Induction of 16alpha-/2-hydroxyestrone metabolite ratios in MCF-7 cells by pesticides, carcinogens, and antiestrogens does not predict mammary carcinogens.

The effects of several pesticides, mammary carcinogens, and antiestrogens on 17beta-estradiol (E2), 16alpha- and 2-hydroxylase activities, and 16alpha-/2-hydroxyestrone (OHE1) ratios were investigated in MCF-7 cells using a radiometric assay. The mammary carcinogens 7,12-dimethylbenz[a]anthracene (DMBA) and benzo[a]pyrene (BaP), respectively, increased and decreased 16alpha-/2-OHE1 ratios at some concentrations. The 16alpha-/2-OHE1 metabolite ratios for 10(-5) M kepone, atrazine, p,p'-DDE, o,p'-DDE, o,p'-DDT, and beta-hexachlorocyclohexane were 1.82 +/- 0.060, 0.71 +/- 0.027, 0.66 +/- 0.030, 1.56 +/- 0.089, 1.14 +/- 0.059, and 0.69 +/- 0.052 (mean +/- standard error), respectively, and did not show any specific trend. The effects of a series of direct and indirect acting antiestrogens on 16alpha-/2-OHE1 metabolite ratios were also investigated, and the results were compound specific. Indole-3-carbinol, tamoxifen, 4'-hydroxytamoxifen, and 9-cis,retinoic acid decreased the ratio; the effects of all trans-retinoic acid and 2,3,7,8-tetrachlorodibenzo-p-dioxin were concentration dependent; the antiestrogen ICI 182,780 increased the 16alpha-/2-OHE1 metabolite ratio. The results indicate that in MCF-7 cells treated with pesticides, mammary carcinogens, and antiestrogens, there were both increased and decreased 16alpha-/2-OHE1 metabolite ratios for each class of chemicals and the assay did not predict mammary carcinogens.

Epidemiology studies have identified several risk factors for breast cancer in women; these include early age at menarche, parity, late age of menopause, and late age at first pregnancy [reviewed in (1,2)]. All of these conditions are associated with increased lifetime exposure to estrogens, which is an important risk factor for this disease. The precise role of 17,B-estradiol (E2) in development of mammary cancer is unknown; however, it is possible that this hormone or its metabolites may influence more than one step in the complex pathways that lead to tumor formation (3)(4)(5)(6)(7).
It has been suggested that metabolism of E2 into 2-and 16a-hydroxy metabolites, such as 2-hydroxyestrone (2-OHE1) and 16a-OHE1, may play a role in mammary carcinogenesis and that 2-OHEI and 16a-OHEI may serve as biomarkers for this process (8)(9)(10). Several in vivo and in vitro studies have demonstrated that altered metabolism of E2 to give increased 16a-and lower 2-hydroxy metabolites is correlated with event(s) associated with tumor development (8)(9)(10)(11)(12)(13)(14)(15). For example, metabolism of E2 to 16a-OHE1 in mammary terminal duct lobular units, a target for mammary carcinogens, was higher in breast cancer patients versus controls (13,14). These data were consistent with laboratory animal studies in mice with different susceptibilities for development of mammary tumors (11), and it was reported that the 16a-/2-OHE1 metabolite ratio was higher in breast cancer patients versus controls (15).
It was also reported that the 16a-/2-OHEI metabolite ratio in MCF-7 cells treated with various compounds was a predictor of mammary carcinogens (8). For example, 10-5 M concentrations of various pesticides and the mammary carcinogen 7,12-dimethylbenz[a]anthracene (DMBA) increased this ratio, whereas the anticarcinogen indole-3-carbinol (I3C) and omega-3 fatty acids decreased the ratio in MCF-7 cells (8). These results were surprising for two reasons. 13C induced CYPlAl at concentrations greater than 10-4 M (16-18); therefore, it was surprising that CYPlAldependent E2 2-hydroxylase activity was induced by 10-5 M 13C. Also, most of the organochlorine compounds used in the prior study enhance CYP2B gene expression, and this response is not usually induced in cancer cell lines (19). Results of initial studies on induction of E2 2-hydroxylase activity in MCF-7 cells by pesticides, mammary carcinogens, and anticarcinogens gave results (20) that were both similar and in contrast to those previously reported (8). For example, 10-5 M 13C and DMBA, respectively, increased and decreased 2hydroxylation of E2 using the radiometric assay as previously reported; however, the mammary carcinogen benzo[a]pyrene (BaP) increased and the antiestrogens ICI 164,384 and 182,780 decreased E2 2-hydroxylase activity (20). It is also possible that variability between studies may also be related to the instability of this cell line (21,22). In this study, low passage MCF-7 cells were used and monitored for both their estrogen and aryl hydrocarbon responsiveness. This study reports the effects of pesticides, mammary carcinogens, and antiestrogens on E2 16ahydroxylase activity and 16a-/2-OHE1 metabolite ratios in MCF-7 cells. After treatment of MCF-7 cells with 10-5 M concentrations of the test compounds, the 16a-/2-OHE1 metabolite ratios in MCF-7 cells did not predict mammary carcinogens.

Materials and Methods
Chemicals. [16a-3H]E2 was custom synthesized by New England Nuclear (Boston, MA) with specific activity of 1.00 mCi/ml and a concentration of 27.9 pmol/ml, and was certified 99% pure by thin-layer chromatography   for 15 min at 1500 x g (40C). Supernatant was analyzed for radioactivity in five aliquots (0.99 ml) by scintillation counting. [16ac-3H]E2 background radioactivity was caused by the slow spontaneous loss of tritium. The background levels were determined in parallel experiments by ethanol-fixing MCF-7 cells prior to the addition of DMSO + [16ct-3H]E2 for each experiment and were subtracted from the total radioactivity of each treatment to determine the net E2 hydroxylase activity (20). The background was 0.65% of the total radioactivity and did not increase significantly during the course of these studies. The average net metabolism of control cells was 2% of the total [16a-3H]E2 substrate added. For each treatment, four plates of cells were processed at the same time; results are presented as mean ± standard error (SE) and as a percent of metabolism or metabolite ratios in controls (DMSO treatment). The value for each plate is the average of the quantitation of five aliquots of its charcoal-washed media. Each experiment was repeated at least once on a separate occasion. The standard error for the 16a-/2-OHEI ratio was determined by propagating the error of the two individual measurements, and treatments were compared against controls using Student's t-test.
Radiometric E2 2-hydroxylase activity in MCF-7 cells. To determine the E2-2hydroxylase activity, a protocol identical to the one above was utilized, using [2-3H]E2 as the radiolabeled substrate, as previously described (20). 166a-Hydroxylase activity and 2-hydroxylase activity were analyzed with analysis of variance (ANOVA) and Scheffe's test.

Discussion
Bradlow and co-workers (8) previously suggested that the effect of various pesticides on the 16a-/2-OHE1 metabolite ratios in MCF-7 cells was a biologic marker of breast cancer risk. These studies were carried out by treating MCF-7 cells for 48 hr with [3H]E2 substrates and 10-5 M concentrations of the various compounds. Previous reports have demonstrated that E2 2-hydroxylase activity is a CYPlAI-dependent response in in vitro cancer cell lines (23)(24)(25)(26), whereas CYP1A2 is the major catalyst for this enzyme activity in vivo (27)(28)(29). CYP lAI and CYP1A2 are important phase I drug-metabolizing enzymes, which are induced by various xenobiotics that bind the aryl hydrocarbon receptor. Our initial interest in reexamining this assay and its predictive utility stemmed from two observations: 1) it was reported that 10-5 M I3C induced E2 2-hydroxylase activity (8), whereas induction of CYP1A1 by 13C in MCF-7 cells requires concentrations greater than 100 pM (16,17); and 2) most of the organochlorine pesticides used in the study resemble phenobarbital as inducers of CYP2B-dependent activity, and this induction response is not observed in most cancer cell lines. Using the radiometric assay for E2 2-hydroxylase activity, the reported up regulation of this response by 13C was confirmed (20) and was also observed in the present study (Fig. 2). 13C increased E2 2-hydroxylase activity after treatment for 2 hr (20); however, the amount of metabolite formed was low. BaP also induced this response (20); this contrasted with the marked decrease of E2 2-hydroxylase activity in MCF-7 cells treated with DMBA (8,20). Previous studies on induction of E2 2-hydroxylase activity in this laboratory (20) showed some differences and similarities with data reported by Bradlow and co-workers (8). This study has reexamined induction of both E2 16a-and 2-hydroxylase activities and metabolite ratios after treatment of MCF-7 cells with diverse pesticides, mammary carcinogens, and anticarcinogens in MCF-7 cells.
The report which suggested that the 16a-/2-OHEI metabolite ratios in MCF-7 cells could be used as a predictive biologic marker for mammary carcinogens showed that the ratio was highest (>5) for atrazine, p,p'-DDE, o,p'-DDT, and p,p'-DDE and was also increased for the mammary carcinogen DMBA (8). In this study, elevated metabolite ratios were also observed for 10-5 M o,p'-DDE (1.56 ± 0.09), kepone (1.81 ± 0.06), and DMBA (2.47 ± 0.21), whereas these ratios were not significantly greater than 1 for the remaining compounds. The most striking difference in 16c-/2-OHE1 metabolite ratios was observed for the mammary carcinogens DMBA and BaP (Fig. 2). In cells treated with DMBA, metabolites ratios were higher than 1 for all concentrations, whereas the ratios for 10-6 and 10-5 M BaP were 1.17 ± 0.08 and 0.08 ± 0.009, respectively. The 16a-/2-OHEI ratios obtained for 10-5 M BaP and DMBA were the lowest and highest, respectively, for all compounds used in this study. The ratio was also investigated for a series of antiestrogens that act through the estrogen receptor (ER) (tamoxifen, 4'-hydroxytamoxifen, and ICI 182,780) or via crosstalk with the ER signaling pathway (TCDD, trans-retinoic acid, and 9-cis-retinoic acid). The 16a-/2-OHE1 ratios for most of these compounds were <1.0; however, values from 1.59 to 2.19 were observed for the antiestrogen ICI 182,780 (10-8-_10-6 M), which is being developed for treatment for mammary cancer (32).
Thus, results from these studies show that mammary carcinogens and antiestrogens both increase and decrease 16a-/2-OHEI metabolite ratios in MCF-7 cells, and the ratio was not a biologic marker of breast cancer risk. The reasons for differences between this study and the prior report (8) are unknown; however, MCF-7 cell lines can be highly variable with respect to wild-type and variant ER expression and E2 responsiveness (21,22), as well as inducibility of CYPlAI-dependent activities (23). Retention of CYPlAI inducibility was periodically monitored in this study by confirming induction of CYPlAl-dependent E2 2-hydroxylase ( Fig. 1) and ethoxyresorufin O-deethylase activities (data not shown) by TCDD. A recent study by Telang et al. (10) summarized the effects of DMBA and BaP on 16a-/2-OHE1 metabolite ratios in explant cultures of human mammary terminal duct lobular units and human mammary epithelial 1 84-B5 cells. They reported that both DMBA and BaP increased 16a-and decreased 2-hydroxylation of E2; this was accompanied by increased 16a0-/2-OHE1 metabolite ratios. Thus, in contrast to MCF-7 cells, BaP did not induce CYPlAI-dependent E2 2hydroxylase activity in these in vitro systems. It is possible that assays using explant cell cultures may be more useful than breast cancer cell lines for predicting some classes of mammary carcinogens. However, it is clear from results of this study that a diverse group of compounds should be used to test the predictive utility ofan in vitro bioassay.