[CANCER RESEARCH 47, 2583-2588, May 15, 1987] Correlation of Aromatic Hydroxylation of llß-substituted Estrogens with Morphological Transformation in Vitro but not with in VivoTumor Induction by These Hormones1

Joachim G. Liehr,2 Robert H. Purdy, John S. Baran, Erhard F. Nutting, Frank Colton, Erika Randerath, and Kurt Randerath Department of Pharmacology, The University of Texas Medical Branch, Galveston, Texas 77550 [J. G. L.¡;Southwest Foundation for BiomédicalResearch, San Antonio, Texas 78284 [R. H. PJ; G. D. Scarte & Co., Skokie, Illinois 60077 [J. S. B., E. F. N., F. C.J; and Department of Pharmacology, Baylor College of Medicine, Houston, Texas 77030 [E. R., K. R.]

ABSTRACT ster embryo cells (6) or in a subclone of BALB/c 3T3 embryo- derived mouse fibroblasts (7) which have both been used to In estrogen-induced cancer, catechol formation from administered study the mechanism of estrogen-induced cell transformation, steroids has been postulated to be a necessary event for estrogen activa the transformation efficiency did not increase with increasing tion and subsequent damage to cellular macromolecules.In the present hormonal potency of the estrogen tested. This lack of correla study, this hypothesis has been tested using two homologousseries of structurally related estrogens:estradiol, 11¿¡-methylestradio!, 110-cthyl- tion of hormone potency with transformation efficiency might estradiol, H/S-methyl-lTa-ethinylestradiol, 110-ethyl-17a-ethinylestra- imply that hormonal effects are not implicated in transforma diol, 110-methoxy-17a-ethinylestradiol,and 17a-ethinylestradiol.In the tion; i.e., avidity of receptor binding does not correlate with Syrian hamster renal carcinoma model, only 110-methylestradiol and transformation efficiency. Rather, these experimental results 17a-ethinylestradiolwereweakcarcinogens(2 of 20 and 2of 24 hamsters could be explained by postulating (7-9) a role for estrogen with tumors, respectively). The other estrogens tested induced renal metabolism in estrogen-induced cell transformation or carcin carcinomawithin6 to 9 months with an incidencein the 80-100% range. ogenesis. In particular, catechol estrogens, formed by 2- and/ The tumor incidencein vivodid not correlate with the rates of catechol or 4-hydroxylation. were invoked to be causally related to tumor formationby hamster kidney microsomesin vitro.Comparedto estradiol initiation. These metabolites of estrogens have been shown to (relative rate, 100), catechol formation by the substituted estrogens was significantly lower, ranging from 48 (110-methylestradioI) to 2 (110- bind to DNA (10). This elevated binding of catechol estrogens methoxy-17a-ethinylestradiol).Kidney DNÕof hamsters treated with versus parent hormones to DNA has been postulated (8, 9) to the four 17catechol estrogen biosynthesis and of which had previously been found to precede estrogen-inducedrenal cell transformation by a variety of estrogens was indeed ob carcinogenesisin vivo. served in partially transformed BALB/c 3T3 cells (7), subclone In contrast, relative rates of catechol estrogen formation by BALB/c A-31-1-13, an immortal mouse fibroblast cell line. Further 3T3 microsomescorrelated with inductionof morphologicaltransforma more, renal carcinoma in male Syrian hamsters could not be tion of BALB/c 3T3 cells and decreased in the followingorder: 110- induced with 2-fluoroestradiol at a dose hormonally equivalent methylestradiol> 17 a-ethinylestradiol > estradiol > 110-ethylestradiol to the positive control E2 (4, 5). Catechol formation from this > 110-methoxy-17a-ethinylestradiol.The hormonalpotenciesof several fluorinated steroid by hamster kidney microsomes in vitro was estrogen derivatives studied by various assays did not correlate with in vivo carcinogenicor in vitro cell-transformingactivities. It is concluded reduced compared to E2 (7) due to partial blockage of hydrox from these experiments that in cell culture catechol formation and ylation by the fluorine substituent, and this inhibition was morphological transformation are directly related. In vivo, aromatic presumed (4, 5) to be responsible for the inhibition of renal hydroxylationof administered estrogens did not correlate with the inci carcinoma in hamsters. denceof estrogen-inducedrenal carcinomain Syrian hamsters. To investigate the relationships among metabolic catechol formation, carcinogenicity, and capacity for morphological transformation of cells in culture, the biological properties of INTRODUCTION two homologous series of 11^'-substituted steroid estrogens In estrogen-induced carcinogenesis, no correlation of hor were investigated. For E2, Me-E2, Et-E2, EE, Me-EE, Et-EE, monal potency and carcinogenic activity of administered sub and Mox (structures are shown in Fig. 1), these and related stances has ever been reported (1-3). In the estrogen-induced biological properties were measured and the results are reported renal carcinoma model in male Syrian hamsters, the high tumor here. incidence elicited routinely (1) by administration of estradiol (E2)3 was not achieved when EE (1) or 2-fluoroestradiol (4, 5) MATERIALS AND METHODS was implanted into animals. Correspondingly, in Syrian ham- Chemicals. E2, EE, cholesterol, progesterone, hydrocortisone, Received6/19/86;revised1/5/87;accepted2/20/87. NADPH, ascorbic acid, magnesium chloride, catechol-0-methyltrans- Thecostsof publicationofthisarticleweredefrayedinpan bythepayment of pagecharges.Thisarticlemustthereforebeherebymarkedadvertisementin ferase (EC 2.1.1.6), and estrone were obtained from Sigma Chemical accordancewith18U.S.C.Section1734solelytoindicatethisfact. Co., St. Louis, MO. Mox was a gift from Dr. Jean-Pierre Raynaud, 1This work was supported in part by grants from the National Cancer Institute Roussel-Uclaf, Paris, France. Me-E2, Et-E2, Me-EE, and Et-EE were (CA 43232, CA 43233, CA 32157, CA 41569 and CA 24629) and by a Du Pont synthesized as described (11-13). The purity of the estrogens used was Occupational and Environmental Health Grant. verified by mass spectrometric analysis. S-[mefA.y/-:>H]Adenosyl-L-me- 2To whom all requests for reprints should be addressed, at Department of thionine (16 mCi/mmol) was purchased from New England Nuclear, Pharmacology, The University of Texas Medical Branch, Galveston, TX 77550- Boston, MA, and 12,4,6,7.-'IIjestradiol was from Amersham Inc., Ar 2774. 'The abbreviations used are: E2, estradiol; EE, 17a-ethinylestradiol; Me-E2, lington Heights, IL. Materials and chemicals required for the 32P- 11/3-methylestradiol; Et-E2, 11/3-ethylestradiol; Me-EE, ll/3-methyl-17a-ethi- nylestradiol; Et-EE, 110-ethyl-17a-ethinylestradiol; Mox, 110-methoxy-17a-eth- adduct assaywere the same as describedbefore (14). Polyethyleneimine inylestradiol (moxestrol); i.g., intragastric; LH, luteinizing hormone; IMl. folli cellulose thin-layer sheets were prepared as described (15). cle-stimulating hormone. Animals. Male Syrian hamsters at 4-6 weeks of age were obtained 2583 Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1987 American Association for Cancer Research. AROMATIC HYDROXYLATION OF ESTROGENS AND HORMONAL CARCINOGENESIS from Harlan/Sprague-Dawley, Madison, WI. The animals were kept in the day of ovariectomy and continued daily for 30 days. Groups of stainless steel cages (4 or 5 hamsters/group) with rodent chow and animals were killed on the day following the last injection. The pituitary water available ad libitum. After an acclimatization of 3-5 days, treat glands were stored frozen until analyzed for LH and FSH. LH activity ment was begun. Female CD-I mice, 21 days old, and sexually mature was determined by the ovarian ascorbic acid depletion test described by female Sprague-Dawley rats were purchased from Charles River Breed Parlow (31). FSH assays were performed by the ovarian augmentation ing Laboratories, Wilmington, MA. Immature uteri from New Zealand test described by Steelman and Pohley (32). White rabbits were obtained frozen in bulk from Pel-Freeze Biologicals, Relative Rate of Catechol Estrogen Formation. Washed microsomes Rogers, AR. from male Syrian hamster kidneys were prepared by differential cen- In Vivo Carcinogenicity. The carcinogenic activity of the various trifugation (33) and stored in 0.25 M sucrose at -70°C prior to use. estrogens was evaluated in male Syrian hamsters using the procedure Washed microsomes from about 2x10' nonconfluent BALB/c 3T3 of Kirk man (1) as described previously (4, 16). At day 0 and after 3.5 cells were similarly prepared (34) and were used immediately. Protein months each animal received s.c. one estrogen implant [25-mg pellet concentration was determined by the method of Bradford (35). The containing 10% cholesterol and prepared by a press procedure (17)]. radioenzymatic assays for catechol formation were carried out as de Untreated hamsters served as a negative control, and E2-treated ani scribed previously (7). mals served as a positive control. After the indicated time periods, the Morphological Transformation of BALB/c 3T3 Cells. The A-31-1-13 hamsters were killed by decapitation, and the kidneys were excised, subclone of the BALB/c 3T3 embryo-derived mouse fibroblast cell line inspected visually to determine the number of renal carcinoma nodules, isolated by Kakunaga et al. (36) was used in the assays for cell trans and then placed in formalin solution for histológica! examination. formation. Assays were carried out in duplicate as recommended by the Kidneys from animals treated with the 17<.cthmyl estrogens were cut IARC/NCI/EPA Working Group (37). in half; one half was placed in formalin solution, and the other half of the kidney was immediately frozen on Dry Ice and stored at —75°Cfor RESULTS later DNA adduct analysis. A histological examination was carried out on two sections from each kidney. Livers which looked diseased were Carcinogenic Activity of Modified Estrogens. The estrogen- also dissected for histological examination. induced renal carcinoma of male Syrian hamsters (1) is an Estrogen-induced DNA Adducts. The estrogen-induced DNA adduct established animal model that has been used to measure the analysis was carried out as described previously (14, 18), following the capacity for tumor induction of a large number of modified general procedure of Randerath et al. (19, 20). The DNA was isolated by a solvent extraction procedure (21). Adducts were purified by treat synthetic estrogens [for a recent review see Liehr and Sirbasku (38)]. With E2, a tumor incidence of 80-100% is obtained after ment of the DNA digest with nuclease PI (22). Only tissues of hamsters treated with one of the four 17a-ethinyl estrogens were tested by 32P 6-9 months of estrogen treatment (Ref. 38; Table 1). When Et- postlabeling analysis. The Chromatographie systems for the polyethyl- E2, Me-EE, Et-EE, and Mox were assayed in this animal model, eneimine cellulose thin-layer analysis of estrogen-induced DNA adducts the same high tumor incidence of 80-100% was achieved (Table have been described by Liehr et al. (14, 18). 1). Kirk man and Bacon (39) reported abdominal métastasesin Hormonal Activities of Me-E2 and Et-E2. Estrogen receptor binding hamsters with renal tumors after prolonged exposure to E- activity was measured by the method of Korenman (23-25). The binding assay was performed for 16 h at 4°Cat a concentration of 5 x IO"10M diethylstilbestrol. Abdominal métastaseswere also found in 33-39% of the hamsters treated for 9 months with the three [2,4,6,7,7-3H]estradiol, with 105,000 x g uterine cytosol prepared from ethinyl derivatives (Table 1). In many of these animals, large immature rabbits. To increase the specificity of the assay, the recom mendations of Tirenius (26) were followed by adding 1 x 10 '' M tumors had nearly destroyed the kidneys. The liver of one unlabeled progesterone and hydrocortisone to each tube. The standard hamster treated with Et-EE also contained a hepatoma. In for all assays was E2. All compounds were evaluated in duplicate at 8 contrast, two of the estrogens tested, Me-E2 and EE, were doses ranging from 10 to 2000 pg. weakly carcinogenic (2 of 20 and 2 of 24 animals with tumors, Estrogenic activity in vivo in the mouse uterine weight assay was respectively). The low carcinogenic activity of EE had been measured by the method of Rubin et al. (Il) as modified by Edgren reported before by Kirk man (1). The higher tumor incidence of (28). Compounds, dissolved or suspended in corn oil, were administered 20% reported by Li et al. (2) for his widely used steroid may s.c. or i.g. for 3 consecutive days. The control group received corn oil have been due to a more frequent s.c. implantation of EE in alone. The increase in uterine weight, compared to that of the oil- treated controls (/'•0.01 ). was used as the index of estrogenic activity. their experiment. Estrogenic Activity. In an attempt to find a cause for the weak Estrogenic activity in vivo in the rat vaginal smear assay was mea carcinogenic activities of Me-E2 and EE, the hormonal poten sured by the method of Biggers and Claringbold (29). One week after priming, the test compounds, dissolved or suspended in corn oil, were cies of these synthetic estrogens were assessed. Those of EE are administered over a 2-day treatment period. Vaginal smears were known (40,41) and were not repeated here. Biological responses performed 72 h after the first injection and those which were composed of Me-E2 and Et-E2 were substantially enhanced relative to E2 predominantly of cornified and/or round nucleated epithelial cells were (Table 2), particularly in their inhibition of the postcastration deemed to indicate estrogenic response. The 50% effective dose of a rise in pituitary LH and FSH. Superior binding of Me-E2 and test compound was calculated according to the method of Berkson (30). Et-E2 to estrogen receptors of rabbit uteri was also demon- The postovariectomy rise in pituitary LH and FSH was measured in young adult female rats which were ovariectomized and placed into TablehamstersTreatment 1 Carcinogenic activity of modified estrogens in male Syrian groups, such that the average of the body weights for each group was approximately 250 g. The compound was administered s.c. starting on of animals with tumors after7mo8(no. of animals examined) Abdominal incidence (no. ofanimals)E2 mo8 mo2(24)' métastases(%)0(18) (20) (9) (9)" Me-E2 (20) 2(10) 0(10)5 2(20) 0(20) 10 Et-E2 (20) 8(8)0 (5)"0 13(13) 0(13) 100 EE (25) 2(24) 0(24) 8 Me-EE (25) 17(21)" 17(21) 7(21) 81 Et-EE (25) 21 (23)° HO' 21 (23) 8(23) 91 Mox (23) 16 (18)" 16(18) 7(18) 89 Fig. 1. Structures of estrogens examined. Left, Me-E2 (R, = CH3) and Et-E2 Untreated (24)No. (8)8 (8)9 0 (8)Total16(18)0(24)Tumor0(24)89 0 (R, = CH2CHj); righi, Me-EE (R2 = CH3), Et-EE (R2 = CH2CH3), and Mox (R2 " This number includes kidneys of dead animals whenever these organs could = OCH,). be recovered and were usuable for histological analysis. 2584 Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1987 American Association for Cancer Research. AROMATIC HYDROXYLATION OF ESTROGENS AND HORMONAL CARCINOGENESIS Table 2 Ratios of the estrogenic potencies ofE2 and lip-substituted estradiol Syrian hamsters treated with EE, Me-EE, Et-EE, and Mox, derivatives, administered s.c. and i.g., relative to E2 respectively, was analyzed by 32P postlabeling assay (14, 18, Vaginal cell comification"CompoundE2 wt»s.c.1.0 inhibition'FSH uterine 20). Covalent DNA adducts that were not detected in DNA of receptor binding activity (invitro)111.00 untreated animals were identified by this method in kidney (s.c.)1.0 (S.C.)1.0 DNA from .E-diethylstilbestrol- or E2-treated hamsters. In all

Me-E2 4.4 1.210.0Uterine2.0 10.0 22.0 29.0 1.23 analyses with hamster kidney DNA exposed to each of the four Et-E2s.c.1.01.8i-g-1.0 0.9i-g-1.013.0Pituitary43.0LH 44.0Relative 1.44 17«-ethinyl estrogens, the adduci patterns found here were " Estrogen-induced stimulation of cornificai inn of vaginal epithelial cells mea indistinguishable with respect to number of adducts and Chro sured by the method of Biggers and Qaringbold (29). The 50% effective dose matographie characteristics (Fig. 2). They were also chromat- values for each compound were estimated from the pooled data of 4-12 doses (N=\5 rats/dose) administered over varying periods of time. The relative potency ographically identical to adducts induced by £-diethylstilbes- of each compound was calculated from the 50% effective dose values using estrone trol, E2, or hexestrol in cochromatography experiments (18). injected s.c. at total doses ranging from 0.2 to 5.0 ng as the common standard. * Estrogen-induced increases in mouse uterine weights measured by the method Thus, the indirect mechanism of induction of DNA adducts by of Edgren (28). Potency ratios were calculated from the relative potencies deter estrogens postulated previously (18) is supported by data with mined by least squares regression analysis from 6-8 different bioassays (5 points/ the additional synthetic estrogens tested here. assay,/V = 10 mice/point, P < 0.01). Estrone injected s.c. at 0.03 and 0.1 Mg/day Catechol Estrogen Formation by BALB/c 3T3 Cell Micro was included as the common standard in each assay. ' Bioassay of pituitary tissue from mature female rats following 30 days of somes. In view of the lack of correlation between tumor inci treatment. FSH concentration was determined by augmentation of ovarian weight dence in vivo (Table 1) and the relative rates of catechol for gain according to the method of Steelman and Pohley (32), and the LH concen tration was determined by the depletion of ovarian ascorbic acid concentration mation by hamster kidney microsomes (Table 3) from 11,8- method of Parlow (31). substituted estrogens, these E2 derivatives were tested for met J Binding of compounds to uterine cytosol receptors of rabbits by method of abolic 2- and/or- 4-hydroxylation and for their ability to induce Korenman (23-25). Relative potencies of each compound were calculated by comparison with the amount of E2 required for reducing specific binding by 50% morphological transformation of BALB/c 3T3 cells. The cor of maximal value. relation between these properties in cell culture noted previ ously for other estrogens (7) needed to be demonstrated for the Table 3 Catechol estrogen formation by hamster kidney microsomes E2 derivatives used here to establish a potential divergence of The method of Purdy et al. (7) was used with 50 MMconcentrations of substrate properties of estrogens in culture and in vivo. and 250 #ig of microsomal protein per assay. The data for catechol formation represent the mean ±SE of four measurements. The relative rates of catechol estrogen formation from \\ß- formation substituted estrogens catalyzed by BALB/c 3T3 cell micro SubstrateE2 (pmol/min/mgprotein)2.579 rate100 somes was measured using microsomes (34). As illustrated in ±0.286 Me-E2 1.233 ±0.158 482228 Et-E2 0.559 ±0.096 EE 0.71 3 ±0.065 A Me-EE 0.440 ±0.035 17 B Et-EE 0.367 ±0.036 142 MoxCatechol 0.061 ±0.012Relative

sinned. In addition, Ft-H2 exhibited at least 10 times the i.g. activity of E2 when measured by both the rat vaginal cornifi- cation and the mouse uterine weight assay and compared well to EE (40, 41). Because of the high hormonal potency of Me- E2, it was concluded that its reduced carcinogenic effect was not due to hormonal factors. Alterations in metabolism were therefore investigated as another possibility for changes in tumor induction and tumor growth stimulation. Catechol Estrogen Formation by Hamster Kidney Microsomes HI Vitro. The capacity of Syrian hamster kidney microsomes to metabolize the 1Ip'-substituted estrogens of catechols was mea a « sured in vitro using a 50 /¿Mconcentration of substrate and a microsomal content of 250 ¿igprotein/assay by analogy to published experiments (7). The rates of total catechol forma tion, shown in Table 3, are those of 2- and 4-hydroxylation. To assess the rates of 2-versus 4-hydroxylation, each of the catechol isomers were measured for E2 and EE as substrates. In these two examples, the 2-hydroxyestrogens were the major products: 2-hydroxyestradiol, 90%; 4-hydroxyestradiol, 10%, 2-hydroxy- 17-a-ethinyl estradiol, 67%, 4-hydroxyl-17a-ethinyl estradiol, Fig. 2. 32P postlabeling analysis of DNA adducts induced in kidneys of male 19% of total product in each reaction. Compared to E2 (relative Syrian hamster by treatment with EE (A), Me-EE (A), Et-EE (C), and Mox (D), DNA was digested to deoxyribonucleoside 3'-monophosphates, which were con rate, 100), total catechol formation by the modified estrogens verted to "P-labeled 3',5'-bisphosphate derivatives by incubating with [-y-"P]- was significantly reduced. The relative rates ranged from 48 for ATP and T4 polynucleotide kinase (15, 18, 20). Adducts were mapped by Me-E2 to 2 for Mox. As illustrated in Tables 1 and 3, catechol polyethyleneimine cellulose thin-layer chromatography (14, 18) and visualized by formation from the modified estrogens was not correlated with autoradiography for 1 day. The solvents for two-dimensional chromatography were 3.8 M lithium formate, 6.8 M urea (pH 3.4) (from bottom to top) and 0.7 M their carcinogenic activities. The relative rates of aromatic sodium phosphate. 7 M urea (pH 6.4) (from left to right). The major adducts have hydroxylation were lowest in three strongly carcinogenic estro been marked a, b, and d. Weaker adducts c and e (18) required longer exposure gens, Mox, Et-EE, and Me-EE. times to become detectable. The dark spot below adduci b, as well as a number of fainter spots, represent background material that became labeled also when DNA Adduct Analysis. Kidney DNA isolated from male kidney DNA from untreated male hamster was analyzed (not shown). 2585 Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1987 American Association for Cancer Research. AROMATIC HYDROXYLATION OF ESTROGENS AND HORMONAL CARCINOGENESIS Fig. 3, Me-E2 was a better substrate for estrogen 2- and/or 4- transformation was measured with the A-31-1-13 subclone of hydroxylase activity than was E2 or EE. Et-E2 was converted BALB/c 3T3 embryo-derived mouse fibroblast cells (36). Using to a mixture of 2- and 4-hydroxyI-ll/î-ethylestradiol at a re this cell line, morphological transformation by estrogens (7,37) duced rate compared with E2. Mox was a very poor substrate has previously been correlated with rates of 2- and/or 4-hy- for 2- and/or 4-hydroxylase activity. Cell transformation ex droxylation (7). Morphological transformation of BALB/c 3T3 periments were carried out subsequently with these same cells. cells by I In-substituted estrogens is reported in Table 4. 3- Morphological Transformation of BALB/c 3T3 Cells. The Methylcholanthrene and E2 served as positive controls. In these capacity of 11.;'-substitut ed estrogens to effect morphological assays, Mox did not elicit a significant increase in numbers of type III transformed foci over those found with dimethyl sulf oxide alone (solvent control). EE, E2, Me-E2, and Et-E2 1.5J showed significant increases of type III foci/104 survivors with increasing concentrations of test substances. In the course of these experiments, it was noted that a good correlation existed between the frequency of induction of morphological transfor mation by estrogens and their relative rates of aromatic hy droxylation. Thus, the correlation between these two properties of estrogens, observed previously in cell culture (7), was dem onstrated also for these two homologous series of l 1/3-substi- tuted estrogens.

DISCUSSION The comparative properties of the two homologous series of estrogens investigated in this study are summarized in Table 5. With respect to tumor incidence in vivo, it is difficult to 10 recognize any structure-activity relationships in these two series Incubation Time (min) of synthetic estrogens. Two of the six estrogens examined, Me- Fig. 3. Radioenzymatic assay for catechol estrogen Formation. Relative rates E2 and EE, were weak inducers of renal carcinoma, while the of aromatic hydroxylation of Me-E2 (•),E2(O), Et-E2, (X), EE (v), and Mox (*) other four structurally related steroids had a carcinogenic activ using microsomes (608 ¡igprotein) from BALB/c 3T3 cells. ity comparable to that of E2. Furthermore, a steroid that combines the substituents of the two weak carcinogens, Me- Table 4 Morphological transformation of subclone A-31-1-IÌofBALB/c 3T3 EE, was highly carcinogenic in the Syrian hamster renal carci cells by estrogens noma model. The optimal structural features of a "noncarcin- ogenic estrogen" can therefore not be derived from the present formation tration type III frequency/ IO4 foci/dish*0.311.910.250.730.670.831.291.181.001.091.422.330.180.330.270.330.550.670.920.551.081.100.250.170.180.000.11Transsurvivors'0.713.60.7"2.03.48.23.02.63.45.915.60.5"1.1"1.51.4"2.92.03.42.76.88.90.2'0.5"0.4"0.0"0.3"analyses. CompoundDimethyl (MM)1102030405010203040SO102030405010203040SO1020304050Survival*(%)(100)32808345231490867455349171425543756247362810810310810296Av.The earlier observations (2, 4, 5) of the lack of correlation (0.25%)3-MethylcholanthreneE2Me-E2EI-E2EEMoxConcensulfoxide between hormonal potency and carcinogenic activity of various estrogens in the Syrian hamster renal carcinoma model have been confirmed here. Me-E2 and Et-E2 have hormonal activi ties consistently greater than those of E2 with respect to recep tor binding and other hormonal responses (Table 2); yet the former estrogen derivative is weakly and the latter is strongly carcinogenic in an in vivo test system (Table 1). Furthermore, EE, a highly potent synthetic estrogen (40, 41), was a poor inducer of renal carcinoma in Syrian hamsters. The difficulties experienced when comparing hormonal and carcinogenic activities of estrogens previously led to suggestions (7-9) of a role for metabolic activation, particularly microsomal aromatic hydroxylation, in tumor induction. The chemical reac tivity of catechol estrogens and their binding to DNA (10) was thought (8, 9) to initiate hormonal cancer. However, in this larger sample of modified steroids, which were chosen for testing because of their structural similarity (Fig. 1), relative rates of catechol estrogen formation with kidney microsomes (Table 3) could not be correlated with carcinogenic activity (Table 1). Estrogen-induced DNA adduct formation (14, 18) (Fig. 2) was investigated to further probe DNA modification as a mech anism for the induction of hormonal cancer. Previous work demonstrated (18) the formation of an identical set of DNA " Average cloning efficiency for dimethyl sulfoxide control (44%) taken as adducts when E2, f-diethylstilbestrol, or hexestrol were used 100%. as tumor-inducing estrogens. Based on these results, an indirect * Average type III foci/dish in 12 dishes. c Type III foci/(% survival/44). mechanism of hormonal genotoxicity was postulated. Estrogens " Not significant. are capable of inducing unknown endogenous substances to 2586

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1987 American Association for Cancer Research. AROMATIC HYDROXYLATION OF ESTROGENS AND HORMONAL CARCINOGENESIS

Table 5 Summary of properties of two homologous series of estrogens taken from Tables 1-4 and Figs. 2 and 3 of formation DNA of by BALB/c 3T3 activity in hamster formation in hamster activity adducts BALB/c 3T3 cells cells EstrogenE2 (Table1)+++" (Table3)+++ (Table2)++ (Fig.2)+* (Table4)++ (Fig.3)++

Me-E2 + ++ +++ ND +++ +++ Et-E2 +++ ++ +++ ND+* + + EE + -H-+ +++c ++ ++ Me-EE +++ ND +» ND ND Et-EE +++ +±Estrogenic ND +» NDCatechol ND +++'Induction +*Transformation MoxCarcinogenic +++Catechol ± °Biological activity: +++, high; ++, medium; +, low; -, not significant; ND, not determined. *Quantitative correlations need further investigation (see "Discussion"). '' As summarized in Refs. 40 and 41. ' As described in Ref. 42. bind covalently to DNA. An identical set of modified nucleo- McLachlan, J. A., Wong, A., Degen, G. H., and Barrett, J. C. Morphological and neoplastic transformation of Syrian hamster embryo fibroblasts by tides was detected (Fig. 2) with the 11/3-substituted estrogens diethylstilbestrol and its analogs. Cancer Res. 42: 3040-3045, 1982. used here. These results support the postulate that estrogens Purdy, R. H., Maclusky, N. J., and Naftolin. Does estrogen-induced geno- toxicity require receptor-mediated events? In: 3. A. McLachlan (ed.), Estro are not part of the adduct structure. Because our DNA prepa gens in the Environment, pp. 146-167. New York: Elsevier/North Holland, rations contained varying amounts of tumor DNA, which had 1986. previously been demonstrated (14, 18) to be free of detectable Metzler, M., and McLachlan, J. A. Oxidative metabolism of diethylstilbestrol and steroidal estrogens as a potential factor in their fetotoxicity. In: D. DNA adducts, quantitative correlations cannot yet be made. Neubert, H. J. Merker, H. Nau, and J. Langman (eds.), Role of Pharmaco- The precise relationship of DNA adduct concentrations to kinetics in Prenatal and Perinatal Pharmacology, pp. 157-163. Stuttgart: tumor incidence is currently under investigation. Georg Thieme Verlag, 1978. Liehr, J. G., DaGue, B. B., Saltatore, A. M., and Sirbasku, D. A. Multiple The lack of correlation between the frequency of tumor roles of estrogen in estrogen-dependent renal clear-cell carcinoma of Syrian induction by the modified estrogens (Table 1) and rates of hamster. In: G. H. Sato, A. B. Pardee, and D. A. Sirbasku (eds.), Growth of Cells in Hormonally Defined Media, Cold Spring Harbor Conferences on catechol formation (Table 3) prompted an investigation of their Cell Proliferation, Vol. 9, Book A, pp. 445-458. Cold Spring Harbor, NY: ability to induce morphological transformation in cell culture. Cold Spring Harbor Laboratory Press, 1982. In the Permanent BALB/c 3T3 mouse fibroblast cell line (35), 10. Metzler, M., and Zeeh, J. V. Different metabolism by mammalian and horseradish peroxidases in vitro of steroidal estrogens and their catechol a correlation had previously been found (7) between morpho metabolites. J. Steroid Biochem. 24:653-655, 1986. logical transformation by estrogens and their relative rates of 11. Baran, J. S., Lennon, H. D., Mares, S. E., and Nulling, E. F. 11/3-Methyl- 2- and/or 4-hydroxylation. With the 11/3-substituted estrogens 19-norsteroids: novel progestational hormones. Experientia (Basel). 26:762- examined here, a close correlation between these properties was 763, 1970. 12. Van den Broek, A. J., Broess, A. I. A., van den Heuvel, M. J., de Jongh, H. also observed (Fig. 3; Table 4). It is therefore concluded that P., Leemhuis, J., Schonemann, K. H., Smits, J., de Visser, J., van Vliet, N. aromatic hydroxylation of l l/3-substituted estrogens correlates P., and Zeelen, F. J. Strategy in drug research. Synthesis and study of the progestationa) and ovulation inhibitory activity of a series of 110-substituted- with the frequency of cell transformation in culture but not with 17a-ethynyl-4-estren-17/3-ols. Steroids. 30:481-510, 1977. in vivo tumor induction. In male Syrian hamsters, kidney is the 13. Baran, J. S., Lungford. D. D., Laos, I., and Liang, C. D. The synthesis of target organ for hormonal cancer, and the tumors are carcino llfü-alkyl-19-norsteroids: a novel class of potent steroid hormones. I. The synthesis of 11/3-methyl- and l l/3-ethylestradiol. Tetrahedron. 33: 609-616, mas derived from epithelial cells (1). In contrast, the target cell 1977. in the in vitro experiments is a fibroblast. This might explain 14. Liehr, J. G., Randerath, K., and Randerath, E. Target organ-specific covalent DNA damage preceding diethylstilbestrol-induced carcinogenesis. Carcino the discrepancies observed here between in vivo carcinogenesis genesis (Lond.). 6:1067-1069, 1985. and the in vitro transformation induced by the modified estro IS. Gupta, R. C., Reddy, M. V., and Randerath, K. 32P-postlabeling analysis of gens. non-radioactive aromatic carcinogen-1 )N A adducts. Carcinogenesis (Lond.). 3:1081-1092, 1982. 16. Liehr, J. G., Saltatore, A. M., McLachlan, J. A., and Sirbasku, D. A. Mechanism of diethylstilbestrol carcinogenicity as studied with the fluori ACKNOWLEDGMENTS nated analogue £-3',3",5',5"-tetrafluorodiethylstilbestrol. Cancer Res. 43: The authors thank Dr. Jean-Pierre Raynaud, Roussel-Uclaf, Paris, 2678-2682, 1983. 17. McGregor, R. F., Putch, J. D., and Ward, D. N. Estrogen-induced kidney France, for a gift of Mox. We gratefully acknowledge the technical tumors in the Golden hamster. I. Biochemical composition during tumori- assistance of Tommie A. Avitts in the DNA adduct analysis, of Connie genesis. J. Nati. Cancer Inst. 24: 1057-1066, 1960. K. Durocher in the transformation assays, of Perry H. Moore, Jr. in 18. Liehr, J. G., Avitts, T. A., Randerath, E., and Randerath, K. Estrogen- induced endogenous DNA adduction: possible mechanism of hormonal can the radiochemical assays, and of Donald J. Garcia, Jr. in the DNA cer. Proc. Nati. Acad. Sci. USA. 83: 5301-5305,1986. isolations. The histológica! examinations were carried out by Dr. W. 19. Randerath, K., Reddy, M. V., and Gupta, R. C. 32P-labeling test for DNA F. Mackenzie, University of Texas Medical School at Houston. damage. Proc. Nati. Acad. Sci. USA. 78:6126-6129, 1981. 20. Randerath, E., Agrawal, H. P., Weaver, J. A., Bordelon, C. B., and Rander ath, K. "P-l'ostlabeling analysis of DNA adducts persisting for up to 42 REFERENCES weeks in the skin, epidermis, and dermis of mice treated topically with 7,12- dimethylbenz[a]anthracene. Carcinogenesis (Lond.). 6: 1117-1126, 1985. Gupta, R. C. Non-random binding of the carcinogen /V-hydroxy-2-acetyla- 1. K¡rkman.H. Growth characteristics in the Syrian hamster. Nati. Cancer Inst. 21. Monogr. /: 1-57, 1959. minofluorene to repetitive sequences of rat liver DNA in vivo. Proc. Nati. 2. Li, J. J., Li, S. A., Klicka, J. K., Parsons, J. A., and Lam, L. K. T. Relative Acad. Sci. USA. 81:6943-6947, 1984. carcinogenic activity of various synthetic and natural estrogens in the Syrian 22. Reddy, M. V., and Randerath, K. Nuclease Pi-mediated enhancement of hamster kidney. Cancer Res. 43: 5200-5204, 1983. sensitivity of "P-postlaheling test for structurally diverse DNA adducts. 3. Furth, J. Hormones as enologica! agents in neoplasia. In: F. F. Becker (ed.), Carcinogenesis (Lond). 7:1543-1551, 1986. Cancer—A Comprehensive Treatise, Ed. 2, Vol. 1, pp. 89-134. New York: 23. Korenman, S. G. Use of receptor proteins for steroid-hormone assays. Plenum Publishing Corp., 1982. Methods Enzymol. 36A: 49-52, 1975. 4. Liehr, J. G. 2-Fluoroestradiol: separation of estrogenicity from carcinogen- 24. Korenman, S. G. Radio-ligand binding assay of specific estrogens using a icity. Mol. Pharmacol. 23: 278-281, 1983. soluble uterine macromolecule. J. Clin. Endocrino!. Menili. 28: 127-130, 5. Liehr, J. G., Stance!, G. M., Chorich, L. P., Bousfield, G. R., and Ulubelen, 1968. A. A. Hormonal carcinogenesis: separation of estrogenicity from carcinogen- 25. Korenman, S. G. Comparative binding of estrogen and its relation to estro- icity. Chem.-Biol. Interact. 59: 173-184,1986. genie potency. Steroids. 13: 163-177, 1969. 2587 Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1987 American Association for Cancer Research. AROMATIC HYDROXYLATION OF ESTROGENS AND HORMONAL CARCINOGENESIS

26. Tirenius, T. Affinities of progestogen and estrogen receptors in rabbit uterus 35. Bradford, M. M. A rapid and sensitive method for the quantitation of for synthetic progestogens. Steroids. 23: 909-919, 1974. microgram quantities of protein utilizing the principle of protein-dye binding. 27. Rubin, I H. Dorfman, A. s.. Black, L., and Dorfman, R. I. Bioassay of Anal. Biochem. 72:248-254, 1976. estrogens using the mouse uterine response. Endocrinology. 49: 429-439, 36. Kakunaga, T. Chemical transformation of human and rodent cells. In: U. 1951. Saffiotti and H. Autrup (eds.), In VitroCarcinogenesis, NCI Technical Report 28. Edgren, R. A. Notes on impeded estrogens. Prue. Soc. Exp. Biol. Med. 92: Series No. 44, pp. 84-99. Bethesda, MD: US National Cancer Institute, 569-571,1956. 1978. 29. Biggers, J. 1).. and Claringbold, P. J. Criteria of the vaginal response to 37. IARC/NCI/EPA Working Group. Cellular and molecular mechanisms of estrogens. J. Endocrino!. 11: 277-284, 1951. cell transformation and standardization of transformation assays of estab 30. Berkson, J. A statistically precise and relatively simple method of estimating lished cell lines for the prediction of carcinogenic chemicals: overview and the bioassay with quanta) response, based on the logistic function. J. Am. recommended protocols. Cancer Res. 45: 2395-2399, 1985. Stat. Assoc. 48: 565-599, 1953. 38. Liehr, J. G., and Sirbasku, D. A. Estrogen-dependent kidney tumors. In: M. 31. Parlow, A. F. Bioassay of pituitary luteinizing hormone by depletion of Taub (ed.), Tissue Culture of Epithelial Cells, pp. 205-234. New York: ovarian ascorbic acid. In: A. Albert (ed.), Human Pituitary Gonadotropins, Plenum Publishing Corp., 1985. pp. 300-310. Springfield, IL: Charles C Thomas, Publishers, 1961. 39. Kirkman, H. J., and Bacon, R. L. Estrogen-induced tumors of the kidney. I. 32. Steelman, S. L., and Pohley, F. M. Assay of the follicle stimulating hormone Incidence of renal tumors in intact and gonadectomized male golden hamsters based on the augmentation with human chorionic gonadotropi. Endocri treated with diethylstilbestrol. J. Nati. Cancer Inst. 13, 745-755, 1952. nology. S3: 604-616, 1953. 40. Brotherton, J. Sex Hormone Pharmacology. New York: Academic Press, 33. Purdy, R. II. and Marshall, M. V. Enhancement of the mutagenicity of 1976. carcinogenic arylamines by ethinyl estradiol. Carcinogenesis (Lond.). 5: 41. Emmens, C. W., and Martin, L. Estrogens. In: R. I. Dorfman (ed.), Methods 1709-1715, 1984. in Hormone Research. Vol. 3, Part A, pp. 1-80. New York: Academic Press, 34. Purdy, R. H., Goldzieher, J. W., LeQuesne, P. W., Abdel-Baky, S., Durocher, 1964. C. K.. Moore, P. H., Jr., and Rhim, J. S. Active intermediates and Carcino 42. Raynaud, J.I'.. Bouton, M. M., Gallet-Bourquin, D., Philibert, D., Tourne- genesis. In: G. Merriam and M. B. Lipsett (als.). Catechol Estrogens, pp. mine, C., and Azadian-Boulanger, G. Comparative study of estrogen action. 123-140. New York: Raven Press, 1983. Mol. Pharmacol. 9: 520-533, 1973.

2588 Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1987 American Association for Cancer Research. Correlation of Aromatic Hydroxylation of 11β-substituted Estrogens with Morphological Transformation in Vitro but not with in Vivo Tumor Induction by These Hormones

Joachim G. Liehr, Robert H. Purdy, John S. Baran, et al.

Cancer Res 1987;47:2583-2588.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/47/10/2583

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/47/10/2583. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1987 American Association for Cancer Research.