Oxidation of Raloxifene to Quinoids: Potential Toxic Pathways via a

Chem. Res. Toxicol. 2004, 17, 879-888

879

Oxidation of Raloxifene to Quinoids: Potential Toxic Pathways via a Diquinone Methide and o-Quinones Linning Yu, Hong Liu, Wenkui Li,† Fagen Zhang,‡ Connie Luckie, Richard B. van Breemen, Gregory R. J. Thatcher, and Judy L. Bolton* Department of Medicinal Chemistry and Pharmacognosy, College of Pharmacy, University of Illinois at Chicago, 833 South Wood Street, M/C 781, Chicago, Illinois 60612-7231 Received December 31, 2003

Raloxifene was approved in 1997 by the FDA for the treatment of osteoporosis in postmenopausal women, and it is currently in clinical trials for the chemoprevention of breast cancer. Before widespread use as a chemopreventive agent in healthy women, the potential cytotoxic mechanisms of raloxifene should be investigated. In the current study, raloxifene was incubated with GSH and either rat or human liver microsomes, and the metabolites and GSH conjugates were characterized using liquid chromatography-tandem mass spectrometry. Raloxifene was converted to raloxifene diquinone methide GSH conjugates, raloxifene o-quinone GSH conjugates, and raloxifene catechols. For comparison, three raloxifene catechols were synthesized and characterized. In particular, 7-hydroxyraloxifene was found to oxidize to the 6,7-o-quinone. As compared with raloxifene diquinone methide, which has a half-life of less than 1 s in phosphate buffer, the half-life of raloxifene 6,7-o-quinone was much longer at t1/2 ) 69 ( 2.5 min. The stability offered by raloxifene 6,7-o-quinone implies that it may be more toxic than raloxifene diquinone methide. Cytotoxicity studies in the human breast cancer cell lines S30 and MDA-MB-231 showed that 7-hydroxyraloxifene was more toxic than raloxifene in both cell lines. These results suggest that raloxifene could be metabolized to electrophilic and redox active quinoids, which have the potential to cause toxicity in vivo.

Introduction Tamoxifen (Figure 1) is the prototypical SERM,1 which has been extensively used for the treatment of hormonedependent breast cancer and more recently as a chemopreventive agent in women at risk of developing breast cancer. However, reports have shown that tamoxifen can lead to an increased endometrial cancer risk (1-8) and induce the formation of DNA adducts (9-11). One of the proposed bioactivation pathways leading to the toxicity of tamoxifen suggests that the generation of reactive intermediates such as the tamoxifen carbocation (1214), o-quinone (15, 16), and quinone methide (17-19) could cause DNA damage directly through the formation of DNA adducts or indirectly through the generation of reactive oxygen species, which oxidize DNA. Raloxifene (Evista) is a second generation SERM that has been approved by the Food and Drug Administration for the treatment and prevention of osteoporosis in postmenopausal women (Figure 1) (20). Raloxifene also * To whom correspondence should be addressed. Tel: 312-996-5280. Fax: 312-996-7107. E-mail: [email protected]. † Current address: Covance Laboratories, Inc., 3301 Kinsman Blvd., Madison, WI 53704. ‡ Current address: The Dow Chemical Company 47 Building, Midland, MI 48667. 1 Abbreviations: 3′-OHRA, 3′-hydroxyraloxifene; 3′-OHRA-di-SG, diglutathionyl-3′-hydroxyraloxifene; 3′-OHRA-SG, glutathionyl-3′-hydroxyraloxifene; 5-OHRA, 5-hydroxyraloxifene; 7-OHRA, 7-hydroxyraloxifene; 7-OHRA-di-SG, di-glutathionyl-7-hydroxyraloxifene; 7-OHRA-SG, glutathionyl-7-hydroxyraloxifene; APCI, atmospheric pressure chemical ionization; CID, collision-induced dissociation; ER, estrogen receptor; LC-MS-MS, liquid chromatography-tandem mass spectrometry; MEME, minimal essential media with Earle’s salts; PPA, polyphosphoric acid; SERM, selective estrogen receptor modulator.

Figure 1. Structures of tamoxifen and raloxifene.

inhibits the development of mammary tumors in animal models (21, 22) and is currently in clinical trials for the chemoprevention of breast cancer (STAR trial; Study of Tamoxifen and Raloxifene) (23). The development of additional SERMs is ongoing, and the knowledge of mechanisms of toxicity of existing drugs should lead to less toxic SERMs in the future. In this study, we describe the formation of raloxifene diquinone methide, raloxifene catechols, and raloxifene o-quinones in incubations of raloxifene with rat liver microsomes. In addition to the three raloxifene monoGSH conjugates, which have already been identified and described by Chen et al. (24), we detected three hydroxylated raloxifenes including the raloxifene catechols 7- and 3′-OHRA; two di-GSH conjugates; and three monoGSH conjugates of 7- and 3′-OHRA. Similar results were observed in incubations of raloxifene with human liver microsomes. Three raloxifene catechols were synthesized as standards and used for further study. Cytotoxicity studies in S30 (ERR-positive) and MDA-MB-231 (ERRnegative) cell lines indicated that 7-OHRA is a toxic

10.1021/tx0342722 CCC: $27.50 © 2004 American Chemical Society Published on Web 05/12/2004

880

Chem. Res. Toxicol., Vol. 17, No. 7, 2004

metabolite. The raloxifene diquinone methide and 6,7o-quinone were generated chemically by silver oxide oxidation, and comparison of their reactivities revealed that raloxifene 6,7-o-quinone is likely to be a more toxic metabolite than the raloxifene diquinone methide.

Materials and Methods Caution: Raloxifene diquinone methide, raloxifene catechols, and raloxifene o-quinones were handled in accordance with the NIH Guidelines for the Laboratory Use of Chemical Carcinogens. All solvents and chemicals were purchased from Aldrich Chemical (Milwaukee, WI), Fisher Scientific (Itasca, IL), or Sigma (St. Louis, MO) unless stated otherwise. Raloxifene was synthesized according to a literature procedure (20). Instrumentation. NMR spectra were obtained using a Bruker DRX 500, AVANCE 360, or AVANCE 300. UV spectra were obtained on a Hewlett-Packard 8452A photodiode array UV-vis spectrophotometer. LC-MS-MS spectra were obtained using a Micromass Quattro II triple quadrupole mass spectrometer equipped with a Waters 2690 HPLC system and 2487 UV detector and electrospray ionization. CID was carried out using MS-MS using a range of collision energies from 25 to 70 eV and an argon collision gas pressure of 1.2 µbar. LC-MS analyses were carried out using an Agilent LC/MSD ion trap equipped with a model 1100 HPLC system and APCI. High-resolution electrospray mass spectra were obtained using a Micromass Q-TOF-2 hybrid mass spectrometer. HPLC Methodology. Two general methods were used to analyze and separate the various metabolites and conjugates. All retention times reported in the text were obtained using method A unless stated otherwise. Method A consisted of an agilent Zorbax Rx-C8 column (4.6 mm × 250 mm, 5 µm) and was used for LC-MS-MS analysis of raloxifene catechols and GSH conjugates. The mobile phase consisted of a linear gradient from 10 to 30% acetonitrile over 60 min, increasing to 90% over 15 min. The counter solvent was water containing 10% methanol and 0.4% formic acid (v/v/v). In method B, a Beckman Ultrasphere C18 column (10 mm × 250 mm, 5 µm) was used for the isolation of 7-OHRA GSH conjugates using a Shimadzu LC-10A gradient HPLC system equipped with a SIL-10A auto injector and a SPD-10AV detector set at 284 nm. The solvent system was the same as that used for method A except that 14% acetonitrile was used for 30 min followed by a linear gradient from 14 to 20% over 5 min. Preparation of Rat Liver Microsomes. Female SpragueDawley rats (200-220 g) were obtained from Sasco Inc. (Omaha, NE). To induce P450 3A isozymes, rats were pretreated with 100 mg/kg dexamethasone (i.p. in corn oil) 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 (25). Incubation of Raloxifene and Raloxifene Catechols with Rat and Human Liver Microsomes and GSH. A solution containing raloxifene (50 µM), rat liver microsomes (1 nmol P450/mL), GSH (0.5 mM), and an 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, 1 mL total volume) was incubated for 30 min at 37 °C. For control incubations, either NADP+ or GSH was omitted. The reactions were terminated by chilling in an ice bath followed by the addition of 50 µL of perchloric acid. The protein was removed using ultrafiltration through a 30 000 molecular weight cutoff regenerated cellulose filter (Millipore, Bedford, MA). Each ultrafiltrate was concentrated to 200 µL under nitrogen, and a 100 µL aliquot was analyzed using LC-MS-MS. A similar procedure was used for the incubation of raloxifene catechols. Human liver microsomes (pooled from 10 individuals) were purchased from In Vitro Technologies (Baltimore, MD). Human liver microsomal incubations were carried out using similar procedures except that the concentration of P450 was set at 0.34 nmol P450/mL.

Yu et al. Synthesis of Raloxifene Catechols. 7-OHRA was synthesized as shown in Scheme 1. 4-Methoxy-a-[(2,3-dimethoxyphenyl)thio]acetophenone 3. To a freshly prepared solution of 25 mL of ethanol, 10 mL of water, and 1.6 g of KOH at room temperature, 2,3dimethoxybenzenethiol (1) (3.8 g, 0.022 mol) was added in one portion, and the solution was cooled to 0 °C. Next, 2-bromo-4′methoxyacetophenone (2) (5.0 g, 0.022 mol) in 15 mL of ethyl acetate was added dropwise to this solution and the reaction was stirred at room temperature overnight. Most of the solvents were removed under reduced pressure, and a mixture of yellow oil and white solid was obtained. The oil and the white solid were dissolved in 40 mL of water and 40 mL of ethyl acetate, and the aqueous layer was extracted two times with additional 40 mL portions of ethyl acetate. The organic layers were combined and washed with consecutive portions of 10% HCl, water, saturated NaHCO3, water, and saturated NaCl. After it was dried over Na2SO4, the solvent was evaporated under reduced pressure. The crude product 3 was recrystallized from methanol to give 4 g, 56% yield. 1H NMR (300 MHz, acetoned6): δ 3.77 (s, 3H), 3.85 (s, 3H), 3.91 (s, 3H), 4.44 (s, 2H), 6.90 (dd, J ) 1.8, 7.9 Hz, 1H), 6.96 (dd, J ) 1.8, 7.9 Hz, 1H), 7.02 (t, J ) 7.9 Hz, 1H), 7.06 (d, J ) 9.0 Hz, 2H), 8.06 (d, J ) 9.0 Hz, 2H). APCI-MS m/z 319 (100%) [M + H]+. 6,7-Dimethoxy-2-(4-methoxyphenyl)benzo[b]thiophene 4. PPA (40 g) was heated to 70 °C, and then, compound 3 (4.0 g, 12.6 mmol) was added over a period of 10 min. After the addition was complete, the mixture was stirred at 70-75 °C for 1 h and then poured into rapidly stirred ice/water. The yellow precipitate was recovered by filtration and was washed with water. Product 4 was crystallized from acetone in 80% overall yield. 1H NMR (300 MHz, CDCl3): δ 3.85 (s, 3H), 3.95 (s, 3H), 4.05 (s, 3H), 6.95 (d, J ) 8.6 Hz, 2H), 7.04 (d, J ) 8.5 Hz, 1H), 7.34 (s, 1H), 7.42 (d, J ) 8.5 Hz, 1H), 7.62 (d, J ) 8.6 Hz, 2H). APCI-MS m/z 301 (100%) [M + H]+. 6,7-Dimethoxy-2-(4-methoxyphenyl)-3-bromo-benzo[b]thiophene 5. A solution of compound 4 (2.5 g, 8.33 mmol) in 150 mL of CH2Cl2 was added to a beaker, which was wrapped with aluminum foil and put into a dry ice/acetone bath. N-bromoacetamide (1.15 g, 8.33 mmol) was dissolved in 20 mL of ethanol and added dropwise over 20 min to the solution of 4. The reaction mixture was allowed to warm to room temperature with stirring. The solvents were removed under reduced pressure, and product 5 (1.6 g, 50%) was crystallized from ethyl acetate and methanol (1:6 v/v). 1H NMR (300 MHz, CDCl3): δ 3.89 (s, 3H), 3.99 (s, 3H), 4.06 (s, 3H), 7.02 (d, J ) 8.8 Hz, 2H), 7.18 (d, J ) 8.7 Hz, 1H), 7.53 (d, J ) 8.7 Hz, 1H), 7.71 (d, J ) 8.8 Hz, 2H). APCI-MS m/z 379 (100%) [M + H]+ (79Br), 381 (97%) [M + H]+ (81Br), 300 (50%) [MH - Br]•+. 6,7-Dimethoxy-2-(4-methoxyphenyl)-3-(4-fluorobenzoyl)benzo[b]thiophene 6. A solution of 5 (4.33 g, 11.42 mmol) in 150 mL of dry THF was cooled to -78 °C in a dry ice/acetone bath. To this was added butyllithium (14.3 mL, 1.6 M in hexane, 22.88 mmol) followed by the addition of 4-fluorobenzoyl chloride (1.5 mL, 12.5 mmol) in 20 mL of dry THF, and the reaction mixture was allowed to warm to room temperature. The mixture was diluted with 200 mL of diethyl ether, washed with 200 mL of saturated NH4Cl, followed by 200 mL of 5% w/v NaHCO3. After they were dried over Na2SO4, the solvents were removed under reduced pressure giving a yellow oil. The oil was subjected to column chromatography using a mobile phase consisting of hexane and ethyl acetate (5:1, v/v) to give the expected product 6 (2.1 g, 44%). 1H NMR (300 MHz, methanol-d4): δ 3.75 (s, 3H), 3.96 (s, 3H), 4.05 (s, 3H), 6.81 (d, J ) 8.8 Hz, 2H), 7.03 (t, J ) 8.7 Hz, 2H), 7.22 (d, J ) 8.7 Hz, 1H), 7.31 (d, J ) 8.8 Hz, 2H), 7.41 (d, J ) 8.7 Hz, 1H), 7.77 (dd, J ) 8.7, 5.5 Hz, 2H). APCIMS m/z 423 (100%) [M + H]+. [6,7-Dimethoxy-2-(4-methoxyphenyl)benzo[b]thien-3-yl][4-[2-(1-piperidinyl)ethoxy]phenyl]methanone 7. To a solution of NaH (312 mg, 60% in mineral oil, 7.8 mmol) in 5 mL of dry DMF was added 1-piperidine ethanol (917 mg, 7.1 mmol) dropwise. A large volume of hydrogen was evolved during the

Quinoids Formed from Raloxifene

Chem. Res. Toxicol., Vol. 17, No. 7, 2004 881 Scheme 1. Synthesis of Raloxifene Catecholsa

a Reagents and conditions: (a) KOH, overnight. (b) PPA, 70-75 °C, 1 h for the synthesis of 4 and 12; 60 °C, 1 h followed by 80-85 °C, 1 h for the synthesis of 20. (c) N-Bromoacetamide, -78 to 20 °C. (d) Butyllithium, 4-fluorobenzoyl chloride, -78 to 20 °C. (e) NaH, DMF, 1-piperidinylethanol, 15 min, room temperature. (f) AlCl3, EtSH, -78 to 20 °C, overnight.

addition. Compound 6 (1.5 g, 3.55 mmol) in 10 mL of dry DMF was added to this solution, and the mixture was stirred at room temperature for 15 min. The mixture was then poured into 100 mL of water and extracted with ethyl acetate (3 × 100 mL). The extracts were combined, dried over Na2SO4, and evaporated under reduced pressure. The mixture was subject to column chromatography with chloroform followed by methanol. The methanol was evaporated under reduced pressure to give 1.78 g (3.35 mmol, 94% yield) of product 7 as a yellow oil. 1H NMR (300 MHz, acetone-d6): δ 1.39 (m, 2H), 1.51 (m, 4H), 2.42 (m, 4H), 2.65 (t, J ) 4.8 Hz, 2H), 3.76 (s, 3H), 3.94 (s, 3H), 4.05 (s, 3H), 4.08 (t, J ) 4.8 Hz, 2H), 6.84 (d, J ) 8.7 Hz, 2H), 6.87 (d, J ) 8.9 Hz, 2H), 7.17 (d, J ) 8.4 Hz, 1H), 7.26 (d, J ) 8.4 Hz, 1H), 7.40 (d, J ) 8.7 Hz, 2H), 7.75 (d, J ) 8.9 Hz, 2H). APCIMS m/z 532 (100%) [M + H]+. [6,7-Dihydroxy-2-(4-hydroxyphenyl)benzo[b]thien-3-yl][4-[2-(1-piperidinyl)ethoxy]phenyl]methanone Hydrochloride (7-OHRA) 8. A solution of 6 (1.5 g, 2.83 mmol) in 25 mL of dry CH2Cl2 under Ar was cooled to -78 °C in a dry ice/acetone bath. To this was added AlCl3 (3.4 g, 25.4 mmol) followed by ethanethiol (1.68 mL, 22.6 mmol). The reaction mixture was allowed to warm to room temperature and stirred overnight. The reaction was quenched by the addition of 16 mL of THF followed by 4 mL of 20% aqueous HCl and 16 mL of degassed water at a rate such that the temperature remained below 25

°C. The precipitate was recrystallized from degassed and nitrogen-purged methanol and water (1:1, v/v) to give 7-OHRA, 8 (1.1 g, 80%), as a yellow powder. The purity was estimated to be >97% by HPLC using method A. 1H NMR (360 MHz, methanol-d4): δ 1.44 (m, 1H), 1.76 (m, 3H), 1.86 (m, 2H), 2.95 (m, 2H), 3.45 (t, J ) 4.8 Hz, 2H), 3.50 (m, 2H), 4.29 (t, J ) 4.8 Hz, 2H), 6.55 (d, J ) 8.7 Hz, 2H), 6.83 (d, J ) 8.4 Hz, 1H), 6.84 (d, J ) 8.9 Hz, 2H), 6.89 (d, J ) 8.4 Hz, 1H), 7.13 (d, J ) 8.7 Hz, 2H), 7.66 (d, J ) 8.9 Hz, 2H). 13C NMR (90.6 MHz, methanol-d4): δ 22.47, 24.01, 54.89, 56.92, 63.25, 115.22, 115.43, 116.44, 116.65, 126.27, 128.85, 131.38, 131.49, 132.64, 133.50, 135.55, 140.05, 142.72, 144.87,159.22, 163.17, 195.62. Positive ion electrospray HRMS m/z calculated for C28H28NO5S, 490.1688; found, 490.1703. The procedure for the synthesis of 5-OHRA was the same as that of 7-OHRA except that the starting material was 3,4-dimethoxybenzenethiol (Scheme 1). 4-Methoxy-r-[(3,4-dimethoxyphenyl)thio]acetophenone 11. Compound 11 was obtained in 82% yield. 1H NMR (300 MHz, CDCl3): δ 3.82 (s, 3H), 3.86 (s, 3H), 3.87 (s, 3H), 4.12 (s, 2H), 6.78 (d, J ) 8.3 Hz, 1H), 6.92 (d, J ) 8.9 Hz, 2H), 6.95 (d, J ) 2 Hz, 1H), 7.01 (dd, J ) 2, 8.3 Hz, 1H), 7.91 (d, J ) 8.9 Hz, 2H). APCI-MS m/z 319 (100%) [M + H]+. 5,6-Dimethoxy-2-(4-methoxyphenyl)benzo[b]thiophene 12. Compound 12 was obtained in 86% yield. 1H NMR (300 MHz, CDCl3): δ 3.84 (s, 3H), 3.95 (s, 6H), 6.94 (d, J ) 8.5

882


Hz, 2H), 7.18 (s, 1H), 7.25 (s, 1H), 7.30 (s, 1H), 7.59 (d, J ) 8.5 Hz, 2H). APCI-MS m/z 301 (100%) [M + H]+. 5,6-Dimethoxy-2-(4-methoxyphenyl)-3-bromo-benzo[b]thiophene 13. Compound 13 was obtained in 91% yield. 1H NMR (300 MHz, CDCl3): δ 3.86 (s, 3H), 3.96 (s, 3H), 4.02 (s, 3H), 7.01 (d, J ) 8.7 Hz, 2H), 7.22 (s, 1H), 7.245 (s, 1H), 7.68 (d, J ) 8.7 Hz, 2H). APCI-MS m/z 379 (100%) [M + H]+ (79Br), 381 (97%) [M + H]+ (81Br), 300 (50%) [MH - Br]•+. 5,6-Dimethoxy-2-(4-methoxyphenyl)-3-(4-fluorobenzoyl)benzo[b]thiophene 14. Compound 14 was obtained in 44% yield. 1H NMR (300 MHz, methanol-d4): δ 3.75 (s, 3H), 3.90 (s, 3H), 4.02 (s, 3H), 6.75 (d, J ) 8.8 Hz, 2H), 6.91 (t, J ) 8.7 Hz, 2H), 7.27 (d, J ) 8.8 Hz, 2H), 7.31 (s, 2H), 7.77 (dd, J ) 8.7, 5.5 Hz, 2H). APCI-MS m/z 423 (100%) [M + H]+. [5,6-Dimethoxy-2-(4-methoxyphenyl)benzo[b]thien-3-yl][4-[2-(1-piperidinyl)ethoxy]phenyl]methanone 15. Compound 15 was obtained in 94% yield. 1H NMR (300 MHz CDCl3): δ 1.28 (m, 2H), 1.43 (m, 4H), 2.30 (s, 4H), 2.55 (t, J ) 5.7 Hz, 2H), 3.54 (s, 3H), 3.70 (s, 3H), 3.78 (s, 3H), 3.90 (t, J ) 5.7 Hz, 2H), 6.67 (d, J ) 8.7 Hz, 2H), 6.61 (d, J ) 8.9 Hz, 2H), 7.07 (s, 1H), 7.13 (s, 1H), 7.17 (d, J ) 8.7 Hz, 2H), 7.63 (d, J ) 8.9 Hz, 2H). APCI-MS m/z 532 (100%) [M + H]+. [5,6-Dihydroxy-2-(4-hydroxyphenyl)benzo[b]thien-3-yl][4-[2-(1-piperidinyl)ethoxy]phenyl]methanone Hydrochloride (5-OHRA) 16. The final product was recrystallized from nitrogen-saturated methanol and water (1:1, v/v). The yield was 80%. The purity was estimated to be >97% by HPLC using method A. 1H NMR (360 MHz, methanol-d4): δ 1.54 (m, 2H), 1.86 (m, 4H), 3.04 (m, 2H), 3.51 (t, J ) 4.8 Hz, 2H), 3.53 (m, 2H), 4.34 (t, J ) 4.8 Hz, 2H), 6.58 (d, J ) 8.6 Hz, 2H), 6.87 (d, J ) 8.9 Hz, 2H), 7.03 (s, 1H), 7.12 (d, J ) 8.6, 2H), 7.23 (s, 1H), 7.69 (d, J ) 8.9 Hz, 2H). Positive ion electrospray HRMS m/z calculated for C28H28NO5S, 490.1688; found, 490.1708. The procedure for the synthesis of 3′-OHRA was similar to that of 7-OHRA except as indicated below (Scheme 1). 3,4-Dimethoxy-r-[(3-methoxyphenyl)thio]acetophenone 19. The yield was 73%. 1H NMR (300 MHz, CDCl3): δ 3.79 (s, 3H), 3.93 (s, 3H), 3.97 (s, 3H), 4.27 (s, 2H), 6.77 (m, 1H), 6.90 (d, J ) 8.4 Hz, 1H), 6.96 (m, 1H), 7.00 (m, 1H), 7.21(t, J ) 8.1 Hz, 1H), 7.52 (d, J ) 2.0 Hz, 1H), 7.59 (dd, J ) 2.0, 8.4 Hz, 1H). APCI-MS m/z 319 (100%) [M + H]+. 6-Methoxy-2-(3,4-dimethoxyphenyl)benzo[b]thiophene 20. Compound 20 was synthesized using an analogous procedure to the synthesis of 4 except that the reaction was maintained at 60-65 °C for 1 h, and then, the temperature was increased to 80-85 °C for 1 h. The yield was 64%. 1H NMR (360 MHz, CDCl3): δ 3.90 (s, 3H), 3.94 (s, 3H), 3.97 (s, 3H), 6.92 (d, J ) 8.3 Hz, 1H), 7.00 (dd, J ) 2.3, 8.7 Hz, 1H), 7.19 (d, J ) 2.1 Hz, 1H), 7.24 (dd, J ) 2.1, 8.3 Hz, 1H), 7.31 (d, J ) 2.3 Hz, 1H), 7.36 (s, 1H), 7.64 (d, J ) 8.7 Hz, 1H). 13C NMR (90.6 MHz, CDCl3): δ 55.77, 56.16 (overlapped), 105.10, 109.65, 111.67, 114.55, 118.19, 119.12, 124.10, 127.76, 135.01, 140.82, 141.80, 149.28, 149.38, 157.44. APCI-MS m/z 301 (100%) [M + H]+. 6-Methoxy-2-(3,4-dimethoxyphenyl)-3-bromobenzo[b]thiophene 21. The yield was 90%. 1H NMR (300 MHz, CDCl3): δ 3.91 (s, 3H), 3.96 (s, 3H), 3.976 (s, 3H), 6.97 (d, J ) 8.1 Hz, 1H), 7.09 (dd, J ) 2.3, 8.8 Hz 1H), 7.30 (m, 3H), 7.74 (d, J ) 8.8 Hz, 1H). APCI-MS m/z 379 (100%) [M + H]+ (79Br), 381 (97%) [M + H]+ (81Br), 300 (50%) [MH - Br]•+. 6-Methoxy-2-(3,4-dimethoxyphenyl)-3-(4-fluorobenzoyl)benzo[b]thiophene 22. The yield was 45%. 1H NMR (300 MHz, CDCl3): δ 3.74 (s, 3H), 3.82 (s, 3H), 3.89 (s, 3H), 6.74 (d, J ) 8.3 Hz, 1H), 6.86 (d, J ) 2.2 Hz, 1H), 7.00 (m, 4H), 7.32 (d, J ) 2.3 Hz, 1H), 7.65 (d, J ) 8.7 Hz, 1H), 7.82 (dd, J ) J ) 8.7, 5.5 Hz, 2H). APCI-MS m/z 423 (100%) [M + H]+. [6-Methoxy-2-(3,4-dimethoxyphenyl)benzo[b]thien-3-yl][4-[2-(1-piperidinyl)ethoxy]phenyl]methanone 23. The yield was 92%. 1H NMR (300 MHz, methanol-d4): δ 1.44 (m, 2H), 1.59 (m, 4H), 2.50 (m, 4H), 2.76 (t, J ) 5.9 Hz, 2H), 3.73 (s, 3H), 3.83 (s, 3H), 3.88 (s, 3H), 4.09 (t, J ) 5.9 Hz, 2H), 6.74 (d, J ) 8.3 Hz, 1H), 6.76 (d, J ) 8.9 Hz, 2H), 6.89 (d, J ) 2.0 Hz, 1H), 6.97 (dd, J ) 2.4, 8.9 Hz 1H), 7.03 (dd, J ) 2.0, 8.3 Hz,

Yu et al. 1H), 7.32 (d, J ) 2.4 Hz, 1H), 7.55 (d, J ) 8.9 Hz, 1H), 7.77 (d, J ) 8.9 Hz, 2H). APCI-MS m/z 532 (100%) [M + H]+. [6-Hydroxy-2-(3,4-dihydroxyphenyl)benzo[b]thien-3-yl][4-[2-(1-piperidinyl)ethoxy]phenyl]methanone Hydrochloride (3′-OHRA) 24. The final product was crystallized from nitrogen-saturated methanol and water (1:1, v/v). The yield was 70%. The purity was estimated to be >80% using HPLC by method A. 1H NMR (360 MHz, methanol-d4): δ 1.55 (m, 1H), 1.80 (m, 3H), 1.90 (m, 2H), 3.02 (m, 2H), 3.52 (t, J ) 4.8 Hz, 2H), 3.56 (m, 2H), 4.37 (t, J ) 4.8 Hz, 2H), 6.57 (d, J ) 8.2 Hz, 1H), 6.68 (dd, J ) 8.2, 2.2 Hz, 1H), 6.78 (d, J ) 2.2 Hz, 1H), 6.86 (dd, J ) 8.8, 2.3 Hz, 1H), 6.91 (d, J ) 8.9 Hz, 2H), 7.25 (d, J ) 2.3 Hz, 1H), 7.43 (d, J ) 8.8 Hz, 1H), 7.73 (d, J ) 8.9 Hz, 2H). Positive ion electrospray HRMS m/z calculated for C28H28NO5S, 490.1688; found, 490.1715. Preparation of Raloxifene Diquinone Methide and 6,7o-Quinone. Activated silver oxide (26) 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, and the precipitate was washed with water, filtered, and dried. Raloxifene diquinone methide was prepared by silver oxide oxidation. Briefly, raloxifene (5 mg) was dissolved in anhydrous acetonitrile (5 mL) and was preheated to 60 °C. Activated silver oxide (500 mg) was added to this mixture and stirred for 10 s. The mixture was then filtered and either subjected to UV analysis or added to rapidly stirred GSH solution as described below. Raloxifene 6,7-o-quinone was prepared from 7-OHRA using a similar procedure except that the reaction mixture was stirred at room temperature for 15 min. UV (acetonitrile) 290, 390 nm. Electrospray MS m/z 488 (100%) [M + H]+. Kinetic Studies of Raloxifene Diquinone Methide, 7OHRA, and Raloxifene 6,7-o-Quinone. Freshly prepared raloxifene diquinone methide was added to phosphate buffer (pH 7.4) at 5 °C. The decrease in the UV absorbance at 556 nm was monitored at a rate of 0.1 s/scan. Because the reaction was too fast to be monitored by this method, an accurate rate measurement was not obtained; however, the half-life of the raloxifene diquinone methide was estimated to be less than 1 s. The rate of formation of raloxifene 6,7-o-quinone was determined by adding 7-OHRA (25 µL, 2 mM in methanol) to phosphate buffer (975 µL, pH 7.4, 37 °C). The increase in the UV absorbance at 390 nm was monitored at a rate of 2 s/scan. The reactivity of raloxifene 6,7-o-quinone was measured by adding freshly prepared 6,7-o-quinone (25 µL, 2 mM in acetonitrile) to phosphate buffer (975 µL, pH 7.4, 37 °C). The decrease in the UV absorbance at 390 nm was monitored at a rate of 2 min/scan. Reaction of Raloxifene Diquinone Methide and 6,7-oQuinone with GSH. A solution of raloxifene 6,7-o-quinone (2 mM) in acetonitrile (250 µL) was combined with GSH (0.5 mM) in phosphate buffer (10 mL, 50 mM, pH 7.4), and the mixture was stirred at room temperature for 5 min. Perchloric acid (0.5 mL) was added, and an aliquot (10 µL) of the mixture was analyzed using LC-MS-MS. The remaining solution was concentrated under reduced pressure and purified using semipreparative HPLC. A similar procedure was used for the reaction of raloxifene diquinone methide with GSH. Cytotoxicity Studies in S30 and MDA-MB-231 Cells. The trypan blue exclusion assay was conducted to determine cell viability (27). MDA-MB-231 cells were maintained in MEME supplemented with 5% fetal bovine serum (Atalanta Biologicals, Norcross, GA), 1% glutamax, 1% penicillin, streptomycin, fungizome, 1% nonessential amino acids, 6 µg/mL insulin (Sigma), and 5% CO2 at 37 °C. The S30 cell line was maintained in the same medium as MDA-MB-231 cells supplemented with 5% 3× charcoal-dextran-treated fetal bovine serum and 500 µg/mL G418. All media products were purchased from Invitrogen (Carlsbad, CA) unless stated otherwise. The medium was routinely changed every 3 or 4 days, and it was changed 24 h prior to each experiment to maintain logarithmic growth. The cells were treated with test compound and assayed at 18 h. The test compounds were added in DMSO to give final concentrations from 5 to 100 µM. Negative controls (cells treated with


Figure 2. Positive ion electrospray LC-MS chromatograms of GSH conjugates in rat liver microsomal incubations of (A) 3′OHRA and (B) 7-OHRA. The [M + 2H]2+ ions of the di-GSH conjugates were monitored at m/z 550.5, and the [M + H]+ ions of the mono-GSH conjugates were monitored at m/z 795. 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 data represent the means ( SD of three determinations.

Results Incubations of Raloxifene and Raloxifene Catechols with Liver Microsomes and GSH. We previously reported the detection of raloxifene metabolites in rat liver microsomes by using LC-MS-MS, which was consistent with raloxifene catechols and their o-quinone GSH conjugates (28). Similarly, Chen et al. (24) found three mono-GSH conjugates of raloxifene in incubations of raloxifene with human liver microsomes. In the current study, we confirmed the metabolism of raloxifene to catechols and o-quinones by chemical synthesis of three of the potential raloxifene catechols: 7-OHRA, 3′-OHRA, and 5-OHRA (Scheme 1). These catechols were fully characterized by using both two-dimensional NMR including HMBC, COSY, HMQC, and mass spectrometry. Co-injections during LC-MS of these three raloxifene catechols with raloxifene incubated with microsomes indicated that both 7-OHRA and 3′-OHRA were formed whereas no 5-OHRA was detected. GSH conjugates formed in incubations of 7-OHRA with rat liver microsomes in the presence of GSH were identified as 7-OHRA-di-SG and 7-OHRA-SG based on the detection of [M + 2H]2+ at m/z 551 at 14.5 min and [M + H]+ at m/z 795 at 24.5 min (Figure 2). MS-MS product ion scanning with CID of the protonated molecule of 7-OHRASG at m/z 795 produced the fragment ions at m/z 720 and 666 corresponding to the loss of glycine and pyroglutamate moieties, respectively (Figure 3), which are characteristic of GSH conjugates (29). The product ion of m/z 522 was derived from the cleavage of the thioether moiety, and the ion of m/z 112 corresponded to a vinyl-

Chem. Res. Toxicol., Vol. 17, No. 7, 2004 883

piperidine ion. Loss of water, [MH - H2O]+, produced the fragment ion of m/z 777. MS-MS product ion scanning with CID of the analysis of the 7-OHRA-di-SG ion of m/z 551 (eluting at 14.5 min, Figure 2) produced ions of m/z 513 and 477 corresponding to doubly charged ions derived from the loss of one and two glycines, respectively (Figure 4). The fragment ions at m/z 486 and 422 were doubly charged ions derived from the loss of one and two pyroglutamates moieties. When one glycine and one pyroglutamate were lost simultaneously, the ion at m/z 448 was derived. The doubly charged ion of m/z 542 corresponded to a loss of water. Incubation of 3′-OHRA with rat liver microsomes in the presence of GSH produced two mono-GSH conjugates with retention times of 23.1 and 25.7 min and one diGSH conjugate with a retention time 14.2 min in the LCMS chromatogram shown in Figure 2A, which were tentatively identified as 3′-OHRA-SG-1, 3′-OHRA-SG-2, and 3′-OHRA-diSG, respectively. During positive ion electrospray LC-MS, the mono-GSH conjugates produced protonated molecules of m/z 795, and the di-GSH conjugate produced [M + 2H]2+ of m/z 551. The fragmentation patterns of these GSH conjugates (Figure 3) were similar to those of 7-OHRA GSH conjugates. Comparison of these GSH conjugates with those formed in microsomal incubations of raloxifene showed that the two raloxifene catechol di-GSH conjugates that we reported previously (28) were 3′-OHRA-di-SG and 7-OHRA-di-SG, and the three raloxifene catechol mono-GSH conjugates were 3′OHRA-SG-1, 7-OHRA-SG, and 3′-OHRA-SG-2. The two major raloxifene catechol metabolites, 7OHRA and 3′-OHRA, as well as one catechol mono-GSH conjugate, 7-OHRA-SG, were also detected in human liver microsomal incubations. In the human liver microsomal incubations of 7-OHRA and 3′-OHRA, the three catechol mono-GSH conjugates that we described previously were also detected. The catechol di-GSH conjugates were formed in minimal amounts or were not detected in the incubations of 3′-OHRA. Kinetic Studies of Raloxifene Diquinone Methide, 7-OHRA, and Raloxifene 6,7-o-Quinone. Activated silver oxide is used to oxidize phenols and catechols to quinone methides and o-quinones, respectively (19) and was chosen in this study as the oxidant for the generation of raloxifene diquinone methide and 6,7-o-quinone. UV spectra of oxidized raloxifene in anhydrous acetonitrile showed absorbance maxima at 440 and 556 nm and that of raloxifene 6,7-o-quinone showed an absorbance maximum at 390 nm. Furthermore, the positive ion electrospray mass spectrum of raloxifene 6,7o-quinone showed a protonated molecule of m/z 488. Because this ion was two mass units less than that of 7-OHRA, this provided additional evidence of o-quinone formation. The addition of raloxifene diquinone methide to phosphate buffer at 37 °C resulted in the immediate disappearance of the UV chromophore. Although the reaction was repeated at 5 °C, it was still too fast to be monitored by conventional UV spectroscopy. As a result, an accurate reaction rate was not obtained; however, the half-life could be conservatively estimated to be less than 1 s at 5 °C and would be expected to be much shorter under physiological conditions. In contrast to the raloxifene diquinone methide, the rate constant for the disappearance of raloxifene 6,7-o-quinone in phosphate buffer at 37 °C was determined to be kobs ) 1.7 ( 0.1 × 10-4 s-1,

884


Yu et al.

Figure 3. Positive ion electrospray product ion tandem mass spectrum with CID of the protonated molecule of 5-SG-7-OHRA eluting at 24.5 min in Figure 2B.

Figure 4. Positive ion electrospray product ion tandem mass spectrum with CID of the diprotonated molecule of 4,5-di-SG-7OHRA eluting at 14.5 min in Figure 2B.

corresponding to a half-life of approximately 70 min. 7-OHRA was found to autoxidize to raloxifene 6,7-oquinone in phosphate buffer (pH 7.4, 37 °C), and its autoxidation rate was calculated to be kobs ) 2.7 ( 0.2 × 10-3 s-1, corresponding to a half-life of 4 min. Reaction of Raloxifene Diquinone Methide and 6,7-o-Quinone with GSH. The reaction of raloxifene diquinone methide with GSH produced three major products (data not shown), which were identical to the three raloxifene mono-GSH conjugates detected in the rat liver microsomal incubations. These conjugates appeared to be identical to those characterized by Chen et al. (24) in human liver microsomal incubations. The reaction of raloxifene 6,7-o-quinone with GSH produced one di-GSH conjugate, 7-OHRA-di-SG, and one monoGSH conjugate, 7-OHRA-SG (Figure 5). These two products were isolated using HPLC and analyzed using 1H NMR. The partial NMR data and that of 7-OHRA are

Figure 5. Positive ion ESI LC/MS detection of GSH conjugates of 7-OHRA in the reaction of raloxifene 6,7-o-quinone and GSH. The [M + 2H]2+ ions of the di-GSH conjugates were monitored at m/z 550.5, and the protonated molecules of the mono-GSH conjugates were monitored at m/z 795.


Chem. Res. Toxicol., Vol. 17, No. 7, 2004 885 Table 1. 1H NMR of 7-OHRA and Its GSH Conjugates chemical shifts (ppm), multiplicity, and integral, DMSO-d6

compounds 7-OHRA 7-OHRA-di-SG 7-OHRA-SG

H4 6.30, d, 1H

H5

H2′6′

H3′5′

H2′′6′′

H3′′5′′

6.46, d, 1H

6.73, d, 2H 7.34, m, 2H 7.33, d, 2H

6.24, d, 2H 6.79, d, 2H 6.78, d, 2H

7.24, d, 2H 7.70, m, 2H 7.69, d, 2H

6.52, d, 2H 7.01, m, 2H 7.02, d, 2H

7.26, 7.18, 1H

Table 2. Cytotoxicity of Raloxifene and Its Catechol Metabolites in Breast Cancer Cell Linesa LC50 (µM) substrate

MDA-MB-231

S30

raloxifene 7-OHRA 3′-OHRA

57 ( 1 35 ( 2 >100

55 ( 1 24 ( 1 >100

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 the Materials and Methods. Results represent the mean ( SD of at least three determinations.

summarized in Table 1. In the spectrum of 7-OHRA, the two doublet signals at 6.73 and 6.24 ppm were assigned to H-2′6′ and H-3′5′, and the two doublets at 7.24 and 6.52 ppm were assigned to H-2′′6′′ and H-3′′5′′. The two doublets, each integrating to one proton, at 6.30 and 6.46 ppm, were assigned to H-4 and H-5. In both GSH conjugates, H-2′6′, H-3′5′, H-2′′6′′, and H-3′′5′′ remained unaffected with the exception of a slight downfield shift; thus, all substitutions were determined to occur in the benzothiophene moiety. In 7-OHRA-di-SG, both H-4 and H-5 disappeared suggesting that GSH substitution had occurred at these positions giving 4,5-di-SG-7-OHRA. On the basis of NMR analysis, the identity of 7-OHRA-SG is suggested to be the 5-isomer, 5-SG-7-OHRA. Cytotoxicity Studies in MDA-MB-231 and S30 Cells. Toxicities of raloxifene and its catechol metabolites, 7-OHRA and 3′-OHRA, in MDA-MB-231 and S30 cell lines are summarized in Table 2. 7-OHRA was shown to be toxic in both cell lines and somewhat more toxic in S30 cells (IC50 ) 24 µM) than in MDA-MB-231 cells (IC50 ) 35 µM). Raloxifene was less toxic than 7-OHRA and did not show a significant difference between the two cell lines. Surprisingly, 3′-OHRA was not toxic in either cell line up to the solubility limit.

Discussion Tamoxifen, an antiestrogen used in the treatment and prevention of breast cancer, is known to increase the risk of endometrial cancer, and the formation of quinoids has been postulated to play an important role (28). The quinones and quinone methides formed from the metabolic activation of antiestrogens can cause damage in cells through alkylation of cellular macromolecules and through generation of reactive oxygen species (28). Raloxifene is in clinical use for the treatment and prevention of osteoporosis in postmenopausal women, and it is currently in clinical trials for the chemoprevention of breast cancer. The current clinical use of raloxifene and, more importantly, the potential future long-term use in chemoprevention and treatment of postmenopausal symptoms raise concerns. Raloxifene is a second generation SERM that retains similar structural motifs to 4-hydroxytamoxifen, the active metabolite of tamoxifen (20). In particular, the polyaromatic phenolic core (Figure 1) of raloxifene should

render this molecule redox reactive. We have shown that raloxifene autoxidizes in the presence of NO, and reaction with the biological oxidant peroxynitrite yields nitroaromatic and quinoid products that can form conjugates with GSH (30). Most of the earlier pharmacokinetic studies on raloxifene suggested that it was primarily metabolized by glucuronidation and not by P450 oxidation (31-33). However, a few published reports as well as the current study indicate otherwise (24, 34). For example, mono- and dihydroxylated raloxifene metabolites were detected by using LC-MS-MS in incubations of raloxifene with rat liver microsomes (34), although these metabolites were not characterized. GSH and N-acetylcysteine conjugates were identified (24) during raloxifene metabolic studies, which were possibly formed through the intermediacy of an electrophilic raloxifene diquinone methide (Figure 1). In addition, the drug was shown to be associated with decreases in P450 aromatase activities in human colon carcinoma cells (35) and importantly was shown to irreversibly inhibit P450 3A4 (24). These data suggest that raloxifene might participate in similar toxic pathways to those proposed for tamoxifen. In a recent study, raloxifene mono-GSH conjugates were identified during raloxifene metabolism (24). In addition to these raloxifene mono-GSH conjugates, we provided a preliminary report of the formation of raloxifene catechols and catechol GSH conjugates (28). Such evidence suggests that raloxifene might cause toxicity through the formation of both quinones and a diquinone methide. Chen et al. (24) proposed two possible mechanisms for the formation of raloxifene mono-GSH conjugates either through a raloxifene arene oxide intermediate or a raloxifene diquinone methide, followed by nucleophilic attack of GSH. In this study, we have successfully generated the raloxifene diquinone methide by chemical oxidation using silver oxide to study the mechanism of raloxifene GSH conjugate formation. When this diquinone methide was reacted with GSH, three GSH conjugates were obtained, which were identical to the raloxifene mono-GSH conjugates detected in microsomal incubations. These results provide strong support for the raloxifene diquinone methide pathway in the formation of raloxifene GSH conjugates. Kinetic studies showed that the raloxifene diquinone methide was highly reactive in phosphate buffer with a half-life of less than 1 s at 5 °C, which would be expected to be much shorter under physiological conditions. According to the relationship between toxicity and reactivity of quinone methides (36), if a quinone methide is too reactive, it is less toxic due to its poor selectivity between nucleophilic cellular macromolecules and solvent or GSH. The high reactivity of the raloxifene diquinone methide makes it less likely to cause toxicity by alkylation of cellular macromolecules; therefore, this metabolite may not make a major contribution to the overall toxicity of raloxifene. In addition to the three raloxifene mono-GSH conjugates, we detected three hydroxylated raloxifenes, as well as two di-GSH and three mono-GSH conjugates of

886


Yu et al.

Scheme 2. Mechanism of the Formation of 7-OHRA GSH Conjugates

hydroxylated raloxifenes. To identify these metabolites, three possible raloxifene catechols were synthesized as standards. Two of the three raloxifene catechols, 7-OHRA and 3-OHRA, are novel structures. Although 5-OHRA was first synthesized by Grese et al. (37), we used a different synthetic procedure. The synthesis of differently substituted benzo[b]thiophene moieties was performed by the cyclization rearrangement induced by PPA, which has been described in the literature by Kost et al. (38). Using the three raloxifene catechols as standards, two of the three hydroxylated raloxifenes detected in the microsomal incubations were determined to be 7-OHRA and 3′-OHRA; however, no 5-OHRA was detected. The two catechols were then incubated with rat liver microsomes in the presence of GSH. Microsomal incubations of 7-OHRA gave one di-GSH conjugate, 7-OHRA-di-SG, and one mono-GSH conjugate corresponding to 7-OHRASG. Similar incubations with 3′-OHRA gave one di-GSH conjugate, 3′-OHRA-di-SG, and two mono-GSH conjugates, 3′-OHRA-SG-1 and 3′-OHRA-SG-2. These GSH conjugates were identical to those formed in the incubations of raloxifene with rat liver microsomes as determined by comparison with mass spectra and co-injection experiments with authentic standards. The raloxifene catechol GSH conjugates were probably formed through the initial formation of the raloxifene o-quinones followed by reaction with GSH. To demonstrate this proposed mechanism, 7-OHRA was used as an example and oxidized to raloxifene 6,7-o-quinone using silver oxide oxidation. Reaction of this o-quinone with GSH produced

one di-GSH conjugate, 7-OHRA-di-SG, and one monoGSH conjugate, 7-OHRA-SG. On the basis of the subsequent proton NMR analysis, they were determined to be 4,5-di-SG-7-OHRA and 5-SG-7-OHRA, respectively. The potential mechanism of the formation of the o-quinone GSH conjugates is shown in Scheme 2. After oxidation of raloxifene to 7-OHRA, the catechol could be further oxidized to raloxifene 6,7-o-quinone either by P450 or through autoxidation, followed by nucleophilic attack by GSH. Further oxidation of 5-SG-7-OHRA followed by the attack of another GSH molecule would yield 4,5-di-SG7-OHRA. A similar mechanism can be proposed for 3′OHRA. Kinetic studies showed that the half-life of the raloxifene 6,7-o-quinone is 69 ( 2.5 min. As compared with the highly reactive diquinone methide, raloxifene 6,7-oquinone is relatively stable. The stability offered by this o-quinone implies that it may be more toxic than the raloxifene quinone methide. To investigate the relevance of the results obtained in rat liver microsomal incubations to humans, human liver microsomal incubations with raloxifene were also carried out (data not shown). In addition to the three raloxifene diquinone methide monoGSH conjugates, the GSH conjugates of 7-OHRA and 3′OHRA were also detected. We examined the relative toxicity of raloxifene and its two major catechol metabolites, 7-OHRA and 3′-OHRA, in two breast cancer cell lines, S30 (ERR-positive) and MDA-MB-231 (ERR-negative). The MDA-MB-231 cell line is an ER-negative cell line, and the S30 cell line is an ER-positive cell line obtained from the MDA-MD-231 cell


line stably expressing ERR. We have previously shown that ERR positive cell lines are considerably more sensitive to the toxic effects of equine catechol estrogen metabolites (39), and as a result, it was of interest to determine if similar effects were observed with raloxifene and its catechol metabolites. Our data showed that 7-OHRA was more toxic than raloxifene and that the S30 cells were more sensitive to the toxicity. In summary, raloxifene was metabolized by both rat and human liver microsomes to electrophilic diquinone methide and o-quinones, which were trapped by GSH to form mono- and/or di-GSH conjugates. These electrophilic species have the potential to alkylate cellular macromolecules to cause toxicity. Among these electrophilic species, raloxifene diquinone methide seemed to be too reactive to be a major toxic metabolite while raloxifene 6,7-oquinone was relatively stable and could be the major alkylating species. These findings might be very important to evaluate the potential toxicity of raloxifene and might be useful guides for further investigations toward the safe long-term use of raloxifene.

Acknowledgment. This work was supported by NIH Grants CA79870 (J.L.B.) and CA83124 (R.B.vB). We thank Dr. Xiangjun Yue for advice on the synthesis of raloxifene catechols. Supporting Information Available: Chromatograms of 3-OHRA, 5-OHRA, and 7-OHRA in MeOD and standard C13cpd (220 ppm, 360 MHz) chromatograms for 3-OHRA, 5-OHRA, and 7-OHRA 13C in MeOD. This material is available free of charge via the Internet at http://pubs.acs.org.

References (1) Magriples, U., Naftolin, F., Schwartz, P. E., and Carcangiu, M. L. (1993) High-grade endometrial carcinoma in tamoxifen-treated breast cancer patients. J. Clin. Oncol. 11, 485-490. (2) Seoud, M. A., Johnson, J., and Weed, J. C., Jr. (1993) Gynecologic tumors in tamoxifen-treated women with breast cancer. Obstet. Gynecol. 82, 165-169. (3) Phillips, D. H. (2001) Understanding the genotoxicity of tamoxifen? Carcinogenesis 22, 839-849. (4) Fisher, B., Constantino, J. P., Redmond, C. K., Fisher, E. R., Wickerham, D. L., and Cronin, W. M. (1994) Endometrial cancer in tamoxifen-treated breast cancer patients: findings from the National Surgical Adjuvant Breast and Bowel project (NSABP) B-14. J. Natl. Cancer Inst. 86, 527-537. (5) van Leeuwen, F. E., Benraadt, J., Coebergh, J. W., Kiemeney, L. A., Gimbrere, C. H., Otter, R., Schouten, L. J., Damhuis, R. A., Bontenbal, M., Diepenhorst, F. W., van den Belt-Dusebout, A. W., and van Tinteren, H. (1994) Risk of endometrial cancer after tamoxifen treatment of breast cancer. Lancet 343, 448-452. (6) Cook, L. S., Weiss, N. S., Schwartz, S. M., White, E., McKnight, B., Moore, D. E., and Daling, J. R. (1995) Population-based study of tamoxifen therapy and subsequent ovarian, endometrial, and breast cancers. J. Natl. Cancer Inst. 87, 1359-1364. (7) Cuenca, R. E., Giachino, J., Arredondo, M. A., Hempling, R., and Edge, S. B. (1996) Endometrial carcinoma associated with breast carcinoma: low incidence with tamoxifen use. Cancer 77, 20582063. (8) Curtis, R. E., Boice, J. D., Jr., Shriner, D. A., Hankey, B. F., and Fraumeni, J. F., Jr. (1996) Second cancers after adjuvant tamoxifen therapy for breast cancer. J. Natl. Cancer Inst. 88, 832834. (9) White, I. N. H., 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 Fisher F344/n rats, DBA/2 and C57BL/6 and in human MCL-5 cells. Carcinogenesis 13, 2197-2203. (10) Han, X. L., and Liehr, J. G. (1992) Induction of covalent DNA adducts in rodents by tamoxifen. Cancer Res. 52, 1360-1363.

Chem. Res. Toxicol., Vol. 17, No. 7, 2004 887 (11) Randerath, K., Moorthy, B., Mabon, N., and Sriram, P. (1994) Tamoxifen: evidence by 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 proximate metabolite. Carcinogenesis 15, 2087-2094. (12) Hardcastle, I. R., Horton, M. N., Osborne, M. R., Hewer, A., Jarman, M., and Phillips, D. H. (1998) Synthesis and DNA reactivity of R-hydroxylated metabolites of nonsteroidal antiestrogens. Chem. Res. Toxicol. 11, 369-374. (13) Shibutani, S., Dasaradhi, L., Terashima, I., Banoglu, E., and Duffel, M. W. (1998) a-Hydroxytamoxifen is a substrate of hydroxysteroid (alcohol) sulfotransferase, resulting in tamoxifen DNA adducts. Cancer. Res. 58, 647-653. (14) Dasaradhi, L., and Shibutani, S. (1997) Identification of tamoxifen-DNA adducts formed by R-sulfate tamoxifen and R-acetoxytamoxifen. Chem. Res. Toxicol. 10, 189-196. (15) 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. (16) Dehal, S. S., and Kupfer, D. (1999) Cytochrome P450 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. (17) Beland, F. A., McDaniel, L. P., and Marques, M. M. (1999) Comparison of the DNA adducts formed by tamoxifen and 4-hydroxytamoxifen in vivo. Carcinogenesis 20, 471-477. (18) Osborne, M. R., Davis, W., Hewer, A. J., Hardcastle, I. R., and Phillips, D. H. (1999) 4-Hydroxytamoxifen gives DNA adducts by chemical activation, but not in rat liver cells. Chem. Res. Toxicol. 12, 151-158. (19) Marques, M. M., and Beland, F. A. (1997) Identification of tamoxifen-DNA adducts by 4-hydroxytamoxifen quinone methide. Carcinogenesis 18, 1949-1954. (20) Jones, C. D., Jevnikar, M. G., Pike, A. J., Peters, M. K., Black, L. J., Thompson, A. R., Falcone, J. F., and Clemens, J. A. (1984) Antiestrogens. 2. Structure-activity studies in a series of 3-aroyl2-arylbenzo[b]thiophene derivatives leading to [6-hydroxy-2-(4hydroxyphenyl)benzo[b]thien-3-yl] [4-[2-(1-piperidinyl)ethoxy]phenyl]methanone hydrochloride (LY156758), a remarkably effective estrogen antagonist with only minimal intrinsic estrogenicity. J. Med. Chem. 27, 1057-1066. (21) Wakeling, A. E., Valcaccia, B., Newboult, E., and Green, L. R. (1984) Nonsteroidal antioestrogenssreceptor binding and biological response in rat uterus, rat mammary carcinoma and human breast cancer cells. J. Steroid Biochem. 20, 111-120. (22) Clemens, J. A., Bennett, D. R., Black, L. J., and Jones, C. D. (1983) Effects of a new antiestrogen, keoxifene (LY156758), on growth of carcinogen-induced mammary tumors and on LH and prolactin levels. Life Sci. 32, 2869-2875. (23) Kelminski, A. (2002) The Study of Tamoxifen and Raloxifene (STAR trial) for the prevention of breast cancer. Hawaii Med. J. 61, 209-210. (24) Chen, Q., Ngui, J. S., Doss, G. A., Wang, R. W., Cai, X., DiNinno, F. P., Blizzard, T. A., Hammond, M. L., Stearns, R. A., Evans, D. C., Baillie, T. A., and Tang, W. (2002) Cytochrome P450 3A4mediated bioactivation of raloxifene: irreversible enzyme inhibition and thiol adduct formation. Chem. Res. Toxicol. 15, 907914. (25) 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. (26) 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. (27) Freshney, R. (1987) Culture of Animal Cells: A Manual of Basic Techniques, p 117, Alan R. Liss, Inc., New York. (28) Bolton, J. L., Yu, L., and Thatcher, G. R. J. (2004) Quinoids formed from estrogens and antiestrogens. Methods Enzymol. 378, 110123. (29) Baillie, T. A., and Davis, M. R. (1993) Mass spectrometry in the analysis of glutathione conjugates. Biol. Mass Spectrom. 22, 319325. (30) Toader, V., Xu, X., Nicolescu, A., Yu, L., Bolton, J. L., and Thatcher, G. R. (2003) Nitrosation, nitration, and autoxidation of the selective estrogen receptor modulator raloxifene by nitric oxide, peroxynitrite, and reactive nitrogen/oxygen species. Chem. Res. Toxicol. 16, 1264-1276.

888


(31) Morello, K. C., Wurz, G. T., and DeGregorio, M. W. (2003) Pharmacokinetics of selective estrogen receptor modulators. Clin. Pharmacokinet. 42, 361-372. (32) Snyder, K. R., Sparano, N., and Malinowski, J. M. (2000) Raloxifene hydrochloride. Am. J. Health-Syst. Pharm. 57, 16691675. (33) Hochner-Celnikier, D. (1999) Pharmacokinetics of raloxifene and its clinical application. Eur. J. Obstet. Gynecol. Reprod. Biol. 85, 23-29. (34) Lim, H. K., Stellingweif, S., Sisenwine, S., and Chan, K. W. (1999) Rapid drug metabolite profiling using fast liquid chromatography, automated multiple-stage mass spectrometry and receptor-binding. J. Chromatogr. A 831, 227-241. (35) Fiorelli, G., Picariello, L., Martineti, V., Tonelli, F., and Brandi, M. L. (1999) Estrogen synthesis in human colon cancer epithelial cells. J. Steroid Biochem. Mol. Biol. 71, 223-230. (36) Thompson, D. C., Perera, K., Krol, E. S., and Bolton, J. L. (1995) o-Methoxy-4-alkylphenols that form quinone methides of

Yu et al. intermediate reactivity are the most toxic in rat liver slices. Chem. Res. Toxicol. 8, 323-327. (37) Grese, T. A., Cho, S., Finley, D. R., Godfrey, A. G., Jones, C. D., Lugar, C. W., 3rd, Martin, M. J., Matsumoto, K., Pennington, L. D., Winter, M. A., Adrian, M. D., Cole, H. W., Magee, D. E., Phillips, D. L., Rowley, E. R., Short, L. L., Glasebrook, A. L., and Bryant, H. U. (1997) Structure-activity relationships of selective estrogen receptor modulators: modifications to the 2-arylbenzothiophene core of raloxifene. J. Med. Chem. 40, 146-167. (38) Kost, A. N., Budylin, V. A., Matveeva, E. D., and Sterligov, D. O. (1970) Isomerization of 3-substituted indoles, benzofurans, and benzo[b]thiophenes. Zh. Org. Khim. 6, 1516-1517. (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.

TX0342722

Oxidation of Raloxifene to Quinoids: Potential Toxic Pathways via a

Recommend Documents