Equine estrogen metabolite 4-hydroxyequilenin induces anchorage-independent growth of human mammary epithelial MCF-10A cells: differential gene expression.

Long-term exposure to synthetic and endogenous estrogens has been associated with the development of cancer in several tissues. One potential mechanism of estrogen carcinogenesis involves catechol formation and these catechols are further oxidized to electrophilic/redox active o-quinones, which have the potential to both initiate and promote the carcinogenic process. Previously we showed that 4-hydroxyequilenin (4-OHEN) autoxidized to an o-quinone and caused a variety of damage to DNA. Since these deleterious effects could contribute to gene mutations, we investigated the Chinese hamster V79 cells to ascertain the relative ability of estradiol, 4-hydroxyestradiol, 17beta-hydroxyequilenin, 4,17beta-hydroxyequilenin, estrone, 4-hydroxyestrone, equilenin, and 4-hydroxyequilenin to induce the mutation of the hypoxanthine-guanine phosphoribosyltransferase (hprt) gene. All the 4-hydroxylated catechols induced significantly more colony formations in V79 cells as compared to the parent phenols at 100nM, suggesting that the catechol estrogen metabolites are more mutagenic towards the hprt gene than estrogens. Since 4-OHEN induced the highest mutation frequency, we examined a biomarker for transformation potential of this compound in MCF-10A cells using an anchorage-independent growth assay. Although 4-OHEN induced anchorage-independent growth of these cells, the isolated clones were not able to grow as tumors in vivo when injected into nude mice. These cells were assayed for genetic changes using cDNA microarrays. Real time RT-PCR confirmation of some of the differentially expressed genes showed down-regulation of metallothionein 2A, p53, BRCA1, and c-myc. Moreover, we showed the involvement of other genes important in cell transformation and oxidative stress, strengthening the hypothesis that this mechanism plays a considerable role in 4-OHEN-induced anchorage-independent growth.

Antiestrogenic and DNA damaging effects induced by tamoxifen and toremifene metabolites.

The antiestrogen, tamoxifen, has been extensively used in the treatment and prevention of breast cancer. Although tamoxifen showed benefits in the chemotherapy and chemoprevention of breast cancer, epidemiological studies in both tamoxifen-treated breast cancer patients and healthy women indicated that treatment caused an increased risk of developing endometrial cancer. These troubling side effects lead to concerns over long-term safety of the drug. Therefore, it is important to fully understand the relationship between the antiestrogenic and the genotoxic mechanisms of tamoxifen, other antiestrogens, and their metabolites. Previously, we have shown that o-quinone formation from tamoxifen and its analogues, droloxifene and 4-hydroxytoremifene, may not contribute to the cytotoxic effects of these antiestrogens; however, these o-quinones can form adducts with deoxynucleosides and this implies that the o-quinone pathway could contribute to the genotoxicity of the antiestrogens in vivo. To further investigate this potential genotoxic pathway, we were interested in the role of estrogen receptor (ER)(1) alpha and beta since work with catechol estrogens has shown that ERs seem to enhance DNA damage in breast cancer cell lines. As a result, we investigated the binding affinities of 4-hydroxy and 3,4-dihydroxy derivatives of tamoxifen and toremifene to ER alpha and beta. The antiestrogenic activities of the metabolites using the Ishikawa cells were also investigated as well as their activity in ERalpha and ERbeta breast cancer cells using the transient transfection reporter, estrogen response element-dependent luciferase assay. The data showed that the antiestrogenic activities of these compounds in the biological assays mimicked their activities in the ER binding assay. To determine if the compounds were toxic and if ERs played a role in this process, the cytotoxicity of these compounds in ERbeta41(2) (ERbeta), S30 (ERalpha), and MDA-MB-231 (ER(-)) cell lines was compared. The results showed that the cytotoxicity differences between the metabolites were modest. In addition, all of the metabolites showed similar toxicity patterns in both ER positive and negative cell lines, which means that the ER may not contribute to the cytotoxicity pathway. Finally, we compared the amount of DNA damage induced by these metabolites in these cell lines using the comet assay. The catechols 3,4-dihydroxytoremifene and 3,4-dihydroxytamoxifen induced a greater amount of cellular single strand DNA cleavage as compared with the phenols in all cell lines. The different amounts of DNA damage in ER positive and negative cell lines suggested that the ERs might play a role in this process. These data suggest that the formation of catechols represents a minor role in cytotoxic and antiestrogenic effects in cells as compared with their phenol analogues. However, catechols induced more DNA damage at nontoxic doses in breast cancer cells, which implies that o-quinones formed from catechols could contribute to genotoxicity in vivo, which is ER-dependent.

Effect of halogenated substituents on the metabolism and estrogenic effects of the equine estrogen, equilenin.

Estrogen replacement therapy has been correlated with an increased risk for developing breast and endometrial cancers. One potential mechanism of estrogen carcinogenesis involves metabolism of estrogens to 2- and 4-hydroxylated catechols, which are further oxidized to electrophilic/redox active o-quinones that have the potential to both initiate and promote the carcinogenic process. Previously, we showed that the equine estrogens, equilin and equilenin, which are major components of the estrogen replacement formulation Premarin (Wyeth-Ayerst), are primarily metabolized to the catechol, 4-hydroxyequilenin. This catechol was found to autoxidize to an o-quinone causing oxidation and alkylation of DNA in vitro and in vivo. To block catechol formation from equilenin, 4-halogenated equilenin derivatives were synthesized. These derivatives were tested for their ability to bind to the estrogen receptor, induce estrogen sensitive genes, and their potential to form catechol metabolites. We found that the 4-fluoro derivatives were more estrogenic than the 4-chloro and 4-bromo derivatives as demonstrated by a higher binding affinity for estrogen receptors alpha and beta, an enhanced induction of alkaline phosphatase activity in Ishikawa cells, pS2 expression in S30 cells, and PR expression in Ishikawa cells. Incubation of these compounds with tyrosinase in the presence of GSH showed that the halogenated equilenin compounds formed less catechol GSH conjugates than the parent compounds, equilenin and 17beta-hydroxyequilenin. In addition, these halogenated compounds showed less cytotoxicity in the presence of tyrosinase than the parent compounds in S30 cells. Also, as stated above, the 4-fluoro derivatives showed similar estrogenic effects as compared with parent compounds; however, they were less toxic in S30 cells as compared to equilenin and 17beta-equilenin. Because 17beta-hydroxy-4-halogenated equilenin derivatives showed higher estrogenic effects than the halogenated equilenin derivatives in vitro, we studied the relative ability of the 17beta-hydroxy-4-halogenated equilenin derivatives to induce estrogenic effects in the ovariectomized rat model. The 4-fluoro derivative showed higher activity than 4-chloro and 4-bromo derivatives as demonstrated by inducing higher vaginal cellular differentiation, uterine growth, and mammary gland branching. However, 17beta-hydroxy-4-fluoroequilenin showed a lower estrogenic activity than 17beta-hydroxyequilenin and estradiol, which could be due to alternative pharmacokinetic properties for these compounds. These data suggest that the 4-fluoroequilenin derivatives have promise as alternatives to traditional estrogen replacement therapy due to their similar estrogenic properties with less overall toxicity.

Inhibition of cellular enzymes by equine catechol estrogens in human breast cancer cells: specificity for glutathione S-transferase P1-1.

Glutathione S-transferases (GSTs) are a family of detoxification isozymes that protect cells by conjugating GSH to a variety of toxic compounds, and they may also play a role in the regulation of both cellular proliferation and apoptosis. We have previously shown that human GST P1-1, which is the most widely distributed extrahepatic isozyme, could be inactivated by the catechol estrogen metabolite 4-hydroxyequilenin (4-OHEN) in vitro [Chang, M., Shin, Y. G., van Breemen, R. B., Blond, S. Y., and Bolton, J. L. (2001) Biochemistry 40, 4811-4820]. In the present study, we found that 4-OHEN and another catechol estrogen, 4,17beta-hydroxyequilenin (4,17beta-OHEN), significantly decreased GSH levels and the activity of GST within minutes in both estrogen receptor (ER) negative (MDA-MB-231) and ER positive (S30) human breast cancer cells. In addition, 4-OHEN caused significant decreases in GST activity in nontransformed human breast epithelial cells (MCF-10A) but not in the human hepatoma HepG2 cells, which lack GST P1-1. We also showed that GSH partially protected the inactivation of GST P1-1 by 4-OHEN in vitro, and depletion of cellular GSH enhanced the 4-OHEN-induced inhibition of GST activity. In addition, 4-OHEN GSH conjugates contributed about 27% of the inactivation of GST P1-1 by 4-OEHN in vitro. Our in vitro kinetic inhibition experiments with 4-OHEN showed that GST P1-1 had a lower K(i) value (20.8 microM) compared to glyceraldehyde-3-phosphate dehydrogenase (GAPDH, 52.4 microM), P450 reductase (PR, 77.4 microM), pyruvate kinase (PK, 159 microM), glutathione reductase (GR, 230 microM), superoxide dismutase (SOD, 448 microM), catalase (562 microM), GST M1-1 (620 microM), thioredoxin reductase (TR, 694 microM), and glutathione peroxidase (GPX, 1410 microM). In contrast to the significant inhibition of total GST activity in these human breast cancer cells, 4-OHEN only slightly inhibited the cellular GAPDH activity, and other cellular enzymes including PR, PK, GR, SOD, catalase, TR, and GPX were resistant to 4-OHEN-induced inhibition. These data suggest that GST P1-1 may be a preferred protein target for equine catechol estrogens in vivo.

Oxidative DNA damage induced by equine estrogen metabolites: role of estrogen receptor alpha.

Excessive exposure to synthetic and endogenous estrogens has been associated with the development of cancer in several tissues. 4-Hydroxyequilenin (4-OHEN), a major metabolite of equine estrogens present in estrogen replacement formulations, has been shown to induce cytotoxic/carcinogenic effects. In the present study, we have found that 4-OHEN caused DNA damage in breast cancer cells, and cells that contain estrogen receptor alpha (S30) are more sensitive to 4-OHEN-mediated DNA damage as compared to estrogen receptor negative cells (MDA-MB-231). For example, concentration-dependent increases in 8-oxo-deoxyguanosine (8-oxo-dG), as measured by LC-MS-MS or by the Fpg comet assay, were only detected in the S30 cells, and the amount of this lesion could be enhanced by agents, which catalyze redox cycling (NADH) or deplete GSH (diethyl maleate). The role of the estrogen receptor in modulating DNA damage was further established in incubations with the ER antagonist tamoxifen, where decreases in 8-oxo-deoxyguanosine were observed. Another equine estrogen metabolite, 4,17 beta-hydroxyequilenin (4,17 beta-OHEN), was found to have the same cytotoxicity and a similar ability to induce reactive oxygen species (ROS), and caused the same oxidative DNA damage in S30 cells as compared to 4-OHEN. However, 4,17 beta-OHEN induced twice as much single strand DNA breaks in S30 cells compared to 4-OHEN. Also 4,17 beta-OHEN was more estrogenic than 4-OHEN as demonstrated by a higher binding affinity for ER alpha and an enhanced induction in activity of estrogen-dependent alkaline phosphatase in Ishikawa cells. These data suggest that the mechanism of DNA damage induced by equine estrogen metabolites could involve oxidative stress and that the estrogen receptor may play a role in this process.