Enantioselective Formation of Ibuprofen-S-Acyl-Glutathione in Vitro in

Ibuprofen is metabolized to chemically reactive ibuprofen-1-O-acyl-glucuronide (I-1-O-G) and ibuprofen-S-acyl-CoA (I-CoA) derivatives, which are propo...
1 downloads 0 Views 783KB Size
Chem. Res. Toxicol. 2008, 21, 1749–1759

1749

Enantioselective Formation of Ibuprofen-S-Acyl-Glutathione in Vitro in Incubations of Ibuprofen with Rat Hepatocytes† Mark P. Grillo*,‡ and Fengmei Hua§ Pharmacokinetics, Dynamics, and Metabolism, Pfizer, Inc., Kalamazoo, Michigan 49007-4940 ReceiVed March 11, 2008

Ibuprofen is metabolized to chemically reactive ibuprofen-1-O-acyl-glucuronide (I-1-O-G) and ibuprofenS-acyl-CoA (I-CoA) derivatives, which are proposed to mediate the formation of drug-protein adducts via the transacylation of protein nucleophiles. We examined the ability of ibuprofen to undergo enantioselective metabolism to ibuprofen-S-acyl-glutathione thioester (I-SG) in incubations with rat hepatocytes, where I-CoA formation is known to be highly enantioselective in favor of the (R)-(-)ibuprofen isomer. We proposed that potential enantioselective transacylation of glutathione forming I-SG in favor of the (R)-(-)-isomer would reveal the importance of acyl-CoA formation, versus acyl glucuronidation, in the generation of reactive transacylating-type intermediates of the drug. Thus, when (R)-(-)- and (S)-(+)-ibuprofen (100 µM) were incubated with hepatocytes, the presence of I-CoA and I-SG was detected in incubation extracts by LC-MS/MS techniques. The formation of I-CoA and I-SG in hepatocyte incubations with (R)-(-)-ibuprofen was rapid and reached maximum concentrations of 2.6 µM and 1.3 nM, respectively, after 8-10 min of incubation. By contrast, incubations with (S)-(+)ibuprofen resulted in 8% and 3.9% as much I-CoA and I-SG formation, respectively, compared to that in corresponding incubations with the (R)-(-)-isomer. Experiments with a pseudoracemic mixture of (R)-(-)-[3,3,3-2H3]- and (S)-(+)-ibuprofen showed that >99% of the I-SG detected in hepatocyte incubations contained deuterium and therefore was derived primarily from (R)-(-)-ibuprofen bioactivation. Inhibition of (R)-(-)-ibuprofen (10 µM) glucuronidation with (-)-borneol (100 µM) led to a 98% decrease in I-1-O-G formation; however, no decrease in I-SG production was observed. Coincubation with pivalic, valproic, or lauric acid (500 µM each) was shown to lead to a significant inhibition of I-CoA formation and a corresponding decrease in I-SG production. Results from these studies demonstrate that the reactive I-CoA derivative, and not the I-1-O-G metabolite, plays a central role in the transacylation of GSH in incubations with rat hepatocytes. Introduction Ibuprofen (racemic-2-[4-isobutylphenyl]propionic acid; Figure 1) is a nonsteroidal antiinflammatory drug (NSAID1) that belongs to the 2-phenylpropionic acid (profen) structural class of carboxylic acid-containing NSAIDs widely used for the treatment of inflammatory diseases such as rheumatoid arthritis and osteoarthritis (1). Profen drugs have a chiral center at the carbon alpha to the carboxylic acid, where the antiinflammatory activity, through the inhibition of cyclooxygenase enzymes and subsequent prostaglandin biosynthesis, resides primarily in the (S)-(+)-enantiomer (2). However, ibuprofen is marketed as a racemic mixture where the effective plasma concentration is ∼50 µM for antipyresis in febrile children (3). Chemically † Preliminary accounts of this work were presented at the XIVth World Congress of Pharmacology, San Francisco, CA, July 2002. * To whom correspondence should be addressed. Tel: (650) 244-2032. Fax: (650) 825-7320. E-mail: [email protected]. ‡ Present address: Amgen Inc., Department of Pharmacokinetics and Drug Metabolism, South San Francisco, CA 94080. § Present address: Pfizer PGRD Pharmacokinetics, Dynamics and Metabolism, Chesterfield, MO 63017. 1 Abbreviations: I-1-O-G, ibuprofen-1-O-acyl glucuronide; I-SG, ibuprofen-S-acyl-glutathione thioester; I-CoA, ibuprofen-S-acyl-CoA thioester; D3-I-CoA, [3,3,3-2H3]-I-CoA; D3-I-SG, [3,3,3-2H3]-I-SG; NSAID, nonsteroidal antiinflammatory drug; CBZ, carbamazepine; profen, 2-phenylpropionic-type acidic drugs; 2-PPA, 2-phenylpropionic acid; 2-PPA-CoA, 2-PPA-S-acyl-CoA; ACN, acetonitrile; AUC, area under the curve; CID, collisionally induced dissociation; MRM, multiple reaction monitoring; LCMS/MS, liquid chromatography-tandem mass spectrometry.

reactive metabolites of ibuprofen have been proposed to mediate idiosyncratic allergic reactions associated with its clinical use (4–6). Ibuprofen is metabolized to ibuprofen-1-O-acyl glucuronide (I-1-O-G, Figure 1), an unstable and chemically reactive metabolite that has been implicated as playing a role in the onset of rare, but sometimes serious, allergic reactions to the drug (4). Acyl glucuronide metabolites of acidic drugs are proposed to bind covalently to protein by two mechanisms which include the transacylation of protein nucleophiles by the 1-O-acyl glucuronide isomer, and by glycation of protein amino-groups Via the open-chain aldehyde forms of acyl migration glucuronide isomers (7, 8). Such drug-protein adducts are proposed to be recognized by the immune system as foreign resulting in an immune-response (9, 10). In addition to the generation of chemically reactive acyl glucuronides, the formation of acylCoA thioester derivatives of carboxylic acid-containing drugs also increases the chemical reactivity of the carbonyl-carbon of the carboxylic acid moiety in transacylation-type reactions with endogenous nucleophiles (11). Ibuprofen is known to be converted to ibuprofen-S-acyl-CoA (I-CoA, Figure 1) in vitro in rat hepatocytes and in rat liver homogenate stereoselectively in favor of the (R)-(-)-isomer (12–14). Recent studies with the structurally most basic profen, 2-phenylpropionic acid (2-PPA), showed that the 2-PPA-S-acyl-CoA thioester derivative (2-PPACoA) was 70-fold more reactive than the corresponding 2-PPA1-O-acyl glucuronide metabolite in transacylation-type reactions with GSH (15). Other studies have shown 2-PPA to undergo

10.1021/tx800098h CCC: $40.75  2008 American Chemical Society Published on Web 08/05/2008

1750

Chem. Res. Toxicol., Vol. 21, No. 9, 2008

Figure 1. Proposed scheme for the metabolic activation of ibuprofen by acyl glucuronidation and acyl-CoA formation leading to the transacylation of glutathione forming ibuprofen-S-acyl-glutathione (ISG).

(R)-(-)-2-PPA favored enantioselective covalent binding to protein in rat hepatocyte incubations (16). Such findings have led to the proposal that xenobiotic acyl-CoA thioesters are reactive transacylating intermediates of carboxylic acid-containing drugs that, in addition to acyl glucuronide metabolites, may contribute to covalent binding to protein. The present studies were designed to examine the relative contribution of these two metabolic activation pathways in the metabolic activation of ibuprofen to reactive intermediates that transacylate GSH in vitro in rat hepatocytes. We propose that if I-CoA is more determining than I-1-O-G toward the transacylation of GSH in vitro, then incubations of (R)-(-)-ibuprofen with rat hepatocytes, because of the known highly enantioselective formation of (R)-(-)-I-CoA, should lead to a greater formation of ibuprofen-S-acyl-glutathione (I-SG) compared to that formed in corresponding incubations with the (S)-(+)ibuprofen isomer.

Materials and Methods Materials. (R)-(-)-ibuprofen (99.1% optical purity), (S)-(+)ibuprofen (99.7% optical purity), and (R)-(-)-[3,3,3-2H3]ibuprofen (98.4% optical purity; isotopic purity 97 atom % excess 2H3) were obtained from prior syntheses from the Research Laboratories at The Upjohn Co. (Kalamazoo, MI). Each of these ibuprofen derivatives came from a supply that was used previously for studies on ibuprofen chiral inversion metabolism performed in vitro and in vivo (17, 18). Carbamazepine, diclofenac sodium salt, pivalic acid, valproic acid sodium salt, lauric acid sodium salt, (-)-borneol, and β-glucuronidase (EC 3.2.1.31; type H-1 from Helix pomatia) were purchased from Sigma Chemical Co. (St. Louis, MO). GSHS-acyl-thioesters (R)-(-)-I-SG, (S)-(+)-I-SG, and (R)-(-)-[3,3,32 H3]-I-SG, and (R,S)-I-CoA acyl-CoA thioester were synthesized

Grillo and Hua

as described below. All solvents used for HPLC and LC/MS analyses were of chromatographic grade. Stock solutions of ibuprofen, GSH- and CoA-thioester derivatives were prepared as solutions in 50/50 (v/v) acetonitrile/water and containing 3% formic acid. Instrumentation and Analytical Methods. All analytical LC/MS analyses were performed on a reverse-phase column (Zorbax SB-C18, 3.5 µM, 150 × 2.1 mm, Agilent Technologies, Palo Alto, CA). LC/MS analyses for the detection of I-SG derivatives (positive ion mode) and ibuprofen acyl glucuronide (negative ion scan mode) were accomplished using a gradient system of 0.1% formic acid with elution from 5% acetonitrile (ACN) to 100% over 13 min at flow rate of 0.3 mL/min. Analysis of I-CoA derivatives was performed by gradient elution from 5% to 100% ACN in aqueous ammonium acetate (10 mM, pH 5.6) with a flow rate of 0.3 mL/min. LC/MS and LC-MS/ MS analyses were performed on a Finnigan TSQ-7000 equipped with API2 and running Excalibur version 1.2 (San Jose, CA). Electrospray ionization was employed with the needle potential held at ∼4.5 kV. MS/MS conditions used were 2 mTorr argon collision gas and a collision potential of 25 eV. 1H NMR spectra were recorded on a Bruker Avance 400 spectrometer operating at 400 MHz. Chemical shifts are reported in parts per million as referenced to the residual solvent peak (2.49 ppm for 2H6DMSO). Synthesis of S-Acyl-Glutathione Thioester Derivatives. GSH-S-acyl-thioester derivatives were synthesized by conventional methods employing ethyl chloroformate (19) as described previously for the synthesis of clofibryl-S-acyl-glutathione thioester (20) and provided (R)-(-)-I-SG, (S)-(+)-I-SG, and (R)(-)-[3,3,3-2H3]-I-SG derivatives as white solids in yields ranging from 30 to 40%. Product thioester derivatives were characterized by tandem mass spectrometry on a Finnigan-MAT TSQ-7000 tandem mass spectrometer, and LC-MS/MS analysis was performed by gradient elution as described above. All of the synthetic GSH thioesters eluted at a retention time of 9.8 min and showed no detectable impurities when analyzed by both positive and negative ion LC/MS scan modes Via reverse-phase gradient elution (as described above). Tandem LC-MS/MS analysis of (R)-(-)- and (S)-(+)-I-SG were essentially identical: (CID of MH+ ion at m/z 496), m/z (%): m/z 421 ([M + H Gly]+, 15%), m/z 367 ([M + H - pyroglutamic acid]+, 6%), m/z 349 ([M + H - pyroglutamic acid - water]+, 100%), m/z 308 ([glutathione + H]+, 22%), m/z 292 ([ibuprofen-SCH2CH(NH2)CdO]+, 50%), m/z 264 ([ibuprofen-S-CH2CH) NH2]+, 98%), m/z 161 ([4-isobutylethylbenzene]+, 45%). Tandem LC-MS/MS analysis of (R)-(-)-[3,3,3-2H3]-I-SG: (CID of MH+ ion at m/z 499), m/z (%): m/z 424 ([M + H - Gly]+, 15%), m/z 370 ([M + H - pyroglutamic acid]+, 6%), m/z 352 ([M + H - pyroglutamic acid - water]+, 100%), m/z 308 ([glutathione + H]+, 22%), m/z 295 ([ibuprofen-S-CH2CH(NH2)CdO]+, 50%), m/z 267 ([ibuprofen-S-CH2CH)NH2]+, 98%), m/z 164 ([4-isobutylethylbenzene]+, 45%). 1H NMR analysis of (R)-(-)-I-SG (2H6-DMSO): δ 0.84-0.86 (d, 6H, isopropyl -CH3 groups), δ 1.33-1.35 (d, 3H, R-CH3), δ 1.77-1.80 (m, 1H, isopropyl-CH), 1.82-1.90 (m, 1H, Glu-β, Glu-β′), 2.26 (m, 2H, Glu-γ,γ′), 2.91 (m, 2H, Cys-β,β′), 2.40-2.42, 2H, isopropyl -CH2), 3.26-3.29 (m, 2H, Glu-R), 3.65-3.67 (m, 2H, Gly-R,R′), 3.92-3.97 (q, 1H, ibuprofen R-CH-), 4.34-4.37 (m, 1H, Cys-R), 7.10-7.20 (m, 4H, phenyl ring), 8.40-8.43 (m, 1H, Cys NH), 8.55-8.75 (br, 1H, Gly NH). Synthesis of (R,S)-I-CoA Thioester. The synthesis and purification of the acyl-CoA thioester of (R,S)-ibuprofen was

EnantioselectiVe BioactiVation of Ibuprofen

accomplished by a procedure using ethyl chloroformate (19) and as previously reported for the synthesis and purification of (R,S)-[ring-2H4]ibuprofen-CoA (18) and clofibryl-S-acyl-CoA (20). The acyl-CoA thioester (R,S)-I-CoA eluted at a retention time of 9.6 min and showed no detectable impurities when analyzed by both positive and negative ion LC/MS scan modes Via reverse-phase gradient elution (as described above). Under these chromatographic conditions, the (R)-(-)- and (S)-(+)enantiomers of I-CoA were not resolved. Tandem LC-MS/MS analysis of (R,S)-I-CoA standard yielded a product ion mass spectrum under collision-induced dissociation (CID) of the protonated molecular ion at MH+ m/z 956, m/z (%): m/z (MH+, 97%), m/z 547 ([M+H-409]+, 7%), m/z 449 ([M+H-507]+, 78%), m/z 428 ([adenosine diphosphate + 2H]+, 49%), m/z 347 ([M+H-609]+, 18%), m/z 161 ([4-isobutylethylbenzene]+, 2%), m/z 136 ([adenine + H]+, 1%). In Vitro Studies with Rat Hepatocytes. Freshly isolated rat (250-300 g, male Sprague-Dawley) hepatocytes were prepared from one rat for each experiment according to the method of Mo´ldeus et al. (1978) and greater than 90% viability was achieved routinely as assessed by trypan blue exclusion testing (21). Incubations of hepatocytes (2 million viable cells/ mL) with (R)-(-)-, (S)-(+)-ibuprofen isomers, or with a pseudoracemic mixture of (R)-(-)-[3,3,3-2H3]-(R)-(-)- and (S)(+)-ibuprofen (100 µM) were performed in Krebs-Henseleit buffer (pH 7.4) in 20 mL glass vials capped with a plastic screw cap containing a small hole (1/8 in. diameter) to avoid evaporation. Incubations (5 mL total volume for all experiments, n ) 3) were performed with continuous rotation and under an atmosphere of 5% CO2 at 37 °C in a VWR Model 1927 incubator (Willard, OH). For time-dependent studies, freshly isolated hepatocytes were incubated with (R)-(-)-, (S)-(+)-ibuprofen, or a pseudoracemic mix of (R)-(-)-[3,3,3-2H3]-ibuprofen and nondeuterated (S)(+)-ibuprofen (100 µM total concentration), and aliquots of the incubation mixtures analyzed for corresponding I-SG formation over a 40 min time-period. Aliquots (200 µL) of the incubation mixture were taken at 0.2, 5, 10, 20, and 40 min of incubation and added directly to microcentrifuge tubes (2 mL) containing a quench solution (200 µL) consisting of methanol, 3% formic acid and 10 µM carbamazepine (CBZ) internal standard. Samples were centrifuged (14,000 rpm, 5 min) and the supernatant fraction (300 µL) transferred to 0.4 mL polypropylene autosampler vials (Sun International, Wilmington, NC) prior to chromatographic analysis. For the analysis of I-CoA formation, aliquots (200 µL) from the same incubations (described above for the detection of I-SG) were taken and added to ACN (400 µL) in microcentrifuge tubes, followed by the addition of hexane (600 µL). The samples were vortex-mixed (1 min), centrifuged (as above), and aliquots (300 µL) of the aqueous layer transferred to autosampler vials. Concentration-dependent experiments were performed with increasing concentrations of (R)-(-)-ibuprofen or (S)-(+)ibuprofen (2, 4, 8, 16, 32, 63, 125, 250, 500, and 1000 µM) incubated with rat hepatocytes (2 × 106 cells/mL) for an incubation time of 3 min. Inhibition experiments were performed with (R)-(-)-ibuprofen (10 µM) incubated with rat hepatocytes (2 × 106 cells/mL) in the presence or absence of (-)-borneol (100 µM) for the inhibition of ibuprofen acyl glucuronidation, and pivalic acid, valproic acid, or lauric acid (500 µM each) for the inhibition of acyl-CoA formation (16, 22). Incubations (n ) 4) were performed as described above for 5 min and terminated with appropriate quenching solutions as described above for the

Chem. Res. Toxicol., Vol. 21, No. 9, 2008 1751 Table 1. Effect of Inhibitors of Ibuprofen Acyl Glucuronidation and Acyl-CoA Formation on the Production of I-SG in Incubations of (R)-(-)-Ibuprofen (10 µM) with Rat Hepatocytes (2 × 106 Cells/mL, 5 min Incubation Time)a inhibition of metabolite formation (% formed vs control)

b

inhibitor

I-1-O-G

I-CoA

I-SG

(-)-borneol (0.1 mM) pivalic acid (0.5 mM) valproic acid (0.5 mM) lauric acid (0.5 mM)

1.7 ( 0. 5 n.d. n.d. 9.2 ( 1.3

n.d.b 53 ( 2.1 36 ( 4.7 7.3 ( 2.7

98 ( 9.0 46 ( 5.1 34 ( 0.5 7.0 ( 2.5

a Values are expressed as the mean ( SD from four incubations. n.d., not determined.

analysis of I-SG and I-CoA. A stock solution of (-)-borneol (100 mM) was prepared in absolute ethanol, and control incubations included the same final concentration of ethanol (0.1%, v/v). Stock solutions of the pivalic acid, valproic acid sodium salt, and lauric acid sodium salt, were prepared as 100 mM solutions in distilled water (pH 7). We did not quantify the absolute amounts of metabolites formed in these inhibition experiments, but instead we measured the relative formation of each derivative by their chromatographic peak area ratios compared to the internal standard. The extent of inhibition of metabolite formation shown in Table 1 is indicated by the percent metabolite formed (as measured by the ratio of derivative/internal standard LC-MS/MS peak area ratios) in the presence of inhibitor relative to that measured in the absence of inhibitor in the incubations. Experiments to evaluate the metabolic stability of (R)-(-)and (S)-(+)-I-SG thioesters were performed with a rat hepatocyte concentration of 2 × 106 cells/mL and an initial I-SG concentration of 1 µM. Aliquots (200 µL) of the incubation mixtures (5 mL volume, n ) 2) were taken at 0.2, 1, 2, 3, 4, 5, 6, 8, and 10 min, added to quench solution and processed as described above for the LC-MS detection of I-SG. Analysis of the amount of I-SG remaining was detected and quantified by LC/MS (positive ion scan mode) and by linear standard curves of (R)-(-)- or (S)-(+)-I-SG (MH+ m/z 496) and CBZ (MH+ m/z 237) peak area ratios from extracted ion chromatograms using the corresponding enantiomer. Identification and Quantification of I-SG. Extracts of ibuprofen-treated rat hepatocytes were analyzed by LC-MS/MS detection for I-SG and CBZ by using the multiple reaction monitoring (MRM) transitions MH+ m/z 496 to m/z 161, m/z 264, m/z 292 and m/z 349 for I-SG detection, and MH+ m/z 237 to m/z 194 for CBZ detection in the positive ion mode and using the chromatographic method described above. For studies with (R)-(-)-[3,3,3-2H3]-ibuprofen, the MRM transitions MH+ m/z 499 to m/z 164, m/z 267, m/z 295 and m/z 352 were used. Authentic I-SG standards eluted at a retention time of 9.8, while CBZ eluted at retention time 10 min. No LC-MS/MS chromatographic peak corresponding to I-SG was detected in incubation extracts lacking either ibuprofen isomer (data not shown). The concentration of I-SG thioesters was determined from a linear standard curve generated from peak area ratios using the corresponding derivative. Standard curves for (R)(-)- and (S)-(+)-I-SG derivatives showed no significant difference in derivative/internal standard peak area ratios. Identification and Quantification of I-CoA. Extracts of ibuprofen-treated rat hepatocytes were analyzed by LC-MS/MS detection for I-CoA and CBZ by using the MRM transitions MH+ m/z 956 to m/z 449 and MH+ m/z 237 to m/z 194, respectively, in the positive ion mode and using the chromatographic method described above. For studies with (R)-(-)-

1752

Chem. Res. Toxicol., Vol. 21, No. 9, 2008

Grillo and Hua

Figure 2. Representative reverse-phase gradient LC-MS/MS MRM chromatograms of I-CoA from the analysis of (A) extract from rat hepatocytes incubated for 6 min with 100 µM (R)-(-)-ibuprofen, (B) extract from rat hepatocytes incubated for 6 min with 100 µM (S)-(+)-ibuprofen, and (C) extract from control rat hepatocytes spiked with I-CoA authentic standard. The LC-MS/MS MRM transition used for this analysis was MH+ m/z 956 to m/z 449 for I-CoA detection. The relative abundance ranges on the y-axes for each chromatogram are equal.

[3,3,3-2H3]-ibuprofen, the selected ion monitoring transition MH+ m/z 959 to m/z 452 was used. Authentic (R,S)-I-CoA standard eluted at a retention time of 9.6, while CBZ eluted at retention time 10 min. No LC-MS/MS chromatographic peak corresponding to I-CoA was detected in incubation extracts lacking either ibuprofen isomer (data not shown). The concentration of I-CoA thioester was determined from a linear standard curve generated from peak area ratios. Identification of I-1-O-G. The (R)-(-)- and (S)-(+)-1-Oacyl glucuronide isomers of ibuprofen were not obtained as purified standards for these studies, but their formation in vitro was confirmed by treatment of (R)-(-)- and (S)-(+)-ibuprofen (100 µM) rat hepatocyte incubation (2 million cells/mL, 30 min) extracts (50 µL) with β-glucuronidase (1 mL total volume) at pH 5.0 and 37 °C (as per the manufacturer’s instructions), followed by HPLC analysis (as above, UV analysis at 254 nm) at 0 and 30 min of incubation. Results from these analyses indicated that the chromatographic peaks corresponding to (R)(-)-I-1-O-G (retention time 13.8 min) and (S)-(+)-I-1-O-G (retention time 12.9 min), which were completely absent in the 30 min β-glucuronidase treated extract, were indeed the I-1O-G isomers (data not shown). In addition, the identification of the (R)-(-)- and (S)-(+)-ibuprofen acyl glucuronide isomers was determined on the basis of the products formed during separate incubations with individual enantiomers. Tandem LCMS/MS analysis of I-1-O-G: (CID of [M - H]- ion at m/z 381), m/z (%): m/z 193 ([glucuronic acid]-, 45%), m/z 175 ([glucuronic acid - OH]-, 27%), m/z 161 ([4-isobutylethylbenzene]-, 90%), m/z 113 ([m/z 175 fragment - CO2 - H20]-, 27%), m/z 85 (100%), m/z 73 (45%), m/z 59 (30%). Analysis for the formation of I-1-O-G in incubations of ibuprofen with rat hepatocytes was performed by LC/MS detection (as described above) in the negative ion scan mode and using diclofenac ([M - H]- m/z 294) as an internal standard (retention time 13.5 min).

Reactions of (R,S)-I-CoA with GSH in Buffer. Incubations (0.5 mL total incubation volume, n ) 3 incubations/time-point) containing (R,S)-I-CoA (10 µM) were performed in phosphate buffer (0.1 M, pH 7.4) at 37 °C with 10 mM GSH. Incubations were stopped by the addition of quench solution (methanol, 3% formic acid and 10 µM CBZ, 0.5 mL), and the entire quenched mixtures transferred to microcentrifuge tubes, centrifuged (14,000 rpm, 5 min), and the supernatant fractions analyzed for I-SG concentration by LC-MS/MS detection as described above.

Results Identification of I-CoA. Analysis of rat hepatocyte incubation extracts by LC-MS/MS MRM detection allowed for the identification of I-CoA formed in incubations with both (R)(-)- and (S)-(+)-ibuprofen (Figure 2). The transition used for this analysis was MH+ m/z 956 to m/z 449, which was chosen because of it being a major fragmentation pathway for I-CoA as assessed by positive ion LC-MS/MS collision-dissociation (CID) of the MH+ ion at m/z 956 of authentic I-SG (Figure 3A). Reverse-phase LC-MS/MS analysis showed the presence of I-CoA in incubations of racemic or (R)-(-)-ibuprofen (100 µM) with freshly isolated rat hepatocytes, and which coeluted with authentic (R,S)-I-CoA standard at a retention time of 9.6 min (Figure 2A). I-CoA also was detected in incubations with (S)-(+)-ibuprofen, but to a much lower extent (Figure 2B). LCMS/MS analysis of I-CoA by CID of the MH+ ion at m/z 956 provided a product ion spectrum that was identical to the authentic (R,S)-I-CoA standard and consistent with its chemical structure (Figure 3). Identification of I-SG. Analysis of extracts by a sensitive LCMS/MS MRM detection technique allowed for the identification of I-SG formed in rat hepatocyte incubations (Figure 4). The transitions used for these analyses were MH+ m/z 496 to m/z 161, m/z 264, m/z 292 and m/z 349 and were chosen because of them being major fragmentation pathways for I-SG authentic standard

EnantioselectiVe BioactiVation of Ibuprofen

Chem. Res. Toxicol., Vol. 21, No. 9, 2008 1753

Figure 3. LC-MS/MS tandem mass spectra of (A) authentic I-CoA standard and (B) biologically formed I-CoA from rat hepatocyte incubations with (R)-(-)-ibuprofen (100 µM, 6 min) obtained by CID of the protonated molecular ion MH+ at m/z 956. The origins of the characteristic fragments are as shown.

Figure 4. Representative reverse-phase gradient LC-MS/MS MRM chromatograms of I-SG from analysis of (A) extract from rat hepatocytes incubated for 6 min with 100 µM (R)-(-)-ibuprofen, (B) extract from rat hepatocytes incubated for 6 min with 100 µM (S)-(+)-ibuprofen, and (C) extract from control rat hepatocytes spiked with I-SG authentic standard. The transitions used for this LC-MS/MS MRM analyses were MH+ m/ 496 to m/z 161, m/z 264, m/z 292, and m/z 349 for I-SG detection. The relative abundance ranges on the y-axes for each chromatogram are equal.

as assessed by positive ion LC-MS/MS collision-dissociation (CID) of the MH+ ion at m/z 496 of authentic I-SG (Figure 5A). Reversephase LC-MS/MS analysis showed the presence of I-SG in incubations of (R)-(-)- or (S)-(+)-ibuprofen (100 µM) with freshly isolated rat hepatocytes and which coeluted with authentic I-SG standard at a retention time of 9.8 min (Figure 4). The GSH conjugate I-SG was detected in extracts from incubations with (S)-

(+)-ibuprofen (Figure 4B) at a level that was ∼5% as much as detected at the 6 min time-point compared to that detected in extracts from incubations with (R)-(-)-ibuprofen. LC-MS/MS analysis of I-SG formed in hepatocyte incubations provided a product ion spectrum that showed fragment ions identical to the authentic I-SG standard and consistent with its chemical structure (Figure 5) (23).

1754

Chem. Res. Toxicol., Vol. 21, No. 9, 2008

Grillo and Hua

Figure 5. LC-MS/MS tandem mass spectra of (A) authentic I-SG standard and (B) biologically formed I-SG from rat hepatocyte incubations with (R)-(-)-ibuprofen (100 µM, 6 min) obtained by CID of the protonated molecular ion MH+ at m/z 496. The origins of the characteristic fragments are as shown.

Figure 6. Representative reverse-phase gradient LC-MS chromatograms of I-1-O-G from the analysis of (A) extract from rat hepatocytes incubated (10 min) with 100 µM (R)-(-)-ibuprofen and (B) extract from rat hepatocytes incubated (10 min) with 100 µM (S)-(+)-ibuprofen. Glucuronides were detected by LC/MS analysis in the negative ion scan mode. The figure shows the extracted ion chromatograms for I-1-O-G at [M - H]- m/z 381. The relative abundance ranges on the y-axes for each chromatogram are equal.

LC-MS Detection of I-1-O-G. The 1-O-acyl glucuronide of ibuprofen was detected by LC/MS analysis of extracts from incubations of (R)-(-)- or (S)-(+)-ibuprofen (100 µM) with rat hepatocyte (2 × 106 cells/mL, 10 min, Figure 6). Results from these analyses showed the chiral isomers of I-1-O-G eluting at

retention times 13.8 and12.9 min from incubations with (R)-(-)or (S)-(+)-ibuprofen, respectively. The relative retention times of these isomeric glucuronides determined in the present work are consistent with other reports showing that the glucuronide formed from (S)-(+)-ibuprofen eluted earlier than the glucuronide of the

EnantioselectiVe BioactiVation of Ibuprofen

Chem. Res. Toxicol., Vol. 21, No. 9, 2008 1755

Figure 7. Time course for the formation of I-CoA and I-SG in freshly isolated rat hepatocytes (2 × 106 cells/mL) incubated with 100 µM (R)-(-)or (S)-(+)-ibuprofen. Values are expressed as the mean ( SD of three incubations.

(R)-(-)-ibuprofen isomer during reverse phase chromatographic analysis (24). LC-MS analysis of extracts from incubations with (R)-(-)-ibuprofen indicated the presence of (S)-(+)-I-O-G, however no (R)-(-)-I-O-G was detected in incubations with (S)-(+)ibuprofen which reflects the known enantioselectivity of chiral inversion of the (R)-antipode to the (S)-antipode. Time Course of I-CoA and I-SG Formation in Incubations with Rat Hepatocytes. When (R)-(-)- or (S)-(+)ibuprofen (100 µM) were incubated with hepatocytes, the formation of I-SG in hepatocyte incubations with (R)-(-)ibuprofen was rapid and reached a maximum concentration of 1.3 nM after 10 min of incubation (Figure 7B). By contrast, incubations with (S)-(+)-ibuprofen showed significantly less area under the curve (AUC) of I-SG to 3.9% of that formed with (R)-(-)-ibuprofen over the 40 min incubation time-period. From these same incubations, the maximum concentration of I-CoA was 2.6 µM and occurred at the 8 min time-point (Figure 7A). Formation of I-CoA from (S)-(+)-ibuprofen was substantially less, where the AUC0-40 min was 8% of I-CoA formation in incubations with (R)-(-)-ibuprofen. Incubations with a pseudoracemic mixture of (R)-(-)-[3,3,3-2H3]- and (S)-(+)-ibuprofen (100 µM total) showed that after 6 min of incubation, as detected by LC-MS/MS analysis, >99% of I-SG formed contained deuterium (Figure 8). No I-CoA was detected that did not contain deuterium over the entire 40 min incubation time-period (Figure 9A). In these incubations, [3,3,3-2H3]-I-CoA (D3-I-CoA) rapidly reached a maximum concentration of 1.1 µM at the 4 min time-point and slowly decreased to ∼0.7 µM after 40 min of incubation (Figure 9A). The formation of [3,3,3-2H3]-I-SG (D3-I-SG) increased rapidly and in a linear fashion reaching a concentration of ∼0.9 nM at the 10 min time-point (Figure 9B). The D3-ISG/D0-ISG AUC0-40 min ratio was ∼341, therefore I-SG formation was derived primarily from (R)-(-)-ibuprofen bioactivation. These results are consistent with the results from hepatocyte incubations with the individual isomers, where the formation of I-SG was highly enantioselective for the (R)-(-)ibuprofen isomer (Figure 7). From the same incubation extracts used to determine I-SG concentration, we analyzed for the formation of I-1-O-G. Results showed that the formation of I-1-O-G from incubations with both (R)-(-)-ibuprofen and (S)-(+)-ibuprofen (100 µM) was time-dependent and increased throughout the 40 min incubation (Supporting Information, Figure 1). The ratio of I-1-O-G formation over 40 min of incubation for the two isomers was determined to be RAUC0-40 min/SAUC0-40 min ∼0.8 (Supporting Information, Figure 1), which was expected due to the known enantioselectivity for ibuprofen glucuronidation cross-species in favor of the (S)-(+)-ibuprofen isomer (24). From the analysis

Figure 8. Representative reverse-phase gradient LC-MS/MS MRM chromatograms of (A) extract from rat hepatocytes incubated for 6 min with pseudoracemic D3-(R)-(-)-/D0-(S)-(+)-ibuprofen (100 µM total); (B) extract from rat hepatocytes incubated for 6 min with racemic D0(R)-(-)-/D0-(S)-(+)-ibuprofen (100 µM total); and (C) extract from rat hepatocytes spiked with a pseudoracemic mixture of D3-(R)-(-)-/ D0-(S)-(+)-I-SG thioester. The transitions used for these LC-MS/MS analyses were MH+ m/z 496 to m/ 161, m/z 264, m/z 292, and m/z 349 for I-SG, and MH+ m/z 499 to m/ 164, m/z 267, m/z 295, and m/z 352 for D3-I-SG. The relative abundance ranges on the y-axes for each chromatogram are equal.

of incubations with the pseudoracemic mixture of (R)-(-)-[3,3,32 H3]-ibuprofen and (S)-(+)-ibuprofen (100 µM total), the ratio of I-1-O-G formation over 40 min of incubation for the two isomers was determined to be RAUC0-40 min/SAUC0-40 min ∼0.5 (Supporting Information, Figure 2). In addition to these data, the time course of degradation of authentic (R)-(-)- and (S)-(+)-I-SG (1 µM) from incubations with rat hepatocytes revealed both thioester derivatives to be degraded at identical rates resulting in ∼80% consumption from the incubations by the 10 min time-point (Figure 10, t1/2 ∼4 min). The products of I-SG degradation were not characterized in this experiment, but the degradation is presumably due to hydrolysis of the thioester, similar to that determined for diclofenac-S-acyl-glutathione (25), and not due to γ-glutamyltranspeptidase activity which is known to be negligible in rat liver tissue (26). (R)-(-)-Ibuprofen Concentration-Dependent Formation of I-CoA and I-SG in Incubations with Freshly Isolated Rat Hepatocytes. When rat hepatocytes (2 × 106 cells/mL)

1756

Chem. Res. Toxicol., Vol. 21, No. 9, 2008

Grillo and Hua

Figure 9. Time course for the formation of D3-I-CoA or D0-I-CoA and D3-I-SG or D0-I-SG in freshly isolated rat hepatocytes (2 × 106 cells/mL) incubated with pseudoracemic D3-(R)-(-)-/D0-(S)-(+)-ibuprofen (100 µM total). Values are expressed as the mean ( SD of three incubations.

valproic, or lauric acid led to a 47, 54, and 93% inhibition of I-CoA formation and also caused a corresponding 53, 66, and 93% inhibition I-SG formation, respectively (Table 1). Reaction of GSH with (R)-(-)-I-CoA in Buffer. Incubation of (R)-(-)-I-CoA (10 µM) in the presence of GSH (10 mM) in buffer (0.1 M potassium phosphate, pH 7.4, 37 °C, 0.5 mL incubation volume) over a 30 min period showed linear timedependent transacylation of GSH forming I-SG at a rate of formation of ∼0.007 nmol/min (Figure 12).

Discussion

Figure 10. Time-dependent degradation of (R)-(-)- and (S)-(+)-I-SG (1 µM) in incubations with freshly isolated rat hepatocytes (2 × 106 cells/mL). Values are expressed as the average of duplicate incubations.

were incubated for 3 min with increasing concentrations of (R)(-)-ibuprofen (2, 4, 8, 16, 32, 63, 125, 250, 500, and 1000 µM), results showed a sharp concentration-dependent formation of I-CoA from incubations containing 2 to 125 µM (R)-(-)ibuprofen, reaching ∼2.3 µM I-CoA at 125 µM (R)-(-)ibuprofen (Figure 11A). The rate of formation of I-CoA from 125 to1000 µM (R)-(-)-ibuprofen was approximately zero order, where the concentration of I-CoA was measured at 2.4 ( 0.2 µM at 1000 µM (R)-(-)-ibuprofen. From these same incubations, I-SG formation increased in a biphasic pattern with increasing concentration of (R)-(-)-ibuprofen (Figure 11B). The concentration of I-SG increased sharply reaching ∼0.8 nM I-SG at the 8 µM (R)-(-)-ibuprofen concentration. At concentrations higher than 8 µM, I-SG concentration increased in a roughly linear fashion reaching 3.8 nM I-SG at 1000 µM (R)-(-)ibuprofen. Inhibition Studies. To determine the importance of acyl glucuronidation or acyl-CoA formation of ibuprofen on I-SG formation in rat hepatocytes, the effect of inhibition of I-1-O-G and I-CoA formation on the production of I-SG was examined. Inhibition experiments were performed with (R)-(-)-ibuprofen (10 µM) in incubations with rat hepatocytes (2 × 106 cells/ mL, 5 min) in the presence or absence of (-)-borneol (100 µM) for the inhibition of I-1-O-G formation (25), and valproic acid (500 µM), pivalic acid (500 µM), or lauric acid (500 µM) for the inhibition of I-CoA formation (13, 22). Results showed that the inhibition of (R)-(-)-ibuprofen (10 µM) glucuronidation by (-)-borneol led to a 98% decrease in I-1-O-G formation, however no decrease in I-SG production was observed (Table 1). By contrast, coincubation of (R)-(-)-ibuprofen with pivalic,

Carboxylic acids containing NSAIDs are a class of drugs that have been subject to a disproportionally high incidence of withdrawal from the market (27, 28). The most frequent types of adverse reactions leading to their discontinuation were idiosyncratic toxicities, such as hepatotoxicity, allergic skin reactions, and renal toxicity sometimes associated with fever, rash, and eosinophilia. Profen-type NSAIDs that were withdrawn from use include benoxaprofen, flunoxaprofen, indoprofen, suprofen, and pirprofen (27). The mechanisms responsible for the initiation of immune-based side effects remain poorly understood. However, the covalent modification of cellular proteins by chemically reactive metabolites formed during the biotransformation of carboxylic-acid-containing drugs has been proposed to be a potential mechanism mediating some of these idiosyncratic allergic reactions. For profen-type drugs such as ibuprofen, acyl glucuronidation and acyl-CoA thioester formation are two common metabolic pathways leading to the formation of chemically reactive metabolites (Figure 1). Many recent studies have indicated that both acyl glucuronides and acyl-CoA thioester metabolites of carboxylic acid-containing drugs are able to transacylate GSH and proteins leading to thioester-, ester- or amide-linked covalent adducts (6, 20, 25, 29). Therefore, it has been proposed that these metabolites may play a role in the observed immune-based reactions for drugs in this chemical class. Covalent binding of carboxylic acid-containing drugs to plasma proteins mediated by acyl glucuronide metabolites has been well documented during the last two decades (8, 30). In addition, increasing reports in the literature concerning the chemical reactivity of xenobiotic acyl-CoA conjugates toward proteins have been appearing (10, 30–32). The present studies provide evidence that ibuprofen can be bioactivated in rat hepatocyte preparations leading to the transacylation GSH and that this transacylation mechanism is mediated by the I-CoA intermediate and not by the acyl glucuronide metabolite. The metabolism of ibuprofen has been

EnantioselectiVe BioactiVation of Ibuprofen

Chem. Res. Toxicol., Vol. 21, No. 9, 2008 1757

Figure 11. Concentration-dependent formation of I-CoA and I-SG in freshly isolated rat hepatocytes (2 × 106 cells/mL) incubated with increasing concentrations of (R)-(-)-ibuprofen for 3 min. Values are expressed as the mean ( SD of three incubations.

Figure 12. Time-dependent formation of I-SG from the reaction of (R,S)-I-CoA (10 µM) with GSH (10 mM) in potassium phosphate buffer (0.1 M, pH 7.4, 37 °C). Values are expressed as the mean ( SD of three incubations.

well characterized in rats, where a major route of metabolism is by acyl glucuronidation (33). Ibuprofen is also known to undergo extensive chiral inversion from the (R)-(-)- to the (S)-(+)-enantiomer Via an acyl-CoA thioester intermediate in animals and in humans (12–14). AcylCoA synthetases are a family of enzymes that catalyze the formation of acyl-CoA thioesters and that require the cofactors ATP, CoA, and Mg2+ Via an adenylate intermediate (34). The acyl-CoA synthetase enzyme that metabolizes (R)-(-)-ibuprofen to (R)-(-)-I-CoA is the same fatty acyl-CoA synthetase enzyme (EC 6.2.1.3) that functions in the metabolism of long-chain fatty acids found primarily in microsomal, mitochondrial, and potentially peroxisomal subcellular fractions, for example in the formation of palmitoyl-CoA from palmitic acid (35). The formation of I-CoA thioester is enantioselective for the (R)(-)-enantiomer (Figure 1). Conversely, acyl glucuronidation exhibits preference for the (S)-(+)-enantiomer cross-species (24). Both of these metabolic activation pathways, namely acyl-CoA formation and acyl glucuronidation, are potentially involved in covalent binding of ibuprofen to cellular nucleophiles such as proteins and GSH. Since enantioselective differences in the bioactivation of (R)-(-)- and (S)-(+)-ibuprofen isomers occur, we predicted that these differences would lead to enantioselec-

tive differences in transacylation of GSH in studies with hepatocyte preparations. We proposed that if bioactivation by I-CoA formation were important, then results from in vitro GSHconjugate detection studies would show I-SG adduct formation preferentially in incubations with the (R)-(-)-ibuprofen isomer. Conversely, if acyl glucuronidation was more important for the transacylation of GSH, then we predicted that incubations with the (S)-(-)-ibuprofen isomer would form more I-SG adducts. We based this hypothesis on results from previous studies by Li et al. (2002) (16) which showed that 2-PPA (the most structurally basic profen) was enantioselectively bioactivated by (R)-2-PPA-CoA formation in rat hepatocyte incubations leading to covalent binding to protein. In those studies, it was also shown that inhibition of 2-PPA-CoA formation led to a corresponding inhibition of covalent binding to rat hepatocyte proteins, whereas a complete inhibition of 2-PPA acyl glucuronidation only led to a small decrease in covalent binding to hepatocyte protein. Ibuprofen was selected for the present studies as a model profen-type drug to investigate the relative importance of acyl glucuronide and acyl-CoA thioester derivatives toward reactions with GSH in rat hepatocyte preparations. In the present study, we used a combination of rat hepatocytes as an in vitro tool, optically pure ibuprofen (R)-(-)- and (S)-(+)-enantiomers, stable isotope labeling, inhibitors of acyl-CoA formation and acyl glucuronidation, and LC-MS/MS to reveal that I-SG formation was enantioselective in favor of the(R)-(-)-ibuprofen enantiomer. The time course for the formation of I-SG determined in these studies was consistent with the time course for I-CoA formation, where the time of maximum concentration of I-CoA and I-SG formation were between 8 and 10 min of incubation (Figure 7). The time course of I-1-O-G formation was not consistent with the time course of I-SG formation, where I-1-O-G production increased throughout the 40 min incubation timeperiod (Supporting Information, Figure 1). After 10 min of incubation, the concentration of I-CoA and I-SG decreased rapidly, presumably due to metabolism by thioesterases, and much less due to chemical hydrolysis (29). As has been shown for the (R)-(-)- and (S)-(+)-I-CoA isomers (36), the degradation of I-SG in rat hepatocyte incubations is not stereoselective (Figure 10) and therefore cannot explain the enantioselective differences in the observed I-SG concentrations produced in the incubations. Results from these quantitative analyses of the I-SG metabolite formed in rat hepatocytes incubations showed it to be a very minor metabolic pathway where it represented only ∼0.05% of ibuprofen-CoA formed in the incubation (Figure 7). However, it is nonetheless a metabolite of mechanistic

1758

Chem. Res. Toxicol., Vol. 21, No. 9, 2008

significance with regard to understanding the important potential pathway(s) of reactive metabolite formation of the drug. Species differences in I-CoA formation are known, for example rat liver homogenate is ∼4-fold more efficient at forming ibuprofenCoA than human whole liver homogenate (13). However, since it has been shown that as much as 63% of a dose of (R)-(-)ibuprofen is stereospecifically inverted to the (S)-(+)-isomer in human subjects, it is obvious that I-CoA formation represents a pathway of significant quantitative importance in humans (37). Our hepatocyte studies with purified (R)-(-)- and (S)-(+)ibuprofen enantiomers showed that the transacylation of GSH forming I-SG is enantioselective for the (R)-(-)-ibuprofen enantiomer. The extent of I-SG formation, as indicated by a ratio of (R)- and (S)-ibuprofen AUC0-40 min, was 25.7-fold greater for (R)-(-)-ibuprofen than for incubations with (S)-(+)ibuprofen (Figure 7B). Such enantioselectivity of I-SG formation correlated more closely with the enantioselectivity of acyl-CoA formation (RAUC 0-40 min/SAUC 0-40 min ) 11.3, Figure 7A) than with the enantioselectivity of acyl glucuronidation (RAUC 0-40 min/SAUC 0-40 min ) ∼0.8, Supporting Information, Figure 1) of the (R)-(-)- and (S)-(+)-ibuprofen enantiomers. The enantioselective formation of I-CoA observed in these studies is in agreement with reports showing that I-CoA formation in incubations with rat hepatocytes, rat liver microsomes, rat liver mitochondria, or rat liver homogenate is highly enantioselective for the (R)-(-)-ibuprofen isomer (12, 13). These results strongly indicate the important role of acyl-CoA thioester for I-SG adduct formation in rat hepatocytes, which is consistent with results obtained from enzyme inhibition studies (discussed below, Table 1). In the present work, we also conducted experiments with a pseudoracemic mixture of (R)-(-)-[3,3,3-2H3]- and nonlabeled (S)-(+)-ibuprofen (100 µM total) which showed that almost all of the I-SG formed contained deuterium (D3-I-SG/D0-I-SG AUC0-40 min ratio determined to be ∼341) and therefore was derived primarily from (R)-(-)-[3,3,3-2H3]-ibuprofen bioactivation. The time course of D3-I-SG adduct formation (Figure 9B) did not correspond to the time course of I-1-O-G formation which was shown to increase throughout the incubation timeperiod (Supporting Information, Figure 2). No I-CoA could be detected from incubations with the D0-(S)-(+)-ibuprofen antipode in these studies (Figure 9A), however we did determine that the ratio of AUC0-40 min for I-1-O-G formation was RAUC0-40 min/SAUC0-40 min ) 0.5 (Supporting Information, Figure 2), which again is consistent with the known enantioselectivity of ibuprofen glucuronidation, but not consistent with measured enantioselectivity of I-SG formation. Concentration-dependent experiments with (R)-(-)-ibuprofen and rat hepatocytes were performed to determine an appropriate concentration of (R)-(-)-ibuprofen forming I-CoA and I-SG for use in subsequent inhibition experiments. Results showed a sharp increase in I-CoA formation up to the 125 µM (R)-(-)-ibuprofen incubation concentration reaching ∼2.3 µM I-CoA, and that 50% maximal formation of I-CoA occurred at ∼6.3 µM (R)(-)-ibuprofen (Figure 11A). No significant further increase in I-CoA formation was observed to occur above the 125 µM concentration. By contrast, there was an observed increase in I-SG formation with increasing (R)-(-)-ibuprofen concentration above 125 µM which is speculated to be due to an inhibition of degradation of I-SG at higher concentrations of (R)-(-)ibuprofen. We do not believe that the increase in I-SG formation is due to (R)-(-)-I-1-O-G formation at higher concentrations of (R)-(-)-ibuprofen, since corresponding experiments with (S)(+)-ibuprofen, which is also metabolized to the acyl glucuronide

Grillo and Hua

in incubations with rat hepatocytes (Figure 6; Supporting Information Figures 1 and 2), did not show a similar concentration-dependent increase in I-SG (data not shown). Based on the results from concentration-dependent I-CoA formation, selective inhibition of each of the metabolic pathways was performed in rat hepatocyte incubations with (R)-(-)ibuprofen (10 µM). Results showed that a decrease in I-SG adduct formation occurred only in incubations cotreated with inhibitors of I-CoA formation (Table 1), indicating that acylCoA formation was primarily involved in the transacylation of GSH forming I-SG. Thus, inhibition of acyl-CoA synthetases by known inhibitors of I-CoA formation such as valproic acid, pivalic acid (38), and lauric acid (22) led to a corresponding decrease in I-SG production. By contrast, an almost complete inhibition of acyl glucuronidation (∼98%) had no significant effect on I-SG adduct formation (Table 1). These results are similar to those found in similar inhibition experiments in hepatocytes with diclofenac, a phenylacetic acid-type NSAID, where a complete inhibition of diclofenac acyl glucuronidation had no effect on diclofenac GSH-thioester formation (25). These results, together with the time-dependent enantioselective I-SG formation, strongly suggest that the I-CoA intermediate plays a more important role in the transacylation of GSH by ibuprofen in hepatocyte preparations than the respective acyl glucuronide metabolite. Xenobiotic acyl-CoA thioesters are chemically reactive species that are able to transacylates endogenous nucleophiles. Examples of other known chemically reactive xenobiotic-acylCoA derivatives include salicyl-CoA (39), clofibryl-CoA (20), zomepirac-CoA (40), tolmetin-CoA (41), and nafenopin-CoA (31). In experiments with human liver homogenate, nafenopinCoA thioester mediated covalent binding to cysteine sulfhydryl groups which accounted for nearly 50% of the total covalent binding to hepatic protein (31). In the present work, (R)-(-)I-CoA was shown to react with GSH in a time-dependent fashion in incubations in phosphate buffer (pH 7.4, 37 °C), directly indicating the ability of I-CoA to transacylate GSH (Figure 12). The prediction of acyl-CoA thioester intermediates for their ability to covalently bind to protein is based on their chemical structures (29), which is analogous to covalent binding predictions for acyl glucuronide metabolites (6). In general, the chemical reactivity of acyl-CoA thioesters with biological nucleophiles, such as GSH and human serum albumin, decreases with increasing alkyl-substitution at the carbon adjacent to the carbonyl-carbon of the carboxylic acid moiety due to steric hindrance. In summary, in the present studies we performed enantioselective and selective enzyme inhibition studies in rat hepatocytes to determine that the metabolic activation of ibuprofen occurs by acyl-CoA formation, and not by acyl glucuronidation, leading to I-CoA that is able to transacylate GSH forming I-SG. Finally, from the results presented here, we propose that metabolic activation of ibuprofen by acyl-CoA formation may also lead to covalent binding of the drug to protein. Such drug-protein adducts might then function as antigens and contribute to immune-based reactions known to occur for ibuprofen and for other drugs in its chemical class (10, 42). We feel that it is important to consider the potential for metabolic activation by acyl-CoA formation when studying the covalent binding properties of carboxylic acid-containing drugs or drug candidates in early drug discovery. Acknowledgment. We thank Terri L. VandeGiessen, Lori R. Norris, and Brian W. Jones from the department of

EnantioselectiVe BioactiVation of Ibuprofen

Pharmacokinetics, Dynamics, and Metabolism, Pfizer, Inc., Kalamazoo, Michigan for assistance in the preparation of rat hepatocytes. Supporting Information Available: Time course for the formation of I-1-O-G from incubations with (R)-(-)- or (S)(+)-ibuprofen in freshly isolated rat hepatocytes and time course for the formation of I-1-O-G from incubations with a pseudoracemic misture of (R)-(-)-[3,3,3,-2H3]-ibuprofen and (S)-(+)ibuprofen in freshly isolated rat hepatocytes. This material is available free of charge via the Internet at http://pubs.acs.org.

References (1) Altman, R. D. (1984) Review of ibuprofen for osteoarthritis. Am. J. Med. 77, 10–18. (2) Evans, A. M. (1996) Pharmacodynamics and pharmacokinetics of profens: enantioselectivity, clinical implications, and special references to S-(+)-ibuprofen. J. Clin. Pharmacol. 36, 7S–15S. (3) Benet, L. Z., Øie, S., and Schwartz, J. B. (1996) Design and optimization of dosage regimens; Pharmacokinetic data. In Goodman and Gilman’s The Pharmacological Basis of Therapeutics (Wonsiewicz, M. J., and McCurdy, P., Eds.) 9th ed., p 1749. (4) Castillo, M., and Smith, P. C. (1995) Disposition and reactivity of ibuprofen and ibufenac acyl glucuronides in vivo in rhesus monkey and in vitro with human serum albumin. Drug Metab. Dispos. 23, 566–572. (5) Castillo, M., Lam, Y. W., Dooley, M. A., Stahl, E., and Smith, P. C. (1995) Disposition and covalent binding of ibuprofen and its acyl glucuronide in the elderly. Clin. Pharmacol. Ther. 57, 636–644. (6) Benet, L. Z., Spahn-Langguth, H., Iwakawa, S., Volland, C., Misuma, T., Mayer, S., Mutschler, E., and Lin, E. T. (1993) Predictability of covalent binding of acidic drugs in man. Life Sci. 53, 141–146. (7) Faed, E. M. (1984) Properties of acyl glucuronides: Implications for studies of the pharmacokinetics and metabolism of acidic drugs. Drug Metab. ReV. 15, 1213–1249. (8) Spahn-Langguth, H., and Benet, L. Z. (1992) Acyl glucuronides revisited: Is the glucuronidation process a toxification was well as a detoxification mechanism? Drug Metab. ReV. 24, 5–48. (9) Zia-Amirhosseini, P., Harris, R. Z., Brodsky, F. M., and Benet, L. Z. (1995) Hypersensitivity to nonsteroidal anti-inflammatory drugs. Nature Med. 1, 2–4. (10) Boelsterli, U. (2002) Xenobiotic acyl glucuronides and acyl CoA thioesters as protein-reactive metabolites with the potential to cause idiosyncratic drug reactions. Curr. Drug Metab. 3, 439–450. (11) Huxtable, R. (1986) Thiols, disulfides and Thioesters, in Biochemistry of Sulfur (Frieden, E., Ed.) pp 230-245, Plenum Press, New York. (12) Knadler, M. P., and Hall, S. D. (1990) Stereoselective arylpropionylCoA thioester formation in vitro. Chirality 2, 67–73. (13) Tracy, T. S., Wirthwein, D. P., and Hall, S. D. (1993) Metabolic inversion of (R)-ibuprofen. Formation of ibuprofenyl-coenzyme A. Drug Metab. Dispos. 21, 114–120. (14) Hall, S. D., and Xiaotao, Q. (1994) The role of coenzyme A in the biotransformation of 2-arylpropionic acids. Chem.-Biol. Interact. 90, 235–251. (15) Li, C., Benet, L. Z., and Grillo, M. P. (2002) Studies on the chemical reactivity of 2-phenylpropionic acid 1-O-acyl glucuronide and S-acylCoA thioester metabolites. Chem. Res. Toxicol. 15, 1309–1317. (16) Li, C., Benet, L. Z., and Grillo, M. P. (2002) Enantioselective covalent binding of 2-phenylpropionic acid to protein in vitro in rat hepatocytes. Chem. Res. Toxicol. 15, 1480–1487. (17) Sanins, S. M., Adams, W. J., Kaiser, D. G., Halstead, G. W., and Baillie, T. A. (1990) Studies on the metabolism and chiral inversion of ibuprofen in isolated rat hepatocytes. Drug Metab. Dispos. 18, 527– 533. (18) Shirley, M. A., Guan, X., Kaiser, D. G., Halstead, G. W., and Baillie, T. A. (1994) Taurine conjugation of ibuprofen in humans and in rat liver in vitro. Relationship to metabolic chiral inversion. J. Pharmacol. Exp. Ther. 269, 1166–1175. (19) Stadtman, E. R. (1957) Preparation and assay of acyl coenzyme A and other thiol esters; use of hydroxylamine. Methods Enzymol. 3, 931–946. (20) Grillo, M. P., and Benet, L. Z. (2002) Studies on the reactivity of clofibryl-S-acyl-CoA thioester with glutathione in vitro. Drug Metab. Dispos. 30, 55–62.

Chem. Res. Toxicol., Vol. 21, No. 9, 2008 1759 (21) Mo´ldeus, P., Hogberg, J., and Orrenius, S. (1978) The isolation and use of the use of liver cells. Methods Enzymol. 52, 69–71. (22) Li, C., Olurinde, M. O., Hodges, L. M., Grillo, M. P., and Benet, L. Z. (2003) Covalent binding of 2-phenylproionyl-S-acyl-CoA thioester to tissue proteins in vitro. Drug Metab. Dispos. 31, 727– 730. (23) Baillie, T. A., and Davis, M. R. (1993) Mass spectrometry in the analysis of glutathione conjugates. Biol. Mass Spectrom. 22, 319– 325. (24) el Mouelhi, M., Ruelius, H. W., Fenselau, C., and Dulik, D. M. (1987) Species-dependent enantioselective glucuronidation of three 2-arylpropionic acids. Naproxen, ibuprofen, and benoxaprofen. Drug Metab. Dispos. 15, 767–772. (25) Grillo, M. P., Hua, F., Knutson, C. G., Ware, J. A., and Li, C. (2003) Mechanistic studies on the bioactivation of diclofenac: identification of diclofenac-S-acyl-glutathione in vitro in incubations with rat and human hepatocytes. Chem. Res. Toxicol. 16, 1410–1417. (26) Hinchman, C. A., and Ballatori, N. (1990) Glutathione-degrading capacities of liver and kidney in different species. Biochem. Pharmacol. 40, 1131–1135. (27) Fung, M., Thornton, A., Mybeck, K., Wu, J. H., Hornbuckle, K., and Muniz, E. (2001) Evaluation of the characteristics of safety withdrawal of prescription drugs from world wide pharmaceutical markets-1960 to 1999. Drug Inform. J. 35, 293–317. (28) Bakke, O. M., Wardell, W. M., and Lasagna, L. (1984) Drug discontinuation in the United Kingdom and the United States: 19641983. Clin. Pharmacol. Ther. 35, 559–567. (29) Sidenius, J. M., Skonberg, C., Olsen, J., and Hansen, S. H. (2004) In vitro reactivity of carboxylic acid-CoA thioesters with glutathione. Chem. Res. Toxicol. 17, 75–81. (30) Skonberg, C., Olsen, J. O., Madsen, K., Honore´, S., and Grillo, M. P. (2008) Metabolic activation of carboxylic acids. Expert Opin. Drug Metab. Toxicol. 4, 425–438. (31) Sallustio, B. C., Nunthasomboon, S., Drogemuller, C. J., and Knights, K. M. (2000) In vitro covalent binding of nafenopin-CoA to human liver proteins. Toxicol. Appl. Pharmacol. 163, 176–182. (32) Mano, N., Uchida, M., Okuyama, H., Sasaki, I., Ikegawa, S., and Goto, J. (2001) Simultaneous detection of cholyl adenylate and coenzyme A thioester utilizing liquid chromatography/electrospray ionization mass spectrometry. Anal. Sci. 17, 1037–1042. (33) Mills, R. F. N., Adams, S. S., Cliffe, E. E., Dickinson, W., and Nicholson, J. S. (1973) The metabolism of ibuprofen. Xenobiotica 3, 589–598. (34) Menzel, S., Waibel, R., Brune, K., and Geisslinger, G. (1994) Is the formation of R-ibuprofenyl-adenylate the first stereoselective step of chiral inversion? Biochem. Pharmacol. 48, 1056–1058. (35) Bruggera, R., Reichel, C., Garcia, A. B., Brune, K., Yamamoto, T., Tegeder, I., and Geisslinger, G. (2001) Expression of rat liver longchain acyl-CoA synthetase and characterization of its role in the metabolism of R-ibuprofen and other fatty acid-like xenobiotics. Biochem. Pharmacol. 61, 651–656. (36) Tracy, T. S., and Hall, S. D. (1992) Metabolic inversion of (R)-(-)ibuprofen. Epimerization and hydrolysis of ibuprofenyl-coenzyme A. Drug Metab. Dispos. 20, 322–327. (37) Lee, E. J. D., Williams, K., Day, R., Graham, G., and Champion, D. (1985) Stereoselective disposition of ibuprofen enantiomers in man. Br. J. Clin. Pharmacol. 19, 669–674. (38) Xiaotao, Q., and Hall, S. D. (1993) Modulation of enantioselective metabolism and inversion of ibuprofen by xenobiotics in isolated rat hepatocytes. J. Pharmacol. Exp. Ther. 266, 845–851. (39) Tishler, S. L., and Goldman, P. (1970) Properties and reactions of salicyl-coenzyme A. Biochem. Pharmacol. 19, 143–150. (40) Olsen, J., Li, C., Bjørnsdottir, I., Sidenius, U., Benet, L. Z., Hansen, S. H., and Benet, L. Z. (2005) In Vitro and in vivo studies on acylcoenzyme A-dependent bioactivation of zomepirac in rats. Chem. Res. Toxicol. 18, 1729–1736. (41) Olsen, J., Li, C., Skonberg, C., Bjørnsdottir, I., Sidenius, U., Benet, L. Z., and Hansen, S. H. (2007) Studies on the metabolism of tolmetin to the chemically reactive acyl-coenzyme A thioester intermediate in rats. Drug Metab. Dispos. 35, 758–764. (42) Zimmerman, H. J. (1994) Hepatic Injury Associated with Nonsteroidal Anti-Inflammatory Drugs, in Nonsteroidal Antiinflammatory Drugs. Mechanisms and Clinical Use (Lewis, A. J., and Furst, D. E., Eds.) pp 171-194, Marcel Dekker Inc., New York.

TX800098H