Identification of Quinone Imine Containing ... - ACS Publications

Nov 5, 2010 - Protein adducts have been detected in rats treated with diclofenac (9, 10) ...... Anthony P. Dickie , Claire E. Wilson , Kay Schreiter ,...
0 downloads 0 Views 264KB Size
Download PDF
Chem. Res. Toxicol. 2010, 23, 1947–1953

1947

Identification of Quinone Imine Containing Glutathione Conjugates of Diclofenac in Rat Bile Daniel J. Waldon,† Yohannes Teffera,† Adria E. Colletti,† Jingzhou Liu,† Danielle Zurcher,‡ Katrina W. Copeland,‡ and Zhiyang Zhao*,† Pharmacokinetics and Drug Metabolism and Medicinal Chemistry, Amgen, Inc., 360 Binney Street, Cambridge, Massachusetts 02142, United States ReceiVed August 31, 2010

High-resolution accurate MS with an LTQ-Orbitrap was used to identify quinone imine metabolites derived from the 5-hydroxy (5-OH) and 4 prime-hydroxy (4′-OH) glutathione conjugates of diclofenac in rat bile. The initial quinone imine metabolites formed by oxidation of diclofenac have been postulated to be reactive intermediates potentially involved in diclofenac-mediated hepatotoxicity; while these metabolites could be formed using in vitro systems, they have never been detected in vivo. This report describes the identification of secondary quinone imine metabolites derived from 5-OH and 4′-OH diclofenac glutathione conjugates in rat bile. To verify the proposed structures, the diclofenac quinone imine GSH conjugate standards were prepared synthetically and enzymatically. The novel metabolite peaks displayed the identical retention times, accurate mass MS/MS spectra, and the fragmentation patterns as the corresponding authentic standards. The formation of these secondary quinone metabolites occurs only under conditions where bile salt homeostasis was experimentally altered. Standard practice in biliary excretion experiments using bile duct-cannulated rats includes infusion of taurocholic acid and/or other bile acids to replace those lost due to continuous collection of bile; for this experiment, the rats received no replacement bile acid infusion. High-resolution accurate mass spectrometry data and comparison with chemically and enzymatically prepared quinone imines of diclofenac glutathione conjugates support the identification of these metabolites. A mechanism for the formation of these reactive quinone imine containing glutathione conjugates of diclofenac is proposed. Introduction Diclofenac (Figure 1) is a nonsteroidal anti-inflammatory drug that is widely prescribed for the treatment of osteoarthritis, rheumatoid arthritis, ankylosing spondylitis, and acute muscle pain conditions (1). Treatment with diclofenac has been associated with a rare, but severe, cases of hepatic injury (2, 3), which is often described as idiosyncratic toxicity. Although the mechanism of diclofenac hepatotoxicity is not yet completely understood, evidence supports the hypothesis that the toxicity is initiated by the formation of reactive metabolites. Subsequent covalent binding to protein may be involved in the activation of an immune response. Reactive metabolites of diclofenac, such as diclofenac-1-O-acyl glucuronide, have been shown to bind covalently to cellular proteins via glycation and/or transacylation (4, 5). Another reactive metabolite of diclofenac, a quinone imine intermediate generated via oxidation by cytochrome P450s, also has been associated with the formation of protein adducts (6, 7). While this quinone imine has been postulated to be involved in diclofenac toxicity, it has never been detected in vivo and only identified indirectly from in vitro methods (8). Protein adducts have been detected in rats treated with diclofenac (9, 10); however, a direct link between covalent modifications and diclofenac hepatotoxicity has not been clearly established (11, 12). Herein, we report the identification of quinone imine * To whom correspondence should be addressed. Tel: 617-444-5171. Fax: 617-444-9913. E-mail: [email protected]. † Pharmacokinetics and Drug Metabolism. ‡ Medicinal Chemistry.

metabolites derived from diclofenac glutathione (GSH)1 conjugates but only in rat bile under conditions in which bile acid homeostasis was altered.

Materials and Methods Chemicals and Reagents. Diclofenac sodium was purchased from Sigma-Aldrich (St. Louis, MO). The 5-OH standard was purchased from Toronto Research Chemicals, Inc. (North York, ON, Canada), and the 4′-OH diclofenac standard was purchased from EMD Chemicals, Inc. (Gibbstown, New Jersey). Dichloro 1,1′bis(diphenylphosphino)ferrocene palladium, silver(II) oxide was purchased from Strem Chemicals Inc. (Newburyport, MA). Bis(pinacolato)diboron, potassium acetate, hydrogen peroxide, and GSH were purchased from Sigma-Aldrich. N,N′-Bis(salicylidene)ethylenediaminecobalt(II) was purchased from Alfa Aesar (Ward Hill, MA). Phosphoric buffer was purchased from Molecular Toxicology, Inc. (Boone, NC). All solvents used for high-pressure liquid chromatography (HPLC) and LC-MS analyses were of chromatographic grade (Burdick and Jackson, Muskegon, MI, or SigmaAldrich). Synthesis of Diclofenac Metabolites. Synthetic routes for all of the diclofenac metabolites are shown in Figure 1 and are described below. Synthesis of 5-Hydroxy-4-(GSH-S-yl)diclofenac (VI) and 5-Hydroxy-6-(GSH-S-yl)diclofenac (V). 5-Iododiclofenac (I) (1.00 g, 2.37 mmol), dichloro 1,1′-bis(diphenylphosphino)ferrocene pal1 Abbreviations: ACN, acetonitrile; DCM, dichloromethane; DMSO, dimethyl sulfoxide; EtOAc, ethyl acetate; GSH, glutathione; HPLC, highpressure liquid chromatography; HRP, horseradish peroxidase; LTQ, linear trap; MeOH, methanol; MPLC, medium-pressure liquid chromatography; TFA, trifluoroacetic acid.

10.1021/tx100296v  2010 American Chemical Society Published on Web 11/05/2010

1948

Chem. Res. Toxicol., Vol. 23, No. 12, 2010

Waldon et al.

Figure 1. Synthesis of diclofenac metabolites. (A) Synthesis of 5-hydroxy diclofenac and GSH conjugates. (B) Synthesis of 4′-hydroxy diclofenac and GSH conjugates.

ladium (II) (0.35 g, 0.47 mmol), bis(pinacolato)diboron (0.90 g, 3.55 mmol), and potassium acetate (0.70 g, 7.11 mmol) were dissolved under nitrogen in dioxane (1 M, 2.37 mL) in a 10 mL sealed glass tube. This mixture was heated to 85 °C and stirred for 16 h until the reaction was complete as measured by LC-MS. The material was cooled to ambient temperature and then acidified with HCl (1 M in diethyl ether, 7.1 mL) to pH ∼ 3. After it was concentrated directly onto silica, the mixture was purified by medium-pressure liquid chromatography (MPLC) and eluted with a gradient hexanes:ethyl acetate (EtOAc) 90:10 to 50:50 to afford 5-(tetramethyldioxaboron)diclofenac (II) (0.47 g, 47% yield). m/z (ESI, +eV ion) 422 (M + H)+. 1H NMR [400 MHz, dimethyl sulfoxide (DMSO)-d6] ppm: 1.28 (s, 12 H), 3.69 (s, 2 H), 6.18 (d, J ) 8.02 Hz, 1 H), 7.26 (dd, J ) 8.41, 7.83 Hz, 1 H), 7.35 (dd, J ) 8.02, 1.47 Hz, 1 H), 7.49 (d, J ) 1.37 Hz, 1 H), 7.50 (br. s., 1 H), 7.56 (d, J ) 8.12 Hz, 2 H), 12.52 (br. s., 1 H). A solution of 5-(tetramethyldioxaboron)diclofenac (II) (0.84 g, 1.99 mmol) in THF (1M, 2.0 mL) was cooled to 0 °C to which 30% w/w hydrogen peroxide (0.61 mL, 5.97 mmol) was added dropwise. After this addition, the solution was warmed to room temperature and stirred for 3 h. The resulting mixture was diluted with 40 mL of dichloromethane (DCM) and neutralized with 20% aqueous sodium thiosulfate (3 × 15 mL). The aqueous phase was then acidified with 1 M HCL to a pH of ∼3 and extracted with EtOAc (3 × 15 mL). The organic layers were dried over Na2SO4, filtered, and concentrated. This mixture was purified by MPLC, eluting with a gradient DCM:methanol (MeOH) 100:0 to 90:10 to afford 5-hydroxydiclofenac (III) (0.34 g, 55% yield). m/z (ESI, +eV ion): 312 (M + H)+. 1H NMR (400 MHz, MeOD) ppm: 3.74 (s, 2 H), 6.41 (d, J ) 8.61 Hz, 1 H), 6.59 (dd, J ) 8.56, 2.89 Hz, 1 H), 6.75 (d, J ) 2 0.64 Hz, 1 H), 6.97-7.03 (m, 1 H), 7.38 (d, J ) 8.02 Hz, 2 H).

5-Hydroxydiclofenac (III) (0.34 g, 1.09 mmol) was dissolved in THF (0.5 M, 2 mL) and N,N′-bis(salicylidene)ethylenediaminecobalt(II) (0.035 g, 0.11 mmol) was added to the solution. The reaction was stirred for 3 h at ambient temperature until complete as measured by LC-MS. The reaction material was concentrated directly onto silica and purified by MPLC with a gradient hexanes: EtOAc 90:10 to 65:35 to afford (E)-2-(6-(2,6-dichlorophenylimino)3-oxocyclohexa-1,4-dienyl)acetic acid (IV) (0.11 g, 33% yield). m/z (ESI, +eV ion) 310 (M + H)+. 1H NMR (400 MHz, MeOD) ppm: 3.76 (s, 2 H), 6.52 (dd, J ) 10.12, 2.20 Hz, 1 H), 6.70-6.76 (m, 2 H), 7.17 (dd, J ) 8.36, 7.87 Hz, 1 H), 7.47 (d, J ) 8.12 Hz, 2 H). (E)-2-(6-(2,6-Dichlorophenylimino)-3-oxocyclohexa-1,4-dienyl)acetic acid (IV) (0.05 g, 0.161 mmol) and GSH (0.05 g, 0.161 mmol) were dissolved in a two-phase reaction mixture of 0.2 M phosphoric buffer (pH 7.4) (4 mL) and MeOH (8 mL). The reaction mixture was then stirred for 4 h at 37 °C (12). The reaction was concentrated, and the two isomers were separated by HPLC (Gilson, Inc., Middleton, WI) [mobile phase, aqueous acetonitrile (ACN) containing 0.1% trifluoroacetic acid (TFA)] using a linear increase from 10 to 70% ACN during a 20 min period. Retention times for the two isoforms: 9.4 min for 5-hydroxy-4-(GSH-S-yl)diclofenac (VI) and 9.95 min for 5-hydroxy-6-(GSH-S-yl)diclofenac (V) (0.045 g (V), 0.006 g (VI), 51% yield). m/z (ESI, +eV ion): 617.1/618.9 (M + H)+ for both isomers. 1H NMR (400 MHz, MeOD) ppm: (V) 2.04-2.25 (m, 2 H), 2.43-2.56 (m, 2 H), 3.01 (dd, J ) 13.69, 8.61 Hz, 1 H), 3.21 (dd, J ) 13.64, 5.23 Hz, 1 H), 3.72 (s, 2 H), 3.79-3.93 (m, 2 H), 3.97-4.03 (m, 1 H), 4.44 (dd, J ) 8.56, 5.23 Hz, 1 H), 6.58 (s, 1 H), 6.82 (s, 1 H), 6.98-7.05 (m, 1 H), 7.38 (d, J ) 8.12 Hz, 2 H); (VI) 2.12 (q, J ) 6.48 Hz, 2 H), 2.46-2.63 (m, 2 H), 3.15-3.20 (m, 1 H), 3.38-3.44 (m, 1 H), 3.58-3.67 (m, 1 H), 3.81 (br s, 2 H), 4.22 (br s, 2 H), 4.41 (dd, J ) 8.58, 4.31 Hz,

Diclofenac Quinone Imine Glutathione Conjugates

Chem. Res. Toxicol., Vol. 23, No. 12, 2010 1949

Figure 2. Structures of diclofenac and metabolites showing a basic pathway for quinone imine formation and GSH conjugation. The bracketed compounds have not been identified but are presumed intermediates identified from in vitro studies. GSH addition is shown in only one possible position. The exact position of the GSH cannot be verified by the MS/MS spectra.

1 H), 6.45 (d, J ) 8.70 Hz, 1 H), 6.70 (d, J ) 8.70 Hz, 1 H), 6.97 (t, J ) 8.05 Hz, 1 H), 7.35 (d, J ) 8.09 Hz, 2 H). Synthesis of Corresponding Quinone Imines from the 5-OH 6 GSH (V) and 5-OH 4 GSH (VI) Conjugates. 5-Hydroxy4-(GSH-S-yl)diclofenac (VI) (0.045 g, 0.07 mmol) was dissolved in THF (1M, 0.7 mL), to which N,N′-bis(salicylidene)ethylenediaminecobalt(II) (0.003 g, 0.007 mmol) was added. This mixture was stirred at ambient temperature for 16 h until the reaction was complete as measured by LC-MS. Attempts to isolate the product (VII) directly by MPLC, HPLC, or with an aqueous workup were all unsuccessful. Alternatively, the larger isomer peak on HPLC, 5-hydroxy-6-(GSH-S-yl)diclofenac (V) (0.040 g, 0.064 mmol) was dissolved in THF (0.06M, 1 mL), and silver oxide (0.044 g, 0.19 mmol) was added to the solution. The mixture was stirred at ambient temperature for 16 h. Attempts to isolate the product (VIII) directly via MPLC, or with an aqueous acid/basic workup, or employing metal scavengers or ion-exchange columns were all unsuccessful. Synthesis of 4′-Hydroxy-3′(GSH-S-yl)diclofenac (XI). 4′Hydroxydiclofenac (IX) (0.05 g, 0.16 mmol) was dissolved in THF (0.64 mL), Ag2O (0.186 g, 0.80 mmol) was added to the solution (13), and the reaction was stirred for 3 h at room temperature. The reaction mixture was concentrated directly onto silica and purified by MPLC using a gradient of hexanes:EtOAc 90:10 to 0:100 to afford 2-(2-(2,6-dichloro-4-oxocyclohexa-2,5-dienylideneamino)phenyl)acetic acid (X) (0.012 g, 24%). m/z (ESI, +eV ion): 310 (M + H)+. 1H NMR (600 MHz, MeOD) ppm: 3.52 (s, 2 H), 6.64 (dd, J ) 7.74, 1.18 Hz, 1 H), 6.87 (s, 2 H), 7.17 (td, J ) 7.50, 1.26 Hz, 1 H), 7.23 (td, J ) 7.61, 1.41 Hz, 1 H), 7.37 (dd, J ) 7.70, 1.20 Hz, 1 H). 2-(2-(2,6-Dichloro-4-oxocyclohexa-2,5-dienylideneamino)phenyl)acetic acid (0.05 g, 0.23 mmol) (X) and GSH (1.5 g, 0.161 mmol) were dissolved in a two-phase reaction mixture of 0.2 M phosphoric buffer (pH 7.4, 4 mL) and MeOH (8 mL). The reaction mixture was then stirred for 4 h at 37 °C (12). The reaction was concentrated and purified by HPLC (Gilson, Inc.). The HPLC conditions used mobile phases of water and ACN both containing

0.1% TFA and a linear gradient from 10 to 70% ACN over a 20 min period. The retention time was 8.8 min for 4′-hydroxy-3′(GSHS-yl)diclofenac (XI) (0.027 g, 27% yield). m/z (ESI, +eV ion): 617 (M + H)+. 1H NMR (600 MHz, MeOD) ppm: 2.09-2.26 (m, 2 H), 2.51-2.60 (m, 2 H), 3.10 (dd, J ) 13.77, 8.96 Hz, 1 H), 3.45 (dd, J ) 13.77, 5.15 Hz, 1 H), 3.73 (s, 2 H), 3.88 (d, J ) 1.14 Hz, 2 H), 3.99 (t, J ) 6.41 Hz, 1 H), 4.40 (dd, J ) 8.89, 5.15 Hz, 1 H), 6.29 (dd, J ) 8.13, 0.95 Hz, 1 H), 6.83 (td, J ) 7.40, 1.07 Hz, 1 H), 7.04 (s, 1 H), 7.06 (td, J ) 7.40, 1.10 Hz, 1 H), 7.18 (dd, J ) 7.55, 1.45 Hz, 1 H). In Vitro Incubations. The known metabolic pathway for the diclofenac quinone imine formation and GSH conjugation is shown in Figure 2. To form diclofenac quinone imine from 5-OH diclofenac (III), 4′-OH diclofenac (IX), and GSH conjugates 5-OH 4-GSH (VI), 5-OH 6-GSH (V) or 4′-OH 3-GSH diclofenac (XI), a solution of 100 µM substrate [diclofenac metabolite(s), ( 5 mM GSH] and 30 units/mL of horseradish peroxidase (HRP) in PBS buffer at pH 7.4 was prepared. The reaction was initiated by adding H2O2 to a final concentration of 500 µM and was allowed to run for 15 min at 37 °C. An equal volume of ACN was added to quench the reaction. After centrifugation, the supernatant was diluted to 20% with water, and 100 µL was injected on to the column. A simple check for the formation of the diclofenac quinone imines used by Kang et al. was to reduce the quinone imines by adding ascorbic acid (14). Following this procedure, ascorbic acid was added to bile samples to make a final concentration of 2 mM. Samples were mixed thoroughly and immediately analyzed by LCMS. Animal Studies. All animal procedures were conducted under protocols approved by the Amgen (Cambridge) Institutional Animal Care and Use Committee. Male Sprague-Dawley rats weighing 300-350 g were purchased from Charles River Laboratories (Wilmington, MA). The rats were housed in a humidity and temperature-controlled environment subject to a 12 h:12 h light: dark cycle and had access to water and a standard laboratory rodent diet ad libitum. Following a 1 week acclimation period, the rats

1950

Chem. Res. Toxicol., Vol. 23, No. 12, 2010

Waldon et al.

Figure 3. Diclofenac quinone imine GSH conjugates were detected in rat bile: (a) selected ion chromatogram of quinone imine conjugates, (b) mass spectrum of quinone imine conjugate, and (c) product ion mass spectrum of the ion at m/z 615.

had silicon catheters implanted in the femoral vein, the bile duct, and the proximal duodenum. The catheters were externalized between the scapulae, protected with a Covance infusion harness, and connected to permit recirculation of bile. The rats were placed in Nalge metabolism cages (Nalge Company, Rochester, NY) and were given access to food and water throughout the experiment. Approximately 16 h prior to dosing, the catheters were disconnected and bile was allowed to drain. Rats were administered a single dose of diclofenac sodium (50 mg/kg, 20 mg/mL) via the intravenous catheter. Bile samples were collected at 0-2, 2-4, 4-6, 6-8, 8-12, and 12-24 h into preweighed bottles. The volume of bile was determined gravimetrically, assuming a density of 1 g/mL. Bile samples were centrifuged at 13000 rpm, and the supernatants were transferred into HPLC vials for LC-MS analysis. Analytical Methods. Identification of metabolites from the in vitro and in vivo samples was carried out using an LC-MS system that consisted of a Shimadzu SIL-20AC autosampler, a Shimadzu CBM controller, two Shimadzu LC-20AD pumps (Shimadzu USA, Palo Alto, CA), and Finnegan LTQ or Finnegan LTQ-Orbitrap (Thermo Electron Corp., San Jose, CA) mass spectrometer. The mass spectrometers were equipped with an API2 source and Xcalibur version 2.0 software (Thermo Electron Corp.). The LTQOrbitrap mass spectra were acquired in the high-resolution mode using 30000 resolving power. Exact mass measurements were accomplished using external calibration. Samples (10-50 µL) were loaded onto a Varian Pursuit C18 reverse-phase column (250 mm × 4.6 mm, 5 µm pore size; Varian, Palo Alto, CA). The analytes were separated using a gradient solvent system consisting of two components, solvent A (0.1% acetic acid in water) and solvent B (0.1% acetic acid in ACN). The gradient was increased in a linear fashion from 5 to 100% B over 50 min at a flow rate of 1 mL/min. The LC eluate was split 1/20, with 1 part directed to the mass spectrometer and the remainder to waste. Electrospray ionization was used with the needle potential held at 4.5 kV.

Results and Discussion Analysis of bile from rats dosed with diclofenac and receiving no taurocholate replacement revealed the expected diclofenac metabolites and unchanged parent drug, as well as the corresponding GSH conjugates and diclofenac S-glutathionyl thioesters. The experimental rationale for not replacing the lost bile with new taurocholate was simply to reduce the background ions for possibly better MS detection of metabolites. In these low bile salt samples, the background was lowered, but more interesting new peaks were detected at m/z 615, which were postulated to be diclofenac-related material based on their characteristic chlorine isotopic pattern. The parent mass of the newly observed metabolite peaks did not match any previously reported biliary metabolites of diclofenac. The extracted ion chromatogram for m/z 615 from rat bile is shown in Figure 3, along with the mass spectrum (Figure 3B) and the product ion spectra of m/z 615 (Figure 3C) for the base peak. The metabolite peaks display the loss of 129 amu that is characteristic of GSH conjugates (15); however, little additional structural information can be deduced from the spectra. On the basis of the protonated molecular ions of m/z 615, 2 amu less than the protonated molecular ions of known diclofenac GSH conjugates, the metabolites were postulated to be the corresponding diclofenac quinone imine GSH conjugates. Authentic standards of the diclofenac 4′-O-3GSH (XII), diclofenac 5-O4GSH (VII), and diclofenac 5-O-6GSH (VIII) quinone imines were prepared by treatment with HRP and GSH using 4′-OHdiclofenac or 5-OH diclofenac as the starting material. The standards displayed the same retention times and MS/MS accurate mass spectra as the novel metabolite peaks observed in rat bile (Figure 4). The most intense diclofenac quinone imine

Diclofenac Quinone Imine Glutathione Conjugates

Chem. Res. Toxicol., Vol. 23, No. 12, 2010 1951

Figure 4. Selected ion chromatograms of quinone imine containing GSH conjugates of diclofenac (M + H)+ m/z 615.0721 shown for comparison of the retention times from LC-MS analysis of extracts from HRP incubations of diclofenac quinone imine GSH conjugate standards and bile from rats dosed with diclofenac: (a) diclofenac 4′-OH-3GS after treatment with HRP, (b) diclofenac 5-OH-4GS after treatment with HRP, (c) diclofenac 5-OH after treatment with HRP, and (d) diclofenac quinine imine GSH conjugates identified from bile.

Figure 5. Comparison of product ion spectra at m/z 615.07 of (a) the synthetic diclofenac 5-O-6GS and (b) the major m/z 615 peak observed in rat bile.

GSH conjugate peak detected in bile and the diclofenac quinone imine GSH conjugate generated synthetically from 5-OH diclofenac produced identical MS/MS accurate mass spectra (Figure 5), which suggests that the major peak is likely the diclofenac 5-O-6GSH (VIII). These metabolites have now been identified as the quinone imines of diclofenac GSH conjugates. The diclofenac quinone imine GSH conjugates were unstable under acidic conditions, with complete degradation observed within 24 h after the addition of 0.01% formic acid to the bile sample. Reduction of the HRP-generated diclofenac quinone imine GSH conjugate following treatment with 2 mM ascorbic acid has been reported (9). This published method was used on the rat bile and following treatment with ascorbic acid, the original peaks of m/z 615 could no longer be detected (Figure

Figure 6. Effect of ascorbic acid treatment on diclofenac quinone imine GSH conjugates. (A) Selected ion peaks at m/z 615.0687 and 312.0190 in rat bile before ascorbic acid treatment. (B) Selected ion peaks at m/z 615.0687 and 312.0190 after ascorbic acid treatment.

6), providing further evidence that the novel metabolite peaks were indeed the quinone imine GSH conjugates of diclofenac. The hydroxyl metabolites of diclofenac with a mass of m/z 312 are shown and as a sample stability control and retention time comparison between the repeat injections. Figure 2 shows the proposed metabolic pathway of diclofenac leading to the formation of the quinone imines of diclofenac GSH conjugates. The oxidative metabolism of diclofenac leading

1952

Chem. Res. Toxicol., Vol. 23, No. 12, 2010

to formation of reactive quinone imines (IV and X) and the subsequent formation of the hydroxyl diclofenac GSH adducts have been well established (16, 17). Identification of quinone imines containing GSH conjugates of diclofenac has not been previously reported from in vivo studies. It is important to note that these quinone imine containing GSH conjugates of diclofenac were observed only when diclofenac was administered to rats that did not receive the bile acid replacement. Under these conditions, the GSH adducts of diclofenac appeared to be further oxidized to the corresponding quinone imine containing GSH conjugates of diclofenac. The unconjugated quinone imines of diclofenac (IV and X) were not detected in bile. However, evidence for the formation of the quinone imines (IV and X) have been well documented from in Vitro studies (14). The reactive quinone imines (IV and X) have also been prepared enzymatically in vitro and can be formed in large quantities using the HRP strong oxidation procedures. The apparent transient existence of the diclofenac quinone imines (IV and X) in vivo may be understandable if they are very reactive molecules. It is possible that depleting the bile acid pool in these rats altered the oxidative environment in the hepatocytes, which facilitated the formation of stable diclofenac quinone imine GSH conjugates; however, we have no direct supporting evidence for such a change. The quinone imine containing GSH conjugates of diclofenac observed in the bile from taurocholate depleted rats were stable and could be detected after months of storage at -20 °C and multiple freeze thaw cycles. Any attempt to separate the quinone imine containing GSH conjugates of diclofenac from the bile using LC conditions resulted in their decomposition, which prevented the isolation of sufficient material for NMR analysis. The diclofenac quinone imine GSH conjugate did not appear to degrade to form the hydroxylated GSH compounds due to no apparent increase in a peak at m/z 617. The diclofenac 5-O quinone imine GSH conjugates (VII and VIII) could be produced through synthetic methods by oxidation with silver oxide or cobalt salen complex (Figure 1); however, attempts to isolate the compound from the metal complex, which required lowering the pH, resulted in degradation of the product. Because we were unable to produce pure synthetic standards of the quinone imine, the identity of the degradation product is unknown. Standards of the diclofenac quinone imine GSH conjugates could be generated in vitro by incubation of commercially available diclofenac metabolite standards (III and IX) with HRP and GSH. The HRP and GSH conjugation steps using 5-OH diclofenac (III) produced a mixture of the various diclofenac quinone imine GSH conjugates, 5-O-4GS (VII) and 5-O-6GS (VIII). On the basis of the retention times and the accurate mass MS/MS spectra, it is likely that the largest peak with mass of m/z 615 is the diclofenac 5-O-6GS (VIII). In this report, we have postulated that the diclofenac quinone imine GSH conjugates form as oxidation products of the hydroxyl diclofenac GSH conjugates. Alternatively, GSH conjugation may occur by a direct Michael addition reaction, which may be favored in a strong oxidizing environment such as occurs with HRP treatment. Analysis of the bile without tauocholate infusion showed the bile to have a basic pH of 8.5-9. This pH is the same as the bile from the tauocholate infused bile. Forming or reforming quinone imines after GSH conjugation could potentially lead to further GSH conjugation. Such di-GSH adducts have been identified and can be formed in vitro and possibly in vivo with our detection of small peaks in bile that

Waldon et al.

match the predicted mass of diglutathione conjugates (m/z 920.1390, 0.5 ppm) by accurate mass LC-MS (18). The di-GSH conjugates of diclofenac were also found as minor peaks in the in vitro HRP treatment, which may indicate that this additional conjugation is a possible pathway. The diclofenac quinone imine GSH conjugates could only be observed in bile when bile acid homeostasis was altered, suggesting that bile acid depletion might lead to a change in the oxidative environment of the hepatocytes or bile. Studies aimed at understanding the effect of taurocholate depletion on rat hepatic function are ongoing. These studies will try to answer questions on where changes in the oxidative environment occur, in the hepatocytes or possibly the canaliculi or bile. It is yet unclear whether lowering bile acid levels could affect the metabolism of diclofenac in humans. If a reduction in bile acid levels can lead to changes in diclofenac metabolism and alter the profile of reactive metabolites of diclofenac in humans, then the resulting reactive metabolites could be linked to the hepatic toxicity and idiosyncratic reaction. Because traditional toxicological assessment is conducted in normal healthy volunteers, any evaluation of toxicity due to perturbations in bile acid homeostasis would not be assessed in a typical clinical setting. Acknowledgment. We thank John Roberts and Meghan Langley for their excellent in vivo support with the bile duct catheterization studies and Mark Grillo for his discussion and review of this manuscript.

References (1) Small, R. E. (1989) Diclofenac sodium. Clin. Pharmacol. 8, 545. (2) Banks, A. T., Zimmerman, H. J., Ishak, K. G., and Harter, J. G. (1995) Diclofenac-associated hepatotoxicity: Analysis of 180 cases reported to the Food and Drug Administration as adverse reactions. Hepatology 22, 820. (3) Boelsterli, U. A. (2003) Diclofenac-induced liver injury: a paradigm of idiosyncratic drug toxicity. Toxicol. Appl. Pharmacol. 192, 307– 322. (4) Kretz-Rommel, A., and Boelsterli, U. A. 1994a Mechanism of covalent adduct formation of diclofenac to rat hepatic microsomal proteins. Retention of the glucuronic acid moiety in the adduct. Drug Metab. Dispos. 22, 956. (5) Kretz-Rommel, A., and Boelsterli, U. A. 1994b Selective protein adducts to membrane proteins in cultured rat hepatocytes exposed to diclofenac: Radiochemical and immunochemical analysis. Mol. Pharmacol. 45, 237. (6) Shen, S., Marchick, M. R., Davis, M. R., Doss, G. A., and Pohl, L. R. (1999) Metabolic activation of diclofenac by human cytochrome P450 3A4: Role of 5-hydroxydiclofenac. Chem. Res. Toxicol. 12, 214. (7) Poon, G. K., Chen, Q., Teffera, Y., Ngui, J. S., Griffin, P. R., Braun, M. P., Doss, G. A., Freeden, C., Stearns, R. A., Evans, D. C., Baillie, T. A., and Tang, W. (2001) Bioactivation of diclofenac via benzoquinone imine intermediates-identification of urinary mercapturic acid derivatives in rats and humans. Drug Metab. Dispos. 29, 1608. (8) Miyamoto, G., Zahid, N., and Uetrecht, J. P. (1997) Oxidation of diclofenac to reactive intermediates by neutrophils, myeloperoxidase, and hypochlorous acid. Chem. Res. Toxicol. 10, 414–419. (9) Shen, S., Hargus, S. J., Martin, B. M., and Pohl, L. R. (1997) Cytochrome P4502C11 is a target of diclofenac covalent binding in rats. Chem. Res. Toxicol. 10, 420. (10) Jones, J. A., Kaphalia, L., Treinen-Moslen, M., and Liebler, D. C. (2003) Proteomic characterization of metabolites, protein adducts, and biliary proteins in rats exposed to 1,1-dichloroethylene or diclofenac. Chem. Res. Toxicol. 16, 1306. (11) Tang, W. (2003) The metabolism of diclofenac - enzymology and toxicology perspectives. Curr. Drug Metab. 4, 319. (12) Tang, W., Stearns, R. A., Bandiera, S. M., Zhang, Y., Raab, C., Braun, M. P., Dean, D. C., Pang, J., Leung, K. H., Doss, G. A., Strauss, J. R., Kwei, G. Y., Rushmore, T. H., Chiu, S.-H., L., and Baillie, T. A. (1999) Studies on hepatic cytochrome P450-mediated bioactivation of diclofenac in rats and in human hepatocytes: Identification of glutathione conjugated metabolites. Drug Metab. Dispos. 27, 365– 372. (13) Kenny, J. R., Maggs, J. L., Meng, X., Sinnott, D., Clarke, S. E., Park, K., and Stachulski, A. V. (2004) Syntheses and characterization of

Diclofenac Quinone Imine Glutathione Conjugates the acyl glucuronide and hydroxy metabolites of diclofenac. J. Med. Chem. 47, 2816–2825. (14) Kang, P., Deepak, D., Smith, E., and Renner, M. (2009) Bioactivation of lumiracoxib by peroxidases and human liver microsomes: Identification of multiple quinone imine intermediates and GSH adducts. Chem. Res. Toxicol. 22 (1), 106–117. (15) Baillie, T. A., and Davis, M. R. (1993) Mass spectrometry in the analysis of glutathione conjugates. Biol. Mass Spectrom. 22, 319– 325. (16) Yohannes, T., Waldon, D. J., Colletti, A. E., Albrecht, B. K., and Zhao, Z. (2008) Identification of a novel glutathione conjugate of diclofenac by LTQorbitrap. Drug Metab. Lett. 2 (1), 35–40.

Chem. Res. Toxicol., Vol. 23, No. 12, 2010 1953 (17) Grillo, M. P., Knutson, C. G., Sanders, P. E., Waldon, D. J., Hua, F., and Ware, J. A. (2003) Studies on the chemical reactivity of diclofenac acyl glucuronide with glutathione: identification of diclofenac-S-acyl-glutathione in rat bile. Drug Metab. Dispos. 31 (11), 1327–1336. (18) Madsen, K. G., Skonberg, C., Jurva, U., Cornett, C., Hansen, S. H., Johansen, T. N., and Olsen, J. (2008) Bioactivation of diclofenac in vitro and in vivo: Correlation to electrochemical studies. Chem. Res. Toxicol. 21, 1107–1119.

TX100296V