(Alcohol) Sulfotransferase, Resulting in Tamoxifen DNA Adducts1

[CANCER RESEARCH 58. 647-653, February 15, 1998] a-Hydroxytamoxifen Is a Substrate of Hydroxysteroid (Alcohol) Sulfotransferase, Resulting in Tamoxifen DNA Adducts1

Shinya Shibutani,2 Lakkaraju Dasaradhi, Isamu Terashima, Erden Banoglu, and Michael W. Duffel

Department of Pharmacological Sciences, State University of New York al Stony Brook, Stony Brook, New York 11794-8651 [S. S.. L. D., 1. T.], and Division of Medicinal and Natural Products Chemistry, College of Pharmacy, The University of Iowa. Iowa City, Iowa 52242 [E. B., M. W. D.Â¡

ABSTRACT likely candidates, capable of forming DNA adducts with nucleosides (20-23). However, a-OHTAM has only a low level of reactivity with When a-hydroxytamoxifen (Â«-OHI AM) was incubated with rat liver DNA in vitro (21). Because the formation of TAM-DNA adducts was hydroxysteroid (alcohol) sulfotransferase a (STa) and 3'-phosphoadenosine 5'-phosphosulfate, (E)-a-OHTAM was found to be a better sub inhibited by sulfotransferase inhibitors (24), the a-hydroxyl moiety of TAM was expected to be O-sulfonated prior to reaction with DNA strate for STa than (Z)-a-OHTAM. To explore the formation of tamoxifen (TAM)-derived DNA adducts, DNA was incubated with STa and either (24, 25). Recently, one of us found that a-sulfate TAMs are highly (E)-a-OHTAM or (Z)-a-OHTAM in the presence of 3'-phosphoadenosine reactive with DNA, forming four diastereoisomers of dG-N2-TAM S'-phosphosulfate. Using 32P-postlabeling analysis, the amount of TAM- (26). Similar results were observed for a model activated form, DNA adducts resulting from (E)-a-OHTAM was 29 times higher than that a-acetoxytamoxifen (25, 26). We also found that these dG-N2-TAM observed with (f)-a-OHTAM alone. Using (Z)-a-OHTAM and STa, some adducts display a high miscoding potential, predicting G â€”>T and TAM-DNA adducts were also detected but at levels 6.5 times lower than G â€”>Ctransversions and deletions in mammalian cells (27). that observed with (E)-a-OHTAM and STa. When compared with stand ards of stereoisomers of 2'-deoxyguanosine S'-monophosphate-A^-tamox- Hydroxysteroid sulfotransferases have been purified to homogene ity (28), and the substrate specificities of these enzymes led to their ifen, the major tamoxifen adduct was identified chromatographically as an epimer of the trans form of a-(JV2-deoxyguanosinyl)tamoxifen, and the designation as both hydroxysteroid and alcohol sulfotransferases. minor adduct was identified as an epimer of the civ form. In the reaction Many hydroxymethyl polycyclic aromatic hydrocarbons are activated mixture, a conversion from (E)-a-OHTAM to (Z)-a-OHTAM through the to carcinogenic and/or mutagenic sulfuric acid esters through reac carbocation intermediate was also detected. These results show that sul- tions catalyzed by these enzymes (29, 30). Because STa catalyzes the fation of a-OHTAM catalyzed by STa results in the formation of TAM- sulfation of xenobiotic alcohols, including benzylic alcohols (31), DNA adducts. a-OHTAM is likely a substrate for this enzyme. In this study, we report that dG-N2-TAM adducts are formed through the sulfation of a-OHTAM catalyzed by STa. INTRODUCTION TAM3 (structure in Fig. 1) is widely used in the treatment of breast cancer (1) and is being considered as a prophylactic agent for healthy MATERIALS AND METHODS women with a positive family history of breast cancer (2, 3). How Chemicals. [â€¢y-32P]ATP(specific activity, 6000 Ci/mmol) was obtained ever, this drug has been reported to be a potent hepatocarcinogen in from Amersham Corp. (Arlington Heights. IL). PEI-cellulose plates were rats (4-6). A high frequency of mutations were found in the DNA purchased from Machery-Nagel (Duren, Germany). Calf thymus DNA, pro- from livers of lambda//ac/ transgenic rats treated with TAM (7) and in teinase K, potato apyrase, alkaline phosphatase (type III), dGVP, PAP, and the rat p53 gene in hepatocarcinomas induced by TAM (8). An PAPS were purchased from Sigma Chemical Co. (St. Louis, MO). RNase A, increased incidence of endometrial cancer has been observed in breast RNase Tl, micrococcal nuclease, and spleen phosphodiesterase were obtained cancer patients treated with TAM (9-11). In 1996, TAM was listed as from Worthington Biochemical Co. (Freehold, NJ). T4 polynucleotide kinase a human carcinogen by the IARC (12). was purchased from Stratagene (La Jolla, CA). HPLC analysis was performed TAM is activated in liver microsomes of rats and humans (13-15) on a Waters 990 HPLC instrument, equipped with a photodiode array detector. Purification of STa. STa was purified to apparent homogeneity from and gives rise to DNA adducts in the livers of rodents (16, 17). A female Sprague Dawley rats (9-10 weeks of age), using a modification (31) of TAM-derived DNA adduct has been detected in the endometrial a procedure published previously (32, 33). The resulting STa was homogene tissues obtained from breast cancer patients treated with TAM (18). ous by SDS-PAGE gel electrophoresis with Coomassie Blue staining. Protein Oxidative species such as 4-hydroxytamoxifen quinone methide have concentrations were determined using a modified Lowry procedure (34), with been reported to promote the reaction with DNA (19). Alternatively, BSA as a standard. Hydroxysteroid sulfotransferase activity was determined in a-hydroxylation of TAM and its metabolites, tamoxifen Af-oxide, assay mixtures containing 0.25 M potassium phosphate (pH 7.0), 8.3 mM Af-desmethyltamoxifen, and 4-hydroxytamoxifen, respectively, are mercaptoethanol, 200 /J.MPAPS, and 40 /Â¿Mdehydroepiandrosterone. Enzyme units are expressed as nanomoles of sulfuric acid ester product formed per minute. The specific activity of the STa was 99.2 units/mg protein. Received 9/17/97; accepted 12/18/97. The costs of publication of this article were defrayed in part by the payment of page Kinetic Studies of STa with a-OHTAM. (E)-a-OHTAM or (Z)-a-OH- charges. This article must therefore be hereby marked advertisement in accordance with TAM was synthesized by the established protocol (35). PAPS was prepared 18 U.S.C. Section 1734 solely to indicate this fact. according to a published procedure (36). Both (Â£)-a-OHTAM and (Z)-a- 1This research was supported in part by the Lilly W. Ammann Breast Cancer Research OHTAM were evaluated as substrates of purified STa using a published HPLC Award, the School of Medicine, SUNY at Stony Brook (to S. S.), Grant ES04068 (to S.S. and A. P. Grollman) from NIEHS, and Grants CA17395 (to A. P. G.) and CA38683 (to procedure for determination of the concentration of PAP formed in the reaction M. W. D.) from NIH. I. T. was supported by a postdoctoral fellowship from the Uehara (37). Reaction mixtures (30 jil total volume) contained 0.25 M potassium Memorial Foundation, Japan. phosphate buffer (pH 7.0), 8.3 mM 2-mercaptoethanol, 0.3 mM PAPS, and 2 To whom requests for reprints should be addressed, at Department of Pharmacolog various concentrations of the alcohols in acetonitrile (final concentration of ical Sciences, State University of New York at Stony Brook, Stony Brook, New York 11794-8651. Phone: (516)444-8018; Fax: (516)444-3218. acetonitrile in the assay was no more than 5% v/v). Reactions were initiated by 3 The abbreviations used are: TAM, tamoxifen; dG, 2'-deoxyguanosine; dG3'P, 2'- the addition of 2.4 fig of STa, incubated at 37Â°Cfor 15 min, and terminated by deoxyguanosine 3'-monophosphate; STa, hydroxysteroid (alcohol) sulfotransferase a; (Â£)- the addition of 30 Â¿ilof methanol. The substrate-dependent concentration of and (Z)-a-OHTAM, trans and cii forms of a-hydroxytamoxifen, respectively; dG-N2- TAM, a-(/V2-deoxyguanosinyl)tamoxifen; PEI, poly(ethyleneimine); PAP, adenosine PAP formed in the reaction was determined by HPLC (37), and linear standard 3',5'-diphosphate; PAPS, 3'-phosphoadenosine 5'-phosphosulfate; HPLC, high-perfor curves directly relating the area of HPLC peaks to the concentrations of PAP mance liquid chromatography; hHST, human hydroxysteroid sulfotransferase. were determined daily. Apparent Km and Vmax values for a-OHTAM as a 647

Fig. 1. Structures of diastercoisomers of dG-N2- TAM DNA adduci, fr-1 and fr-2 (Fr-I & 2) are epimers of the trans form of dG-N2-TAM. and fr-3 and fr-4 (Fr-3 & 4) are epimers of the cis form of dG-N2-TAM.

(H,C)2N Fr-1 & 2 Fr-3 & 4

substrate lor STa are presented as Â±SEobtained by nonlinear least squares cellulose plates during the first and second development. The plates were curve-fitting (38) of the velocity data to the Michaelis-Menten equation. washed twice with distilled water for 20 min at room temperature and air dried. Preparation of dGVP-N2-TAM. Pyridinium (Z)-a-sulfate TAM was syn The plates were developed for 16 h in 1.7 M sodium phosphate buffer (pH 6.0; thesized by our established protocol (26). dGVI>(l.O mg) was reacted at 37Â°C Dl buffer) with a paper wick and subsequently developed to 2 cm onto the for 2 h with 2.0 mg of pyridinium (Z)-a-sulfate TAM in 500 fil of 100 mM paper wick in the same direction with 1.7 M lithium formate and 4.14 M urea Tris-HCl buffer, pH 8.0. The reaction mixture was centrifuged. and the (pH 3.5: D2 buffer). The plates were further developed at a right angle to the supernatant was extracted twice with 500 ftl of butanol. To isolate TAM- previous direction of development in 0.6 M LiCI, 0.375 M Tris-HCl, and 6.375 modified dGv,,. the pooled butanol extracts were concentrated at reduced M urea (pH 8.0; D3 buffer). The position of adducts was established by a pressure and subjected to a reverse-phase column. fiBondapak C,s (0.78 X 30 ÃŸ-phosphorimaging (Molecular Dynamics, Inc.) or autoradiography. using cm; Waters), eluted over 30 min at a flow rate of 2.0 ml/min with a linear Kodak X-Omat XAR film. gradient of 0.05 M trielhylammonium acetate (pH 7.0) containing 10â€”Â»50% The radioactive spots on the PEI-cellulose plates were scraped, and the acetonitrile (39). A part of the repurified TAM-modified dGVp was evaporated radioactivities were measured by scintillation counting. The relative adduci to dryness and incubated at 37Â°Cfor l h with alkaline phosphatase (3 units) in levels were calculated according to Levay et al. (41), using dpm instead of 100 fil of 100 mM Tris-HCl buffer (pH 7.0). The reaction mixture was cpm: (total dpm in adducts)/l.21 X 10' ' dpm, assuming that 3 fig of DNA was extracted twice with 100 /Â¿Iofbutanol. The butanol extract was evaporated at 9.09 X 10' pmol of dGVP and the specific activity of the [y-12P]ATP was reduced pressure and subjected to a reverse-phase column, fiBondapak C,8 1.33 X IO7 dpm/pmol. The specific activity of the [y-12P]ATP was corrected (0.39 X 30 cm: Waters), eluted over 30 min at a flow rate of 1.0 ml/min with by calculating the extent of decay. a linear gradient of 0.05 M triethylammonium acetate (pH 7.0) containing Determination of "P-labeled DNA Adducts by HPLC. The standard, 10â€”Â»50%acetonitrile (39). The retention time and UV spectra of the products dG3.p-N2-TAM, was labeled with 32P (42). A part of the 52P-labeled sample or were compared with that of diastereoisomers of dG-N2-TAM prepared previ dG, p-N2-TAM was developed for 16 h on a PEI-cellulose plate using 1.7 M ously by reacting dG with a-acetoxytamoxifen (26). sodium phosphate buffer (pH 6.0) with a paper wick. 32P-labeled products that Reaction of a-OHTAM with DNA in the Presence of STa and PAPS. remained around the original spot were recovered using 4 M pyrimidinium Calf thymus DNA (200 fig) was purified by an established protocol (40). formate (pH 4.3), evaporated to dryness, applied to a Supelcosil LC-18S Aliquots of DNA (6 /ig) were incubated at 37Â°Cfor I h with 200 fiM PAPS column (0.46 X 25 cm; Supelco), and eluted over 30 min at a flow rate of 1.0 and 100 /ÃŒMofeither (Â£")-or (Z)-a-OHTAM in 50 /il of 0.25 M potassium ml/min with a linear gradient of 0.2 M ammonium formate (pH 4.2) containing phosphate (pH 7.0) containing 8.3 mM mercaptoethanol and with or without 2.4 fig of STa. Eight hundred fil of absolute ethanol were added to the reaction samples, which were kept at â€”70Â°Cfor 30 min, and then centrifuged to 0.4 i- recover the DNA. The DNA was washed twice by 500 /il of ethanol and used for analysis of DNA adducts. The supernatant was evaporated to dryness and subjected to HPLC for determination of (Â£>or (Z)-a-OHTAM. These isomers can be separated using a fiBondapak C18 (0.39 X 30 cm; Waters), eluted over 30 min at a flow rate of 1.0 ml/min with a linear gradient of 0.05 M triethylammonium acetate (pH 7.0) containing 10â€”Â»70%acetonitrile.Six fig of e a the DNA were reacted at 37Â°Cfor l h with 50 fig of a-acetoxytamoxifen as O) described previously (26). 0.2 - Digestion of DNA Samples. Three fig of the DNA samples or 0.1 fig of a-acetoxytamoxifen-treated DNA was enzymatically digested at 37Â°Cfor 2 h in 10 fil of 17 mM sodium succinate buffer (pH 6.0) containing 8 mM CaCl,. O using 2.0 units of micrococcal nuclease and 2.0 X 10~3 units of spleen E phosphodiesterase. The samples were dissolved in 100 fil of distilled water and 0.1 extracted twice with 100 fil of butanol. The butanol fractions were dried and used for analysis of TAM-DNA adducts. Detection of DNA Adducts by "P-Postlabeling Method. The digests of pooled extracts were incubated at 37Â°Cfor 40 min with 40 fiCi of |y-'2P]ATP 0.0 10 20 30 and 3 fil of T4 polynucleotide kinase (10 unils/fil: Ref. 34) and further incubated for l h with potato apyrase (8 x \0~- units). A part of the 1/[(E)-a-OHTAM] (mM)'1 '2P-labeled sample was spotted on a 10 X 10-cm PEI-cellulose thin-layer plate Fig. 2. Kinetic study for STa-catalyzed sulfonation of (Â£>a-OHTAM. Determination 2 cm from the bottom and 2 cm from the left edge of the plate and developed of kinetic constants for (E)-a-OHTAM was carried out as described in "Materials and using three different solvents. Paper wicks were attached to the top of PEI- Methods." 648

Fig. 3. HPLC and gel separation of dG3.P-N2-TAM formed by (Z)-a-sulfate TAM. /, dG3.P (1.0 mg) was reacted at 37Â°Cfor 2 h with (Z)-a-sulfate TAM (2.0 mg) in 100 mM Tris-HC] buffer (pH 8.0) and then extracted with butanol, as described in "Materials and Methods." One-tenth of the butanol fraction was dried and subjected to HPLC for the isolation of dGVP-TAMs. II, a trans form (A) or cis form (B) of dG3.p-N2-TAM was labeled with HP and developed on a PEI-cellulose TLC plate as described in "Materials and Methods." 10 20 30 TIME (min) II. A B

10â€”Â»70%methanol.One-minute fractions were collected, and the radioactivity nmol/min/mg protein. Thus, (Â£>a-OHTAM is a much better substrate was measured by scintillation counting. for STa than (Z)-a-OHTAM. Preparation of dG3.P-N2-TAM. To prepare the dG3.P-N2-TAM that was used as a standard for 32P-postlabeling analysis, dG3.P was RESULTS reacted with (Z)-a-sulfate TAM. As shown in Fig. 31, three major Sulfation of a-OHTAM Catalyzed by STa. (E)-a-OHTAM or fractions (#1, rR = 30.5 min; #2, tR = 32.0; and #3, iR = 37.0, where (Z)-a-OHTAM was incubated with STa in the presence of PAPS. iR is the retention time) were separated on HPLC. After repurification (Â£>a-OHTAM was a substrate for STa with an apparent Km of of these fractions using HPLC, a part of each fractions was evaporated 72 Â±29 /XMand a maximal velocity of 12.2 Â±1.8 nmol of product to dryness, incubated with alkaline phosphatase to produce the de- formed/min/mg protein (Fig. 2). In contrast, (Z)-a-OHTAM was a oxynucleoside, and subjected to HPLC. The retention times of prod poor substrate for STa. Significant values of Km and Vmaxfor this ucts obtained from fraction #1 (iR = 37.4 in Fig. 45) and #2 isomer were not obtained due to the low and variable reaction veloc (iR = 38.0 in Fig. 4C) were consistent with that of fr-1 and fr-2, an ities. A comparison of reaction rates at a fixed substrate concentration epimer of trans form of dG-N2-TAM (Fig. 4A), respectively. Fraction of 80 /AMshowed that the rate for sulfation of the cis isomer was 1.8 #3 contained two products: the retention times (iR = 40.4 and 41.1 in nmol/min/mg protein, whereas that of the trans isomer was 6.8 Fig. 4D) were consistent with those of the epimers of the cis form of 649

A-, ? than that observed with (Â£)-a-OHTAM alone (Fig. 5D). When DNA I Â° was treated with a-acetoxytamoxifen (Fig. 5ÃŸ),several TAM adducls

Q dG CO were delected. The pattern of the formation of DNA adducts on the CM ,_ m o TLC plale was similar lo lhal observed with (Â£)-a-OHTAM and STa o fr-4 z (Fig. 5C). The major and minor adducts formed by a-acetoxytamox O) in ifen migrated similarly lo ihe spois a and b, respectively, formed by OCE Â°9 (E)-a-OHTAM. â€¢ B Identification of Tamoxifen-DNA Adducts. Standards of 32P- labeled dG3.P-N2-TAM were subjected lo HPLC. The relenlion limes of #1 (data noi shown) and #2 (Fig. 6ÃŸ),trans forms of dG-N2-TAM, were 34 and 35 min. respectively, whereas the retenlion lime of #3, ds forms of dG-N2-TAM, was 39 min (Fig. 6Q. The 32P-labeled samples obiained from DNA incubated with (Â£>a-OHTAM and with or withoul STa in ihe presence of PAPS were developed on a PEI- cellulose TLC plate using Dl buffer. The 32P-labeled products lhal remained around ihe spot origin were recovered from ihe TLC plale and subjected to HPLC. With (Â£)-a-OHTAM alone, only a small peak representing a trans form of dG-N2-TAM was detected al 35 min (Fig. c.

D. LJÃœ â€”

10 20 30 40 TIME (min) Fig. 4. Dephosphorylation of dG, ,,-N:-TAM. A part of the dG, P-TAM fraction (#l, #2. or #3) shown in Fig. 3/ was evaporated to dryness. incubated with alkaline phospha- lase in IOOmM Tris-HCl buffer (pH 7.0). and extracted with butanol. The butano! extract obtained from #l (B), #2 (Ci. or #3 (D) was subjected to HPLC as described in "Materials and Methods" and compared with standard dG-N3-TAMs (A). ciG-N2-TAM (fr-3 and fr-4 in Fig. 44). The ratio of fr-3 to fr-4 was 45:55. UV spectra of the products were identical to those of standard trans and cis forms of dG-N2-TAM, as described in our previous study (26). Thus, fractions #l and #2 were the trans forms of dGVP- N2-TAM and fraction #3 was a mixture of the cis forms of dGVP- N2-TAM. The 12P-labeled dGVP-N2-TAM was developed on PEI-cellulose TLC plates using three different buffer solutions. Both fractions #l and #2 (Fig. 3//, A) migrated similarly but very differently from fraction #3 (Fig. 3/7. B). Thus, the trans forms of dG3.P-N2-TAM can be resolved from the cis forms by TLC. . Detection of TAM-DNA Adducts. To investigate the reactivities of (Â£)-a-OHTAM and (Z)-a-OHTAM to DNA through O-sulfonation of the a-hydroxyl group of TAM, these compounds were incubated with purified DNA and PAPS, with or without STa. The formation of TAM-DNA adducts were determined using a 32P-posllabeling method combined with butanol extraction. When the DNA was incubated without (E)-a-OHTAM and STa, no DNA adducts were detected (Fig. 5/1). With (Â£)-a-OHTAM alone, a small amount (1.3 adducts/106 normal nucleotides) of a TAM adduci were detected as indicated by an arrow (a; Fig. 5Â£>).When STa and PAPS were incubated with (E)-a-OHTAM, several TAM-DNA adducts were detected (Fig. 5Q. The formation of the major TAM adducts (a; 36.4 adducts/106) was dramatically increased 29 times over that observed with (E)-a-OH- Fig. 5. "P-Postlabeling analysis of DNA incubated with a-OHTAM and STa in the TAM alone. The amount of a minor adduci (b) was 1.1 adducts/106. presence of PAPS. Six fig of calf thymus DNA were incubated at 37Â°Cfor l h with 100 UM(F)-a-OHTAM (C and E) or (Z)-a-OHTAM (D and F) and with (C and D| or without In contrast, when (Z)-a-OHTAM was incubated with STa and PAPS, (D and F| 2.4 /ig of STa in the presence of PAPS, as described in "Materials and small amounts (5.6 adducts/106) of adduci (a) were also detected (Fig. Methods." The DNA was also reacted without tamoxifen derivatives (Ai or with 50 Â¿igof 5E): however, the amount was 6.5 times less than that of (E)-a- a-acetoxytamoxifen (ÃŸ).a. major adduct: b. minor adduci. Three fxg of the recovered OHTAM (Fig. 5C). With (Z)-a-OHTAM alone, only 0.8 adducls/10" DNA was digested enzymatically. extracted with butano!, and labeled with 12P. One-third of the sample was applied to PEI-cellulose TLC chromatography and developed using was delecled (Fig. 5F): the level of this adduci (a) was 1.5 limes less three different buffer solutions, as described in "Materials and Methods." 650

125000 -i 80000 aloneA(E)-ct-OHTAM B. trans-form of dG-N2-TAM C. cfe-form of dG-N2-TAM 100000- 60000-'

75000- 40000 1VE125000-,20000-15000-10000-5000-A.50000-

20000- 25000-

A0 0 Â«*â€¢1 â€¢â€”iâ€¢r"-â€”pâ€”â€¢ 1 1050Retention 20 30 40 10 20 X 40 50 10 20 30 40 50 (min)25000-120000-15000-10000-5000fi.D.time Retention time (min) Retention time (min) 80000- STaII(E)-a-OHTAM + E. ct-acetoxytamoxifen

60000

40000- i

20000- Jl 10 20 30 40 50 10 20 30 40 SO Retention time (min) Retention time (min) Fig. 6. Determination of 12P-labeled DNA adducts by HPLC. A pan of the '2P-labeled sample (A. (E)-a-OHTAM alone; D. (E)-a-OHTAM + STa: Â£.a-aceloxytamoxifen) or a Irans form (5) and r/.v form (C) of dGVp-N2-TAM was developed on a PEI-cellulose TLC plate using 1.7 M sodium phosphate buffer (pH 6.0) with a paper wick. The t2P-labeled products that remained around the original spot were recovered using 4 M pyrimidinium formate (pH 4.31 and subjected to a Supelcosil LC-18S column (0.46 X 25 cm: Supelco). A linear gradient of 0.2 M ammonium formate (pH 4.2) containing 10â€”*10c/cmethanol was eluted over 30 min at a flow rate of 1.0 ml/min. The eluate was collected every minute, and the radioactivity was measured by a scintillation counter.

6A). In contrast, with STa and (E)-a-OHTAM (Fig. 6D), a major peak representing a trans form of dG-N2-TAM was detected at 35 min, and a small peak representing a r/'.s-form of dG-N2-TAM was also ob eCS-St served at 39 min. The HPLC pattern was similar to that observed for (Â£)-a-acetoxytamoxifen (Fig. 6F). Conversion between the trans and cis Form of a-OHTAM. To o"UJUABSORBAr determine whether there is conversion between the trans and cis forms of a-OHTAM when (Â£)-or (Z)-a-OHTAM was incubated with DNA, STa, and PAPS (Figs. 5C and 6D), the amounts of (Â£")-and (Z)-a- OHTAM in the reaction mixture were analyzed. (E)-a-OHTAM (fR = 34.8 min) can be separated from (Z)-a-OHTAM (rR = 36.3) 00.0500.0500.05\_A(E)-a-OHTAM( using HPLC, as shown in Fig. 1A. When purified (E)-a-OHTAM (Fig. IB) was incubated with STa and DNA in the presence of PAPS, 5.8% of (E)-a-OHTAM was converted to (Z)-a-OHTAM (Fig. 1C). In contrast, when (Z)-a-OHTAM was used, no significant conversion from (Z)-a-OHTAM to (E)-a-OHTAM was detected (data not shown).

DISCUSSION .t- B- a-OHTAM has been detected as a metabolite of TAM in the culture media from rat hepatocytes (21 ) and human Hep G2 cells treated with (E)-a-OHTAMt TAM (43) as well as in the extract of human liver homogenates Â°-w incubated with TAM (43). Our experiments showed that the sulfation â€”(Z)-a-OHTiKM(Z)-a-OHTAMM of (E)-a-OHTAM can be catalyzed by rat liver STa in the presence of PAPS, forming the a-sulfuric acid ester of TAM (Fig. 2). Although 10 20 30 40 the Km of (Â£>o-OHTAM was higher and the Vmaxwas lower than that TIME (min) reported for some benzylic alcohols (31), apparently the hydroxyl Fig. 7. Analysis of a-OHTAM in the STa-catalyzed reaction mixture. A mixture of (Â£")- moiety at the allylic position of TAM can be O-sulfonated in a and (Z)-a-OHTAM (A). (E)-a-OHTAM alone (fl). or a supernatant obtained from the reaction catalyzed by STa. Based on the kinetic data. (Â£)-a-OHTAM reaction sample with (Â£)-a-OHTAM and STa in the presence of PAPS (Cl was applied to a Â¿Â¿BondapakC,n(0.39 X 30 cm; Waters), eluted over 30 min at a How rate of 1.0 ml/min was a much better substrate for STa than the Z form. Thus, STa could with a linear gradient of 0,05 M triethylammonium acetate (pH 7.0) containing 10â€”Â»70% distinguish between the stereoisomeric forms of this compound. acetonitrile. 651

(H>C)iN OH OH

(E)-a-OHTAM (Z)-a-OHTAM

(H,C),N O-sulfonation O-sulfonation

(HiC)iN

(H,C),N

|H,C),N (H,C),N- frans-form

DNA/

DNA

|H,C),N

Fig. 8. A proposed mechanism for the formation of TAM-DNA adducis.

When DNA was incubated with STa and a-OHTAM in the pres between the (Â£)and (Z) form of a-OHTAM most likely occurs ence of PAPS, dG-N2-TAM adducts were formed in the DNA (Figs. through a carbocation intermediate (Fig. 8), as proposed in our pre 5C and 6D). As reported previously, a-sulfate TAM is highly reactive vious report (26), the cis form of dG-N2-TAM and (Z)-a-OHTAM can with DNA, forming dG-N2-TAM adducts (26). Thus, the sulfuric acid be formed through this conversion. Similar results were observed ester of a-OHTAM, produced with STa and PAPS, most likely reacts when DNA was reacted with (Â£)-a-sulfate TAM (26) or (E)-a- with DNA through carbocation intermediate to form TAM adducts acetoxytamoxifen (Refs. 25 and 26; Fig. 5ÃŸ)andwhen rat hepatocytes (Fig. 8). The amount of dG-N2-TAM adducts formed via sulfation of were treated with (Â£>a-OHTAM or (Â£>TAM (25). When DNA was (Â£)-a-OHTAM was much higher than that formed through sulfation incubated with (Z)-a-OHTAM and STa, only the trans forms of of (Z)-a-OHTAM (Fig. 5, C and E). This result may be based on the dG-N2-TAM were detected. Generation of the trans form appears to fact that (E)-a-OHTAM, a better substrate for STa, is more readily be preferred over conversion to the CMform (26). Alternatively, the converted to the sulfuric acid ester than is (Z)-a-OHTAM. carbocation of the cis form may be less reactive to DNA than that of When (E)-a-OHTAM is incubated with STa and DNA, the trans the trans form. form of dG-N2-TAM was formed as the major adduci, and the cis Randerath et al. (24) has demonstrated a significant inhibition of form of dG-N2-TAM was formed as the minor adduci (Fig. 5). In mouse hepatic DNA adduci formation by TAM with the coadminis- addition, a small amount of (Z)-a-OHTAM was produced from (Â£> tration of pentachlorophenol, a potent inhibitor of phenol sulfotrans- a-OHTAM in the presence of PAPS (Fig. 7). Because a conversion ferase. In fact, both (Â£)-and (Z)-a-OHTAM are weak substrates for 652

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1998 American Association for Cancer Research. a-HYDROXYTAMOXIFEN-DERIVED DNA ADDUCT FORMED BY STa recombinant rat hepatic aryl sulfotransferase IV.4 In addition to STa, S., Hewer, A., and Phillips, D. H. Genotoxic potential of tamoxifen and analogues in other isoforms of sulfotransferase may also participate in sulfation of female Fischer F344/n rats, DBA/2 and C57B1/6 mice and in human MCL-5 cells. Carcinogenesis (Lond.). 13: 2197-2203. 1992. a-OHTAM, forming DNA adducts in vivo. 18. Hemminki, K., Rajaniemi. H.. Lindahl, B., and Moberger, B. Tamoxifen-induced Four distinct cytosolic sulfotransferases have been identified and DNA adducts in endometrial samples from breast cancer patients. Cancer Res., 56: 4374-4377, 1996. characterized in human tissues (44). Based on the levels of their 19. Moorthy. B., Sriram, P., Pathak, D. N., Bodell, W. J.. and Randerath, K. Tamoxifen expression and their relationship to the rat enzymes that activate metabolic activation: comparison of DNA adducts formed by microsomal and chem benzylic alcohols, hHST may play a significant role in the activation ical activation of tamoxifen and 4-hydroxytamoxifen with DNA adducts formed in vitro. Cancer Res., 56: 53-57, 1996. of benzylic alcohols in human livers (45). Moreover, similarities in 20. Poon, G. K., Walter, B., Lining, P. E., Horton, M. N., and McCague, R. Identifi the substrate specificities of STa and hHST are consistent with the cation of tamoxifen metabolites in human Hep G2 cell line, human liver homogenate, possibility that hHST may also catalyze the sulfation of a-OHTAM. and in patients on long-term therapy for breast cancer. Drug Metab. Dispos., 23: 377-382, 1995. Sulfation of a-OHTAM in human tissues might then result in forma 21. Phillips, D. H.. Carmichael. P. L.. Hewer, A., Cole, K. J., and Poon, G. K. a-Hy- tion of dG-N2-TAM-DNA adducts. Because dG-N2-TAM has a strong droxytamoxifen, a metabolite of tamoxifen with exceptionally high DNA-binding activity in rat hepatocytes. Cancer Res., 54: 5518-5522, 1994. miscoding potential (27), this adduci may be involved in the devel 22. Potter, G. A., McCague, R., and Jarman. M. A mechanism hypothesis for DNA opment of endometrial cancer. More extensive studies on the role of adduct formation by tamoxifen hepatic oxidative metabolism. Carcinogenesis sulfation in TAM adduci formation in human lissues will be necessary (Lond.), /5: 439-442, 1994. 23. Phillips. D. H., Potter. G. A., Horton. M. N.. Hewer. A., Crofton-Sleigh, C., Jarman. lo fully address ihese possibilities. M., and Venitt, S. Reduced genotoxicity of [D5-ethyl|-tamoxifen implicates a-hydroxylation of the ethyl group as a major pathway of tamoxifen activation to a liver carcinogen. Carcinogenesis (Lond.), /5: 1487-1492, 1994. 24. Randerath, K., Moorthy, B., Mabon. N.. and Sriram. P. Tamoxifen: evidence by ACKNOWLEDGMENTS 32P-postlabeling and use of metabolic inhibitors for two distinct pathways leading to mouse hepatic DNA adduct formation and identification of 4-hydroxytamoxifen as a We thank Dr. A. Umemoto for technical advice and Dr. A. P. Grollman for proximate metabolite. Carcinogenesis (Lond.), 15: 2087-2094, 1994. encouragement. 25. Osbome, M. R., Hewer. A.. Hardcastle, I. R., Carmichael, P. L., and Phillips, D. H. Identification of the major tamoxifen-deoxyguanosine adduct formed in the liver DNA of rats treated with tamoxifen. Cancer Res., 56: 66-71, 1996. REFERENCES 26. Dasaradhi, L.. and Shibutani, S. Identification of lamoxifen-DNA adducts formed by a-sulfate tamoxifen and o-acetoxytamoxifen. Chem. Res. Toxicol.. 10: 189-196, 1997. 1. Jordan, V. C. A current view of tamoxifen for the treatment and prevention of breast 27. Shibutani. S., and Dasaradhi. L. Miscoding potential of tamoxifen-derived DNA cancer. Br. J. Pharmacol., 110: 507-517, 1993. adducts: o>(/V2-deoxyguanosinyl)tamoxifen. Biochemistry, 36: 13010-13017, 1997. 2. Powles, T. J., Hardy. J. R., Ashley, S. E., Harrington, G. H., Cosgrove, D., Davey, 28. Lyon, E. S.. Marcus, C. J., Wang, J-L.. and Jakoby. W. B. Hydroxysteroid sulfo J. R., Dowsett, M., McKinna, J. A., Wash, A. G., Sennet, H. D., Tillyer, C. R., and transferase. Methods Enzymol., 77: 206-213, 1981. Treleaven, J. G. A pilot trial to evaluate the acute toxicology and feasibility of 29. Watabe. T., Ishizuka, T., Isobe, M., and Ozawa, N. A 7-hydroxy-methyl sulfate ester tamoxifen for the prevention of breast cancer. Br. J. Cancer, 60: 126-131. 1993. as an active metabolite of 7,12-dimethyl-benz[a]an!hracene. Science (Washington 3. Nayfield, S. G., Karp, J. E., Ford, L. G., Dorr, F. A., and Kramer. B. S. Potential role DC). 2/5: 403-405. 1982. of tamoxifen in prevention of breast cancer. J. Nati. Cancer Inst., 83: 1450-1459. 30. Surh, Y-J.. Liem, A., Miller, E. C., and Miller, J. A. Metabolic activation of the 1991. carcinogen 6-hydroxymethylbenzo|u]pyrene: formation of an electrophilic sulfuric 4. Williams, G. M., latropoulos, M. J., Djordjevic, M. V., and Kaltenberg, O. P. The acid ester and benzylic DNA adducts in rat liver in vivo and in reactions in vitro. triphenylethylene drug tamoxifen is a strong liver carcinogen in the rat. Carcinogen- Carcinogenesis (Lond.), 10: 1519-1528. 1989. esis (Lond.), 14: 315-317. 1993. 31. Chen, G., Banoglu, E., and Duffel, M. W. Influence of substrate structure on the 5. Greaves, P., Goonetilleke, R., Nunn, G.. Topham, J.. and Orlon, T. Two-year catalytic efficiency of hydroxysteroid sulfotransferase STa in the sulfation of alco carcinogenicity study of tamoxifen in Alderley Park Wister-derived rats. Cancer Res.. hols. Chem. Res. Toxicol., 9: 67-74, 1996.- 53: 3919-3924, 1993. 32. Ogura. K., Sohtome, T.. Sugiyama, A.. Okuda, H., Hiratsuka, A., and Watabe, T. Rat 6. Hard, G. C., latropoulos, M. J., Jordan, K., Radi. L., Kaltenberg, O. P., Imondi, A. R., liver cytosolic hydroxysteroid sulfotransferase (sulfotransferase a) catalyzing the and Williams, G. M. Major difference in the hepatocarcinogenicity and DNA adduci formation of reactive sulfate esters from carcinogenic polycyclic hydroxymethyl- forming ability between toremifene and tamoxifen in female Crl:CD(BR) rats. Cancer arenes. Mol. Pharmacol., 37: 848-854, 1990. Res., 53: 4534-4541, 1993. 33. Watabe. T. Ogura, K.. Satsukawa, H., Okuda. H., and Hiratsuka. A. Molecular cloning 7. Davies, R., Oreffo, V. I. C., Martin. E. A., Festing, M. F. W., White, I. N. H., Smith, and functions of rat liver hydroxysteroid sulfotransferases catalysing covalent binding of L. L., and Styles, J. A. Tamoxifen causes gene mutations in the livers of lambda/Latl carcinogenic polycyclic arylmethanols to DNA. Chem. Biol. Interact., 92: 87-105. 1994. transgenic rats. Cancer Res., 57: 1288-1293, 1997. 34. Bensadoum. A-, and Weinstein. D. Assay of proteins in the presence of interfering 8. Vancutsem, P. M., Lazarus, P., and Williams, G. M. Frequent and specific mutations materials. Anal. Biochem.. 70: 241-250, 1976. of the rat p53 gene in hepatocarcinomas induced by tamoxifen. Cancer Res., 54: 35. Foster. A. B.. Jarman, M.. Leung. O-T.. McCague, R.. Leclercq. G.. and Devleeschouwer. 3864-3867, 1994. N. Hydroxy derivatives of tamoxifen. J. Med. Chem., 2Â«:1491-1497. 1985. 9. Seoud, M. A-F., Johnson, J., and Weed, J. C. Gynecologic tumors in tamoxifen- 36. Sekura, R. D. Adenosine 3'-phosphate 5'-phosphosulfate. Methods Enzymol., 77: treated women with breast cancer. Obstet. Gynecol.. 82: 165-169, 1993. 413-415. 1981. 10. Fischer, B.. Costammo, J. P.. Redmond. C. K., Fisher, E. R., Wickerham. D. L., and 37. Duffel, M. W.. Binder, T. P., and Rao. S. I. Assay of purified aryl sulfotransferase Cronin, W. M. Endometrial cancer in tamoxifen-lreated breast cancer patients: suitable for reactions yielding unstable sulfuric acid esters. Anal. Biochem.. 183: findings from the National Surgical Adjuvant Breast and Bowel Project (NSABP) 320-324, 1989. B-14. J. Nati. Cancer Inst., 86: 527-537. 1994. 38. Perella, F. W. EZ-FIT: a practical curve-fitting microcomputer program for the 11. van Leeuwen, F. E., Benraadt, J., Coebergh, J. W. W., Kiemeney, L. A. L. M., analysis of enzyme kinetic data on IBM-PC compatible computers. Anal. Biochem., Diepenhorst, F. W.. van den Belt-Dusebout, A. W.. and van Tinteren. H. Risk of 174: 437-447, 1988. endometrial cancer after tamoxifen treatment of breast cancer. Lancet, 343: 448-452. 39. Shibutani, S., Gentles, R., Johnson. F.. and Grollman, A. P. Isolation and character 1994. ization of oligodeoxynucleotides containing dG-N2-AAF and oxidation products of 12. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Vol. 66, dG-C8-AF. Carcinogenesis (Lond.), 12: 813-818, 1991. Some Pharmaceutical Drugs. Lyon: IARC. 1996. 40. Umemoto. A.. Kajikawa. A., Tanaka. M., Hamada, K., Seraj. M. J.. Kubota, A., 13. Pathak. D. N., and Bodell, W. J. DNA adduci formation by tamoxifen with rat and Nakayama. M.. Kinouchi. T., Ohnishi, Y.. Yamashita. K., and Monden, Y. Presence human liver microsomal activation systems. Carcinogenesis (Lond.), IS: 529-532. of mucosa-specific DNA adduct in human colon: possible implication for colorectal 1994. cancer. Carcinogenesis (Lond.), /5: 901-905. 1994. 14. Pathak, D. N., Pongracz, K., and Bodell, W. J. Microsomal and peroxidase activation 41. Levay, G., Pongracz, K.. and Bodell, W. J. Detection of DNA adducts in HL-60 cells of 4-hydroxy-tamoxifen to form DNA adducts: comparison wilh DNA adducts treated with hydroquinone and p-benzoquinone by '2P-postlabeling. Carcinogenesis formed in Sprague-Dawley rats treated with tamoxifen. Carcinogenesis (Lond.), 16: (Lond.), 12: 1181-1186, 1991. 11-15,1995. 42. Maniatis, T., Fritsch, E. F., and Sambrook, J. Molecular Cloning: A Laboratory 15. Hemminki, K., Widlak, P.. and Hou, S. M. DNA adducts caused by tamoxifen and Manual. Cold Spring Harbor. NY: Cold Spring Harbor Laboratory. 1982. toremifene in human microsomal system and lymphocyte in vitro. Carcinogenesis 43. Poon, G. K., Walter, B., Lining, P. E., Horton, M. N., and McCague. R. Identification (Lond.), 16: 1661-1664, 1995. of tamoxifen metabolites in human Hep G2 cell line, human liver homogenate, and 16. Han, X., and Liehr, J. G. Induction of covalent DNA adducts in rodents by tamoxifen. patients on long-term therapy for breast cancer. Drug. Metab. Dispos., 23: 377-382, 1995. Cancer Res., 52: 1360-1363, 1992. 44. Falany. C. N. Enzymology of human cytosolic sulfotransferases. FASEB J., 11: 17. White, I. N. H., de Matteis, F., Davies, A.. Smith, L. L., Crofton-Sleigh, C., Venitt, 206-216, 1997. 45. Glatt. H., Pauly. K.. Czich. A., Falany, J. L.. and Falany, C. N. Activation of benzylic alcohols to mutagens by rat and human sulfotransferases expressed in Escherichia 4 M. W. Duffel and S. Shibutani, unpublished results. coli. Eur. J. Pharmacol., 293: 173-181. 1995.

653

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1998 American Association for Cancer Research. α-Hydroxytamoxifen Is a Substrate of Hydroxysteroid (Alcohol) Sulfotransferase, Resulting in Tamoxifen DNA Adducts

Shinya Shibutani, Lakkaraju Dasaradhi, Isamu Terashima, et al.

Cancer Res 1998;58:647-653.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/58/4/647

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/58/4/647. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.