Estradiol metabolism: an endocrine biomarker for modulation of human mammary carcinogenesis.

The natural estrogen 17beta-estradiol (E2) has a profound influence on proliferation and neoplastic transformation of mammary epithelium. The role of cellular metabolism of E2 in mammary carcinogenesis, however, remains to be elucidated. Explant culture and cell culture models developed from noncancerous human mammary tissue were used to examine modulation of E2 metabolism in response to treatment with prototype rodent mammary carcinogens and the ability of the naturally occurring phytochemical indole-3-carbinol (13C) to influence E2 metabolism and regulate aberrant proliferation. In the two models, treatment with the chemical carcinogens 7,12-dimethylbenz[a]anthracene and benzo[a]pyrene altered the metabolism of E2 as determined from the radiometric (tritium release) and gas chromatography-mass spectrometry (GC-MS) assays. This alteration in E2 metabolism was accompanied by aberrant proliferation and abrogation of apoptosis as determined by the extent of replicative DNA synthesis, S-phase fraction and Sub G0 (apoptotic) peak. Exposure of carcinogen-initiated cultures to 13C resulted in induction of C2-hydroxylation of E2 and of apoptosis and downregulation of hyperproliferation. Determination of altered cellular metabolism of E2 in response to initiators and modulators of carcinogenesis and evaluation of cell cycle related markers for proliferation and apoptosis may provide a mechanism-oriented approach to validate E2 metabolism as an endocrine biomarker for induction and prevention of human mammary carcinogenesis.


Introduction
Breast cancer is one of the prevalent causes Laboratory investigations on animal of death in women in the United States. models have provided compelling but cir-The American Cancer Society has estimated cumstantial evidence that human mammary a 31% incidence of breast cancer (184,300 carcinogenesis may be a multifactorial and new breast cancer cases) and about 17% multistep process involving early-occurring mortality (44,300 cancer related deaths) in molecular, biochemical, and cellular events 1996 (1). These estimates emphasize a need that represent preneoplastic transformation to identify markers for risk, early detection, and late-occurring epigenetic events that and effective prevention.
The cellular biotransformation of E2 represents a complex enzymatic process by which metabolically competent cells convert the mitogenic estrogen E2 to its less active metabolites. Cytochrome P450 (cyp450)-dependent steroid hydroxylases are critical for E2 metabolism, while estrogen receptor, a nuclear transcription factor, is indispensable for transcriptional activation, expression of early response genes cfos, c-jun, c-myc, and resultant E2-mediated positive regulation of growth (12)(13)(14)(15)(16).
Our studies on murine mammary explant cultures and on immortalized nontumorigenic mammary epithelial cell cultures have demonstrated that treatment with chemical carcinogens and transfection with oncogenes results in altered cellular metabolism of E2 and aberrant hyperproliferation in vitro prior to tumorigenicity in vivo (17)(18)(19). In these studies alteration in E2 metabolism was detected by specific and significant increase in C16a-hydroxylation, with a concomitant decrease in C2-hydroxylation pathways, while aberrant hyperproliferation was quantified by the relative extent of cell proliferation in anchoragedependent and anchorage-independent conditions of growth. Altered cellular metabolism of E2 and aberrant hyperproliferation, therefore, represent biochemical and cellular surrogate end point biomarkers for mammary carcinogenesis (4,9,(17)(18)(19)(20)(21). The clinical relevance of these biochemical and cellular perturbations, however, depends on extrapolation and therefore is largely equivocal. Clinical investigations on breast cancer patients, disease-free subjects, and subjects at risk have demonstrated a correlation between elevated C16a-hydroxylation of E2, presence of atypical hyperplasia or proliferative breast disease and increased risk for developing breast cancer (2,5,(6)(7)(8)13,14,20,21). Thus, altered E2 metabolism may represent a biochemical or endocrine marker for breast cancer development. A systematic investigation of this aspect using appropriate human tissuederived models should provide important leads that will help researchers evaluate the clinical relevance of specific molecular, biochemical, endocrine, and cellular biomarkers for human mammary carcinogenesis and its prevention (4,9,22).
The present report provides an overview of the experiments designed on explant and cell culture models developed from noncancerous human mammary tissue to establish potential clinical relevance of E2 metabolism as a biochemical or an endocrine biomarker for effective chemoprevention of human mammary carcinogenesis. Experimental Systems and Biomarker Assays Human Mammary Explant Culture and Cell Culture Models. The mammary explant culture system and mammary epithelial cell culture system provide useful in vitro models to examine the responsiveness of noncancerous mammary tissue to agents that affect cell proliferation, cytodifferentiation, and neoplastic transformation at the molecular, biochemical, and cellular levels (2,4,(17)(18)(19). The tissue culture technology and biomarker assays established for the murine models have been optimized for human mammary tissue (4,22,23).
The explant cultures were prepared from human mammary terminal duct lobular unit (TDLU) obtained from surgical samples. The TDLU are the endocrine responsive and proliferatively active intact organoids that represent target tissue for carcinogenesis (2,4,8). These organoids were maintained in a chemically defined, serum-free Waymouth's MB 752/1 medium (GIBCO/BRL, Grand Island, NY) supplemented with 5 pg/ml insulin, 1 ng/ml E2, 2 mM L-glutamine, and antibiotics. The medium was routinely changed every 48 hr and the cultures were maintained in a humidified atmosphere of 95% air: 5% CO2 at 37°C.
Chemical Carcinogens and Chemopreventive Agent. The stock solutions (IOOOx) of the chemical carcinogens 7,1 2-dimethylbenz[a] anthracene (DMBA) and benzo[a]pyrene (B[a]P) were made up in dimethyl sulfoxide (DMSO). The stock solution of the naturally occurring phytochemical indole-3-carbinol (I3C), to be used as the chemopreventive test compound, was made up in 100% ethanol. These stock solutions were appropriately diluted with the culture medium to obtain the effective nontoxic concentrations. The selection of chemical carcinogens and of the naturally occurring phytochemical was based on their documented tumorigenic or tumor modulating effects on the rodent models (3,10,(26)(27)(28).
Cellular Metabolism of173-estradiol The metabolism of E2 by TDLU and 184-B5 cultures was determined by the radiometric assay that measures the tritium exchange from specifically labeled E2 to form 3H20 (4,17,19) and by the gas chromatography-mass spectrometry (GC-MS) assay that involves product isolation and identification of the metabolites (29).
For the radiometric assay, cultures were incubated with 8.0 x 10-10 M [C2-3H] E2 or [Cl6a-3H] E2 for 48 hr at 370C. Aliquots of 500 pl of the incubation medium were diluted to 3.0 ml with distilled water and lyophilized to separate 3H20 from the residual radioactive E2. The relative extent of 3H20 formed provided an indirect measure of 2-hydroxyestrone (2-OHE1) or 166a-hydroxyestrone (16a-OHE1) formed via the C2-hydroxylation and C16a-hydroxylation pathways of E2 metabolism, respectively. Based on the stoichiometric conversion and the specific activity of labeled E2, the amounts of the metabolites formed were calculated. The data were expressed as pmol metabolite/48 hr/mg tissue. For the GC-MS assay, cultures were incubated for 48 hr at 37°C with 10-8 M nonradioactive E2. Fifteen milliliters of incubation medium was added with deuterated E2 as an internal standard and processed for solid phase extraction. The extracted sample was derivatized in the presence of 50 1il dry pyridine and 10 Ill bis (trimethylsilyl) trifluoroacetamide (BSTFA). The derivatized samples were analyzed under select ion mode and mass ion, and GC elution times of the metabolites were determined (29). The data were expressed as relative abundance of E2 metabolites per 107 cells.
Cell Cycle Analysis and Cellular Apoptosis. The effect of initiators and modulator of carcinogenesis on alteration in cell cycle progression and on the relative extent of cellular apoptosis was examined on 184-B5 cells using the fluorescenceassisted cell sorting (FACS) flow cytometric assay. For this assay, trypsinized cell suspensions fixed in 2% formaldehyde and subsequently in ice-cold 70% ethanol were stained with propidium iodide according to the published procedure (30). The extent of proliferation was expressed as percent synthesis (S)-phase fraction, while the extent of apoptosis was expressed as the intensity of Sub Go (apoptotic) peak obtained from FACS analysis.

Effect of Chemical Carcinogens on
Terminal Duet Lobular Units. The experiment presented in Table 1 was performed on explant cultures of human mammary TDLU to examine the effects of well known rodent carcinogens DMBA and B[a]P on the cellular metabolism of E2. The relative extent of E2 metabolism via C2-hydroxylation and C16a-hydroxylation pathways was determined using the radiometric assay. The results obtained from The lack of substantial difference in the ratios from the three groups indicates that the two solvents at 0.1% concentrations did not influence the constitutive level of E2 metabolism. In the explant cultures of TDLU the epithelial component is surrounded by intralobular and extraductular stroma. It is therefore not possible to demonstrate whether the epithelial component is directly responsible for E2 metabolism. In addition, the radiometric assay measures the reaction kinetics of C2-and C16ahydroxylation pathways and therefore represents an indirect assay for the formation of 2-OHE, or 16a-OHE,. In an attempt to eliminate the above-mentioned limitations, experiments were conducted  Table  2 demonstrate that treatment of 184-B5 cells with DMBA and with B[a]P results in a decreased abundance of 2-OHE1 relative to that observed in the solvent-treated control cultures. Furthermore, B[a]P was more effective in upregulating 16a-OHEI formation than was DMBA. Because of the observed distinct effects of DMBA and B[a]P on the metabolic pathways, the carcinogen-mediated alteration in the 184-B5 system also resulted in an 87.6 and a 91% inhibition in the C2/C16a-hydroxylation ratio, respectively, relative to that observed in the solvent-treated controls.   3.6 ± 0.9f 6.5 ± 1.6i ¶Explant cultures incubated with 0.1% DMSO or 39 pM B[a]P for 24 hr and were maintained with or without 50 pM 13C for the subsequent 10 days. bPulse labeled with 5 pci/ml of 3H-thymidine between days 9 and 10 of culture. Radioactivity was determined as trichloroacetic acid precipitable counts. cDetermined by the radiometric assay as in Table 1. "1Mean ± SD, n = 6 per treatment group. d' '1-tp = 0.001. P4Mean ± SD, n = 18 per treatment group. 9hp= 0.0001. *4p= 0.005.  Table 4  B5 cell culture system, treatment with 50 pM I3C exhibited 29.7 ± 0.7% S-phase fraction, 15.5 ± 0.6% apoptosis, and a C2/C16a-hydroxylation ratio of 4.1 ± 0.3. These data were comparable to those observed in untreated master controls and in 0.1% ethanol-treated solvent controls (data not shown). Thus, I3C, at the dose level tested, does not appear to influence the constitutive status of the biochemical and cellular quantitative end points. Treatment of B[a]P-initiated cells with I3C exhibited inhibition of the S-phase fraction, increase in cellular apoptosis, and increase in C2/Cl6a-hydroxylation ratio of E2 metabolism. The induction of cellular apoptosis by I3C in B[a]P-initiated 184-B5 cells was also evidenced by a substantial increase in Sub Go (apoptotic) peak in cells treated with B[a]P+I3C relative to those treated with B[a]P alone ( Figure  2A,B). Thus, the ability of I3C to inhibit the effect of B[a]P may, in part, be due to growth regulation by decreased prolifera-tion, increased apoptosis, and increased C2-hydroxylation of E2.
The effect of DMBA appears to be distinct from that of B[a]P in the present experimental system. It will therefore be of considerable interest to examine whether 13C in DMBA initiated cells downregulates the status of S-phasespecific gene expression.

Discussion
The experiments in this study were conducted on in vitro models for human mammary carcinogenesis to examine the role of E2 metabolism in chemical carcinogen-induced initiation of tumorigenic transformation and to validate E2 metabolism as a surrogate end point biomarker for efficacy of chemopreventive agents. The observed increase in replicative DNA synthesis or S-phase fraction in response to the chemical carcinogens may be due to a combination of replicative DNA synthesis and DNA repair synthesis. Mammary tissue has been reported to effectively metabolize the procarcinogens and generate DNA-damaging oxidative metabolites that may induce DNA repair synthesis (2,18,(31)(32)(33)(34)(35).
In the multistep process of mammary carcinogenesis, the natural estrogen E2 functions as a potent tumor promoter, acting prevalently during the late-occurring, postinitial (promotional) stage of tumorigenesis (3,6,7,10). The possible mechanisms responsible for the positive growth regulation by E2 include mitogenic signal transduction via upregulation of early response genes c-jun, c-fos, and cmyc; induction of DNA synthesis in quiescent cells; and activation of nuclear transcriptional factors estrogen receptor and nuclear factor (12,15,16,36). In contrast, the biological activity of various oxidative metabolites of E2 is pleiotropic depending upon the experimental systems (37)(38)(39)(40)(41), and therefore the role of E2 metabolites in mammary carcinogenesis remains equivocal.
The experiments on TDLU explant culture system clearly demonstrated that prototypic rodent carcinogens DMBA and B[a]P decrease the C2/C16a-hydroxylation ratio of E2 metabolism. This observation is consistent with our previous studies that demonstrated a similar perturbation in in vitro models of rodent mammary carcinogenesis in response to DMBA or c-myc (17,18,20,21,31,32). In addition, clinical investigations on patients with breast cancer, as well as laboratory investigations on noncancerous breast tissue, or on human mammary carcinomaderived cell lines, have provided convincing evidence for a role of E2 metabolism in breast cancer development (4,13,14,20,22,23,(40)(41)(42). Taken together, these observations suggest that altered metabolism of E2 may represent an endocrine biomarker for carcinogenic insult to the mammary tissue.
The presence of interlobular and interductular stroma in the TDLU explant culture system represents the nontarget component for mammary carcinogenesis. This cellular heterogeneity therefore compromises the specificity of E2 metabolism as a biomarker assay. In addition, the radiometric assay measuring the reaction kinetics of the two pathways provides an indirect measurement for the formation of E2 metabolites. The two technical limitations were eliminated by experiments that utilized the mammary epithelial 184-B5 cells to determine E2 metabolites by the GC-MS assay. The experiments on the 184-B5 cell culture system demonstrated that while both DMBA and B[a]P decrease 2-OHE, formation, upregulation of 16a-OHE, is observed only in response to B[a]P. Decrease in C2/Cl6a-hydroxylation ratio by the two carcinogens in the two experimental systems is comparable, and the human mammary tissue is more susceptible to B[a]P than it is to DMBA. The observed preferential susceptibility of human tissue to the polycyclic aromatic hydrocarbons is consistent with that reported previously (33)(34)(35). The results obtained from the TDLU explant culture and 184-B5 cell culture systems, taken together, suggest that the target epithelial component responds directly to the carcinogenic insult.
The naturally occurring plant product 13C represents a major phytochemical in such cruciferous vegetables as cabbage, broccoli, and brussels sprouts. Our previous in vitro and in vivo studies on rodent models of mammary carcinogenesis have shown that exposure to I3C protects the target tissue from chemical carcinogeninduced transformation or from mammary tumor virus-induced mammary tumorigenesis (27,31). Furthermore, 13C has been reported to induce C2-hydroxylation of E2 in human mammary carcinoma cells as well as in human subjects (42)(43)(44)(45). We therefore sought to examine whether 13C is also an effective inhibitor of carcinogenesis in human mammary TDLU explant culture and cell culture models. In the experiments utilizing the TDLU explant culture system, treatment of cultures with B[a]P resulted in induction of aberrant proliferation as evidenced by increased replicative DNA synthesis, and in altered E2 metabolism, as seen by inhibition of C2/Cl6ahydroxylation ratio. B[a]P-initiated cultures in the presence of I3C exhibited inhibition of aberrant proliferation and enhancement of C2/Cl6a-hydroxylation ratio. In the experiments on 184-B5 cell culture system, exposure to B[a]P exhibited an increase in S-phase fraction, and a decrease in cellular apoptosis and in C2/Cl6a-hydroxylation ratio. The B[a]Pinitiated cultures in the presence of I3C exhibited downregulation of cell proliferative activity and upregulation of apoptosis and of C2/Cl6a-hydroxylation ratio. The possible mechanisms responsible for the protective effect of I3C against B[a]Pinduced human mammary carcinogenesis, however, remain to be identified. In this context, it is noteworthy that 13C functions as an inducer of cyp4501AI-dependent C2-hydroxylation of estradiol (42) and 2-OHE1 has been reported to antagonize the genotoxic and transforming effects of DMBA (31). These observations, taken together, raise the possibility that increased production of antiproliferative E2 metabolite 2-OHE1 by 13C in carcinogen-initiated human mammary tissue may negatively regulate aberrant hyperproliferation, in part by induction of cellular apoptosis.
In conclusion, the pre §ent study on human mammary tissue-derived in vitro models for carcinogenesis has provided evidence that E2 metabolism and cell cyclerelated markers for proliferation and apoptosis may represent valuable surrogate end point biomarkers to evaluate efficacy of chemopreventive agents for human mammary carcinogenesis.