July 2007, Volume 7, Issue 1

Studies on the Influence of Dietary Isoflavones, Daidzein and Genistein, in Chemical Mutagenesis and Tumor Development in Female Rats

Anane Aidoo 1*, Tao Chen, Michelle E. Bishop, Sharon Shelton, Lascelles E. Lyn-Cook, and Mugimane G. Manjanatha

National Center for Toxicological Research (NCTR),
U.S. Food and Drug Administration
Division of Genetic & Reproductive Toxicology
Jefferson, Arkansas 72079

1 Corresponding Author’s email: anane.aidoo@fda.hhs.gov

Tel: 870.543.7331
Fax: 870.543.7373

* The views presented in this article do not necessarily reflect those of the Food and Drug Administration.

Abstract

The use of hormone replacement therapy (HRT) is known to confer several benefits, including a reduction of hot flashes, but it is linked to increased risks for heart attack, stroke, and breast cancer. Many women are switching to alternative sources, such as isoflavone phytoestrogens, in the belief that naturally occurring substances may be safer than HRT. Despite the large body of compelling evidence on the health benefits of isoflavones, it is possible that these compounds may also be toxic in that they can alter endogenous hormones and thus influence the pathogenesis of hormone-dependent cancers. Because the Western lifestyle normally precludes the consumption of adequate soy foods to meet health needs, many women are currently taking dietary supplements of phytoestrogens on a regular basis to forestall the complications of menopause. Earlier studies indicated that daidzein (DZ) and genistein (GE), major components of soy isoflavones presumed to be responsible for health-promoting effects, are mutagenic indicating that isoflavones may play a role in the carcinogenesis process. Although mutation alone may not be able to directly explain the possible carcinogenic role of isoflavones, understanding the effect of isoflavones on chemical mutagenesis provides great potential for future risk assessment strategies. At the heart of current mutation in vivo assays is the capacity to measure mutations in every tissue of Big Blue (BB) transgenic mice or rats. Thus using BB rats, ovariectomized to model menopause and to exclude the effects of intrinsic sex hormones, we investigated the effects of feeding DZ, GE, or 17β-estradiol (E2) on the genotoxicity of the potent rodent carcinogen 7,12-dimethylbenz(α)anthracene (DMBA) in several estrogen-responsive tissues, including mammary, uterus, heart, and liver. Mammary glands and uterus were also analyzed for tumorigenicity.

Introduction

The undesirable symptoms of menopause and other disorders, such as cardiovascular disease and osteoporosis, are linked to decreased levels of estrogen (1-3). To alleviate these problems, hormone replacement therapy (HRT) is often administered to menopausal women. However, in light of recent studies that call into question the safety of HRT [1, 2], many women are switching to naturally occurring estrogens such as soy isoflavones believed to exhibit beneficial effects in the prevention of menopausal symptoms and other related disorders [3, 4]. Structurally related to steroidal estrogens (Figure 1 after this paragraph), DZ and GE, the major components of soy isoflavones, exhibit similar properties for receptor affinity [5], but like tamoxifen appear to be selective estrogen modulators without untoward estrogenic side effects 6]. Thus, isoflavones have been proposed as effective chemopreventive agents for certain types of cancer, particularly breast and prostate cancers [6-8]. Evidence also points to the beneficial effects of isoflavones in preventing cardiovascular disease and osteoporosis [9, 10]. In addition, there are other constituents of soy, including lignans, protease inhibitors, saponins, phytosterols, coumestans, and phytates that might also possess health-promoting benefits [11].

Figure 1. Structures of Estrogen and Isoflavone.

Given the potential role of soy isoflavones in decreasing the risk of certain hormone-dependent cancers and ameliorating menopausal symptoms, it is not surprising that isoflavones are ingested on a daily basis as nutritional supplements or as constituents of other preparations, including tofu, soy milk, tempeh, and textured soy protein. The isoflavones can exert their effects by genomic mechanism involving estrogen receptors or through a variety of nongenomic mechanisms, including tyrosine kinase and topoisomerase inhibition [8, 12, 13]. Genistein is reported to be a potent inhibitor of topoisomerase II by stabilizing a cleavable complex that results in DNA strand breaks [14]. Thus it is conceivable that the biological effects of DZ or GE may also be associated with DNA damage and potentiate the carcinogenesis process [15]. It has been reported that isoflavones are clastogenic, both in vitro and in vivo [16-18], and are genotoxic in vivo [18, 19]. In addition, dietary genistein has been shown to enhance chemical carcinogenesis in the colon and in mammary glands of rodents [20-22]. Administration of an isoflavone mixture to a p53-deficient mouse that develops early spontaneous tumors, however, showed no effect on the incidence and types of tumors produced [18]. Although consumption of soy products in the Western countries is much less than that in Asian countries where soybean consumption is associated with beneficial effects in women [23-25], dietary supplements of isoflavones are on the rise in this country suggesting a need to evaluate their potential mutagenicity and carcinogenicity in detail.

Mutations, which are heritable alterations in the structure of a gene, are not only associated with physiological or biochemical changes in an organism, but are also involved in many diseases, including cancer. However, the assessment of mutation in certain tissues such as mammary or uterus has been hampered by a lack of relevant animal models. The use of the rat lymphocyte Hprt assay [26, 27] or the Big Blue® (BB) transgenic rat system, which harbors mutational target [28], has facilitated the measurement of in vivo mutations that can provide genetically useful information for chemical risk assessment. While the Hprt assay is useful for both animal and human studies, it is limited to circulating blood lymphocytes. On the other hand, the transgenic mutation assay can detect mutations in any tissue, including the mammary gland, heart, liver, and uterus. Thus, the BB transgenic assay readily permits direct comparison of cancer and mutation induction in the same tissues.

A significant advance in our understanding of chemical carcinogenesis in hormone-dependent tissues has come from the use of the model multi-organ carcinogen 7,12-dimethylbenz(α)anthracene (DMBA) [29-31]. DMBA is a synthetic polycyclic aromatic hydrocarbon (PAH) that has been used extensively as a prototype carcinogen in mutation and cancer research. Major target organs of this agent in rodents are the skin and mammary gland. Many other tissues are susceptible to DMBA insult [32]. Therefore, to better evaluate the genotoxic/carcinogenic effects of isoflavones, we treated transgenic BB rats with a single dose of DMBA, with or without dietary DZ, GE, or 17β-estradiol (E2) as positive control, and conducted mutagenesis and carcinogenesis experiments in the mammary and uterus. The heart and liver, which are estrogen-responsive tissues, were also included in the mutagenesis study; the lymphocytes were added as a surrogate tissue for evaluating responses in nonestrogenic tissue. Due to the existence of estrogen deficiency at menopause, both OVX and INT rats were used in the study.

Animals, diets, and carcinogen treatment

The experiments were conducted in female Big Blue® transgenic rats obtained from Taconic Farm (Germantown, N.Y.). Animals were acclimatized for two weeks at NCTR before treatment. The Institutional Animal Care and Use Committee at NCTR approved animal handling, maintenance, and the study protocol. Starting at five weeks of age, rats were fed ad libitum isoflavone-free diet (NIH-31C) and had free access to water. NIH-31C diet has the same basic formulation as standard NIH-31, except that the protein contributed by soy meal and alfalfa was replaced by casein and the soy oil by corn oil. The feed was analyzed by LC/MS and shown to be free of isoflavones at the detection limit of 0.5 μg. Beginning at 7 weeks of age, animals were fed ad libitum either NIH-31C or NIH-31C containing either 0.25 or 1.0 g/kg of the diet isoflavones: DZ and GE, respectively, or a mixture (DZG) containing 1.0 g/kg of diet DZ and 1.0 g/kg of diet GE. Other rats received 0.005 g/kg E2 as positive control in some experiments. The doses of the isoflavones used were biologically active, as evidenced by an increase in uterine cell proliferation as determined by proliferating cell nuclear antigen (PCNA) immunohistochemistry in the OVX rats. Also, in a similar experiment, 0.25 g/kg intake of GE increased the expression of Bcl2 gene while 1.0 g/kg decreased this response in the pancreatic endocrine tissue of intact rats (Lyn-Cook, L.E., unpublished information). At postnatal day (PND) 50, the rats were gavaged with a single dose of 80 mg/kg DMBA suspended in sesame oil (Figure 2 following this paragraph). This dose of DMBA has been shown to produce tumors in female Fischer 344 rats that are infrequently used for DMBA carcinogenesis compared to female Sprague-Dawley rats. The PND50 treatment was based on carcinogenesis studies indicating that rats at this age have a high density of terminal end buds: ductal structures that are more sensitive to chemically induced mammary tumors [33]. The animals were maintained on isoflavone/estradiol-supplement or the isoflavone/estradiol-free diets until the experiments were terminated.

Figure 2. Experimental Design

Animals were divided into two groups two weeks after DMBA treatment (Figure 2 above this paragraph). Group I rats were bilaterally ovariectomized (OVX) under ketamine/xylazine (100 mg/kg and 15 mg/kg, respectively) anesthesia. Group II rats were kept intact (INT). The rationale for treating the animals two weeks prior to ovariectomy is based on the fact that sensitivity of the rat-to-mammary tumor induction by DMBA is in part dependent on the hormonal state of the animal [34, 35]. Animals continued to have free access to food and water. Food consumption and body weight were recorded weekly. Animals were killed at 16 or 20 weeks following DMBA treatment by CO2 anesthesia, and tissues or organs were aseptically harvested. The spleens were used immediately for Hprt mutagenesis assay. Portions of other tissues were either frozen in liquid nitrogen and stored at - 80ºC for the lacI or cII mutagenesis assay or preserved in 10% neutral buffered formalin for histopathological analysis.

For the mammary and uterus histopathology, tissues were examined grossly, removed, and preserved in 10% neutral buffered formalin. Lesion descriptions were recorded on the IANR form (Individual Animal Necropsy Record). Tissues were trimmed, processed, and embedded in Tissue Prep II, sectioned at 4-6 microns, and stained with hematoxylin and eosin. Slides were microscopically examined and, when applicable, nonneoplastic lesions were graded for severity.

Mortality

In the OVX group, two animals died early in the study (DMBA plus 0.25 and 1.0 g/kg DZ); the cause of death was due to either surgery or gavage error, and the animals were excluded from the study. In ovary-intact animals, one rat in the group treated with DMBA and fed E2 died early in the study, but the cause of death was undetermined. Six other animals from the INT group treated with DMBA were found to be moribund between the 11th and 16th week after DMBA treatment and therefore euthanized. All of these DMBA-treated INT rats bore mammary gland adenocarcinomas and were included in the analysis of tumor-bearing animals.

Assessment of isoflavone levels in serum

Isoflavones undergo extensive metabolism in the intestinal tract prior to absorption (35). Following absorption, the metabolites and/or the parent compounds are transported to the liver where they are removed from the portal blood. However, a percentage of the isoflavones in the portal blood can escape uptake by the liver and enter the peripheral circulation. The effectiveness of hepatic first-pass clearance influences the amount that reaches peripheral tissues [36]. Therefore the concentration of DZ, GE, and equol, a metabolite of DZ, were measured by HPLC/MS. Table 1 (following this paragraph) shows the mean serum concentrations of DZ, GE, and equol, 16 weeks after the commencement of isoflavone feeding. The levels detected in the serum were physiologically relevant and biologically active, because they were within the range of isoflavone concentrations that significantly modify clinical markers of cardiovascular disease and osteoporosis [37]. The biological activity of these isoflavone levels was also demonstrated by an increase in cell proliferation and a reduction in the severity of atrophy (discussed under histopathological changes) observed in OVX rats fed DZ or GE.

Table 1. Mean Values of Serum Isoflavones (µM).

Food intake, body, and organ weights

Food intake and body weight were measured weekly during the course of the study. DZ or GE supplementation alone had no effect on body weight gain (Figure 3 following this paragraph). However, the body weight gains of E2-fed rats in both OVX (Figure 3A following this paragraph) and INT (Figure 3B following this paragraph) groups with or without DMBA treatment were markedly diminished. The E2-mediated reduction in body weight gain was not statistically significant. However, it is intriguing to note that this response by E2 has been associated with the suppression of the expression of neuropeptide-Y in the hypothalamus that regulates appetite in rats [38, 39]. We did not investigate this brain substance in the present study; however, we observed that the amount of food consumed per rat was essentially similar in all the treatment groups, and it is possible that the suppression of the neuropeptide may not be the sole determinant of E2-mediated decrease in body weight gain.

Figure 3A.

Figure 3B

Figure 3. Body weights of OVX and INT BB rats fed DZ, GE, or E2 with or without DMBA treatment.

In order to determine organ weights, all necropsy livers, kidneys, hearts, brains, uteri, thymuses, and adrenal glands were removed, examined, and weighed wet as soon as possible after dissection, while the thyroid/parathyroid and pituitary glands were weighed after fixation. In the OVX rats, the only gross observation other than decreased uterine weight was the occurrence of mammary gland (lymph node) and clitoral gland adenomas in rats fed the E2 diet. Table 2 (following this paragraph) shows the mean organ weights. While most of the treatment groups, including daidzein and genistein, demonstrated organ weights comparable to the control diet group, E2 treatment markedly increased uterine weight of OVX rats (Table 2A following this paragraph). The E2 related effects in the OVX group clearly suggest exogenous estrogen-induced dysplasia. In the INT animals, the organ weights, including uterine, were essentially similar in all the treatment groups (Table 2B following this paragraph).

Table 2A. Mean organ weights (grams) derived from OVX rats. All organ weights were essentially similar with the exception of E-2-treated rats where the uterine weight was increased.