Synthesis and Reactivity of Potential Toxic Metabolites of Tamoxifen

Chem. Res. Toxicol. 2001, 14, 1643-1653

1643

Synthesis and Reactivity of Potential Toxic Metabolites of Tamoxifen Analogues: Droloxifene and Toremifene o-Quinones Dan Yao, Fagen Zhang, Linning Yu, Yanan Yang, Richard B. van Breemen, and Judy L. Bolton* Department of Medicinal Chemistry and Pharmacognosy (M/C 781), College of Pharmacy, University of Illinois at Chicago, 833 South Wood Street, Chicago, Illinois 60612-7231 Received August 15, 2001

Tamoxifen remains the endocrine therapy of choice in the treatment of all stages of hormonedependent breast cancer. However, tamoxifen has been shown to increase the risk of endometrial cancer which has stimulated research for new effective antiestrogens, such as droloxifene and toremifene. In this study, the potential for these compounds to cause cytotoxic effects was investigated. One potential cytotoxic mechanism could involve metabolism of droloxifene and toremifene to catechols, followed by oxidation to reactive o-quinones. Another cytotoxic pathway could involve the oxidation of 4-hydroxytoremifene to an electrophilic quinone methide. Comparison of the amounts of GSH conjugates formed from 4-hydroxytamoxifen, droloxifene, and 4-hydroxytoremifene suggested that 4-hydroxytoremifene is more effective at formation of a quinone methide. However, all three substrates formed similar amounts of o-quinones. Both the tamoxifen-o-quinone and toremifene-o-quinone reacted with deoxynucleosides to give corresponding adducts. However, the toremifene-o-quinone was shown to be considerably more reactive than the tamoxifen-o-quinone in terms of both kinetic data as well as the yield and type of deoxynucleoside adducts formed. Since thymidine formed the most abundant adducts with the toremifene-o-quinone, sufficient material was obtained for characterization by 1H NMR, COSY-NMR, DEPT-NMR, and tandem mass spectrometry. Cytotoxicity studies with tamoxifen, droloxifene, 4-hydroxytamoxifen, 4-hydroxytoremifene, and their catechol metabolites were carried out in the human breast cancer cell lines S30 and MDA-MB-231. All of the metabolites tested showed cytotoxic effects that were similar to the parent antiestrogens which suggests that o-quinone formation from tamoxifen, droloxifene, and 4-hydroxytoremifene is unlikely to contribute to their cytotoxicity. However, the fact that the o-quinones formed adducts with deoxynucleosides in vitro implies that the o-quinone pathway might contribute to the genotoxicity of the antiestrogens in vivo.

Introduction Tamoxifen1 is currently the standard treatment for hormone dependent breast cancer in postmenopausal women (1-4). However, tamoxifen has been shown to cause an increased incidence of hepatocellular tumors in rats (5, 6) and an increase in the risk of endometrial cancer in women (7). A potential genotoxic mechanism might involve metabolic activation by oxidative enzymes in vivo to an electrophile(s) that binds irreversibly to * To whom correspondence should be addressed. Phone: (312) 9965280. Fax: (312) 996-7107. E-mail: [email protected]. 1 Abbreviations: tamoxifen, Z-2-[4-(1,2-diphenyl-1-butenyl)-phenoxy]N,N-dimethylethanamine; toremifene, Z-2-[(4-chloro-1,2-diphenyl-1butenyl)-phenoxy]-N,N-dimethylethanamine; 4-OHTAM, 4-hydroxytamoxifen; 3,4-di-OHTAM, 3,4-dihydroxytamoxifen; 4-OHTAM-SG, glutathione conjugates of 4-hydroxytamoxifen; 3,4-di-OHTAM-diSG, di-glutathione conjugates of 3,4-dihydroxytamoxifen; 4-OHTOR, 4-hydroxytoremifene; 3,4-di-OHTOR, 3,4-dihydroxytoremifene; 4-OHTORdiSG, di-glutathione conjugates of 4-hydroxytoremifene; 3,4-di-OHTORdiSG, di-glutathione conjugates of 3,4-dihydroxytoremifene; 3,4-diOHTOR-triSG, tri-glutathione conjugates of 3,4-dihydroxytoremifene; 3,4-di-OHTOR-T, thymidine adducts of 3,4-di-OHTOR; 3,4-di-OHTORdG, deoxyguanosine adducts of 3,4-di-OHTOR; 3,4-di-OHTOR-dA, deoxyadenosine adducts of 3,4-di-OHTOR; 3,4-di-OHTOR-dC, deoxycytosine adducts of 3,4-di-OHTOR; P450, cytochrome P450; GSH, glutathione; electrospray-MS, electrospray mass spectrometry; MSMS, tandem mass spectrometry; ER, estrogen receptor.

DNA (8). Concerns about the side effects of tamoxifen have stimulated a search for new effective agents that do not form such genotoxic species. Toremifene has been recently approved by the United States Food and Drug Administration for the treatment of advanced breast cancer in postmenopausal women (9). There have been several recent reports comparing the efficacy and metabolism of tamoxifen and toremifene (10-19). Although toremifene is structurally similar to tamoxifen, differing only by a single chlorine atom in the ethyl side chain, toremifene does not cause hepatocarcinogenesis in rats in vivo and it produces considerably fewer DNA adducts (20). Droloxifene (3-hydroxytamoxifen) is another antiestrogen, which has shown promise in the treatment of breast cancer (21-23). There is no evidence that droloxifene produces DNA adducts or hepatocelluar carcinoma in rats (7, 24). There are several metabolic activation pathways of tamoxifen and toremifene leading to DNA adducts such as formation of carbocations (20, 25) and electrophilic quinone methides (26-28). In addition, an o-quinone has been shown to be a potential reactive intermediate involved in the metabolism of both droloxifene and toremifene (29, 30). In the present study, we compared

10.1021/tx010137i CCC: $20.00 © 2001 American Chemical Society Published on Web 11/27/2001

1644

Chem. Res. Toxicol., Vol. 14, No. 12, 2001

the potential toxicity of quinoid intermediates formed from tamoxifen analogues, including those of 4-hydroxytamoxifen, droloxifene, and 4-hydroxytoremifene. Two human breast cancer cell lines, S30 (ER-positive) and MDA-MB-231 (ER-negative), were used to study the relative cytotoxicity of 4-hydroxytamoxifen, droloxifene, 3,4-dihydroxytamoxifen, 4-hydroxytoremifene, and 3,4dihydroxytoremifene. The data suggest that o-quinone formation might not contribute to the cytotoxic mechanism of the antiestrogens. However, the formation of DNA adducts implies that the o-quinone pathway might contribute to the genotoxic effects of these antiestrogens in vivo.

Materials and Methods Caution: Toremifene-o-quinone was handled in accordance with the NIH Guidelines for the Laboratory Use of Chemical Carcinogens (31). All solvents and chemicals were purchased from either Aldrich Chemical Co. (Milwaukee, WI) or Fisher Scientific (Itasca, IL) unless stated otherwise. [3H]GSH (glycine2-3H) was obtained from New Life Science Products Inc. (Boston, MA) and diluted to a specific activity of 40 nCi/nmol. 4-Hydroxytamoxifen and 4-hydroxytoremifene were synthesized as described previously (32). Droloxifene was synthesized according to the literature procedure (33). Instrumentation. 1H NMR and 13C NMR spectra were obtained with a Bruker (Billerica, MA) Avance DPX300 spectrometer at 300 MHz. UV spectra were measured on a HewlettPackard (Palo Alto, CA) 8452A photodiode array UV-vis spectrophotometer. Exact mass measurements were obtained using EI or CI on a Fininigan (Bremen, Germany) MAT90 magnetic sector mass spectrometer at a resolving power of 10 000. LC-MS was carried out with a Hewlett-Packard 5989B mass spectrometer equipped with an Analytica (Branford, CT) electrospray ion source and ion guide. LC-MS-MS spectra were obtained using a Micromass (Manchester, U.K.) Quattro II triple quadrupole mass spectrometer equipped with a Waters (Milford, MA) 2487 UV detector and a Waters 2690 HPLC system. Collision-induced dissociation (CID) was carried out using a range of collision energies from 25 to 70 eV and an argon collision gas pressure of 2.7 µbar. HPLC experiments were performed on a Shimadzu (Columbia, MD) LC-10A gradient HPLC system equipped with a SIL-10A auto injector, an SPDM10AV UV-vis photodiode array detector, and an SPD-10AV detector. HPLC Methodology. Six general methods were used to analyze and separate metabolites, GSH conjugates, and deoxynucleoside adducts. All retention times mentioned in the text were obtained using Method A unless stated otherwise. Method A: Analytical HPLC analysis was performed using a 4.6 × 150 mm Ultrasphere C18 column (Beckman, San Ramon, CA) using a Shimadzu HPLC system with UV detection at 280 nm. The mobile phase consisted of 20% methanol in 0.25% acetic acid and 0.25% perchloric acid (pH 3.5) at a flow rate of 1.0 mL/min for 5 min, increased to 35% methanol over 1 min, then to 60% methanol over the course of the next 39 min, and finally increased to 95% methanol over the remaining 5 min. Method B: A semipreparative method was developed using a 10 × 250 mm Ultrasphere C18 column and a mobile phase consisting of 40% methanol in 0.25% acetic acid and 0.25% perchloric acid (pH 3.5) at a flow rate of 3.0 mL/min for 5 min, increased to 50% methanol over the next 6 min, isocratic for the following 29 min, increased to 54% methanol over the next 15 min, and finally to 95% methanol over the last 5 min of the run. Method C: LC-MS analysis was performed using a 4.6 × 150 mm Ultrasphere C18 column with a photodiode array UV-vis absorbance detector set at 230-350 nm and a electrospray mass spectrometer. The mobile phase consisted of 20% methanol in 0.5% ammonium acetate (pH 3.5) at a flow rate of 1.0 mL/min for 5 min, increased to 35% methanol over 1 min, then to 60%

Yao et al. methanol over the next 39 min, and finally increased to 95% methanol over the remaining 5 min. Method D: LC-MS-MS analysis was performed using a 4.6 × 150 mm Ultrasphere C-18 column with on-line UV absorbance detection at 280 nm and electrospray MS-MS detection. The mobile phase consisted of 25% methanol in 0.5% ammonium acetate (pH 3.5) at a flow rate of 1.0 mL/min for 5 min, which was increased to 35% methanol over 1 min, then to 50% methanol over the next 39 min, and finally to 95% methanol over the remaining 10 min. Method E: LC-MS-MS analysis was the same as described in Method D with a modified mobile phase. The mobile phase consisted of 30% methanol in 0.5% ammonium acetate (pH 3.5) at a flow rate of 1.0 mL/min for 10 min, which was increased to 48% methanol over the next 35 min, then to 95% methanol over the remaining 20 min. Method F: A semipreparative isocratic method was developed using a 10 × 250 mm Ultrasphere C18 column with a mobile phase consisting of 46% methanol in 0.5% acetic acid (pH 3.5) at a flow rate of 3.0 mL/min and detection at 280 nm. Synthesis of 3,4-Dihydroxytoremifene (Scheme 1). 4-[2(N,N-Dimethylamino)ethoxy]-3,4-methylenedioxybenzophenone 1. 4-(2-chloroethoxy)-3,4-methylenedioxybenzophenone was synthesized as described previously (34). To a 250 mL flask were combined 4-(2-chloroethoxy)-3,4-methylenedioxybenzophenone (3.55 g, 11.6 mmol) and dimethylamine (2.0 M solution in methanol, 60 mL, 0.12 mol), and the mixture was refluxed under N2 for 5 days. The reaction mixture was concentrated and the residue was partitioned between diethyl ether (80 mL) and 0.1 M aqueous NaOH (50 mL). The aqueous layer was extracted with ether, and the combined organic layers were washed with water and dried over anhydrous Na2SO4. After concentration, the residue was purified by column chromatography (silica gel) using ethyl acetate/triethylamine (3:1, v/v) as eluent to give compound 1 (2.60 g, 70% yield). 1H NMR (CDCl3) δ 2.37 (s, 6H, CH3), 2.78 (t, 2H, J ) 5.7 Hz, NCH2), 4.16 (t, 2H, J ) 5.7 Hz, OCH2), 6.07 (s, 2H, OCH2O), 6.87 (d, 1H, J ) 8.9 Hz, ArH), 6.99 (d, 2H, J ) 8.9 Hz, ArH), 7.32 (d, 2H, J ) 8.9 Hz, ArH), 7.77(d, 2H, J ) 8.9 Hz, ArH); 13C NMR (CHCl3) δ 46.3, 58.5, 66.6, 102.1, 108.0, 110.3, 114.4, 126.6, 131.0, 132.6, 132.9, 148.2, 151.5, 162.6, 194.4. (E,Z)-1-{4-[2-(N,N-Dimethylethylamino)ethoxy]phenyl}-1-(3,4-methylenedihdroxyphenyl)-2-phenyl-4-chlorobut-1ene 2. TiCl4 (2.85 g, 15.0 mmol) was added dropwise to a stirred suspension of zinc powder (1.96 g, 30.0 mmol) in THF (35 mL) at -10 °C under N2. The resulting mixture was heated at reflux for 1.5 h. The suspension was cooled to room temperature and a mixture of compound 1 (1.10 g, 3.5 mmol) and 3-chloropropiophone (0.60 g, 3.5 mmol) in dry THF (30 mL) was added dropwise. The mixture was refluxed for 4 h, cooled to room temperature, and combined with 10% aqueous potassium carbonate (80 mL). The resulting mixture was stirred for 5 min and extracted with ethyl acetate (3 × 150 mL). The combined organic extracts were washed with saturated NaCl, dried over anhydrous Na2SO4, and concentrated. The final residue was purified by column chromatography (silica gel) using ethyl acetate/triethylamine (20:1, v/v) as eluent to give compound 2 as a mixture of E (trans) and Z (cis) isomers (2:5 E:Z) (0.85 g, 54% yield): 1H NMR (CD3OD) δ 2.29 (s, 6H, trans N(CH3)2), 2.35 (s, 6H, cis N(CH3)2), 2.68 (t, 2H, J ) 5.4 Hz, trans CH2N), 2.79 (t, 2H, J ) 5.4 Hz, cis CH2N), 2.94 (m, 4H, cis and trans CH2), 3.42 (m, 4H, cis and trans CH2Cl), 3.96 (t, 2H, J ) 5.4 Hz, trans CH2O), 4.11 (t, 2H, J ) 5.4 Hz, cis CH2O), 5.77 (s, 1H, trans OCHO), 5.95 (s, 1H, cis OCHO), 6.34 (d, 1H, J ) 1.3 Hz, ArH), 6.37 (d, 1H, J ) 8.8 Hz, ArH), 6.48 (d, 1H, J ) 8.8 Hz, ArH), 6.82 (d, 2H, J ) 8.8 Hz, ArH), 6.97 (d, 2H, J ) 8.8 Hz, ArH), 7.15-7.23 (m, 5H, ArH); EI-MS (m/z) 449 (100%) [M]+., 451 (38%) [M + 2]+. (E,Z)-1-{4-[2-(N,N-Dimethylethylamino)ethoxy]phenyl}-1-(3,4-dihyoxyphenyl)-2-phenyl-4-chlorobut-1ene 3 (34). To a flame-dried flask were added compound 2 (0.84 g, 1.87 mmol), 1 M solution of boron trichloride in methylene chloride (15.0 mL, 15.0 mmol), and 1,2-dichloroethane (35 mL) under N2. The solution was stirred for 48 h at room temperature. MeOH (45 mL) was gradually added to the solution, and the

Droloxifene- and Toremifene-o-quinones

Chem. Res. Toxicol., Vol. 14, No. 12, 2001 1645

Scheme 1. Synthesis of 3,4-Dihydroxytoremifenea

a Reagents and conditions: (a) dimethylamine, reflux, 5 days; (b) TiCl , Zn, 3-chloropropiophenone, THF, reflux, 4 h; (c) BCl , 1,24 3 dichloroethane, rt, 48 h.

mixture was stirred for 40 min. The solvent was removed under reduced pressure and a saturated NaHCO3 solution (50 mL) and MeOH (7 mL) was added to the residue. The pH of the solution was adjusted to 9.0 with 1 N NaOH and the resulting solution was extracted with ethyl acetate (2 × 150 mL). The combined organic layers were washed with saturated NaHCO3 (50 mL), H2O (100 mL), and dried over anhydrous sodium sulfate. After filtration, the solvent was removed and the final residue was purified by flash chromatography (silica gel) using methanol/ 2-propanol/acetic acid (12:2:1, v/v/v) as eluent to give compound 3 as a mixture of E (trans) and Z (cis) isomers (5:8 E:Z) (0.35 g, 43% yield): 1H NMR (CD3OD) δ 2.71 (s, 6H, trans N(CH3)2), 2.78 (s, 6H, cis N(CH3)2), 2.90 (t, 2H, J ) 7.3 Hz, trans CH2Cl), 2.96 (t, 2H, J ) 7.3 Hz, cis CH2Cl), 3.40 (t, 4H, cis and trans CH2N), 4.12 (t, 2H, J ) 5.1 Hz, trans CH2O), 4.30 (t, 2H, J ) 5.1 Hz, cis CH2O), 6.18 (d, 1H, J ) 8.1 Hz, ArH), 6.32 (d, 1H, J ) 2.0 Hz, ArH), 6.38 (d, 1H, J ) 8.1 Hz, ArH), 6.64 (d, 2H, J ) 8.6 Hz, ArH), 6.77 (d, 1H, J ) 8.0 Hz, ArH), 6.83 (d, 1H, J ) 8.7 Hz, ArH), 7.02 (d, 2H, J ) 8.7 Hz, ArH), 7.14-7.20 (m, 13H, ArH), 7.24 (d, 2H, J ) 8.7 Hz, ArH); CI-MS (positive ion, methane) m/z 438 (100%) [M + H]+. Preparation of Toremifene-3,4-quinone. Activated silver oxide (35) was prepared by adding KOH (0.72 g in 20 mL of H2O) to AgNO3 solution (2.0 g in 20 mL of H2O). The mixture was stirred for 15 min, the precipitate was washed with H2O, filtered, and dried. A mixture of 3,4-dihydroxytoremifene (4.4 mg), fresh silver oxide (440 mg), and acetonitrile (4 mL) was stirred for 15 min at 60 °C. After filtration, the solution was concentrated to a final volume of 1 mL. An aliquot (25 µL) of this solution was analyzed by LC-MS (Method A). Toremifeneo-quinone: UV (CH3OH/H2O) 227, 265, 375, 460 nm; positive ion electrospray MS m/z 436 (100%) [M + H]+; retention time 40 min. Reaction of Toremifene-o-quinone with GSH. A solution of toremifene-o-quinone (0.01 M) in acetonitrile was combined with GSH (0.1 M) in incubation buffer (6 mL, 50 mM phosphate buffer, pH 7.4), and the mixture was stirred at room temperature for 5 min. Perchloric acid (1.2 mL) was added, and the solution was concentrated to remove the remaining acetonitrile under a stream of N2. The final solution was centrifuged for 6 min at 13 000 rpm. The conjugates were purified by using

Method B. 3,4-di-OHTOR-SG 1: UV (CH3/H2O) 245, 270, 340 nm; 1H NMR (CD3OD) δ 2.10 (m, 2H, Gluβ), 2.44 (m, 2H, Gluγ), 2.80 (m, 1H, cysβ), 2.92 (t, 2H, J ) 7.6 Hz, CH2), 3.03 (s, 6H, CH3N), 3.07 (m, 1H, cysβ), 3.42 (t, 2H, J ) 7.6 Hz, CH2Cl), 3.60 (t, 2H, J ) 4.9 Hz, CH2N), 3.64 (m, 1H, GluR), 3.78 (d, 2H, GlyR), 4.23 (m, 1H, cysR), 4.40 (t, 2H, J ) 4.9 Hz, OCH2), 6.34 (d, 1H, J ) 2.0 Hz, ArH), 6.41 (d, 1H, J ) 2.0 Hz, ArH), 7.06 (d, 2H, J ) 8.7 Hz, ArH), 7.13-7.27 (m, 5H, ArH), 7.29 (d, 2H, J ) 8.7 Hz, ArH); positive ion electrospray MS m/z 743 (100%) [M + H]+; MS-MS with CID of m/z 725 (30%) [MH - H2O]+, 614 (100%) [MH - Glu]+, 511 (5%) [3,4-dihydroxytoremifene + SCH2CHdNH2]+, 470 (60%) [3,4-dihydroxytoremifene + SH]+; retention time 34 min. 3,4-di-OHTOR-SG 2: UV (CH3OH/H2O) 248, 277, 325, 343 nm; 1H NMR (CD3OD) δ 2.10 (m, 2H, Gluβ), 2.44 (m, 2H, Gluγ), 2.92 (s, 6H, CH3N), 2.99 (t, 2H, J ) 7.2 Hz, CH2), 3.04 (m, 1H, cysβ), 3.41 (t, 2H, J ) 7.2 Hz, CH2Cl), 3.50 (t, 2H, J ) 4.7 Hz, CH2N), 3.62 (m, 1H, GluR), 3.72 (d, 2H, GlyR), 4.23 (t, 2H, J ) 4.7 Hz, CH2O), 4.35 (m, 1H, cysR), 6.69 (d, 2H, J ) 8.8 Hz, ArH), 6.75 (d, 1H, J ) 2 Hz, ArH), 6.81 (d, 1H, J ) 2 Hz, ArH), 6.87 (d, 2H, J ) 8.8 Hz, ArH), 7.04-7.19 (m, 5H, ArH); positive ion electrospray MS m/z 743 (100%) [M + H]+; MS-MS with CID of m/z 725 (25%) [MH - H2O]+, 614 (97%) [MH - Glu]+, 511 (8%) [3,4-dihydroxytoremifene + SCH2CHd NH2]+, 470 (100%) [3,4-dihydroxytoremifene + SH]+; retention time 36 min. 3,4-di-OHTOR-SG 3: UV (CH3OH/H2O) 243, 263, 322, 342 nm; positive ion electrospray MS m/z 743 (100%); MSMS with CID of m/z 725 (8%) [MH - H2O]+, 614 (75%) [MH Glu]+, 511 (3%) [3,4-dihydroxtoremifene + SCH2CHdNH2]+, 470 (45%) [3,4-dihydroxytoremifene + SH]+; retention time 40 min. 3,4-di-OHTOR-diSG: UV (CH3OH/H2O) 234, 273, 314 nm; positive ion electrospray MS m/z 525 (100%) [M + 2H]2+; retention time 30 min. Oxidation of 3,4-Dihydroxytoremifene by Tyrosinase. A mixture of 3,4-dihydroxytoremifene (0.5 mM), tyrosinase (0.5 mg/mL), and GSH (5 mM) in 50 mM phosphate buffer (pH 7.4, 2.0 mL of total volume) was incubated for 30 min at 37 °C. Reactions were terminated by chilling in an ice bath followed by the addition of perchloric acid (50 µL/mL). Control incubations were conducted without tyrosinase or GSH. The reaction mixtures were centrifuged at 13 000 rpm for 6 min and the solutions were extracted using solid-phase extraction cartridges

1646


(Waters Oasis, HLB), eluted with methanol, and concentrated to a final volume of 250 µL. Aliquots (50 µL) were analyzed directly by LC-MS using Method C. In addition to the monoand di-GSH conjugates described above, tri-GSH conjugates were also obtained. 3,4-di-OHTOR-triSG 1: UV (CH3OH/H2O) 225, 270, 325 nm; positive ion electrospray MS m/z 678 (100%) [M + 2H]2+; retention time 14 min. 3,4-di-OHTOR-triSG 2: UV (CH3OH/H2O) 232, 272, 320 nm; positive ion electrospray MS m/z 678 [M + 2H]2+; retention time 15 min. Preparation of Rat Liver Microsomes. Female SpragueDawley rats (200-220 g) were obtained from Sasco Inc. (Omaha, NE). The rats were pretreated with dexamethasone to induce P450 3A isozymes. They were given intraperitoneal injections of dexamethasone in corn oil (100 mg/kg) daily for 3 consecutive days and were sacrificed on day 4. Rat liver microsomes were prepared and protein and P450 concentrations were determined as described previously (36). Oxidation of 4-Hydroxytamoxifen, 4-Hydroxytoremifene, and Droloxifene by Rat Liver Microsomes. A solution containing the substrate (0.1 mM), rat liver microsomes (1 nmol P450), [3H]GSH (specific activity of 40 nCi/nmol, 2.5 mM), and a NADPH-generating system (including 1.0 mM NADP+, 5.0 mM MgCl2, 5.0 mM isocitric acid, and 0.2 unit/mL isocitric dehydrogenase) in 50 mM phosphate buffer (pH 7.4, 0.5 mL total volume) was incubated for 30 min at 37 °C (37). For control incubations, NADP+ was omitted. The reactions were terminated by chilling in an ice bath followed by the addition of perchloric acid (25 µL). Radioactivity eluting from the HPLC column was measured in fractions collected at 18 s intervals. Reaction of Toremifene-o-quinone with Deoxynucleosides. A solution of toremifene-o-quinone (0.5 mM) in acetonitrile was combined with four deoxynucleosides (25 mM each) in phosphate buffer (pH 7.4) in a total volume of 5.0 mL. The solution was incubated for 5 h at 37 °C, extracted using Oasis solid-phase extraction cartridges, eluted with methanol, and concentrated to a final volume of 250 µL. Aliquots (50 µL) were analyzed directly by LC-MS-MS (Methods D and E). Toremifeneo-quinone was shown to react with all four deoxynucleosides. Three adducts were obtained with thymidine, five adducts were detected with deoxyguanosine, two adducts were observed with deoxyadenosine, and eight adducts were obtained with deoxycytosine. The 3,4-dihydroxytoremifene thymidine adducts all gave similar positive ion electrospray mass spectra with protonated molecules at m/z 678. The retention times were 24.0, 28.5, and 29.8 min (Method D). 3,4-di-OHTOR-T 1 was purified by semipreparative HPLC (Method F). 3,4-di-OHTOR-T 1: 1H NMR (DMSO-d6) δ 1.91 (s, 3H, CH3-T), 2.23 (s, 6H, N(CH3)2), 2.63 (t, 2H, J ) 5.7 Hz, NCH2), 2.85 (t, 2H, CH2), 3.74 (m, 2H, 5′-H-T), 3.83 (m, 1H, 4′-H-T), 4.08 (m, 1H, 3′-H-T), 4.25 (t, 1H, 5′-OH-T, D2O exchangeable), 5.63 (t, 1H, J ) 6.9 Hz, 1′-H-T), 6.27 (s, 1H, 2-OH, D2O exchangeable), 6.28 (s, 1H, 1-OH, D2O exchangeable), 6.36 (d, 1H, J ) 2.1 Hz, ArH), 6.39 (d, 1H, J ) 1.8 Hz, ArH), 6.69 (d, 2H, J ) 8.0 Hz, ArH), 6.86 (d, 2H, J ) 8.0 Hz, ArH), 7.16-7.20 (m, 5H, ArH), 7.58 (s, 1H, 6-H-T); MSMS with CID of m/z 678 (22%) [M + H]+, 562 (100%) [MH deoxyribose]+, 72 (21%) [CH2CH2N(CH3)2]+; retention time 24.0 min (Method D). The 3,4-dihydroxytoremifene deoxyguanosine adducts all gave similar positive ion electrospray mass spectra with protonated molecules at m/z 703. The retention times were 23.8, 29.0, 30.1, 32.0, and 41.5 min (Method E). 3,4-di-OHTORdG 1: MS-MS with CID of m/z 703 (15%) [M + H]+, 587 (100%) [MH - deoxyribose]+. The 3,4-dihydroxytoremifene deoxyadenosine adducts all gave similar positive ion electrospray mass spectra with protonated molecules at m/z 687. The retention times were 23.8 and 25.6 min (Method E). The 3,4-dihydroxytoremifene deoxycytosine adducts also gave similar positive ion electrospray mass spectra with protonated molecules at m/z 663. The retention times were 18.9, 23.1, 25.8, 31.6, 34.2, 35.3, 37.0, and 38.7 min (Method E). Cytotoxicity Studies in S30 and MDA-MB-231 Cells. The trypan blue exclusion assay was conducted to determine cell viability (38). The S30 cell line was provided by Dr. V. C. Jordan

Yao et al.

Figure 1. Time course analysis of incubation of tamoxifen-oquinone (squares) and toremifene-o-quinone (circles) at 37 °C and pH 7.4. Aliquots (200 µL) were combined with 5.0 mM GSH and 10 µL of PCA at various times. The data represent the peak area ratios of the GSH conjugates relative to that of the internal standard (2,4-dihydroxybenzoic acid). (Northwestern University, Evanston, IL) and was maintained in MEME medium supplemented with 5% charcoal-dextrantreated fetal bovine serum, 1% gluta-max, 1% penicillin, streptomycin, fungizome, 1% nonessential amino acids, 500 µg/mL G418, 6 mg/L insulin, and 5% CO2 (39). MDA-MB-231 cells were maintained in Liebovitz L-15 medium supplemented with 1% penicillin, streptomycin, fungizome, 1% gluta-max, 6 mg/L insulin, 5% fetal bovine serum, and 5% CO2 (39). The medium was routinely changed every 3 or 4 days and was changed 24 h prior to each experiment to maintain logarithmic growth. The cells were treated for 18 h with each test compound at final concentrations ranging from 5 to 100 µM. Negative controls (cells treated with DMSO only) were also carried out to define 100% cell viability. After treatment, floating cells were collected, attached cells were first trypsinized, and then floating and attached cells were combined. All cells were harvested by centrifugation at 2000 rpm for 5 min. Trypan blue stain (0.1 mL of 0.4%) was added to each cell suspension at a total volume of 0.6 mL. The cell suspension was transferred to a hemacytometer and the number of cells was counted under a microscope. The dead cells were stained blue and viable cells excluded the stain. The LC50 values were obtained by regression and linear estimation analysis. The data represent the means ( SD of three determinations.

Results and Discussion Reactivity of Toremifene-o-quinone. Among the chemical and enzymatic agents that have been reported to oxidize catechols including manganese dioxide, lead (IV) oxide, and tyrosinase (40, 41), we used silver oxide (35) as the chemical oxidizing agent and tyrosinase as the enzymatic agent to generate the o-quinone of 3,4dihydroxytoremifene. The UV spectra showed typical o-quinone chromophores at 375 and 460 nm. The lifetime of the toremifene-o-quinone under physiological conditions was determined by removing aliquots at various times and “quenching” it with GSH. The reaction system at each time was analyzed by HPLC to give a series of chromatograms with time-dependent decreased peak area of the o-quinone GSH conjugates (Figure 1). Under these conditions, the half-life of the toremifene-o-quinone was determined to be approximately 8 min. Similar studies with tamoxifen-o-quinone gave a half-life of about 80 min. The much shorter half-life of toremifene-oquinone compared to tamoxifen-o-quinone suggests that the toremifene-o-quinone is less stable and more reactive than tamoxifen-o-quinone. The reason for the increased reactivity could be due to the chlorine substituent which makes the toremifene-o-quinone more electrophilic compared to tamoxifen-o-quinone.



Figure 2. Positive ion electrospray CID MS-MS of m/z 743 of 3,4-di-OHTOR-SG 2.

3,4-Dihydroxytoremifene-o-quinone formed using silver oxide reacted with GSH to give four GSH conjugates, one di-GSH conjugate, and three mono-GSH conjugates (Scheme 2). When tyrosinase was used as the oxidizing agent, two additional tri-GSH conjugates were obtained likely due to further oxidation of the mono and di conjugates in the presence of the enzyme. During MSMS with CID of the protonated molecules of m/z 743, all of the mono GSH conjugates showed similar fragmentation patterns. For example, all formed dehydrated ions of m/z 725 (Figure 2). The ion at m/z 614 represented the loss of 129 mass units (pyroglutamic acid), and the ions of m/z 511, 470, and 72 corresponded to [3,4dihydroxytoremifene + SCH2CHdNH2]+, [3,4-dihydroxytoremifene + SH]+, and [dimethylaziridinium, CH2CH2N(CH3)2]+, respectively. The di-GSH conjugate 3,4-diOHTOR-di-SG gave a doubly charged ion at m/z 525 [M + 2H]2+. The mono-GSH conjugates were sufficiently stable for HPLC purification and their structures were elucidated with 1H NMR and COSY-NMR. We predict that the GSH molecule reacts at the C6 position on the catechol ring to give the E/Z isomers 3,4-di-OHTOR-SG 1 and 3,4-di-OHTOR-SG 2. In support of this, we observed two doublets in the aromatic region of the 1H NMR spectra which represent the meta coupling (J ) 2.0 Hz) between the H-3 proton and H-5 protons, respectively. Unfortunately, the di-GSH conjugates were not sufficiently stable for the HPLC purification and NMR characterization. The reason for the instability of the diGSH conjugates is likely due to the additional electrondonating sulfur substituent which facilitates oxidation of the aromatic ring, producing initially the o-quinone di-GSH conjugate which can further react with other nucleophiles in solution or undergo dimerization reactions. The tri-GSH conjugates obtained with tyrosinase were also not stable and only their MS and UV data were obtained. Oxidation of 4-Hydroxytamoxifen, Droloxifene, and 4-Hydroxytoremifene by Rat Liver Microsomes. It is well established that 4-hydroxylation of both tamoxifen and toremifene represents a major phase I metabolic

pathway for these antiestrogens (42, 43). In the current study, it was of interest to determine the relative ability of cytochrome P450 to convert droloxifene, 4-hydroxytamoxifen, and 4-hydroxytoremifene to o-quinones. We used microsomes isolated from female Sprague-Dawley rats treated with dexamethasone to induce P450 3A isozymes. The results showed that both droloxifene and 4-hydroxytamoxifen were oxidized to tamoxifen-o-quinone which reacted with GSH to give GSH conjugates. Previously it has been shown that tamoxifen-o-quinone binds covalently to proteins and DNA (44). In addition to the o-quinone, we and others (45) have shown that 4-hydroxytamoxifen is also oxidized to 4-hydroxytamoxifen quinone methide (Scheme 3). Previous work (46) showed that 4-hydroxytamoxifen quinone methide could react with GSH either through 1,8-Michael addition or 1,6-Michael addition to give GSH adducts. In contrast, droloxifene, which has a hydroxy group at the C3 position of the phenyl ring, cannot be converted to a quinone methide. As a result, only 3,4-dihydroxytamoxifen GSH conjugates and unreacted 3,4-dihydroxytamoxifen were detected in microsomal incubations with droloxifene (Scheme 3). Incubations with 4-hydroxytoremifene gave similar results as 4-hydroxytamoxifen in that two species of quinoid metabolites were formed: 4-hydroxytoremifene quinone methide and toremifene-o-quinone (Scheme 4). We have previously shown that 4-hydroxytoremifene quinone methide reacted with two molecules of GSH to produce the corresponding two isomers of di-GSH conjugates as shown in Scheme 4 (46). The reaction mechanism probably involves formation of an episulfonium ion intermediate which might contribute to the potential cytotoxic effects of toremifene. In the present study, three additional peaks were observed in the HPLC analysis of 4-hydroxytoremifene incubations, which were identified as the 3,4-dihydroxytormifene di-GSH conjugate and 3,4dihydroxytoremifene mono-GSH conjugates. Their retention times were identical to those obtained when 3,4dihydroxytoremifene was used as the substrate. 3,4Dihydroxytoremifene was also observed in the HPLC chromatogram of the 4-hydroxytoremifene incubation

1648


Yao et al.

Scheme 2. Structures of the GSH Conjugates of Toremifene-o-quinone

Scheme 3. Metabolism of 4-Hydroxytamoxifen and Droloxifene to Quinoids and Reaction with GSH

system. We determined the amount of each quinoid formed from 4-hydroxytamoxifen, droloxifene, and 4-hydroxytoremifene in rat liver microsomes by trapping the electrophiles with [3H]GSH. The values shown in Tables 1-3 might underestimate quinoid metabolite formation, because of the possibility that other nucleophiles in the incubation system, such as microsomal proteins, might bind to the quinoid metabolites. The radiochromatograms gave peaks with retention times identical to those obtained from GSH conjugation with the synthetic oquinones or quinone methides (Table 1-3). In the 4-hydroxytamoxifen incubation (Table 1), the amount of quinone methide-derived GSH conjugates was 0.20 nmol (nmol of P450)-1 (10 min)-1 and the amount of o-quinone-

derived GSH conjugates was 0.49 nmol (nmol of P450)-1 (10 min)-1. The increased amount of o-quinone relative to quinone methide may suggest that aromatic hydroxylation followed by oxidation to an o-quinone is preferred for 4-hydroxytamoxifen compared to direct oxidation to a quinone methide. In the 4-hydroxytoremifene incubation (Table 2), the amount of quinone methide-derived GSH conjugates was higher than that of o-quinonederived GSH conjugates. The amount of quinone methide formed from 4-hydroxytoremifene was 3-fold higher than 4-hydroxytamoxifen perhaps due to the β-Cl substitution which increased the acidity of the R-H and eased 2-electron oxidation forming 4-hydroxytormifene quinone methide. Previous work examining the effect of ring sub-



Scheme 4. Metabolism of 4-Hydroxytoremifene to Quinoids and Reaction with GSH

Table 1. Conversion of 4-Hydroxytamoxifen to GSH Conjugates by Rat Liver Microsomesa quinoid

conjugate

retention time (min)

rate of formation [nmol (nmol of P450)-1 (10 min)-1]

quinone methide

4-OHTAM-SG 1 4-OHTAM-SG 2 and 3 4-OHTAM-SG 4

25 28 30

o-quinone

3,4-di-OHTAM-diSG 2 3,4-di-OHTAM-SG 1 3,4-di-OHTAM-SG 2 3,4-di-OHTAM-SG 3

31 34 37 40

0.052 ( 0.007 0.089 ( 0.006 0.058 ( 0.013 total: 0.20 ( 0.03 0.211 ( 0.028 0.073 ( 0.014 0.111 ( 0.019 0.090 ( 0.015 total: 0.49 ( 0.08

a Incubations were conducted for 30 min with 0.1 mM substrate and rat liver microsomes (1.0 nmol of P450/mL) in the presence of an NADPH-generating system and 2.5 mM [3H]GSH (specific activity of 40 nCi/nmol) at 37 °C. Radioactivity eluting from the HPLC column was measured in fractions collected at 18 s intervals. Results are the average ( SD of three incubations.

Table 2. Conversion of 4-Hydroxytoremifene to GSH Conjugates by Rat Liver Microsomesa quinoid

conjugate


rate of formation [nmol (nmol of P450) -1 (10 min) -1]

quinone methide

4-OHTOR-diSG 1 and 2 4-OHTOR-diSG 3 and 4

13 16

o-quinone

3,4-di-OHTOR-diSG 3,4-di-OHTOR-SG 2 3,4-di-OHTOR-SG 3

30 36 40

0.24 ( 0.03 0.42 ( 0.11 total: 0.66 ( 0.14 0.32 ( 0.02 0.13 ( 0.05 0.03 ( 0.01 total: 0.48 ( 0.08


Table 3. Conversion of Droloxifene to GSH Conjugates by Rat Liver Microsomesa quinoid

conjugate


rate of formation [nmol (nmol P450)-1(10 min)-1]

o-quinone

3,4-di-OHTAM-diSG 1 3,4-di-OHTAM-diSG 2 3,4-di-OHTAM-SG 1 3,4-di-OHTAM-SG 2 3,4-di-OHTAM-SG 3

29 31 34 37 41

0.025 ( 0.003 0.160 ( 0.016 0.057 ( 0.006 0.084 ( 0.004 0.084 ( 0.016 total: 0.41 ( 0.05


stitution on the relative hepatotoxic effects of 4-methylphenols showed that the toxic mechanism likely involves biotransformation to a reactive quinone methide (47). The data showed that the presence of electron-withdrawing

substituents (2-chloro or 2-bromo) markedly enhanced both metabolism and toxicity of 4-methylphenol. As far as o-quinone formation was concerned, droloxifene (Table 3) gave similar amounts of o-quinone-derived GSH

1650


Figure 3. LC-MS-MS multiple reaction monitoring chromatograms of toremifene-o-quinone adducts with thymidine showing the chlorine isotope pattern for each species including (a) 3,4di-OHTOR-T 1, (b) 3,4-di-OHTOR-T 2, (c) 3,4-di-OHTOR-T 3.

conjugates when compared to 4-hydroxytamoxifen and 4-hydroxytoremifene. Reaction of Toremifene-o-quinone with Deoxynucleosides. Toremifene-o-quinone was incubated with four deoxynucleosides under physiological conditions, and the solutions were analyzed by HPLC and LC-MS-MS. Toremifene-o-quinone reacted with all four deoxynucleosides to form the corresponding adducts. Three adducts were detected in incubations with thymidine, five adducts in incubations with deoxyguanosine, two adducts in incubations with deoxyadenosine, and eight adducts in incubations with deoxycytosine. Toremifene-o-quinone produced considerably more adducts with thymidine as compared to the other three deoxynucleosides. Our previous work showed that tamoxifen-o-quinone formed adducts with deoxyguanosine and thymidine, but not with deoxyadenosine or deoxycytosine (34). As supported by the kinetic data which showed that the toremifeneo-quinone was considerably more reactive as compared to the tamoxifen-o-quinone, we also observed that the toremifene-o-quinone was much more effective at forming deoxynucleoside adducts. The toremifene-o-quinone deoxynucleoside adducts were analyzed by LC-MS-MS. Three experiments were used to provide evidence for these adducts. First, the protonated molecule of each deoxynucleoside adduct was detected by mass spectrometry. Second, MS-MS with CID showed the product ion of [MH - 116]+ resulting from loss of the sugar and the ion at m/z 72 corresponding to dimethylaziridinium, CH2CH2N(CH3)2+. Finally, the isotopic ratios of the protonated molecule of each deoxynucleoside adduct correlated with the natural abundance ratio between 35Cl and 37Cl (3:1) (Figure 3). The MS-MS spectrum of the 3,4-dihydroxytoremifene thymidine adduct is shown in Figure 4. The ion at m/z 678 (22%) was

Yao et al.

the protonated molecule of 3,4-dihydroxytoremifene thymidine adduct. The ion at m/z 562 (100%) [MH - 116]+ represented the fragment resulting from loss of the deoxyribosyl moiety from the protonated molecule. The ion at m/z 72 (21%) was identified as dimethylaziridinium [CH2CH2N(CH3)2+]. The tandem mass spectrum of the 3,4-dihydroxytoremifene deoxyguanosine adduct is shown in Figure 5. The ion at m/z 703 (15%) was the protonated molecule of the deoxyguanosine adduct. The ion at m/z 587 (100%) [MH - 116]+ represented the fragment resulting from loss of the deoxyribosyl moiety from the protonated molecule. The deoxyadenosine and deoxycytosine adducts of toremifene-o-quinone were not isolated and characterized because the yields were too low to be detected by HPLC with UV detection. As a result, only LC-MS-MS was used to analyze these adducts. Three adducts were generated from reaction of thymidine with toremifene-o-quinone (Figure 3). The major adduct 3,4-di-OHTOR-T 1 (retention time 24.0 min) was isolated and characterized. On the basis of 1H NMR (Figure 6), COSY-NMR, and DEPT-NMR, we predict that toremifene-o-quinone forms a covalent bond between the C6 position on the o-quinone and N3 position of thymidine (Figure 6). In the 1H NMR of 3,4-di-OHTOR-T 1, the peak representing the proton attached to the N3 position of thymidine (δ ) 11.26 ppm) disappeared, which suggested that the binding between thymidine and the o-quinone occurred at this position. Thymidine was also found to form adducts with 3,4-estrone-o-quinone at the N3 position and the C1 position of the steroid (48). The proton NMR spectrum also showed characteristic thymidine peaks for the C5 methyl group (1.91 ppm) and the C6 proton (7.58 ppm). The significant difference between the coupling pattern of aromatic protons of the catechol ring in 3,4-dihydroxytormifene and 3,4-diOHOR-T 1 indicated that the reaction between toremifeneo-quinone and thymidine occurred in the catechol ring of the o-quinone. Proton NMR showed the typical metacoupling between the C2 and C6 protons of the 3,4dihydroxytoremifene moiety (J ≈ 2 Hz), which was also observed in the COSY-NMR. This suggested that the reaction occurred at the C5 position. DEPT-NMR showed 12 aromatic carbons attached to one proton, one from thymidine (C6 position) and 11 from the 3,4-dihydroxytoremifene moiety which supports addition in the catechol ring. We concluded that the 3,4-di-OHTOR-T 1 was a Z-isomer based on its similarity to the proton NMR of the Z-isomer of 3,4-dihydroxytoremifene glutathione conjugates. In the proton NMR of the GSH conjugates, the two meta-coupling aromatic protons in the catechol ring of the Z-isomer (3,4-di-OHTOR-SG 1) had lower chemical shifts (6.34 and 6.41 ppm) than those of the E-isomer (3,4-di-OHTOR-SG 2) (6.75 and 6.81 ppm). The four aromatic protons in the phenyl ring with the dimethylaminoethoxy substituent of the Z-isomer had higher chemical shifts (7.06 and 7.28 ppm) than those of the E-isomer (6.68 and 6.86 ppm) (33, 34). DNA adducts of tamoxifen and toremifene have been detected in experimental animals (15, 49). Binding of tamoxifen and toremifene to DNA in rat liver was studied using 32P-postlabeling methods with HPLC-radioactivity detection (24). Two major adducts and at least six minor adducts were produced in the livers of female SpragueDawley rats. Toremifene was found to produce only one detectable adduct. The major tamoxifen adducts were suggested to be R-(N2-deoxyguanosinyl)tamoxifen and



Figure 4. Positive ion electrospray MS-MS with CID of the protonated molecule of 3,4-di-OHTOR-T 1 at m/z 678.

Figure 5. Positive ion electrospray MS-MS with CID of the protonated molecule of 3,4-di-OHTOR-dG 1 at m/z 703.

R-(N2-deoxyguanosinyl)-N-desmethyltamoxifen. The R-hydroxylation of the ethyl group is proposed to be the major metabolic activation pathway of tamoxifen to form DNA adducts (50). These tamoxifen-DNA adducts were also identified in the endometrium of women treated with tamoxifen (51). Since we have shown that both tamoxifeno-quinone and toremifene-o-quinone reacted with deoxynucleosides to form corresponding adducts, the o-quinone pathway might contribute to the genotoxicity of tamoxifen and toremifene. Cytotoxicity Studies in MDA-MB-231 and S30 Cells. We examined the relative toxicity of tamoxifen, analogues, and metabolites in two breast cancer cell lines, S30 (ER+) and MDA-MB-231 (ER-). The MDA-MB-231 cell line is an estrogen receptor negative cell line and the S30 cell line is an estrogen receptor positive cell line obtained from the MDA-MD-231 cell line stably transfected with ERR. We have previously shown that ERR positive cell lines are considerably more sensitive to the toxic effects of equine catechol estrogens (39), and as a

result, it was of interest to determine if similar effects were observed with the antiestrogens and metabolites. Tamoxifen, 4-hydroxytamoxifen, droloxifene, 3,4-dihydroxytamoxifen, 4-hydroxytoremifene, and 3,4-dihydroxytoremifene were all shown to be cytotoxic in both cell lines (Table 4). Droloxifene, 4-hydroxytamoxifen, and 4-hydroxytoremifene had equivalent cytotoxic potency in the estrogen receptor negative cell line (MDA-MB-231) and were more toxic than tamoxifen, 3,4-dihydroxytamoxifen, and 3,4-dihydroxytoremifene in the MDA-MB-231 cell line. Tamoxifen and 4-hydroxytamoxifen were more toxic than droloxifene, 4-hydroxytoremifene, 3,4-dihydroxytamoxfen, and 3,4-dihydroxytoremifene in the estrogen receptor positive cell line (S30). 3,4-Dihydroxytamoxifen and 3,4-dihydroxytoremifene had equivalent cytotoxic effects in both cell lines. Since the catechols showed less cytotoxic effects in breast cancer cells compared to the phenols, the catechol pathway may play a minor role in cytotoxicity. However, the catechols might contribute to the genotoxic effects of tamoxifen and toremifene since

1652


Figure 6.

1H

Yao et al.

NMR spectrum of 3,4-di-OHTOR-T 1.

Table 4. Cytotoxicity of Antiestrogen Metabolites in Breast Cancer Cell Linesa LC50 (µM) substrate

MDA-MB-231

S30

tamoxifen 4-OHTAM droloxifene 4-OHTOR 3,4-di-OHTAM 3,4-di-OHTOR

34 ( 3 29 ( 3 23 ( 2 28 ( 3 37 ( 1 35 ( 3

23 ( 1 19 ( 2 32 ( 1 33 ( 2 40 ( 2 38 ( 3

a Cells (105 cells/mL) were incubated with various concentrations of tested substrates for 18 h. Cell viability was determined by trypan blue exclusion as described in Materials and Methods. Results represent the mean ( SD of at least three determinations.

we have shown that tamoxifen-o-quinone (34) and toremifene-o-quinone forms adducts with deoxynucleosides. In conclusion, we have found that 3,4-dihydroxytoremifene was oxidized by silver oxide or tyrosinase to an o-quinone which reacted with GSH giving the corresponding conjugates. In microsomal incubations, both 4-hydroxytamoxifen and 4-hydroxytoremifene were metabolized to quinone methides and o-quinones. In contrast, droloxifene was only transformed to an o-quinone although the amount of o-quinone-derived GSH conjugates were similar as compared to 4-hydroxytamoxifen and 4-hydroxytoremifene. Both the tamoxifen-o-quinone and toremifene-o-quinone reacted with deoxynucleosides to give corresponding adducts. However, the toremifeneo-quinone was shown to be considerably more reactive than the tamoxifen-o-quinone in terms of both kinetic data as well as the yield and type of deoxynucleoside adducts formed. In cytotoxicity studies in human breast cancer cell lines, both 3,4-dihydroxytamoxifen and 3,4dihydroxytoremifene were less toxic than their parent phenols, 4-hydroxytamoxifen and 4-hydroxytoremifene, suggesting that o-quinone formation represents a minor cytotoxic pathway. However, the fact that the o-quinones formed adducts with deoxynucleosides in vitro implies that the o-quinone pathway could contribute to the genotoxicity of the antiestrogens in vivo. It should be noted that the major genotoxic pathway for tamoxifen likely involves R-hydroxylation, sulfate ester conjugation, and loss of the sulfate group producing a carbocation which forms covalent DNA adducts at the exocyclic amino groups of adenine and guanine (52, 53). In addition, both

toremifene and droloxifene are considerably less effective at induction of DNA adducts or liver tumors in animal models compared to tamoxifen (7, 15, 23). As a result, o-quinone formation from these antiestrogens likely represents a minor genotoxic pathway in vivo.

Acknowledgment. This work is supported by NIH Grants CA79870 and CA83124. We thank Dr. Yousheng Hua for assistance with mass spectrometry experiments. We also thank Dr. Shaonong Chen and Mr. Jianqiao Gu for their help in the structure characterization of 3,4dihydroxytoremifene thymidine adduct. We acknowledge Dr. John M. Pezzuto for access to the Bioassay Research Facility. We are grateful to Dr. V. C. Jordan (Northwestern University) for the gift of the S30 cell line.

References (1) Kitagawa, M., Ravindernath, A., Suzuki, N., Rieger, R., Terashima, I., Umemoto, A., and Shibutani, S. (2000) Identification of tamoxifen-DNA adducts induced by alpha-acetoxy-N-desmethyltamoxifen. Chem. Res. Toxicol. 13, 761-769. (2) White, I. N. H. (1999) The tamoxifen dilemma. Carcinogenesis 20, 1153-1160. (3) Nayfield, S. G., Karp, J. E., Ford, L. G., Dorr, F. A., and Kramer, B. S. (1991) Potential role of tamoxifen in prevention of breast cancer. J. Natl. Cancer Int. 83, 1450-1459. (4) Green, S., and Furr, B. (1999) Prospects for the treatment of endocrine-responsive tumours. Endocr.-Relat. Cancer 6, 349-371. (5) Potter, G. A., McCague, R., and Jarman, M. (1994) A mechanistic hypothesis for DNA adduct formation by tamoxifen following hepatic oxidative metabolism. Carcinogenesis 5, 439-442. (6) Divi, R. L., Osborne, M. R., Hewer, A., Phillips, D. H., and Poirier, M. C. (1999) Tamoxifen-DNA adduct formation in rat liver determined by immunoassay and 32P-postlabeling. Cancer Res. 59, 4829-4833. (7) Dhingra, K. (1999) Antiestrogensstamoxifen, SERMs and beyond. Invest. New Drugs 17, 285-311. (8) MacGregor, J. I., and Jordan, V. C. (1998) Basic guide to the mechanisms of antiestrogen action. Pharmacol. Rev. 50, 151196. (9) Williams, G. M., Ross, P. M., Jeffrey, A. M., and Karlsson, S. (1998) Genotoxicity studies with the antiestrogen toremifene. Drug Chem. Toxicol. 21, 449-476. (10) Cunningham, A., Klopman, G., and Rosenkranz, H. S. (1996) A study of the structural basis of the carcinogenicity of tamoxifen, toremifene and their metabolites. Mutat. Res. 349, 85-94. (11) Buzdar, A. U., and Hortobagyi, G. N. (1998) Tamoxifen and toremifene in breast cancer: Comparison of safety and efficacy. J. Clin. Oncol. 16, 348-353. (12) Hellmann-Blumberg, U., Cartner, M. G., Wurz, G. T., and DeGregorio, M. W. (1998) Intrinsic reactivity of tamoxifen and toremifene metabolites with DNA. Breast Cancer Res. Treat. 50, 135-141.

Droloxifene- and Toremifene-o-quinones (13) Davies, A. M., Martin, E. A., Jones, R. M., C. K., L., Smith, L. L., and I. N. H., W. (1995) Peroxidase activation of tamoxifen and toremifene resulting in DNA damage and covalently bound protein adducts. Carcinogenesis 16, 539-545. (14) Bartsch, H., Phillips, D. H., Nair, J., Hewer, A., MeybergSolomeyer, G., and Grischke, E. M. (2000) Lack of evidence for tamoxifen- and toremifene-DNA adducts in lymphocytes of treated patients. Carcinogenesis 21, 845-847. (15) Li, D., Dragan, Y., Jordan, V. C., Wang, M., and Pitot, H. C. (1997) Effects of chronic administration of tamoxifen and toremifene on DNA adducts in rat liver, kidney, and uterus. Cancer Res. 57, 1438-1441. (16) Kuramochi, H. (1996) Conformational studies and electronic structures of tamoxifen and toremifene and their allylic carbocations proposed as reactive intermediates leading to DNA adduct formation. J. Med. Chem. 39, 2877-2886. (17) O’Regan, R. M., Cisneros, A., England, G. M., MacGregor, J. I., Muenzner, H. D., Assikis, V. J., Bilimoria, M. M., Piette, M., Dragan, Y. P., Pitot, H. C., Chatterton, R., and Jordan, V. C. (1998) Effects of the antiestrogens tamoxifen, toremifene, and ICI 182,780 on endometrial cancer growth. J. Natl. Cancer Inst. 90, 1552-1558. (18) Rajaniemi, H., Koskinen, M., Mantyla, E., and Hemminki, K. (1998) DNA binding of tamoxifen and its analogues: identification of the tamoxifen-DNA adducts in rat liver. Toxicol. Lett. 102103, 453-457. (19) Hemminki, K., Widlak, P., and Hou, S. M. (1995) DNA adducts caused by tamoxifen and toremifene in human microsomal system and lymphocytes in vitro. Carcinogenesis 16, 1661-1664. (20) Rajaniemi, H., Mantyla, E., and Hemminki, K. (1998) DNA adduct formation by tamoxifen and structurally-related compounds in rat liver. Chem.-Biol. Interact. 113, 145-159. (21) Eppenberger, U., Wosikowski, K., and Kung, W. (1991) Pharmacologic and biologic properties of droloxifene, a new antiestrogen. Am. J. Clin. Oncol. 14, S5-14. (22) Ke, H. Z., Chen, H. K., Simmons, H. A., Qi, H., Crawford, D. T., Pirie, C. M., Chidsey-Frink, K. L., Ma, Y. F., Jee, W. S. S., and Thompson, D. D. (1997) Comparative effects of droloxifene, tamoxifen, and estrogen on bone, serum cholesterol, and uterine histology in the ovariectomized rat model. Bone 20, 31-39. (23) Rivera-Gonzalez, R., Petersen, D. N., Tkalcevic, G., Thompson, D. D., and Brown, T. A. (1998) Estrogen-induced genes in the uterus of ovariectomized rats and their regulation by droloxifene and tamoxifen. J. Steroid Biochem. Mol. Biol. 64, 13-24. (24) Hasmann, M., Rattel, B., and Loser, R. (1994) Preclinical data for droloxifene. Cancer Lett. 84, 101-116. (25) Phillips, D. H., Hewer, A., Horton, M. N., Cole, K. J., Carmichael, P. L., Davis, W., and Osborne, M. R. (1999) N-demethylation accompanies alpha-hydroxylation in the metabolic activation of tamoxifen in rat liver cells. Carcinogenesis 20, 2003-2009. (26) Wiebe, V. J., Benz, C. C., Shemano, I., Cadman, T. B., and DeGregorio, M. W. (1990) Pharmacokinetics of toremifene and its metabolites in patients with advanced breast cancer. Cancer Chemother. Pharmacol. 25, 247-251. (27) Coradini, D., Biffi, A., Cappelletti, V., and Di Fronzo, G. (1991) Effects of toremifene and its main metabolites on growth of breast cancer cell lines. Anticancer Res. 11, 2191-2197. (28) Kangas, L. (1990) Biochemical and pharmacological effects of toremifene metabolites. Cancer Chemother. Pharmacol. 27, 8-12. (29) Dehal, S. S., and Kupfer, D. (1999) Cytochrome P-450 3A and 2D6 catalyze ortho hydroxylation of 4-hydroxytamoxifen and 3-hydroxytamoxifen (droloxifene) yielding tamoxifen catechol: involvement of catechols in covalent binding to hepatic proteins. Drug Metab. Dispos. 27, 681-688. (30) Bolton, J. L., Trush, M. A., Penning, T. M., Dryhurst, G., and Monks, T. J. (2000) Role of quinones in toxicology. Chem. Res. Toxicol. 13, 135-160. (31) NIH Guidelines for the Laboratory Use of Chemical Carcinogens, NIH Publication, 81-2385 (1981) U.S. Government Printing Office, Washington, DC. (32) Gauthier, S., Mailhot, J., and Labrie, F. (1996) New Highly Stereoselective Synthsis of (Z)-4-Hydroxytamoxifen and (Z)-4Hydroxytoremifene via McMurry Reaction. J. Org. Chem. 61, 3890-3893. (33) McCague, R., Leclercq, G., Legros, N., Goodman, J., Blackburn, G. M., Jarman, M., and Foster, A. B. (1989) Derivatives of Tamoxifen. Dependence of Antiestrogenicity on the 4-Substituent. J. Med. Chem. 32, 2527-2533. (34) Zhang, F., Fan, P. W., Liu, X., Shen, L., van Breemen, R. B., and Bolton, J. L. (2000) Synthesis and reactivity of a potential carcinogenic metabolite of tamoxifen: tamoxifen-o-quinone. Chem. Res. Toxicol. 13, 53-62.

Chem. Res. Toxicol., Vol. 14, No. 12, 2001 1653 (35) Curtin, D. Y., and Harder, R. J. (1960) Stereochemistry and reactions with hydroxyl ion and with silver oxide of the 2-bromo4-phenylcyclohexanols and the 1-methyl-2-bromo-4-phenylcyclohexanols. J. Am. Chem. Soc. 82, 2357-2368. (36) Thompson, J. A., Malkinson, A. M., Wand, M. D., Mastovich, S. L., Mead, E. W., Schullek, K. M., and Laudenschlager, W. G. (1987) Oxidative metabolism of butylated hydroxytoluene by hepatic and pulmonary microsomes from rats and mice. Drug Metab. Dispos. 15, 833-840. (37) Iverson, S. L., Shen, L., Anlar, N., and Bolton, J. L. (1996) Bioactivation of estrone and its catechol metabolites to quinoidglutathione conjugates in rat liver microsomes. Chem. Res. Toxicol. 9, 492-499. (38) Freshney, R. (1987) Culture of Animal cells: A manual of basic technique, p 117, Alan R. Liss, Inc., New York. (39) Chen, Y., Liu, X., Pisha, E., Constantinou, A. I., Hua, Y., Shen, L., van Breemen, R. B., Elguindi, E. C., Blond, S. Y., Zhang, F., and Bolton, J. L. (2000) A metabolite of equine estrogens, 4-hydroxyequilenin, induces DNA damage and apoptosis in breast cancer cell lines. Chem. Res. Toxicol. 13, 342-350. (40) Cooksey, C. J., Garratt, P. J., Land, E. J., Ramsden, C. A., and Riley, P. A. (1998) Tyrosinase kinetics: failure of the autoactivation mechanism of monohydric phenol oxidation by rapid formation of a quinomethane intermediate. Biochem. J. 333, 685691. (41) Fenoll, L. G., Rodriguez-Lopez, J. N., Garcia-Sevilla, F., Tudela, J., Garcia-Ruiz, P. A., Varon, R., and Garcia-Canovas, F. (2000) Oxidation by mushroom tyrosinase of monophenols generating slightly unstable o-quinones. Eur. J. Biochem. 267, 5865-5878. (42) Mani, C., Gelboin, H. V., Park, S. S., Pearce, R., Parkinson, A., and Kupfer, D. (1993) Metabolism of the antimammary cancer antiestrogenic agent tamoxifen. I. Cytochrome P450-catalyzed N-demethylation and 4-hydroxylation. Drug Metab. Dispos. 21, 645-656. (43) Berthou, F., and Dreano, Y. (1993) High-performance liquid chromatographic analysis of tamoxifen, toremifene and their major human metabolites. J. Chromatogr. 616, 117-127. (44) Dehal, S. S., and Kupfer, D. (1996) Evidence that the catechol 3,4-Dihydroxytamoxifen is a proximate intermediate to the reactive species binding covalently to proteins. Cancer Res. 56, 12831290. (45) Marques, M. M., and Beland, F. A. (1997) Identification of tamoxifen-DNA adducts formed by 4-hydroxytamoxifen quinone methide. Carcinogenesis 18, 1949-1954. (46) Fan, P. W., Zhang, F., and Bolton, J. L. (2000) 4-Hydroxylated metabolites of the antiestrogens tamoxifen and toremifene are metabolized to unusually stable quinone methides. Chem. Res. Toxicol. 13, 45-52. (47) Thompson, D. C., Perera, K., and London, R. (1996) Studies on the mechanism of hepatotoxicity of 4-methylphenol (p-cresol): effects of deuterium labeling and ring substitution. Chem.-Biol. Interact. 101, 1-11. (48) Akanni, A., and Abul-Hajj, Y. J. (1997) Estrogen-nucleic acid adducts: Reaction of 3,4-estrone-o-quinone radical anion with deoxynucleosides. Chem. Res. Toxicol. 10, 760-766. (49) White, I. N., de Matteis, F., Davies, A., Smith, L. L., CroftonSleigh, C., Venitt, S., Hewer, A., and Phillips, D. H. (1992) Genotoxic potential of tamoxifen and analogues in female Fischer F344/n rats, DBA/2 and C57BL/6 mice and in human MCL-5 cells. Carcinogenesis 13, 2197-2203. (50) Phillips, D. H., Potter, G. A., Horton, M. N., Hewer, A., CroftonSleigh, C., Jarman, M., and Venitt, S. (1994) Reduced genotoxicity of [D5-ethyl]-tamoxifen implicates alpha-hydroxylation of the ethyl group as a major pathway of tamoxifen activation to a liver carcinogen. Carcinogenesis 15, 1487-1492. (51) Shibutani, S., Ravindernath, A., Suzuki, N., Terashima, I., Sugarman, S. M., Grollman, A. P., and Pearl, M. L. (2000) Identification of tamoxifen-DNA adducts in the endometrium of women treated with tamoxifen. Carcinogenesis 21, 1461-1467. (52) Umemoto, A., Monden, Y., Suwa, M., Kanno, Y., Suzuki, M., Lin, C. X., Ueyama, Y., Momen, M. A., Ravindernath, A., Shibutani, S., and Komaki, K. (2000) Identification of hepatic tamoxifenDNA adducts in mice: alpha-(N(2)-deoxyguanosinyl)tamoxifen and alpha-(N(2)-deoxyguanosinyl)tamoxifen N-oxide. Carcinogenesis 21, 1737-1744. (53) Umemoto, A., Komaki, K., Monden, Y., Suwa, M., Kanno, Y., Kitagawa, M., Suzuki, M., Lin, C. X., Ueyama, Y., Momen, M. A., Ravindernath, A., and Shibutani, S. (2001) Identification and Quantification of Tamoxifen-DNA Adducts in the Liver of Rats and Mice. Chem. Res. Toxicol. 14, 1006-1013.

TX010137I

Synthesis and Reactivity of Potential Toxic Metabolites of Tamoxifen

Recommend Documents