproductive tissues. In summary, lower doses of genistein from soy foods would be expected to act similarly to estrogens with a beneficial effect on bone tissue, but at high doses that are unlikely to be consumed in human diets, this soy derivative may have potentially adverse effects on bone cell functions and thereby on bone tissue.

Genistein, a principal soy isoflavone, has been identified as a protein kinase inhibitor that possesses immunosuppressive and anti-inflammatory properties. The aim of the study was to determine if genistein modified chronic ileitis in guinea pigs induced by the hapten trinitrobenzene sulfonic acid (TNBS), and the activity index of cultured macrophages (RAW 264.7 cells) stimulated with lipopolysaccharide (LPS). Genistein at low doses (0.1 mg/kg, s.c.) had mild anti-inflammatory effects in TNBS ileitis. Therapeutic benefit included a reduction in nitric oxide production, granulocyte infiltration and improved mucosal architecture. Genistein, at low doses, also appeared to attenuate immunohistochemical staining for inducible nitric oxide synthase (iNOS) and nitrotyrosine. The beneficial effects of genistein were not apparent at doses above 0.1 mg/kg. We found that genistein also inhibited LPS-induced nitrite production by cultured macrophages and protected against LPS-induced necrosis despite its ability to cause apoptosis. These results indicate that genistein displayed mild antiinflammatory properties which may, in part, involve an attenuation of nitric oxide release via inducible nitric oxide synthase, and the formation of peroxynitrite.

Breast cancer is the most common cancer in US females and is the second leading cause of cancer death among women. By contrast, Asian women consuming a traditional diet high in soy products have a relatively low incidence of breast cancer. Asians who emigrate to the United States and adopt a Western diet lose this protection. Soy-based diets are high in phytoestrogens, and one of these components is genistein. Using the dimethylbenz(a)anthracene (DMBA) mammary cancer rodent model, we have investigated the breast cancer protective potential of genistein. Our results demonstrate that neonatal and prepubertal genistein treatments altered the ontogeny of the mammary gland and rendered the adult animals less susceptible to chemically-induced mammary cancer. Neonatal genistein treatment did not significantly alter the rate of formation and persistence of DMBA-DNA adducts in the mammary gland. While high concentrations of genistein during the neonatal period caused adverse effects on ovarian follicular development, prepubertal genistein treatment did not appear to be toxic in either the female reproductive tract or the endocrine system.

Hormone replacement therapy for postmenopausal women greatly reduces their risk of coronary heart disease. However, current pharmaceutical regimens have a low acceptance rate among postmenopausal women. We have sought to identify an alternative treatment that would retain the beneficial health effects of current standard therapy without its negative aspects. We have concentrated our research on naturally occurring estrogens (called phytoestrogens) found in soybeans, in the belief that delivery of phytoestrogens via the diet would be more acceptable than pharmaceutical regimens. Using a well-established nonhuman primate model of postmenopausal women, we have investigated the effects of soy phytoestrogens on cardiovascular risk factors and the reproductive system. In this review, we summarize the results of our ongoing research.

In this study we report on the content and bioactivity of plant (phyto) estrogens and progestins in various foods, herbs, and spices, before and after human consumption. Over 150 herbs traditionally used by herbalists for treating a variety of health problems were extracted and tested for their relative capacity to compete with estradiol and progesterone binding to intracellular receptors for progesterone (PR) and estradiol (ER) in intact human breast cancer cells. The six highest ER-binding herbs that are commonly consumed were soy, licorice, red clover, thyme, tumeric, hops, and verbena. The six highest PR-binding herbs and spices commonly consumed were oregano, verbena, tumeric, thyme, red clover and damiana. Some of the herbs and spices found to contain high phytoestrogens and phytoprogestins were further tested for bioactivity based on their ability to regulate cell growth rate in ER (+) and ER (–) breast cancer cell lines and to induce or inhibit the synthesis of alkaline phosphatase, an end product of progesterone action, in PR (+) cells. In general, we found that ER-binding herbal extracts were agonists, much like estradiol, whereas PR-binding extracts, were neutral or antagonists. The bioavailability of phytoestrogens and phytoprogestins in vivo were studied by quantitating the ER-binding and PR-binding capacity of saliva following consumption of soy milk, exogenous progesterone, medroxyprogesterone acetate, or wild mexican yam products containing diosgenin. Soy milk caused a dramatic increase in saliva ER-binding components without a concommitant rise in estradiol. Consumption of PR-binding herbs increased the progestin activity of saliva, but there were marked differences in bioactivity. In summary, we have demonstrated that many of the commonly consumed foods, herbs, and spices contain phytoestrogens and phytoprogestins that act as agonists and antagonists in Vivo.

There are several suggested health benefits of phytoestrogens, particularly those found in soy products. Herbal medicines are also widely thought to confer health benefits. Additionally, drugs are prescribed to improve human health, but unlike phytoestrogens and herbal medicines, toxicities are defined in experimental animals and monitored in humans before and after marketing. Knowledge of toxicity is crucial to decrease the risk:benefit ratio; this knowledge defines appropriate conditions for use and strategies for development of safer products. However, our awareness of the toxicity of herbal medicines and phytoestrogen-containing foods is dramatically limited compared to drugs. Some aspects of the toxicity of herbal medicines are briefly reviewed; it is concluded that virtually all of our knowledge is derived from human exposures leading to acute toxicities. Importantly, detection of toxicity is sporadic, and little information is available from prior animal experimentation. Additionally, well-organized monitoring of human populations (as occurs for drugs) is virtually nonexistent. Important toxicities with long latencies are particularly difficult to associate with a causative agent during or even after large scale exposures, as exemplified by tobacco smoking and lung cancer; estrogen replacement therapy and endometrial cancer; diethylstilbestrol and reproductive tract cancers; and fetal alcohol exposure and birth defects. These considerations suggest that much closer study in experimental animals and human populations exposed to phytoestrogen-containing products, and particularly soy-based foods, is necessary. Among human exposures, infant soy formula exposure appears to provide the highest of all phytoestrogen doses, and this occurs during development, often the most sensitive life-stage for induction of toxicity. Large, carefully controlled studies in this exposed infant population are a high priority.

Soy is a unique dietary source of the isoflavones, genistein and daidzein. It has been part of the Southeast Asian diet for nearly five millenia, whereas consumption of soy in the United States and Western Europe has been limited to the 20th century. Heavy consumption of soy in Southeast Asian populations is associated with reduction in the rates of certain cancers and cardiovascular disease. Recent experimental evidence suggests that phytochemicals in soy are responsible for its beneficial effects, which may also include prevention of osteoporosis, a hereditary chronic nose bleed syndrome, and autoimmune diseases. Exposure of soy formula-fed infants to the potential estrogenizing effects of the isoflavones is limited by the first pass effect of the liver following the uptake of isoflavones from the gut. Several mechanisms of action of isoflavones have been proposed—both through estrogendependent and estrogen-independent pathways.