In Vitro Reactivity of Carboxylic Acid-CoA Thioesters with Glutathione

Chemical Research in Toxicology 2015 28 (12), 2292-2303. Abstract | Full Text ... Chemical Research in Toxicology 2011 24 (9), 1345-1410. Abstract | F...
0 downloads 0 Views 147KB Size
Download PDF
Chem. Res. Toxicol. 2004, 17, 75-81

75

In Vitro Reactivity of Carboxylic Acid-CoA Thioesters with Glutathione† Ulrik Sidenius, Christian Skonberg,* Jørgen Olsen, and Steen Honore´ Hansen Department of Analytical Chemistry, The Danish University of Pharmaceutical Sciences, Universitetsparken 2, DK-2100 Copenhagen, Denmark Received June 24, 2003

The chemical reactivity of acyl-CoA thioesters toward nucleophiles has been demonstrated in several recent studies. Thus, intracellularly formed acyl-CoAs of xenobiotic carboxylic acids may react covalently with endogenous proteins and potentially lead to adverse effects. The purpose of this study was to investigate whether a correlation could be found between the structure of acyl-CoA thioesters and their reactivities toward the tripeptide, glutathione (γGlu-Cys-Gly). The acyl-CoA thioesters of eight carboxylic acids (ibuprofen, clofibric acid, indomethacin, fenbufen, tolmetin, salicylic acid, 2-phenoxypropionic acid, and (4-chloro-2methyl-phenoxy)acetic acid (MCPA)) were synthesized, and each acyl-CoA (0.5 mM) was incubated with glutathione (5.0 mM) in 0.1 M potassium phosphate (pH 7.4, 37 °C). All of the acyl-CoAs reacted with glutathione to form the respective acyl-S-glutathione products, with MCPA-CoA having the highest rate of conjugate formation (120 ( 10 µM/min) and ibuprofenCoA having the lowest (1.0 ( 0.1 µM/min). The relative reactivities of the acyl-CoAs were dependent on the substitution at the carbon atom R to the acyl carbon and on the presence of an oxygen atom in a position β to the acyl carbon and were as follows: phenoxyacetic acid > o-hydroxybenzoic acid ∼ phenoxypropionic acid > arylacetic acid derivatives > 2-methyl-2phenoxypropionic acid ∼ 2-phenylpropionic acid. For each acyl-CoA thioester, the overall hydrolysis rate was determined as the time-dependent formation of parent compound. A linear trend was observed when comparing the reactivities of the acyl-CoAs with glutathione with the corresponding overall hydrolysis rates. Thus, the most reactive compound (MCPA-CoA) was also the compound with the highest rate of hydrolysis and the least reactive compounds (ibuprofen-CoA, clofibryl-CoA) were also the compounds least susceptible to hydrolysis.

Introduction A serious problem in drug therapy is the occurrence of hepatotoxic idiosyncratic drug reactions (1, 2). Although the underlying mechanisms for these reactions are poorly understood, it has been hypothesized that formation of chemically reactive metabolites and subsequent reaction with proteins may be important for explaining these idiosyncratic reactions (1). Thus, it may not be the intrinsic toxicity of the drug itself that is responsible for the idiosyncratic reactions but rather the formation and reactivity of the metabolites. The theoretical consequences of drug-protein adduct formation are many and include alteration or inactivation of the functionality of proteins, alterations in cell signaling and homeostasis, and elicitation of an immune response through haptenation of cellular or plasma proteins (1). Several drugs belonging to the NSAID1 class of compounds have been reported to be involved in idiosyncratic reactions (3, 4). In the period between 1974 and 1993, nine NSAIDs were withdrawn from the market, which makes NSAIDs the therapeutic class most frequently † The main part of this work was presented in poster form at the 12th North American ISSX meeting in Providence, RI, October 2003. * To whom correspondence should be addressed. Tel: (+45)35 30 62 26. Fax: (+45)35 30 60 10. E-mail: [email protected]. 1 Abbreviations: NSAID, nonsteroidal antiinflammatory drug; MCPA, (4-chloro-2-methyl-phenoxy)acetic acid; POPA, 2-phenoxypropionic acid; GSH, glutathione; MeCN, acetonitrile; ESI, electrospray ionization; acyl-SG, acyl-S-glutathione.

involved in safety discontinuations (5). Although otherwise a chemically heterogeneous group, most of these drugs contain a carboxylic acid moiety, which frequently is metabolized by conjugation with glucuronic acid. Acyl glucuronides are known to be reactive intermediates that may bind covalently to GSH or proteins (6-9). In addition to acyl glucuronidation, carboxylic acid drugs may undergo metabolism to CoA-thioesters, which may further become metabolized to amino acid conjugates or take part in lipid neogenesis (10-13) or β-oxidation (14-16). Because of the chemical nature of the thioester bond, the CoA thioesters of carboxylic acids are more reactive than the corresponding 1-O-acyl glucuronides (8, 9, 17). As a consequence, acyl-CoA thioesters may also acylate proteins, as recently demonstrated for 2-phenylpropionic-CoA (18), nafenopin-CoA (19), and naproxenCoA (9). In fact, covalent adduct formation via the acylCoA thioester may be of greater importance than via the respective acyl glucuronides, according to recent studies by Li et al. (18) and by Olsen et al. (9). That the reactivity of acyl glucuronides can be predicted from the chemical structure of the corresponding carboxylic acid was shown by Benet et al. (20) and Bolze et al. (7), who reported that in vitro covalent binding to human serum albumin was dependent on the degree of substitution at the carbon atom R to the carbonyl carbon of the acyl group (Rcarbon). For example, a methyl group at the R-carbon decreased the reactivity of the acyl glucuronide as

10.1021/tx034127o CCC: $27.50 © 2004 American Chemical Society Published on Web 12/23/2003

76

Chem. Res. Toxicol., Vol. 17, No. 1, 2004

Figure 1. Chemical structures of the carboxylic acids used in the present study.

compared to the corresponding acetic acid derivative. In a recent study, Grillo and Benet showed a similar correlation between the structures of 11 acyl-SG conjugates and their reactivities with N-acetylcysteine (17). In addition, it has been shown that the degradation rate of the acyl glucuronides correlates with the covalent binding to protein in vitro and in vivo (7, 20). As the reactivity of acyl-CoA thioesters toward nucleophiles has been established, it was the aim of this study to examine the relative reactivities of the acyl-CoA thioesters of eight carboxylic acids (Figure 1) and to investigate the correlation to structure and hydrolysis rate (Scheme 1). The carboxylic acids were selected to represent compounds with different substitution at the R-carbon that potentially affect the electron distribution and steric hindrance close to the R-carbon. The selected carboxylic acids were salicylic acid, ibuprofen, indomethacin, fenbufen, tolmetin, clofibric acid, POPA, and MCPA. As a model, the reaction of the acyl-CoA thioesters with GSH was examined in accordance with previous studies on the reactivity of acyl-CoA thioesters and acyl glucuronides with nucleophiles (8, 9, 17, 21, 22).

Experimental Procedures Chemicals. Clofibric acid, ibuprofen, MCPA, POPA, indomethacin, fenbufen, ketoprofen, tolmetin, zomepirac, GSH, and anhydrous THF were purchased from Sigma-Aldrich Chemie (Steinhem, Germany). Naproxen was obtained from Syntex Research (Palo Alto, CA). Salicylic acid was from Mecobenzon (Copenhagen, Denmark). CoA was purchased from Applichem (Darmstadt, Germany) as a trilithium dihydrate salt. 1,1′Carbonyldiimidazole was obtained from Fluka Chemie AG (Buchs, Switzerland). Ammonium acetate and p-chlorobenzoic acid were from Merck (Hohenbrunn, Germany), and MeCN was from BDH Laboratory Supplies (Poole, England). Instrumentation. Preparative HPLC was preformed on a Shimadzu system equipped with a LC-10AD pump and a SPD10A UV-vis detector (for the purification of synthetic indomethacin-CoA and fenbufen-CoA) or on an Agilent 1100 Series system (for the purification of POPA-CoA, salicyl-CoA, and

Scheme 1. Reaction of an Acyl-CoA with GSH and Hydrolysis of the Thioesters to Carboxylic Acid

Sidenius et al. tolmetin-CoA). LC-MS and LC-MS/MS were performed using an Agilent 1100 Series LC/MSD Trap. Synthesis of Acyl-CoA Thioesters. The acyl-CoA thioesters of the carboxylic acids were synthesized according to the method of Kawaguchi et al. (23) with minor modifications. Briefly, the carboxylic acid (free acid form, 100 µmol), was dissolved in anhydrous THF (1 mL). 1,1′-Carbonyldiimidazole (200 µmol) dissolved in anhydrous THF (1 mL) was added and allowed to react with the carboxylic acid at room temperature in a nitrogen atmosphere. After 1-2 h, the reaction was stopped by addition of water (0.5 mL). To the reaction mixture was added CoA (50 µmol, as the trilithium salt) dissolved in 0.5 mL of water. When the formation of acyl-CoA ceased (as monitored by HPLC analysis with UV detection), the reaction was stopped by addition of 1 M HCl until the final pH in the solution was 2-3. Purification of Acyl-CoA Thioesters. The nonreacted carboxylic acid was removed from the reaction mixture by extraction with 2 × 2 mL heptane and 2 × 2 mL ethyl acetate. Clofibryl-CoA, ibuprofen-CoA, and MCPA-CoA were purified by solid phase extraction (SPE). The SPE columns were conditioned by 1 mL of MeCN and 1 mL of H2O. The water phase was applied to a Varian Bond-Elut C18 column and eluted consecutively with H2O and 10, 25, 50, and 100% MeCN. All of the fractions were analyzed by reversed phase HPLC (Table 1). Fractions containing acyl-CoA were pooled, evaporated to dryness, redissolved in 0.1 M potassium phosphate (pH 7.4) and stored at -20 °C until use. POPA-CoA, indomethacin-CoA, fenbufen-CoA, tolmetin-CoA, and salicyl-CoA were purified from the water phases by preparative HPLC. Indomethacin-CoA and fenbufen-CoA were purified using an Xterra PrepMS C18 5 µm, 19 mm × 100 mm column with a flow of 9 mL/min and a mobile phase consisting of H2O:MeCN:1 M ammonium acetate 65:35:1 (v:v:v). POPACoA was purified on a Spherisorb 5 µm, 8 mm × 250 mm column with a flow of 2.5 mL/min using a gradient (0-50% B 15 min, 50% B for 5 min). The mobile phases were A, H2O:MeCN:1 M ammonium acetate 95:5:1 (v:v:v), and B, H2O:MeCN:1 M ammonium acetate 40:60:1 (v:v:v). Tolmetin-CoA and salicyl-CoA were purified using a Luna C18 5 µm 4.6 mm × 250 mm column with a flow of 1.0 mL/min using the same mobile phases and gradient as were used for purification of POPA-CoA. Fractions containing acyl-CoA were collected, pooled, evaporated to dryness, and redissolved in 0.1 M potassium phosphate (pH 7.4). The solutions were stored at -20 °C until use. The identity and purity of the synthesized and purified acylCoAs were determined by LC-MS and LC-MS/MS with ESI. The LC conditions were as described in Table 1, except LC columns with an i.d. of 2.0 mm and a flow rate of 0.2 mL/min were used. The mass spectrometer was operated in positive ion mode, with the following settings: drying temperature, 350 °C; nebulizer pressure, 30 psi; drying gas, 8 L/min; spray voltage, 3500 V; trap drive, 50-65 (arbitrary units). Determination of Concentrations and UV Response Factors. To quantify acyl-CoA and acyl-SG thioesters, the UV response factors of the individual thioesters relative to the free carboxylic acids were determined. This was done by calculating the ratio between the HPLC peak area of the parent compound, formed by alkaline hydrolysis (5 M NaOH, 60 °C for 1 h) of the acyl-CoA thioester or acyl-SG conjugate, and the corresponding area of the nonhydrolyzed thioester. Using the UV response factors, it was possible to quantify the acyl-CoA and acyl-SG thioesters from a standard curve of the free carboxylic acid. The alkaline lability of the carboxylic acids at elevated temperatures was evaluated, similar to the acyl thioesters, using carboxylic acid instead of acyl-CoA. As indomethacin was unstable in alkaline conditions and the response factors for the thioesters of indomethacin could not be determined, it was assumed that the molar absorptivities for indomethacin and the indomethacin thioesters were identical at 320 nm. Reactivity of Acyl-CoA Thioesters with GSH. Acyl-CoA (final concentration 0.5 mM), GSH (final concentration 5 mM), and internal standard (final concentration 0.2 mM; see Table 1

Reactivity of Acyl-CoA Thioesters with GSH

Chem. Res. Toxicol., Vol. 17, No. 1, 2004 77

Table 1. Chromatographic Methods for Analysis of Incubations of Acyl-CoA with GSH

a

drug

column

wavelength (nm)

internal standard

mobile phase H2O/MeCN/ 1 M ammonium acetate

clofibric acid fenbufen ibuprofen indomethacin MCPA POPA salicylic acid tolmetin

Luna phenyl-hexyla Luna phenyl-hexylb Luna phenyl-hexyla Luna phenyl-hexyla Luna phenyl-hexyla Luna C18c Luna C18a Luna C18b

226 280 220 320 226 226 226 320

fenbufen naproxen ketoprofen tolmetin clofibric acid MCPA p-chlorobenzoic acid zomepirac

77/23/1 (pH 7.0) v/v/v 67.5/32.5/1 (pH 7.0) v/v/v 73/27/1 (pH 6.0) v/v/v 65/35/1 (pH 7.0) v/v/v 80/20/1 (pH 7.0) v/v/v 88/12/1 (pH 7.0) v/v/v 90/10/1 (pH 7.0) v/v/v 80/20/1 (pH 6.0) v/v/v

Column dimension (4.6 mm × 150 mm). b Column dimension (2.0 mm × 150 mm). c Column dimension (4.6 mm × 100 mm).

for the respective internal standards for each experiment) were mixed and incubated in triplicate at 37 °C in 0.1 M potassium phosphate buffer (pH 7.4). Ibuprofen-CoA, clofibryl-CoA, fenbufen-CoA, and indomethacin-CoA were incubated in a temperature-controlled heating block in an Agilent 1100 series autosampler and at regular time points, determined by the duration of the HPLC method, 2 µL of sample was injected and analyzed by HPLC as described in Table 1. MCPA-CoA, POPA-CoA, tolmetin-CoA, and salicyl-CoA were incubated off-line in the same heating block as used above, and at selected time points, a 25 µL aliquot was added to 75 µL of ice cold 1% (v/v) acetic acid to stop the reaction with GSH. (Pilot studies with fenbufen-CoA had shown that the two incubation methods, on-line and off-line, gave essentially the same results and time curves for acyl-SG product formation [data not shown]). The samples were kept at -20 °C until analysis by HPLC as described in Table 1. As described in the previous section, the acyl-CoAs and acylSGs were quantified relative to a series of standards (0.05, 0.10, 0.25, and 0.50 mM of the respective carboxylic acid), all containing 0.2 mM internal standard. The identities of the

respective acyl glutathione thioester products formed were determined by LC-ESI-MS/MS (vide supra). The slopes of the initial, linear part of plots of acyl-SG concentration vs time were used to determine reaction rates of the acyl-CoAs with GSH, and this was used as a measure of the chemical reactivities of the acyl-CoAs with GSH. The stability of the acyl-CoA and the corresponding acyl-SG thioesters were determined from the slope from plots of parent compound formation vs time.

Results Synthesis of Acyl-CoA Thioesters. The identity of the synthesized and purified acyl-CoAs was determined with LC-ESI-MS/MS, and the spectra obtained (Table 2) were in accordance with the fragmentation patterns previously published for related acyl-CoAs (8, 17, 21, 24). Incubation of Acyl-CoA Thioesters with GSH. Synthesized and purified acyl-CoA at a concentration of 0.5 mM was incubated with 5.0 mM GSH in 0.1 M potassium phosphate (pH 7.4) at 37 °C. Two different

Table 2. MS/MS Data for the Synthesized Acyl-CoAsa

a m/z values and relative intensities are given for each compound. *, Loss of one water molecule was observed for this fragment. ()*, Relative intensity of the fragment ion after loss of one water molecule.

78

Chem. Res. Toxicol., Vol. 17, No. 1, 2004

Figure 2. (a) Representative chromatograms obtained from analysis of the incubation of 0.5 mM ibuprofen-CoA with 5 mM GSH in 0.1 M phosphate buffer (pH 7.4, 37 °C). UV trace at 220 nm (see Table 1 for details on chromatographic method). Incubation mixture analyzed after 0 min of incubation (;) and 283 min of incubation (- - -). (b) Concentration vs time plots of ibuprofen-SG (×) and ibuprofen-CoA (9) from incubations of ibuprofen-CoA (0.5 mM) with GSH (5 mM) in 0.1 M phosphate buffer (pH 7.4, 37 °C). Values represent the mean ( SD of three incubations.

methods of incubation were used for these studies, and identical results were obtained when analyzing the reactivity of fenbufen-CoA by both methods. Thus, it seems valid to compare the results obtained by the two methods. HPLC methods were developed for the analysis of the incubation mixtures (Table 1), and baseline separation of carboxylic acid, acyl-SG, acyl-CoA, and internal standard was achieved for each method. A typical chromatogram from the experiments is shown in Figure 2a. The identities of the individual peaks were determined by LC-ESI-MS/MS. MS/MS data obtained for all analyzed compounds are shown in Tables 2 (acylCoAs) and 3 (acyl-SGs). The following fragments were considered characteristic for acyl-CoAs: [M + 2H 409]+, [M + H - 427 (adenosine diphosphate)]+, [M + H - 507 (adenosine triphosphate)]+, [M + H - 609]+, and m/z 428 [adenosine diphosphate + 2H]+ (Table 2). For the acyl-SG thioesters, the characteristic fragments were as follows: [M + H - 75 (Gly)]+, [M + H - 129 (pyroglutamic acid)]+, [M + H - 204 (Gly + Glu)]+, [M + H - 232 (Gly + Glu + CO)]+ (Table 3). The obtained mass spectra of the acyl-CoA and acyl-SG (Table 2) agreed well with the fragmentation patterns observed by others (8, 17, 21, 24). Reactivity of Acyl-CoA Thioesters with GSH. The degradation of the acyl-CoA and the formation of the acylSG took place in a time-dependent manner, as represented by the result from the experiments with ibuprofenCoA in Figure 2b. This profile was similar for all compounds investigated. The different acyl-SGs were formed with different rates, with ibuprofen-SG and clofibryl-SG having the

Sidenius et al.

slowest rates of formation and MCPA-SG having the fastest rate (Figure 3). The other compounds were placed between with no distinct groupings. Calculating the rates of formation of the acyl-SGs from the initial slopes of concentration vs time plots showed an approximate 100-fold difference between the slowest (ibuprofen-SG, 1.0 ( 0.1 µM/min) and the fastest (MCPASG, 120 ( 10 µM/min) formation rates, with fenbufenSG having an intermediary rate (8.1 ( 0.4 µM/min) (Table 4). The rates of formation of the acyl-SGs were of the same magnitude as the degradation rates of the corresponding acyl-CoA, with ibuprofen-CoA having the slowest degradation rate (1.3 ( 0.2 µM/min) and MCPACoA having the fastest degradation rate (99 ( 7 µM/min) (Table 4). Thus, the reactivity of the acyl-CoAs with GSH could be assessed as the initial rate of formation of the acyl-SG and as the initial rate of degradation of acylCoA. Stability of Acyl Thioesters. The formation of significant amounts of parent compound by hydrolysis of acyl-CoAs and acyl-SGs was observed in all of the incubation experiments, except in experiments with ibuprofen-CoA, where only trace amounts of ibuprofen (eluting at 5.9 min) were observed after 283 min of incubation (Figure 2a). The overall hydrolysis rates of the acyl thioesters (acylCoA and acyl-SG) were determined as the rate of formation of parent compound. The overall hydrolysis rates were approximately 50 times lower than the formation rates of acyl-SG (and degradation rates of acylCoA), except for the hydrolysis rate of the fenbufen thioesters, which was 10 times lower than the formation rate of fenbufen-SG and degradation rate of fenbufenCoA (Table 4). Prediction of the Reactivity of Acyl-CoA Thioesters with GSH. When the calculated overall hydrolysis rates were compared with the formation rate of acylSG, a linear trend was observed (Table 4 and Figure 4). Thus, the most reactive compound (MCPA-CoA) was also the compound with the highest rate of hydrolysis, and the least reactive compounds (ibuprofen-CoA, clofibrylCoA) were also the compounds least susceptible for hydrolysis.

Discussion Numerous examples of the in vitro and in vivo reactivity of acyl glucuronides toward proteins and peptides have been documented (7, 22, 25-29), but only recently has interest turned to the relative importance of acylCoA thioesters in similar reactions. Several studies have now shown that acyl-CoA thioesters are also reactive toward GSH and proteins (9, 17-19, 22, 30, 31), and as compared to the acyl glucuronides, the acyl-CoAs have proven much more reactive (9, 17, 18). To investigate the relationship between structure and reactivity for acyl-CoA thioesters, the CoA thioesters of eight structurally different carboxylic acids (Figure 1) were synthesized and the reactivities of the CoA thioesters with GSH were determined in vitro. The carboxylic acids were selected to represent groups of compounds with different substitution at the R-carbon. Thus, methylsubstituted compounds and phenoxy compounds were chosen along with one compound (salicylic acid) with a benzene ring attached directly to the acyl carbon. Comparing the reactivities of the acyl-CoAs with GSH, the

Reactivity of Acyl-CoA Thioesters with GSH

Chem. Res. Toxicol., Vol. 17, No. 1, 2004 79

Table 3. MS/MS Data for Acyl-SGs from Incubations of Acyl-CoA wiht GSHa

a m/z values and relative intensities are given for each compound. *, Loss of one water molecule was observed for this fragment. -, Fragment not observed.

rate of formation of the acyl-SGs was dependent on the structure of the acyl-CoA. This was most evident when comparing the relative reactivities of the phenoxy compounds (clofibryl-CoA, POPA-CoA, and MCPA-CoA). In this series of compounds, the reactivity decreased with increasing substitution at the R-carbon, resulting in the lowest reactivity for clofibryl-CoA, with two methyl groups at the R-carbon, and the highest reactivity for MCPA-CoA, with no substitution at the R-carbon (Figure 1 and Table 4). A similar relationship in reactivity with GSH and substitution of the R-carbon was observed when comparing reactivities of the acyl-CoAs of the arylacetic acid compounds. Indomethacin-CoA, tolmetin-CoA, and fenbufenCoA (with no substitution at the R-carbon) were more reactive than ibuprofen-CoA (one methyl group at the R-carbon). Finally, when comparing the reactivities of acyl-CoAs with identical numbers of methyl substituents on the carbon atom in the R-position to the acyl carbon atom, the reactivities of the phenoxy derivatives were higher than the corresponding reactivities of the aryl compounds. Thus, MCPA-CoA was more reactive than indomethacin-CoA, fenbufen-CoA, and tolmetin-CoA (no

substitution at the R-carbon), and POPA-CoA was more reactive than ibuprofen-CoA (one methyl group at the R-carbon). As a consequence, it may be concluded that when the number of methyl groups at the carbon atom R to the acyl carbon increases, the reactivity of the corresponding acyl-CoA with GSH decreases, and when the aryl group is substituted with a phenoxy group, the reactivity increases. These results are in good agreement with the previously reported relationship between the structure of acyl glucuronides and the covalent binding to protein (7, 20). A similar relationship was observed by Grillo and Benet for the reaction between acyl-SG derivatives and N-acetylcysteine in vitro (17). In the latter study, it was likewise found that the presence of an oxygen atom in the β-position to the carboxylic acid group had an increasing effect on the reactivity (17), and very recently, Li et al. observed the same effect by comparing the reactivities of the acyl-CoAs of 2-phenylpropionic acid and 2,4-dichlorophenoxyacetic acid with GSH. (21). Although no direct investigations have been conducted on the relative reactivities of the glucuronides of phenoxy derivatives, we find it reasonable to assume that the glucuronides of phenoxy derivatives would react in

80

Chem. Res. Toxicol., Vol. 17, No. 1, 2004

Sidenius et al.

Figure 4. Correlation between the formation of acyl-SG and the overall hydrolysis rate of the acyl-thioesters from incubations of acyl-CoA with GSH in 0.1 M phosphate buffer (pH 7, 37 °C). Full line shows the linear regression with 95% confidence intervals (dotted lines).

Figure 3. Time-dependent formation of acyl-SGs from incubations of acyl-CoA (0.5 mM) with GSH (5 mM) in 0.1 M phosphate buffer (pH 7.4, 37 °C). Values represent means ( SD of three incubations. Ibuprofen-SG (9), clofibryl-SG ([), indomethacinSG (2), fenbufen-SG (×), tolmetin-SG (O) salicyl-SG (b), POPASG (+), and MCPA-SG. (4).

the same way, relative to the degree of substitution, as we have found for phenoxy-acyl-CoAs in the present study. In a study by Bolze et al., a good correlation was observed between the relative reactivities of structurally different acyl glucuronides with human serum albumin in vitro and their respective hydrolysis rates, when corrected for the relative propensities of the 1-O-glucuronides to isomerize (7). To investigate a similar correlation in our study, the rates of hydrolysis of the acyl thioesters were determined from the time-dependent formation of parent compound, from hydrolysis of the respective acyl-CoA and acyl-SG (Table 4). We assumed that the rate of hydrolysis of the acyl-CoA and the rate of hydrolysis of the acyl-SG were similar, which was Table 4. Initial Rates of Formation for Acyl-SG, Degradation of Acyl-CoA, and Formation of Parent Compound

MCPA salicylic acid POPA tolmetin fenbufen indomethacin clofibric acid ibuprofen

acyl-SG formation (µM/min) (( SE)

acyl-CoA degradation (µM/min) (( SE)

formation of parent compound (µM/min) (( SE)

120 ( 10 29 ( 1 25 ( 1 13.3 ( 0.3 8.1 ( 0.4 3.9 ( 0.2 1.6 ( 0.1 1.0 ( 0.1

99 ( 7 23 ( 1 19 ( 1 12.1 ( 2.5 7.6 ( 0.3 3.9 ( 0.8 1.7 ( 0.3 1.3 ( 0.2

2.6 ( 0.1 1.1 ( 0.1 0.29 ( 0.02 0.14 ( 0.02 0.85 ( 0.02 0.055 ( 0.004 0.037 ( 0.009 ND

a Values are calculated as the slope of the linear part of the data points from incubations of 0.5 mM acyl-CoA with 5 mM GSH in phosphate buffer (pH 7.4, 37 °C). Values represent means ( SE of three incubations.

supported by our data (not shown) as the initial slopes from plots of concentrations of carboxylic acids vs time (when the concentration of acyl-CoA was high) were not significantly different from the final slopes (when the concentration of acyl-SG was high). As only trace amounts of ibuprofen were detected after 283 min of incubation, the hydrolysis rate of the ibuprofen thioesters could not be determined from this experiment. This was in accordance with the previously reported stability of 2-phenylpropionic-CoA, with a halflife of 12 days (8). Furthermore, we found that the thioesters of MCPA hydrolyzed fast (estimated half-life approximately 2 h), which is consistent with a recent reported half-life of 5.4 h for the hydrolysis of 2,4dichlorophenoxyacetyl-CoA (21). However, the data presented here showed that the rate of hydrolysis of the clofibryl thioesters was faster than the rate of hydrolysis of the ibuprofen thioesters. This was not in accordance with the previously reported half-life of clofibryl-CoA of 21 days (17). The same relationship found between substitution at the R-carbon and reactivity of the acyl-CoA thioesters was also observed for the substitution and the hydrolysis rates of the acyl thioesters. The most reactive compound (MCPA-CoA) was also the compound with the fastest rate of hydrolysis, and the least reactive compounds (ibuprofen-CoA, clofibryl-CoA) were also the compounds least susceptible to hydrolysis. A linear trend was observed in reactivity vs hydrolysis, with only fenbufen-CoA not fitting the line. The reason for this deviation is unclear. However, the data presented here indicate that the reactivity of the acyl-CoA thioesters may be predicted from the rate of hydrolysis, although additional data points may be necessary in order to estimate a statistical correlation. This is in accordance with previous studies on the reactivities of acyl-glucuronides, in which a correlation between the degradation rate of the 1-O-acyl-glucuronide (hydrolysis and acyl migration) and the covalent binding to human serum albumin was demonstrated (20), as well as a correlation between the aglycone appearance rate (overall hydrolysis, cor-

Reactivity of Acyl-CoA Thioesters with GSH

rected for the propensity for acyl-glucuronide isomerization) and the extent of covalent drug binding to serum albumin (7). In this study, a relatively high reactivity of salicyl-CoA, with respect to the structure at the R-carbon, was observed. Previously, a low degree of covalent binding to protein has been reported for the furosemide-glucuronide (7, 20) and the telmisartan-glucuronide (32), which is in accordance with the structure R to the acyl carbon atom (a benzene ring). We find, however, that the salicylCoA thioester is more reactive than would be expected from the substitution at the R-carbon alone. This is similar to the observation by Tishler and Goldman in 1970, when comparing the in vitro reactivity of salicylCoA with glycine, with that of benzoyl-CoA (33). This difference cannot be related to the substitution at the R-carbon alone but is likely due to the aromatic hydroxyl group ortho to the carboxylic acid group. This indicates that the substitution at the R-carbon is not the only determinant for reactivity toward nucleophiles. In conclusion, a correlation between the structure of acyl-CoA thioesters and the reactivities toward the nucleophile GSH has been demonstrated. Furthermore, the data obtained indicate that the reactivities of the acylCoA thioesters may be predicted from the rate of hydrolysis of the thioester.

Acknowledgment. We thank Bente Larsen for skilled technical assistance. This work was supported in part by the Danish Technical Research Council (Grant No. 5601-0014).

References (1) Lee, W. M. (2003) Drug-Induced Hepatotoxicity. N. Engl. J. Med. 349 (5), 474-485. (2) Pirmohamed, M., Breckenridge, A. M., Kitteringham, N. R., and Park, B. K. (1998) Adverse Drug Reactions. Br. Med. J. 316, 1295-1298. (3) Boelsterli, U. A., Zimmerman, H. J., and Kretz-Rommel, A. (1995) Idiosyncratic liver toxicity of nonsteroidal antiinflammatory drugs: molecular mechanisms and pathology. Crit. Rev. Toxicol. 25 (3), 207-235. (4) Zimmerman, H. J. (2002) Hepatic Injury Associated with Nonsteroidal Antiinflammatory Drugs. In Nonsteroidal Antiinflammatory Drugs: Mechanisms and Clinical Uses, 2nd ed. (Lewis, Aj. J., and Gay, G. R., Eds.) pp 171-194, Marcel Dekker, New York. (5) Bakke, O. M., Manocchia, M., de Abajo, F., Kaitin, K. I., and Lasagna, L. (1995) Drug safety discontinuations in the United Kingdom, the United States, and Spain from 1974 through 1993: a regulatory perspective. Clin. Pharmacol. Ther. 58 (1), 108-117. (6) Boelsterli, U. A. (2002) Xenobiotic Acyl Glucuronides and Acyl CoA Thioesters as Protein-Reactive Metabolites With the Potential to Cause Idiosyncratic Drug Reactions. Curr. Drug Metab. 3 (4), 439-450. (7) Bolze, S., Bromet, N., Gay-Feutry, C., Massiere, F., Boulieu, R., and Hulot, T. (2002) Development of an in vitro screening model for the biosynthesis of acyl glucuronide metabolites and the assessment of their reactivity toward human serum albumin. Drug Metab. Dispos. 30 (4), 404-413. (8) 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-acyl-CoA thioester metabolites. Chem. Res. Toxicol. 15 (10), 1309-1317. (9) Olsen, J., Bjornsdottir, I., Tjornelund, J., and Honore, H. S. (2002) Chemical reactivity of the naproxen acyl glucuronide and the naproxen coenzyme A thioester towards bionucleophiles. J. Pharm. Biomed. Anal. 29 (1-2), 7-15. (10) Dodds, P. F. (1991) Incorporation of xenobiotic carboxylic acids into lipids. Life Sci. 49 (9), 629-649. (11) Dodds, P. F., Chou, S. C., Ranasinghe, A., and Coleman, R. A. (1995) Metabolism of fenbufen by cultured 3T3-L1 adipocytes: synthesis and metabolism of xenobiotic glycerolipids. J. Lipid Res. 36 (12), 2493-2503.

Chem. Res. Toxicol., Vol. 17, No. 1, 2004 81 (12) Moorhouse, K. G., Dodds, P. F., and Hutson, D. H. (1991) Xenobiotic triacylglycerol formation in isolated hepatocytes. Biochem. Pharmacol. 41 (8), 1179-1185. (13) Sallustio, B. C., Meffin, P. J., and Knights, K. M. (1988) The stereospecific incorporation of fenoprofen into rat hepatocyte and adipocyte triacylglycerols. Biochem. Pharmacol. 37 (10), 19191923. (14) Browne, G. S., Nelson, C., Nguyen, T., Ellis, B. A., Day, R. O., and Williams, K. M. (1999) Stereoselective and substrate-dependent inhibition of hepatic mitochondria beta-oxidation and oxidative phosphorylation by the nonsteroidal antiinflammatory drugs ibuprofen, flurbiprofen, and ketorolac. Biochem. Pharmacol. 57 (7), 837-844. (15) Foxworthy, P. S., Perry, D. N., and Eacho, P. I. (1993) Induction of peroxisomal beta-oxidation by nonsteroidal antiinflammatory drugs. Toxicol. Appl. Pharmacol. 118 (2), 271-274. (16) Fromenty, B. and Pessayre, D. (1995) Inhibition of Mitochondrial Beta-Oxidation As A Mechanism of Hepatotoxicity. Pharmacol. Ther. 67 (1), 101-154. (17) 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 (1), 55-62. (18) Li, C., Grillo, M. P., and Benet, L. Z. (2003) In vivo mechanistic studies on the metabolic activation of 2-phenylpropionic Acid in rat. J. Pharmacol. Exp. Ther. 305 (1), 250-256. (19) 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 (2), 176182. (20) Benet, L. Z., Spahn-Langguth, H., Iwakawa, S., Volland, C., Mizuma, T., Mayer, S., Mutschler, E., and Lin, E. T. (1993) Predictability of the Covalent Binding of Acidic Drugs in Man. Life Sci. 53 (8), L141-L146. (21) Li, C., Grillo, M. P., and Benet, L. Z. (2003) In vitro studies on the chemical reactivity of 2,4-dichlorophenoxyacetyl-S-acyl-CoA thioester. Toxicol. Appl. Pharmacol. 187 (2), 101-109. (22) Shore, L. J., Fenselau, C., King, A. R., and Dickinson, R. G. (1995) Characterization and Formation of the Glutathione Conjugate of Clofibric Acid. Drug Metab. Dispos. 23 (1), 119-123. (23) Kawaguchi, A., Yoshimura, T., and Okuda, S. (1981) A new method for the preparation of acyl-CoA thioesters. J. Biochem. (Tokyo) 89 (2), 337-339. (24) Olsen, J. (2002) Reactive Drug Metabolites: Formation, metabolism and reactivity of acyl coenzyme A thioesters in vitro and in vivo. Ph.D. Thesis, Department of Analytical and Pharmaceutical Chemistry, The Royal Danish School of Pharmacy. (25) Dickinson, R. G., Baker, P. V., and King, A. R. (1994) Studies on the reactivity of acyl glucuronides- -VII. Salicyl acyl glucuronide reactivity in vitro and covalent binding of salicylic acid to plasma protein of humans taking aspirin. Biochem. Pharmacol. 47 (3), 469-476. (26) Ding, A., Ziaamirhosseini, P., McDonagh, A. F., Burlingame, A. L., and Benet, L. Z. (1995) Reactivity of Tolmetin Glucuronide with Human Serum-Albumin - Identification of Binding-Sites and Mechanisms of Reaction by Tandem Mass-Spectrometry. Drug Metab. Dispos. 23 (3), 369-376. (27) Ojingwa, J. C., Spahn-Langguth, H., and Benet, L. Z. (1994) Irreversible binding of tolmetin to macromolecules via its glucuronide: binding to blood constituents, tissue homogenates and subcellular fractions in vitro. Xenobiotica 24 (6), 495-506. (28) Qiu, Y., Burlingame, A. L., and Benet, L. Z. (1998) Mechanisms for covalent binding of benoxaprofen glucuronide to human serum albumin - Studies by tandem mass spectrometry. Drug Metab. Dispos. 26 (3), 246-256. (29) Smith, P. C., McDonagh, A. F., and Benet, L. Z. (1986) Irreversible binding of zomepirac to plasma protein in vitro and in vivo. J. Clin. Invest. 77 (3), 934-939. (30) Kassahun, K., Hu, P., Grillo, M. P., Davis, M. R., Jin, L., and Baillie, T. A. (1994) Metabolic activation of unsaturated derivatives of valproic acid. Identification of novel glutathione adducts formed through coenzyme A-dependent and -independent processes. Chem.-Biol. Interact. 90 (3), 253-275. (31) Grillo, M. P., and Benet, L. Z. (2001) Interaction of gammaglutamyltranspeptidase with clofibryl-S-acyl-glutathione in vitro and in vivo in rat. Chem. Res. Toxicol. 14 (8), 1033-1040. (32) Ebner, T., Heinzel, G., Prox, A., Beschke, K., and Wachsmuth, H. (1999) Disposition and chemical stability of telmisartan 1-Oacylglucuronide. Drug Metab. Dispos. 27 (10), 1143-1149. (33) Tishler, S. L., and Goldman, P. (1970) Properties and reactions of salicyl-coenzyme A. Biochem. Pharmacol. 19 (1), 143-150.

TX034127O