Metabolic Biotransformation of Estradiol in Human Mammary Explant Cultures" NITIN T. TELANG, DEBORAH M. AXELROD, H. LEON BRAD LOW,^ AND MICHAEL P. OSBORNE Breast Cancer Research Laboratory Department of Surgery Memorial Sloan-Kettering Cancer Center New York, New York 10021 bBiochemical Endocrinology Laboratory The Rockefeller University New York, New York 10021
INTRODUCTION The mammotrophic effects of selected steroid and polypeptide hormones are well documented.'.' The natural estrogen 17&estradiol( E2), in concert with progesterone, is noted to be essential for regulated morphogenesis of the mammary epithelial component. These ovarian steroids induce proliferation of ductal epithelium and its differentiation into acini. The polypeptide hormone prolactin and the glucocorticoid hydrocortisone are responsible for the functional differentiation of acini into secretory lobul~alveoli.'-~ In recent years the ability of these hormones to induce and/or mediate mammary cell proliferation, cytodifferentiation, and neoplastic transformation has been documented on mammary tissue in c ~ l t u r e . ~ ,These ',~~~ in~vifro studies provide strong support for the concept that murine mammary tissue may be responsive to a direct hormonal stimulation. In contrast, very little definitive information is available at present regarding the response of nontransformed human mammary tissue to selected hormones and the role of hormones in tumorigenic transformation? In addition to the mitogenic effects of estrogens, their ability to enhance the growth of murine mammary t u m o r ~ ,and ~ . ~the observation that chemical-carcinogen-induced rodent mammary tumors are predominantly estrogen dependent, that is, regress upon ovariectomy or antiestrogen administration,7.8suggest that estrogens in general or estradiol in particular may be potent mammary tumor promoters. However, evidence for human mammary tumor promotion by estradiol and/or its metabolites is largely based on extrapolation of data from animal studies. To further elucidate the role of estradiol as a cocarcinogen and/or a tumor promoter in human mammary tumorigenesis, several clinical investigations have attempted to identify possible cancer-related metabolic differences in E$ biotransfor=Supported by National Institutes of Health grant R29 CA 44741 and the Wanda Jablonski Fund. 70
TELANG et al.: ESTRADIOL METABOLISM
71
mation.’-’’ These investigations have demonstrated a significant elevation specifically in 16a-hydroxylation of E,. The extent of increase in this metabolic pathway was found to be dependent upon the presence of breast cancer or of identifiable familial risk for the disease. These experiments essentially measured systemic metabolism of E,, which included the biotransformation at nonmammary sites such as liver, kidney, or lungs and, therefore, provided little information about & metabolism in the mammary tissue, the target site of mammary tumors. The ability of mammary gland to metabolize E, and/or to accumulate “genotoxic” metabolites could profoundly influence the tumorigenic transformation of the mammary epithelium. The present report describes our experiments on the explant cultures of human mammary terminal duct lobular units (TDLU). This epithelial component of the mammary gland exhibits high proliferative activity14 and susceptibility to chemical carcinogensI s and, therefore, has been considered as the presumptive target site for tumorigenesis. Using a newly developed mammary explant culture system, we have evaluated the ability of benign (noninvolved) human mammary tissue to metabolize E,, compared tissue specificity for & metabolism, and examined the effect of a prototype chemical carcinogen on E, metabolism.
MATERIALS AND METHODS Patient Selection
The surgical specimens from 15 premenopausal patients were used in this study. The patients ranged in age from 24 to 50 years. Ten patients were diagnosed as having intraductal and/or infiltrating ductal carcinoma, and, therefore, their benign (noninvolved) tissue was considered to be at “high risk” for developing breast cancer. Five patients did not present any clinicopathological evidence for breast cancer and were undergoing surgery for reduction mammoplasty. The tissues from these patients were considered to be the “low-risk” group. All patients had a normal menstrual cycle of 28 2 3 days. To ensure culturing of noninvolved tissue from the “high-risk” mastectomy specimens, the tissue samples were obtained from sites at least 2-4 cm remote from the cancer.
Prepration of Explant Cultures
The surgical specimens were supravitally stained with 25 pg/ml methylene blue for about 90 minutes at 10°C. This procedure specifically stains the viable epithelial component.’6 The TDLU and mammary fat (MF) were microdissected under lox magnification, and intact explants of TDLU and MF were attached to sterile dacron rafts. The individual explant cultures were maintained in 1 ml of phenol red-free Dulbecco’s minimum essential medium (DMEM) in a 24-well tissue culture plate. DMEM containing antibiotics was supplemented with 350 pg/ml glutamine and 5 pg/ml insulin. The cultures were maintained in a humidified atmosphere of 95% air: 5% CO, at 37°C for 10 days prior to the metabolism studies.
72
ANNALS NEW YORK ACADEMY OF SCIENCES
Metabolism of Estradiol The relative extent of estradiol metabolism via the C 17-0xidation, C2-hydroxylation, and C16a-hydroxylation pathways was determined by radiometric assay which measures the loss of 'H from steriospecificallylabeled & to form 'H,O."" The explant cultures were incubated with [3H]E, labeled at C17, C2 or C16a positions for 24, 48, or 72 hours. The radioactive ligand (5-6 x lo6 dpm) was dissolved in sterile propylene glycol and added to the culture medium at a concentration of 10 pl/ml. At the predetermined time points, 500 pl aliquots of the culture medium were diluted with 3 ml of water, lyophilized, and 1 ml of sublimed water was counted for 'H radioactivity. The & metabolism was determined by the extent of 'H,O formed relative to the dose of ['HIE, administered.
RESULTS Biotransformation of Estradiol in TDLU One hundred and eight TDLU explant cultures from six patients undergoing mastectomy for breast cancer were utilized to examine the extent of metabolism. Thirty-six explant cultures from each of the six patients were incubated with stereospecifically labeled E, for 24, 48, and 72 hours, and the metabolism of E, via C17oxidation, C2-hydroxylation, and C16a-hydroxylation was determined (TABLE1). In general, all three metabolic pathways exhibited a time-dependent increase that plateaued between 48 and 72 hours of incubation. The relative extent of C17 oxidation, although was higher at 72 hours than that at 24 hours incubation, did not reach statistical significance. In contrast, the relative extent of C2-hydroxylation and C16ahydroxylation showed a significant increase after 72 hours incubation (p < 0.001), exhibiting a 126% and 800% increase, respectively, relative to that observed at 24 hours incubation.
Injluence of the Menstrual Cycle on E2 lba-Hydroxylation Twenty-four explant cultures of TDLU and MF obtained from mastectomy specimens from four patients were used to examine whether the extent of & 16a-hydroxylation is dependent upon the phase of the menstrual cycle (TABLE2). Of the four patients, two were in the follicular phase (day 2-day 15) and two were in the luteal phase (day 16-day 1) of the menstrual cycle. The MF explants, irrespective of the phase of the menstrual cycle, did not show any change in the extent of E, 16ahydroxylation. In contrast, this metabolic pathway was significantly higher (p < 0.001) in TDLU. The TDLU from luteal phase patients exhibited almost a 400% increase in E, 16a-hydroxylation relative to that observed in the TDLU from follicular phase patients.
TELANC et al.: ESTRADIOL METABOLISM
73
Ez Ida-Hydroxylation in TDLU from Mammoplasty and Mastectomy Specimens Since the extent of E, 16a-hydroxylation is noted to be increased in breast cancer patients ‘OJ’ and in subjects at increased risk for developing breast c a n ~ e r ,it’ ~was of interest to examine whether this increase is also exhibited by benign (noninvolved) mammary tissue. Twenty-four TDLU explants obtained from four mammoplasty specimens (patients lacking clinicopathological evidence of cancer) and an equal number of TDLU explants from four mastectomy specimens (patients exhibiting the presence of intraductal carcinoma or infiltrating ductal carcinoma) were used to compare the relative extent of E, 16a-hydroxylation. The data presented in TABLE 3 clearly demonstrated that TDLU from mastectomy specimens possessed a substantially higher relative extent of E, 16a-hydroxylation than that observed in TDLU obtained from mammoplasty specimens.
TABLE 1. Metabolic Biotransformation of Estradiol in Explant Cultures of Human
Mammary Terminal Duct Lobular Units Percent Estradiol Metabolism“ Duration of Exposure (hours) 24 72 Relative increase‘
C 17-Oxidation 1.21 2.23
f f
0.38’ 0.90’
84.3
C2-Hydroxylation
C16a-Hydroxylation
0.19 ? 0.01‘ 0.43 f 0.12’
0.02 f O.Old 0.18 f 0.02d
126.3
800
“Determined by the extent of loss of ’H to form ’H,O. Cultures incubated with [C17-’H]E2, [C2-’H]E2, or [C16a-’H]E, Values are mean f standard deviation (SD)/mg tissue. Values with the same superscript differed significantly as shown below. b p > 0.5 not significant (NS). ‘ p < 0.001. d p < 0.001. ‘Relative increase = [(metabolism at 72 hours - metabolism at 24 hours)/(metabolism at 24 hours)] x 100.
Effect of Chemical Carcinogen on E2 Ida-Hydroxylation Having obtained the evidence that E, 16a-hydroxylation is elevated, depending upon the presence of breast cancer, it was of interest to examine the response of benign (noninvolved) TDLU to a chemical carcinogen. Twenty-four TDLU explants obtained from four mammoplasty specimens and 24 explants from four mastectomy specimens were treated with the metabolism-dependent procarcinogen benzo( a)pyrene (BP) for 24 hours prior to the determination of E, 16a-hydroxylation. The results of this experiment, presented in TABLE4, revealed that although BP was able to induce increased extent of E, 16a-hydroxylation in both mammoplasty as well as mastectomy specimens, the response of TDLU from mastectomy specimens was almost 165% higher than that seen in TDLU from mammoplasty specimens.
ANNALS NEW YORK ACADEMY OF SCIENCES
74
TABLE 2. Influence of the Phase of Menstrual Cycle on Estradiol Metabolism in Explant Cultures of Human Mammary Tissue
Phase of Menstrual Cycle
Type of Tissue
Follicular
MF TDLU MF TDLU
Luteal
Percent Estradiol 16a-Hydroxylation"
* *
0.001' 0.006 0.021 2 0.007' 0.004 0.001d 0.110 t 0.021"d
Relative Increaseb
250.0 2650.0
423.8
Determined by the extent of loss of 'H to form 'H,O. Cultures incubated with [C16a-'H]E2 for 48 hours. Values are mean 2 SD/mg tissue. Values with the same superscript differed significantly as shown below. Calculated as described in footnote e, TABLE1. c p < 0.001. dp < 0.001. 'p < 0.005.
DISCUSSION Among the three major metabolic pathways for E2 biotransformation, the C16ahydroxylation pathway leading to the conversion of estrone (E,) to 16a-hydroxyestrone ( 16a-OHEl)has been reported to be significantly elevated in patients with breast cancer. The other two pathways, C17-oxidation of & to form El and the C2-hydroxylation pathway converting El to 2-hydroxyestrone (2-OHEI), remain essentially unaltered.'-'' Additionally, elevated levels of E, 16a-hydroxylation have also been found in women with identifiable, familial risk for breast cancer13 as well as in mice that are at high risk of developing breast cancer.".'* These observations indicate that E, 16a-hydroxylation is elevated well before the appearance of breast cancer and, therefore, may constitute an intermediate, endocrine biomarker to identify the risk for developing breast cancer. Although the above-described in vivo studies provide Breast Cancer-Dependent Elevation of Estradiol 16a-Hydroxylation in Explant Cultures of Human Mammary Tissue
TABLE 3.
Type of Tissue MF TDLU MF TDLU
Presence of Breast Cancer
-
+ +
Percent Estradiol 16a-Hydroxylation" 0.039 0.071 0.039 0.310
0.006' t 0.009"' t 0.004d rf: 0.032"' rf:
Relative Increase' -
82.1 694.8
336.6
Determined by the extent of loss of 'H to form 'H,O. Cultures incubated with [C16a-'H]EZ for 48 hours. Values are mean 4 SD/mg tissue. Values with the same superscript differed significantly as shown below. Calculated as described in footnote e, TABLE1. 'p < 0.001. d p < 0.001. = p < 0.001.
15
TELANG et al.: ESTRADIOL METABOLISM
strong evidence for the role of E, 16a-hydroxylation in the pathogenesis of breast cancer, the target tissue specificity of this metabolic biomarker is not established. In situ metabolism of E, in the mammary gland, the target tissue for breast cancer, may be more relevant to the early occurring initiational events of tumorigenic transformation. The explant cultures of TDLU exhibited metabolic competence for E, biotransformation via C17-oxidation, C2-hydroxylation, and C16a-hydroxylation. In general, the relative extent of all three pathways followed essentially similar trends as those These results demonstrated the ability of reported in the earlier in vivo ~tudies.'""~'~ TDLU, the target tissue for tumorigenic transformation of the mammary gland, to metabolize E&. The relative extent of E, metabolism by TDLU was substantially lower than that reported in the in vivo studies where net systemic metabolism, including that occurring at nonmammary sites such as liver, is measured. Furthermore, the extent of in v i m metabolism of E, by individual TDLU expressed per milligram tissue weight will be substantial upon appropriate correction to obtain total conversion in Benzo( a)pyrene Induced Elevation of Estradiol 16a-Hydroxylation in Ex~lantCultures of Human Mammarv Tissue
TABLE 4.
Type of Tissue
Presence of Breast Cancer
TDLU
-
TDLU TDLU TDLU
+ +
-
Treatment' DMSO (solvent control) BP DMSO BP
Percent Estradiol 16a-Hydr~xylation~ 0.054
4
0.005d
0.085 ? 0.CKNd
0.065 2 0.012' 0.225 2 0.059'
Relative Increase'
57.4
246.2
164.7
"Ten-day-old cultures exposed to 0.1% DMSO or 1 pg/ml benzo(a) pyrene (BP) for 24 hours. 'Determined from the extent of loss of 'H to form 'H,O in DMSO or BP-treated cultures incubated with [C16a-'HH]E, for 48 hours. Values are mean SD/mg tissue. Values with the same superscript differed signficantly, as shown below. ' Calculated as described in footnote e, TABLE1. d p < 0.01. ' p < 0.001.
*
the entire mammary gland. Indeed, our recently completed study has demonstrated that & metabolism in mouse liver explant cultures is at least 3- to 10-fold higher than that in the mammary explant cultures. Furthermore, a comparison of relative extent of E& 16a-hydroxylation in liver and mammary explant cultures from low- and highrisk strains of mice indicated that whereas the extent of & 16a-hydroxylation in liver in the two strains was comparable, it was significantly increased in mammary tissue from high-risk mice relative to that observed in the mammary tissue from low-risk mice.I9 This observation suggests that the relative extent of E, 16a-hydroxylation in the mammary tissue may be biologically more relevant in tumorigenic transformation of the mammary tissue. The enhanced E, 16a-hydroxylation that is observed in TDLU from luteal phase patients may reflect an altered responsiveness of the mammary tissue to the plasma levels of ovarian steroids that fluctuate during the phases of the menstrual cycle. Thus, enhanced metabolism of E, may be due to the increased progesterone levels during the luteal phase. Recent studies by Mauvais-Jarvis et al. have shown that in the luteal
76
ANNALS NEW YORK ACADEMY OF SCIENCES
phase the activity of 17P-hydroxysteroid dehydrogenase leading to the conversion of E, to El is increased.’ Similarly progestins have been reported to enhance conversion of E, to E, in human breast epithelial cell cultures:’ and E, 16a-hydroxylation has been observed to be substantially elevated after exposure to progesterone in mouse mammary explant c~ltures.~’ It was of interest to note that whereas the extent of & 16a-hydroxylation was essentially unaltered in MF, it exhibited a substantial increase in TDLU. This selective increase in TDLU, but not in MF, is indicative of a target tissue specificity for the metabolic competence. The comparison of constitutive levels of E, 16a-hydroxylation between TDLU from mammoplasty specimens and those from mastectomy specimens revealed almost a 300% increase in the latter group, while the levels of metabolism in MF from the two groups were low and did not exhibit any change. These results indicate that benign TDLU from the cancerous breast that do not present any morphologic abnormality may already be altered in their metabolic competence. Thus, alteration in the E, 16a-hydroxylation pathway may constitute a metabolic marker for carcinogenesis. The relative risk for developing breast cancer may also be manifested by the response of the tissue to chemical carcinogens. This is clearly demonstrated in the experiment where the effect of metabolism-dependent procarcinogen BP is examined on E, 16a-hydroxylation in TDLU from mammoplasty and mastectomy specimens. Although BP was able to induce elevated levels of E, metabolism in both groups, the magnitude of b16a-hydroxylation was almost 165% higher in TDLU from mastectomy specimens relative to that seen in TDLU from mammoplasty specimens. These results indicate that benign (noninvolved) TDLU from cancerous breast may be more susceptible to the carcinogenic insult. It is interesting to note that BP exposure resulted in a concomitant decrease in the 2-hydroxylation pathway?’ These observations, taken together, suggest that exposure of noninvolved mammary TDLU to a chemical carcinogen may alter E, metabolism in favor of E, 16a-hydroxylation at the expense of the 2-hydroxylation pathway. The 16a-OHE, formed in TDLU may directly influence the epithelial component of TDLU to promote the expression of mutant phenotype induced by BP.
SUMMARY The metabolism of E, via the 16a-hydroxylation pathway is reported to be elevated in breast cancer patients as well as in subjects at high risk for developing breast cancer. The biological relevance of the metabolic pathway during the initiational events that lead to the tumorigenic transformation of mammary epithelium is not fully understood. The results obtained from the in vitro experiments discussed in this report permit the following conclusions:
1. Human mammary TDLU, the presumptive target site for breast cancer, possesses metabolic competence to biotransform &. 2. The biotransformation of E, in TDLU via the 16a-hydroxylation pathway is responsive to endogenous hormonal changes and to the presence of cancer, and is susceptible to carcinogenic insult. 3. The relative extent of E, 16a-hydroxylation may constitute a sensitive metabolic marker for evaluating the susceptibility of noninvolved mammary epithelium to carcinogen-induced transformation.
TELANG el al.: ESTRADIOL METABOLISM
77
REFERENCES 1. BANERJEE,M. R. 1976. Responses of mammary cells to hormones. Int. Rev. Cytol. 4
1-97. N. M. MEHTA,A. P. IYER& R. GANGULY. 1980. 2. BANERJEE,M. R., N. GANGULY, Functional differentiation and neoplastic transformation in an isolated whole mammary organ in vitro. In Cell Biology of Breast Cancer. C. McGrath, M. J. Brennan & M. A. Rich, Eds.: 485-516. Academic Press. New York, N.Y. P., F. KU~TENN & A. GOMPEL.1986. Estradiol/progesterone inter3. MAUVAIS-JARVIS, action in normal and pathologic breast cells. Ann. N.Y. Acad. Sci. 464: 152-167. 4. BANERIEE,M. R. 1986. An in vitro model for neoplastic transformation of epithelial cells in an isolated whole mammary gland of the mouse. In In W r o Models for Cancer Research. M. M. Webber & L. I. Sekeley, Eds.: 69-114. CRC Press. Boca Raton, Fla. E. A., M. J. VOCCI,J. W. COMBS,H. SANEFUJI, T. ROBBINS, D. H. JANSS,C. 5. HILLMAN, C. HARRIS& B. F. TRUMP.1980. Human breast organ culture studies. Methods Cell Biol. 21: 80-106. M. & R. VAN NIE. 1974. Estrogen receptor content and hormone responsive 6. SLUYSER, growth of mouse mammary tumor. Cancer Res. 3 4 3252-3257. 7. WELSCH,C. W. 1986. Interrelationship between fat and endocrine processes in mammary gland tumorigenesis. In Dietary Fat and Cancer. C. Ip, A. E. Rogers, D. F. Birt & C. Mettlin, Eds.: 623-654. A. R. Liss, Inc. New York, N.Y. V. C. 1976. Effect of tamoxifen (ICS 46,474) on initiation and growth of DMBA8. JORDAN, induced rat mammary carcinomata. Eur. J. Cancer 1 2 419-424. J. SCHNEIDER, K. E. ANDERSON & A. KAPPAS.1980. 9. FISHMAN,J., H. L. BRADLOW, Radiometric analysis of oxidation in man: sex differences in estradiol metabolism. Proc. Nat. Acad. Sci. USA 77: 4957-4960. J., D. KINNE,A. FRACCHIA, V. PIERCE,K. E. ANDERSON, H. L. BRADLOW 10. SCHNEIDER, & J. FISHMAN. 1982. Abnormal oxidative metabolism of estradiol in women with breast cancer. Proc. Nat. Acad. Sci. USA 7 9 3047-3051. H.L., R. HERSHCOPF, C. MARTUCCI & J. FISHMAN. 1986. 16a-Hydroxylation 11. BRADLOW, of estradiol: a possible risk marker for breast cancer. Ann. N.Y. Acad. Sci. 464: 138- 151. M. P., N. T. TELANG,H. KURIHARA & H. L. BRADLOW. 1988. Risk enhanced 12. OSBORNE, metabolism of estradiol in explant cultures of human benign mammary tissue. Proc. Am. Assoc. Cancer Res. 2 9 236. M. P., R. A. KARMALI, J. HERSHCOPF, H. L. BRADLOW, I. A. KOURIDES, W. 13. OSBORNE, WILLIAMS, P. P. ROSEN& J. FISHMAN. 1988.Omega-3fatty acid: modulation of estrogen metabolism and potential breast cancer prevention. Cancer Invest. 6 629-63 1. 14. Russo, I., C. CALAF,L. ROI & I. H.Russo. 1987. Influence of age and reproductive history on cell kinetics of normal human breast tissue in vitro. J. Nat. Cancer Inst. 7 8 413-418. & I. H. RUSSO.1988. Expression of phenotypical 15. Russo, J., D. REINA,J. FREDERICK changes by human breast epithelial cells treated with carcinogens in v i m . Cancer Res. 68:2837-2857. 16. BEUHRING, G. C. & M. H. JENSEN.1983. Lack of toxicity of methylene blue chloride to supravitally stained human mammary tissues. Cancer Res. 43:6039-6044. 17. TELANG,N. T., R. S. BOCKMAN, M. J. MODAK& M. P. OSBORNE.1989. The role of fatty acids in murine and human mammary carcinogenesis: an in vitro approach. In Carcinogenesisand Dietary Fat. S. Abraham, Ed.: 427-45 1. Kluwer Academic Publishers. Boston, Mass. 18. BRADLOW,H. L., R. J. HERSHCOPF, C. P. MARTUCCI & J. FISHMAN. 1985. Estradiol 16a-hydroxylationin the mouse correlates with mammary tumor incidence and presence of murine mammary tumor virus: a possible model for the hormonal etiology of breast cancer in humans. Proc. Nat. Acad. Sci. USA 8 2 6295-6299. & M. P. OSBORNE.1989. In vitro 19. TELANG,N. T., H. L. BRADLOW,H. KURIHARA biotransformation of estradiol by explant cultures of murine mammary tissues. Breast Cancer Res. Treat. 13: 173-181.
I8
ANNALS NEW YORK ACADEMY OF SCIENCES
20. PRUDHOMME, J. F., C. MALET,A. GOMPEL,J. P. LA LARDRIE,P. OCHOA,A. BOUE,P. MAUVAIS-JARVIS & F. KUTTENN.1984. 17P-Hydroxysteroiddehydrogenase activity in human breast epithelial and fibroblast cell cultures. Endocrinology 114: 1483- 1489. 21. TELANG,N. T., A. BASU,M. J. MODAK,H. L. BRADLOW & M. P. &BORNE. 1988. Parallel enhancement of ras proto-oncogene expression and estradiol 16a-hydroxylation in human mammary terminal duct lobular units (TDLU) by a carcinogen. Breast Cancer Res. Treat. 11: 138.
INTRODUCTION The mammotrophic effects of selected steroid and polypeptide hormones are well documented.'.' The natural estrogen 17&estradiol( E2), in concert with progesterone, is noted to be essential for regulated morphogenesis of the mammary epithelial component. These ovarian steroids induce proliferation of ductal epithelium and its differentiation into acini. The polypeptide hormone prolactin and the glucocorticoid hydrocortisone are responsible for the functional differentiation of acini into secretory lobul~alveoli.'-~ In recent years the ability of these hormones to induce and/or mediate mammary cell proliferation, cytodifferentiation, and neoplastic transformation has been documented on mammary tissue in c ~ l t u r e . ~ ,These ',~~~ in~vifro studies provide strong support for the concept that murine mammary tissue may be responsive to a direct hormonal stimulation. In contrast, very little definitive information is available at present regarding the response of nontransformed human mammary tissue to selected hormones and the role of hormones in tumorigenic transformation? In addition to the mitogenic effects of estrogens, their ability to enhance the growth of murine mammary t u m o r ~ ,and ~ . ~the observation that chemical-carcinogen-induced rodent mammary tumors are predominantly estrogen dependent, that is, regress upon ovariectomy or antiestrogen administration,7.8suggest that estrogens in general or estradiol in particular may be potent mammary tumor promoters. However, evidence for human mammary tumor promotion by estradiol and/or its metabolites is largely based on extrapolation of data from animal studies. To further elucidate the role of estradiol as a cocarcinogen and/or a tumor promoter in human mammary tumorigenesis, several clinical investigations have attempted to identify possible cancer-related metabolic differences in E$ biotransfor=Supported by National Institutes of Health grant R29 CA 44741 and the Wanda Jablonski Fund. 70
TELANG et al.: ESTRADIOL METABOLISM
71
mation.’-’’ These investigations have demonstrated a significant elevation specifically in 16a-hydroxylation of E,. The extent of increase in this metabolic pathway was found to be dependent upon the presence of breast cancer or of identifiable familial risk for the disease. These experiments essentially measured systemic metabolism of E,, which included the biotransformation at nonmammary sites such as liver, kidney, or lungs and, therefore, provided little information about & metabolism in the mammary tissue, the target site of mammary tumors. The ability of mammary gland to metabolize E, and/or to accumulate “genotoxic” metabolites could profoundly influence the tumorigenic transformation of the mammary epithelium. The present report describes our experiments on the explant cultures of human mammary terminal duct lobular units (TDLU). This epithelial component of the mammary gland exhibits high proliferative activity14 and susceptibility to chemical carcinogensI s and, therefore, has been considered as the presumptive target site for tumorigenesis. Using a newly developed mammary explant culture system, we have evaluated the ability of benign (noninvolved) human mammary tissue to metabolize E,, compared tissue specificity for & metabolism, and examined the effect of a prototype chemical carcinogen on E, metabolism.
MATERIALS AND METHODS Patient Selection
The surgical specimens from 15 premenopausal patients were used in this study. The patients ranged in age from 24 to 50 years. Ten patients were diagnosed as having intraductal and/or infiltrating ductal carcinoma, and, therefore, their benign (noninvolved) tissue was considered to be at “high risk” for developing breast cancer. Five patients did not present any clinicopathological evidence for breast cancer and were undergoing surgery for reduction mammoplasty. The tissues from these patients were considered to be the “low-risk” group. All patients had a normal menstrual cycle of 28 2 3 days. To ensure culturing of noninvolved tissue from the “high-risk” mastectomy specimens, the tissue samples were obtained from sites at least 2-4 cm remote from the cancer.
Prepration of Explant Cultures
The surgical specimens were supravitally stained with 25 pg/ml methylene blue for about 90 minutes at 10°C. This procedure specifically stains the viable epithelial component.’6 The TDLU and mammary fat (MF) were microdissected under lox magnification, and intact explants of TDLU and MF were attached to sterile dacron rafts. The individual explant cultures were maintained in 1 ml of phenol red-free Dulbecco’s minimum essential medium (DMEM) in a 24-well tissue culture plate. DMEM containing antibiotics was supplemented with 350 pg/ml glutamine and 5 pg/ml insulin. The cultures were maintained in a humidified atmosphere of 95% air: 5% CO, at 37°C for 10 days prior to the metabolism studies.
72
ANNALS NEW YORK ACADEMY OF SCIENCES
Metabolism of Estradiol The relative extent of estradiol metabolism via the C 17-0xidation, C2-hydroxylation, and C16a-hydroxylation pathways was determined by radiometric assay which measures the loss of 'H from steriospecificallylabeled & to form 'H,O."" The explant cultures were incubated with [3H]E, labeled at C17, C2 or C16a positions for 24, 48, or 72 hours. The radioactive ligand (5-6 x lo6 dpm) was dissolved in sterile propylene glycol and added to the culture medium at a concentration of 10 pl/ml. At the predetermined time points, 500 pl aliquots of the culture medium were diluted with 3 ml of water, lyophilized, and 1 ml of sublimed water was counted for 'H radioactivity. The & metabolism was determined by the extent of 'H,O formed relative to the dose of ['HIE, administered.
RESULTS Biotransformation of Estradiol in TDLU One hundred and eight TDLU explant cultures from six patients undergoing mastectomy for breast cancer were utilized to examine the extent of metabolism. Thirty-six explant cultures from each of the six patients were incubated with stereospecifically labeled E, for 24, 48, and 72 hours, and the metabolism of E, via C17oxidation, C2-hydroxylation, and C16a-hydroxylation was determined (TABLE1). In general, all three metabolic pathways exhibited a time-dependent increase that plateaued between 48 and 72 hours of incubation. The relative extent of C17 oxidation, although was higher at 72 hours than that at 24 hours incubation, did not reach statistical significance. In contrast, the relative extent of C2-hydroxylation and C16ahydroxylation showed a significant increase after 72 hours incubation (p < 0.001), exhibiting a 126% and 800% increase, respectively, relative to that observed at 24 hours incubation.
Injluence of the Menstrual Cycle on E2 lba-Hydroxylation Twenty-four explant cultures of TDLU and MF obtained from mastectomy specimens from four patients were used to examine whether the extent of & 16a-hydroxylation is dependent upon the phase of the menstrual cycle (TABLE2). Of the four patients, two were in the follicular phase (day 2-day 15) and two were in the luteal phase (day 16-day 1) of the menstrual cycle. The MF explants, irrespective of the phase of the menstrual cycle, did not show any change in the extent of E, 16ahydroxylation. In contrast, this metabolic pathway was significantly higher (p < 0.001) in TDLU. The TDLU from luteal phase patients exhibited almost a 400% increase in E, 16a-hydroxylation relative to that observed in the TDLU from follicular phase patients.
TELANC et al.: ESTRADIOL METABOLISM
73
Ez Ida-Hydroxylation in TDLU from Mammoplasty and Mastectomy Specimens Since the extent of E, 16a-hydroxylation is noted to be increased in breast cancer patients ‘OJ’ and in subjects at increased risk for developing breast c a n ~ e r ,it’ ~was of interest to examine whether this increase is also exhibited by benign (noninvolved) mammary tissue. Twenty-four TDLU explants obtained from four mammoplasty specimens (patients lacking clinicopathological evidence of cancer) and an equal number of TDLU explants from four mastectomy specimens (patients exhibiting the presence of intraductal carcinoma or infiltrating ductal carcinoma) were used to compare the relative extent of E, 16a-hydroxylation. The data presented in TABLE 3 clearly demonstrated that TDLU from mastectomy specimens possessed a substantially higher relative extent of E, 16a-hydroxylation than that observed in TDLU obtained from mammoplasty specimens.
TABLE 1. Metabolic Biotransformation of Estradiol in Explant Cultures of Human
Mammary Terminal Duct Lobular Units Percent Estradiol Metabolism“ Duration of Exposure (hours) 24 72 Relative increase‘
C 17-Oxidation 1.21 2.23
f f
0.38’ 0.90’
84.3
C2-Hydroxylation
C16a-Hydroxylation
0.19 ? 0.01‘ 0.43 f 0.12’
0.02 f O.Old 0.18 f 0.02d
126.3
800
“Determined by the extent of loss of ’H to form ’H,O. Cultures incubated with [C17-’H]E2, [C2-’H]E2, or [C16a-’H]E, Values are mean f standard deviation (SD)/mg tissue. Values with the same superscript differed significantly as shown below. b p > 0.5 not significant (NS). ‘ p < 0.001. d p < 0.001. ‘Relative increase = [(metabolism at 72 hours - metabolism at 24 hours)/(metabolism at 24 hours)] x 100.
Effect of Chemical Carcinogen on E2 Ida-Hydroxylation Having obtained the evidence that E, 16a-hydroxylation is elevated, depending upon the presence of breast cancer, it was of interest to examine the response of benign (noninvolved) TDLU to a chemical carcinogen. Twenty-four TDLU explants obtained from four mammoplasty specimens and 24 explants from four mastectomy specimens were treated with the metabolism-dependent procarcinogen benzo( a)pyrene (BP) for 24 hours prior to the determination of E, 16a-hydroxylation. The results of this experiment, presented in TABLE4, revealed that although BP was able to induce increased extent of E, 16a-hydroxylation in both mammoplasty as well as mastectomy specimens, the response of TDLU from mastectomy specimens was almost 165% higher than that seen in TDLU from mammoplasty specimens.
ANNALS NEW YORK ACADEMY OF SCIENCES
74
TABLE 2. Influence of the Phase of Menstrual Cycle on Estradiol Metabolism in Explant Cultures of Human Mammary Tissue
Phase of Menstrual Cycle
Type of Tissue
Follicular
MF TDLU MF TDLU
Luteal
Percent Estradiol 16a-Hydroxylation"
* *
0.001' 0.006 0.021 2 0.007' 0.004 0.001d 0.110 t 0.021"d
Relative Increaseb
250.0 2650.0
423.8
Determined by the extent of loss of 'H to form 'H,O. Cultures incubated with [C16a-'H]E2 for 48 hours. Values are mean 2 SD/mg tissue. Values with the same superscript differed significantly as shown below. Calculated as described in footnote e, TABLE1. c p < 0.001. dp < 0.001. 'p < 0.005.
DISCUSSION Among the three major metabolic pathways for E2 biotransformation, the C16ahydroxylation pathway leading to the conversion of estrone (E,) to 16a-hydroxyestrone ( 16a-OHEl)has been reported to be significantly elevated in patients with breast cancer. The other two pathways, C17-oxidation of & to form El and the C2-hydroxylation pathway converting El to 2-hydroxyestrone (2-OHEI), remain essentially unaltered.'-'' Additionally, elevated levels of E, 16a-hydroxylation have also been found in women with identifiable, familial risk for breast cancer13 as well as in mice that are at high risk of developing breast cancer.".'* These observations indicate that E, 16a-hydroxylation is elevated well before the appearance of breast cancer and, therefore, may constitute an intermediate, endocrine biomarker to identify the risk for developing breast cancer. Although the above-described in vivo studies provide Breast Cancer-Dependent Elevation of Estradiol 16a-Hydroxylation in Explant Cultures of Human Mammary Tissue
TABLE 3.
Type of Tissue MF TDLU MF TDLU
Presence of Breast Cancer
-
+ +
Percent Estradiol 16a-Hydroxylation" 0.039 0.071 0.039 0.310
0.006' t 0.009"' t 0.004d rf: 0.032"' rf:
Relative Increase' -
82.1 694.8
336.6
Determined by the extent of loss of 'H to form 'H,O. Cultures incubated with [C16a-'H]EZ for 48 hours. Values are mean 4 SD/mg tissue. Values with the same superscript differed significantly as shown below. Calculated as described in footnote e, TABLE1. 'p < 0.001. d p < 0.001. = p < 0.001.
15
TELANG et al.: ESTRADIOL METABOLISM
strong evidence for the role of E, 16a-hydroxylation in the pathogenesis of breast cancer, the target tissue specificity of this metabolic biomarker is not established. In situ metabolism of E, in the mammary gland, the target tissue for breast cancer, may be more relevant to the early occurring initiational events of tumorigenic transformation. The explant cultures of TDLU exhibited metabolic competence for E, biotransformation via C17-oxidation, C2-hydroxylation, and C16a-hydroxylation. In general, the relative extent of all three pathways followed essentially similar trends as those These results demonstrated the ability of reported in the earlier in vivo ~tudies.'""~'~ TDLU, the target tissue for tumorigenic transformation of the mammary gland, to metabolize E&. The relative extent of E, metabolism by TDLU was substantially lower than that reported in the in vivo studies where net systemic metabolism, including that occurring at nonmammary sites such as liver, is measured. Furthermore, the extent of in v i m metabolism of E, by individual TDLU expressed per milligram tissue weight will be substantial upon appropriate correction to obtain total conversion in Benzo( a)pyrene Induced Elevation of Estradiol 16a-Hydroxylation in Ex~lantCultures of Human Mammarv Tissue
TABLE 4.
Type of Tissue
Presence of Breast Cancer
TDLU
-
TDLU TDLU TDLU
+ +
-
Treatment' DMSO (solvent control) BP DMSO BP
Percent Estradiol 16a-Hydr~xylation~ 0.054
4
0.005d
0.085 ? 0.CKNd
0.065 2 0.012' 0.225 2 0.059'
Relative Increase'
57.4
246.2
164.7
"Ten-day-old cultures exposed to 0.1% DMSO or 1 pg/ml benzo(a) pyrene (BP) for 24 hours. 'Determined from the extent of loss of 'H to form 'H,O in DMSO or BP-treated cultures incubated with [C16a-'HH]E, for 48 hours. Values are mean SD/mg tissue. Values with the same superscript differed signficantly, as shown below. ' Calculated as described in footnote e, TABLE1. d p < 0.01. ' p < 0.001.
*
the entire mammary gland. Indeed, our recently completed study has demonstrated that & metabolism in mouse liver explant cultures is at least 3- to 10-fold higher than that in the mammary explant cultures. Furthermore, a comparison of relative extent of E& 16a-hydroxylation in liver and mammary explant cultures from low- and highrisk strains of mice indicated that whereas the extent of & 16a-hydroxylation in liver in the two strains was comparable, it was significantly increased in mammary tissue from high-risk mice relative to that observed in the mammary tissue from low-risk mice.I9 This observation suggests that the relative extent of E, 16a-hydroxylation in the mammary tissue may be biologically more relevant in tumorigenic transformation of the mammary tissue. The enhanced E, 16a-hydroxylation that is observed in TDLU from luteal phase patients may reflect an altered responsiveness of the mammary tissue to the plasma levels of ovarian steroids that fluctuate during the phases of the menstrual cycle. Thus, enhanced metabolism of E, may be due to the increased progesterone levels during the luteal phase. Recent studies by Mauvais-Jarvis et al. have shown that in the luteal
76
ANNALS NEW YORK ACADEMY OF SCIENCES
phase the activity of 17P-hydroxysteroid dehydrogenase leading to the conversion of E, to El is increased.’ Similarly progestins have been reported to enhance conversion of E, to E, in human breast epithelial cell cultures:’ and E, 16a-hydroxylation has been observed to be substantially elevated after exposure to progesterone in mouse mammary explant c~ltures.~’ It was of interest to note that whereas the extent of & 16a-hydroxylation was essentially unaltered in MF, it exhibited a substantial increase in TDLU. This selective increase in TDLU, but not in MF, is indicative of a target tissue specificity for the metabolic competence. The comparison of constitutive levels of E, 16a-hydroxylation between TDLU from mammoplasty specimens and those from mastectomy specimens revealed almost a 300% increase in the latter group, while the levels of metabolism in MF from the two groups were low and did not exhibit any change. These results indicate that benign TDLU from the cancerous breast that do not present any morphologic abnormality may already be altered in their metabolic competence. Thus, alteration in the E, 16a-hydroxylation pathway may constitute a metabolic marker for carcinogenesis. The relative risk for developing breast cancer may also be manifested by the response of the tissue to chemical carcinogens. This is clearly demonstrated in the experiment where the effect of metabolism-dependent procarcinogen BP is examined on E, 16a-hydroxylation in TDLU from mammoplasty and mastectomy specimens. Although BP was able to induce elevated levels of E, metabolism in both groups, the magnitude of b16a-hydroxylation was almost 165% higher in TDLU from mastectomy specimens relative to that seen in TDLU from mammoplasty specimens. These results indicate that benign (noninvolved) TDLU from cancerous breast may be more susceptible to the carcinogenic insult. It is interesting to note that BP exposure resulted in a concomitant decrease in the 2-hydroxylation pathway?’ These observations, taken together, suggest that exposure of noninvolved mammary TDLU to a chemical carcinogen may alter E, metabolism in favor of E, 16a-hydroxylation at the expense of the 2-hydroxylation pathway. The 16a-OHE, formed in TDLU may directly influence the epithelial component of TDLU to promote the expression of mutant phenotype induced by BP.
SUMMARY The metabolism of E, via the 16a-hydroxylation pathway is reported to be elevated in breast cancer patients as well as in subjects at high risk for developing breast cancer. The biological relevance of the metabolic pathway during the initiational events that lead to the tumorigenic transformation of mammary epithelium is not fully understood. The results obtained from the in vitro experiments discussed in this report permit the following conclusions:
1. Human mammary TDLU, the presumptive target site for breast cancer, possesses metabolic competence to biotransform &. 2. The biotransformation of E, in TDLU via the 16a-hydroxylation pathway is responsive to endogenous hormonal changes and to the presence of cancer, and is susceptible to carcinogenic insult. 3. The relative extent of E, 16a-hydroxylation may constitute a sensitive metabolic marker for evaluating the susceptibility of noninvolved mammary epithelium to carcinogen-induced transformation.
TELANG el al.: ESTRADIOL METABOLISM
77
REFERENCES 1. BANERJEE,M. R. 1976. Responses of mammary cells to hormones. Int. Rev. Cytol. 4
1-97. N. M. MEHTA,A. P. IYER& R. GANGULY. 1980. 2. BANERJEE,M. R., N. GANGULY, Functional differentiation and neoplastic transformation in an isolated whole mammary organ in vitro. In Cell Biology of Breast Cancer. C. McGrath, M. J. Brennan & M. A. Rich, Eds.: 485-516. Academic Press. New York, N.Y. P., F. KU~TENN & A. GOMPEL.1986. Estradiol/progesterone inter3. MAUVAIS-JARVIS, action in normal and pathologic breast cells. Ann. N.Y. Acad. Sci. 464: 152-167. 4. BANERIEE,M. R. 1986. An in vitro model for neoplastic transformation of epithelial cells in an isolated whole mammary gland of the mouse. In In W r o Models for Cancer Research. M. M. Webber & L. I. Sekeley, Eds.: 69-114. CRC Press. Boca Raton, Fla. E. A., M. J. VOCCI,J. W. COMBS,H. SANEFUJI, T. ROBBINS, D. H. JANSS,C. 5. HILLMAN, C. HARRIS& B. F. TRUMP.1980. Human breast organ culture studies. Methods Cell Biol. 21: 80-106. M. & R. VAN NIE. 1974. Estrogen receptor content and hormone responsive 6. SLUYSER, growth of mouse mammary tumor. Cancer Res. 3 4 3252-3257. 7. WELSCH,C. W. 1986. Interrelationship between fat and endocrine processes in mammary gland tumorigenesis. In Dietary Fat and Cancer. C. Ip, A. E. Rogers, D. F. Birt & C. Mettlin, Eds.: 623-654. A. R. Liss, Inc. New York, N.Y. V. C. 1976. Effect of tamoxifen (ICS 46,474) on initiation and growth of DMBA8. JORDAN, induced rat mammary carcinomata. Eur. J. Cancer 1 2 419-424. J. SCHNEIDER, K. E. ANDERSON & A. KAPPAS.1980. 9. FISHMAN,J., H. L. BRADLOW, Radiometric analysis of oxidation in man: sex differences in estradiol metabolism. Proc. Nat. Acad. Sci. USA 77: 4957-4960. J., D. KINNE,A. FRACCHIA, V. PIERCE,K. E. ANDERSON, H. L. BRADLOW 10. SCHNEIDER, & J. FISHMAN. 1982. Abnormal oxidative metabolism of estradiol in women with breast cancer. Proc. Nat. Acad. Sci. USA 7 9 3047-3051. H.L., R. HERSHCOPF, C. MARTUCCI & J. FISHMAN. 1986. 16a-Hydroxylation 11. BRADLOW, of estradiol: a possible risk marker for breast cancer. Ann. N.Y. Acad. Sci. 464: 138- 151. M. P., N. T. TELANG,H. KURIHARA & H. L. BRADLOW. 1988. Risk enhanced 12. OSBORNE, metabolism of estradiol in explant cultures of human benign mammary tissue. Proc. Am. Assoc. Cancer Res. 2 9 236. M. P., R. A. KARMALI, J. HERSHCOPF, H. L. BRADLOW, I. A. KOURIDES, W. 13. OSBORNE, WILLIAMS, P. P. ROSEN& J. FISHMAN. 1988.Omega-3fatty acid: modulation of estrogen metabolism and potential breast cancer prevention. Cancer Invest. 6 629-63 1. 14. Russo, I., C. CALAF,L. ROI & I. H.Russo. 1987. Influence of age and reproductive history on cell kinetics of normal human breast tissue in vitro. J. Nat. Cancer Inst. 7 8 413-418. & I. H. RUSSO.1988. Expression of phenotypical 15. Russo, J., D. REINA,J. FREDERICK changes by human breast epithelial cells treated with carcinogens in v i m . Cancer Res. 68:2837-2857. 16. BEUHRING, G. C. & M. H. JENSEN.1983. Lack of toxicity of methylene blue chloride to supravitally stained human mammary tissues. Cancer Res. 43:6039-6044. 17. TELANG,N. T., R. S. BOCKMAN, M. J. MODAK& M. P. OSBORNE.1989. The role of fatty acids in murine and human mammary carcinogenesis: an in vitro approach. In Carcinogenesisand Dietary Fat. S. Abraham, Ed.: 427-45 1. Kluwer Academic Publishers. Boston, Mass. 18. BRADLOW,H. L., R. J. HERSHCOPF, C. P. MARTUCCI & J. FISHMAN. 1985. Estradiol 16a-hydroxylationin the mouse correlates with mammary tumor incidence and presence of murine mammary tumor virus: a possible model for the hormonal etiology of breast cancer in humans. Proc. Nat. Acad. Sci. USA 8 2 6295-6299. & M. P. OSBORNE.1989. In vitro 19. TELANG,N. T., H. L. BRADLOW,H. KURIHARA biotransformation of estradiol by explant cultures of murine mammary tissues. Breast Cancer Res. Treat. 13: 173-181.
I8
ANNALS NEW YORK ACADEMY OF SCIENCES
20. PRUDHOMME, J. F., C. MALET,A. GOMPEL,J. P. LA LARDRIE,P. OCHOA,A. BOUE,P. MAUVAIS-JARVIS & F. KUTTENN.1984. 17P-Hydroxysteroiddehydrogenase activity in human breast epithelial and fibroblast cell cultures. Endocrinology 114: 1483- 1489. 21. TELANG,N. T., A. BASU,M. J. MODAK,H. L. BRADLOW & M. P. &BORNE. 1988. Parallel enhancement of ras proto-oncogene expression and estradiol 16a-hydroxylation in human mammary terminal duct lobular units (TDLU) by a carcinogen. Breast Cancer Res. Treat. 11: 138.
