Chem. Res. Toxicol. 2004, 17, 137-143
137
In Vitro Metabolism of 2-Acetylbenzothiophene: Relevance to Zileuton Hepatotoxicity† Elizabeth M. Joshi, Brian H. Heasley, Mahendra D. Chordia, and Timothy L. Macdonald* Department of Chemistry, University of Virginia, Charlottesville, Virginia 22901 Received July 2, 2003
Zileuton, an inhibitor of 5-lipooxygenase, the initial enzyme in the leukotriene pathway, was marketed as a new treatment for asthma. This drug has been associated with liver toxicity, which has limited its clinical usefulness. We provide evidence here that the liver toxicity likely involves a sequence of biotransformations leading to 2-acetylbenzothiophene (2-ABT), which is subsequently metabolized to give a reactive intermediate(s). In vitro experiments with the human lymphoblast MCL5 cell line demonstrated that 2-ABT is cytotoxic in a P450-dependent manner. Human liver microsome (HLM) incubations with 2-ABT revealed the formation of two short-lived oxidized species, “M + 16” and “M + 32”. Both of these metabolites formed adducts in the presence of GSH or NAC. Singly oxidized M + 16 adducts, from either GSH or NAC, appeared to be unstable in acidic medium and eliminated water readily to form a new compound. Authentic synthetic standards demonstrated that 2-ABT-S-oxide M1 corresponded to the M + 16 metabolite and that the S-oxide underwent nucleophilic addition with GSH and NAC to produce the singly oxidized adducts observed in HLM. The S-oxide adducts readily eliminated water to form a rearomatized 2-ABT-GSH adduct or 2-ABT-NAC adduct. Coelution experiments with the synthetic standard confirmed the structure of the eliminated 2-ABTNAC adduct C1. LC/MS analyses of urine samples collected from rats dosed with zileuton indicate that C1 is a metabolite of zileuton formed in vivo. The in vitro and in vivo data presented here demonstrate the formation of 2-ABT from zileuton and its further bioactivation to a potentially toxic metabolite.
Introduction ADRs1 are a major clinical problem and have been the focus of extensive research over the past few years. ADRs are responsible for >100 000 deaths per year and have been attributed to 2-5% of hospital admissions (1). Estimates suggest that this costs the health care system approximately $136 billion annually (2). These statistics have prompted significant research efforts in understanding the mechanisms related to ADRs prior to widespread distribution of a drug. ADRs that do not occur in most patients at any dose and cannot be explained by any known metabolism or pharmacological effect of the drug are referred to as idiosyncratic drug reactions. It is believed that numerous factors contribute to the development of an idiosyncratic drug reaction, including the metabolic balance between bioactivation and detoxification pathways for an individual (1, 3). It is well-recognized that metabolic activation of some drugs may lead to the formation of a reactive † A preliminary account of this work was presented at the 11th North American International Society for the Study of Xenobiotics Meeting, Orlando, FL, 2002. * To whom correspondence should be addressed. Tel: 434-924-7718. E-mail:
[email protected]. 1 Abbreviations: 2-ABT, 2-acetylbenzothiophene; ADR, adverse drug reaction; APCI, atmospheric pressure chemical ionization; ATCC, American type culture collection; C1, 2-ABT-NAC adduct; CID, collision-induced dissociation; ESI, electrospray ionization; FMO3, flavin monooxygenase 3; GI50, 50% cell growth inhibition; GSH, glutathione; HLMs, human liver microsomes; M1, 2-ABT-S-oxide; NAC, N-acetylcysteine; SIM, single ion monitoring.
species capable of covalently binding to proteins, ultimately culminating in a toxic response (4). Idiosyncratic drug reactions are unpredictable, independent of the therapeutic dose range of a drug, host-dependent, and the least understood of all ADRs. Zileuton (1; Scheme 1) represents the first new class of drugs to be introduced for the treatment of asthma in more than 20 years (5). The metabolism of zileuton has been studied in human and includes the formation of a N-dehydroxylated metabolite (2), a number of P450mediated metabolites (3, 4), and two diastereomeric O-glucuronide conjugates (5) (6, 7). Alvarez and coworkers have also demonstrated the formation of 2-ABT, 6, as a degradation product of zileuton in vitro (8). Adverse events during clinical use of zileuton were limited; however, numerous postmarketing cases have cited elevated levels in liver enzyme tests prompting removal of patients from therapy. The current metabolic profile of zileuton fails to offer a mechanism to explain the observed hepatotoxicity. Biological activation of thiophene-related molecules has received increased attention over the past 10 years. A number of studies have shown the formation of S-oxide metabolites, which act as reactive, electrophilic intermediates in vivo and in vitro (9-11). Most of the work in this area is related to tienilic acid. The administration of tienilic acid to patients led to adverse events culminating in hepatitis and the formation of autoantibodies, prompting detailed studies to elucidate the mechanism for the hepatic toxicity (12). Work with tienilic acid and
10.1021/tx0341409 CCC: $27.50 © 2004 American Chemical Society Published on Web 01/24/2004
138
Chem. Res. Toxicol., Vol. 17, No. 2, 2004
Joshi et al.
Scheme 1. Detailed Metabolism of Zileuton
model thiophenes implicated a mechanism for toxicity involving initial oxidation to the S-oxide, followed by conjugation with GSH or another biological nucleophile, and subsequent dehydration to the conjugated parent thiophene product (9, 11). We propose that the toxicity observed during the use of zileuton may be the result of a sequence analogous to tienilic acid. The purpose of this study was to address the potential metabolic bioactivation of zileuton’s benzothiophene moiety, as well as to examine additional possible mechanism(s) associated with the development of hepatotoxicity. In particular, we were interested in investigating the formation of 2-ABT in vivo and demonstrating the potential of 2-ABT to mediate zileuton toxicity through bioactivation and formation of a reactive species.
Material and Methods Chemicals. Unless otherwise noted, all chemicals were purchased from Sigma (St. Louis, MO), Aldrich (Milwaukee, WI), or Alfa Aesar (Ward Hill, MA) and were of the highest grade available. cDNA-expressed human P450 enzymes and FMO3 expressed from either baculovirus-infected insect cells or human lymphoblast cells were purchased from BD Gentest Corp. (Woburn, MA). HLMs (pooled from 21 livers) were also from BD Gentest Corp. All other reagents and chemicals were of the highest quality available. Stock solutions of 2-ABT (10.7 mM), C1 (9.8 mM), and M1 (10.0 mM) were prepared in 50/50 methanol/water (0.02% TFA). Instruments. Routine 1H spectra were obtained on a Varian Unity Inova 300 (300 MHz, Palo Alto, CA) spectrometer. 13C NMR spectra were recorded on a Varian Unity Inova 300 (75.4 MHz) spectrometer. Chemical shifts are reported in ppm referenced to residual protonated solvent peaks, and coupling constants are in Hz. HPLC analysis was performed on a Waters 2690 separations module equipped with a Waters 486 tunable absorbance detector (Milford, MA). The LC was interfaced to a Thermo-Finnigan LCQ classic ion trap mass spectrometer running Excalibur version 1.1 (San Jose, CA). ESI or APCI ionization was employed as outlined in the Materials and Methods. The cytotoxicity assays were read on a Molecular
Devices Corporation Emax precision microplate reader (Sunnyvale, CA) set at λ ) 490 nm. LC/APCI- or LC/ESI-MS/MS Analysis. Parameters for APCI were as follows: capillary temperature, 150 °C; capillary voltage, 15.0 V; source voltage, 4.0 kV; source current, 4.98 µA; vaporizer temperature, 450 °C; sheath gas flow rate (nitrogen), 80. Parameters for ESI were as follows: capillary temperature, 200 °C; spray voltage, 5.0 kV; capillary voltage, 35.0 V; sheath gas flow rate (nitrogen), 40. All data were collected in the positive ion mode. For MS/MS experiments, full scan spectra from m/z 150-800 were obtained in the positive ion mode, and product ion spectra were generated by CID of the MH+ ions of interest. A 15 µL injection of each sample was separated on a Waters Symmetry C8 reversed phase column (5 µm, 2.1 mm × 150 mm) flowing at 200 µL/min. Synthesis of 1-(1-Oxo-benzo[b]thiophen-2-yl)ethanone (M1). 1-Benzo[b]thiophen-2-yl-ethanone (2-ABT) (150 mg, 0.85 mmol) was dissolved in trifluoroacetic acid (4.0 mL). The bright yellow solution was cooled to -78 °C, and an aqueous solution of 30% hydrogen peroxide (0.060 mL, 0.53 mmol) was then added. The mixture was allowed to warm to 0 °C over 30 min. The reaction was then quenched with a saturated solution of sodium meta-bisulfite (∼2 mL). Upon dilution with ethyl acetate (20 mL), the organic layer was subjected to an aqueous workup. The organic layer was dried over anhydrous sodium sulfate, filtered, and concentrated under reduced pressure to yield the solid residue. The crude yellow solid was purified by flash chromatography (silica gel, 5% methanol/ethyl acetate) to afford 80.2 mg (49%) of M1 as a yellow-tinted solid. The product appeared to be stable both as a solid and in aqueous solution for about 1 h at room temperature. APCI-MS: [M + H]+ m/z ) 193.1. 1H NMR (300 MHz, CDCl3): δ 7.97 (d, 1H, J ) 7.7 Hz), 7.80 (s, 1H), 7.57-7.69 (m, 3H), 2.60 (s, 3H). 13C NMR (CDCl3): δ 191.2, 150.9, 146.2, 140.2, 135.4, 132.6, 131.8, 127.4, 127.1, 27.7. HRMS (FAB): m/z calculated for C10H9O2S [M + H]+, 193.032326; found, 193.032400. Synthesis of 2-Acetylamino-3-(2-acetyl-benzo[b]thiophen3-ylsulfanyl)propionic Acid (C1). M1 (40 mg, 0.21 mmol) was dissolved in trifluoroacetic acid (10.0 mL). To this solution, NAC (45 mg, 0.28 mmol) was added, and the reaction was allowed to stir at room temperature for about 12 h. The reaction was quenched with MeOH (10 mL). The solution was then concen-
Oxidative Metabolism of 2-Acetylbenzothiophene
Chem. Res. Toxicol., Vol. 17, No. 2, 2004 139 Table 1. 1H NMR Spectroscopy Data for M1 and C1
trated under reduced pressure. Ethyl acetate (2.5 mL) was added to the residue, and the crude product was purified using semipreparative HPLC to afford 10.3 mg (14.5%) of C1. ESIMS: [M + H]+ ) 338.2. 1H NMR (300 MHz, DMSO-d6): δ 8.33 (d, 1H, J ) 7.9 Hz), 8.10 (dd, 2H, J ) 9.0 Hz), 7.52-7.59 (m, 2H), 4.21 (m, 1H), 3.18 (dd, 2H, J ) 12.0 Hz), 2.83 (s, 3H), 1.74 (s, 3H). 13C NMR (DMSO-d6): δ 192.7 (keto-carbonyl), 171.7 (acid-carbonyl), 169.4 (amide-carbonyl), 146.1, 140.7, 139.0, 129.6, 128.0, 125.6, 125.2, 123.4, 52.3, 37.2, 30.5, 22.2. HRMS (FAB): m/z calculated for C15H16NO4S2 [M + H]+, 338.052077; found, 338.052100. Synthesis of 1-(1,1-Dioxo-1H-1λ6-benzo[b]thiophen-2-yl)ethanone (7). 2-ABT (100 mg, 0.57 mmol) was dissolved in trifluoroacetic acid (4.0 mL) and cooled to 0 °C. Slowly, an aqueous solution of 30% hydrogen peroxide (0.240 mL, 2.1 mmol) was then added. The reaction was quenched after 2 h with a saturated solution of sodium meta-bisulfite (∼2 mL). Upon dilution with ethyl acetate (20 mL), the organic layer was collected and washed repeatedly with a sodium bicarbonate solution (∼20 mL). The organic layer was dried over anhydrous sodium sulfate, filtered, and concentrated under reduced pressure. The final product was obtained by crystallization of the crude material with ethyl acetate to afford 21.6 mg (18.2%) of 7 as a white solid. APCI-MS: [M + H]+ m/z ) 209.2. 1H NMR (300 MHz, CDCl3): δ 7.85 (s, 1H), 7.80 (d, 1H, J ) 6.6 Hz), 7.737.63 (m, 2H), 7.56 (d, 1H, J ) 7.0 Hz), 2.60 (s, 3H). 13C NMR (CDCl3): δ 219.2, 188.0, 140.8, 137.9, 137.6, 134.0, 133.5, 127.7, 122.0, 28.4. HRMS (FAB): m/z calculated for C10H9O3S [M + H]+, 209.027241; found, 209.027300. Cytotoxicity Assay (GI50). Cells used for the cell proliferation studies, MCL5 and K562 (both human lymphoblast), were obtained from ATCC (Manassas, VA). The cells were grown in media according to conditions supplied by ATCC and incubated at 37 °C in humidified air containing 5% carbon dioxide. For the cytotoxicity assay, the cells were plated into a 96 well plate at a density of approximately 1 × 105 cells per well and allowed to incubate for 24 h. After 24 h, the cells were treated with media containing varying concentrations of 2-ABT ranging from 1 to 150 µM (concentration of MeOH < 0.8%). Following standard procedures, the cells were grown in media containing the drug for 72 h, after which the number of viable cells was determined using the Promega Aqueous One MTS cell proliferation assay (Madison, WI) according to the manufacturer’s instructions. The GI50 was determined by plotting percent growth inhibition vs drug concentration in the linear region. Standard deviations were determined from quadruplicate assays. Microsomal Metabolism of 2-ABT. Incubations containing 125 µM compound, 2 mM NADPH, 12 mM MgCl2‚6H2O, and 1 mg/mL HLM in 0.1 M potassium phosphate buffer (pH 7.4) were incubated in a dri-bath at 37 °C for 30 min. The final concentration of MeOH from 2-ABT stock solutions was no greater than 0.9%. Reactions were stopped in a dry ice/acetone bath or with a 10% TCA solution, vortex-mixed, and centrifuged to precipitate the proteins. The resulting supernatant was used for all successive LC/MS and LC/MS/MS analyses. The formation of GSH or NAC adducts, in the presence of reduced GSH or NAC at a concentration of 1 mM, was monitored as well. Negative controls consisted of the aforementioned system lacking either microsomes, NADPH, GSH, or NAC. Samples were separated on a Waters Symmetry C8 reversed phase column (5 µm, 2.1 mm × 150 mm) and analyzed by LC/ MS. Separation was achieved using a gradient method employing aqueous 0.02% TFA (solvent A) and methanol (solvent B). The initial composition consisted of 80:20 A:B where B was slowly increased to 50% over a 30 min period. Under these gradient conditions, M1, 7, and C1 eluted off the column at 8.4, 11.2, and 25.2 min, respectively. The formation of oxidative products was analyzed using LC/APCI-MS while those experiments identifying conjugate formation were analyzed by LC/ ESI-MS.
chemical shift (ppm)a
a
proton
M1
CH3 C3‚H C4‚H C5,6‚H C7‚H S‚CH2 N‚CH NH N‚COOH3
2.60 (s) 7.80 (s) 7.97 (d) 7.63 (m) 7.63 (m)
C1 2.83 (s) 8.10 (dd) 7.57 (m) 8.10 (dd) 3.18 (dd) 4.21 (m) 8.33 (d) 1.74 (s)
NMR data obtained in CDCl3.
In Vitro 2-ABT Metabolism by cDNA-Expressed Human P450 Enzymes. Incubations with P450 2C8 and P450 2D6 were accomplished using lymphoblast-expressed enzymes. All other incubations were performed using baculovirus-expressed P450s and FMO3. Incubations were prepared as described above for HLMs, except 40 pmol/mL of enzyme was used. To monitor for the formation of GSH or NAC adducts, GSH or NAC was included in the assay at a concentration of 1 mM as in the HLM studies. Incubation samples containing NAC were analyzed as described above, and the amount of C1 was determined by LC/ ESI-SIM/MS integration of the mass chromatogram peaks using Xcalibur software. The concentration of C1 was quantified from a linear calibration curve obtained by using felbamate as an internal standard. The following is a list of human cDNAexpressed enzymes utilized in this study: 1A2, 2A6, 3A4, 4A11, 2B6, 2C8, 2C9, and 2E1. Kinetics Studies of the Reaction of 1-(1-Oxo-benzo[b]thiophen-2-yl)ethanone with GSH. M1 (50 µM) was added to 20 mM potassium phosphate buffer containing 2 mM GSH at various pH values (Table 1). Each solution was incubated at 37 °C. At appropriate time points, 100 µL aliquots were removed and added to 10.0 µL 10% TCA and an internal standard, felbamate. The remaining amount of unreacted M1 was determined by LC/APCI-SIM/MS by integrating the mass chromatogram peaks using XCalibur. The concentration of M1 was quantified from a linear calibration curve by using felbamate as an internal standard. The half-life (t1/2) was determined using a first-order approximation by plotting the natural logarithm of the remaining concentration of M1 vs time to give a slope equal to -k. The t1/2 is equal to 0.693/k. The identity of the GSH conjugate formed was determined using LC/ESI-MS. Its retention time and mass spectral data were compared with that of its corresponding synthetic standard. Metabolism of Zileuton by Rats. Animal experiments were performed according to procedures approved by the University of Virginia Animal Care and Use Committee. Six adult male Sprague-Dawley rats (∼250 g) were obtained from Harlan (Indianapolis, IN). The rats were divided into two groups (n ) 3). Prior to dosing, the rats were acclimated for 24 h. The first group was dosed with an aqueous suspension of 275 mg/kg zileuton in Cerestar b.i.d. via gavage. The second group, used as controls, was gavaged b.i.d. with a control suspension of Cerestar. Throughout the course of the experiment, the animals were maintained on a 12 h light/dark cycle and were provided free access to food and water. The urine samples were collected 18 h postdose in metabolic cages and were stored frozen at -20 °C until analyzed. Preparation and Analysis of Urine Samples. Urine samples were removed from the freezer and thawed to room
140
Chem. Res. Toxicol., Vol. 17, No. 2, 2004
Joshi et al. TFA (solvent A) and methanol (solvent B). Initial composition consisted of 80:20 A:B where the concentration of B was increased to 55% over 60 min. Under these conditions, C1 was found to elute off the column at 42.3 min.
Results
Figure 1. Oxidative cytotoxicity of 2-ABT in MCL5 cells. temperature. One milliliter of urine was applied to a Waters Oasis SEP cartridge (3 cm3, 60 mg) that had been preequilibrated with 3 mL of MeOH and then 3 mL of 0.02% TFA in water. The sample was then eluted, and the column was washed with 500 µL of 0.1% acetic acid in water. The analytes were eluted with 0.5 mL of 50% MeOH:50% (0.02%) TFA, collected, and saved for LC/MS analysis. Partially purified urine samples were analyzed using LC/ESIMS and were separated using a gradient method with 0.02%
Cytotoxicity of 2-ABT. The MCL5 cell line was markedly more sensitive to 2-ABT as compared to the control K562 cell line. Metabolically competent MCL5 cells have been stably transfected with human P450s 1A1, 1A2, 2A6, 2E1, 3A4, and human epoxide hydrolase. On the other hand, K562 is a human lymphoblast cell line not transfected and therefore does not express any exogenous P450s. 2-ABT was found to be toxic to MCL-5 cells with a GI50 of 103.2 µM (Figure 1) as compared to the control, K562 cells, which survived 100% at similar concentrations. These experiments suggest that the observed toxicity is mediated through metabolic activation of 2-ABT by P450 enzymes. In Vitro Metabolism of 2-ABT. Because of the potential role of P450s in activation of 2-ABT by the
Figure 2. (A) Top left: LC chromatogram with mass filter m/z 500 showing peak associated with (M + 16 + GSH) formation in HLM. Top right: MS/MS spectrum (Rt ) 5.0 min) of MH+ at m/z ) 500. The origins of the characteristic fragments are as shown. (B) Bottom left: LC chromatogram with mass filter m/z 516 showing peak associated with (M + 32 + GSH) formation in HLM. Bottom right: MS/MS spectrum (Rt ) 18.6 min) of MH+ at m/z ) 516. The origins of the characteristic fragments are as shown.
Oxidative Metabolism of 2-Acetylbenzothiophene
Chem. Res. Toxicol., Vol. 17, No. 2, 2004 141
Figure 3. Left panel: LC chromatograms with mass filter m/z 356 showing peak associated with oxidized 2-ABT + [O] + NAC formation in HLM (top) and coelution of synthetic S-oxide-NAC with HLM (bottom). Right panel: MS/MS spectrum (Rt ) 11.1 min) of MH+ at m/z ) 356. The origin of the characteristic fragment is as shown.
MCL5 cell line, HLM studies were conducted to probe possible hepatic metabolism. Incubations containing 2-ABT, cofactors, and enzyme resulted in steady loss of 2-ABT. LC/APCI-MS analysis of HLM incubations with 2-ABT revealed the formation of two distinct short-lived, oxidized metabolites: “M + 16” and “M + 32”. These metabolites were found to undergo conjugation with biologically relevant trapping agents, GSH and NAC. LC/ ESI-MS/MS analysis revealed the formation of two adducts of the oxidized products with GSH (Figure 2). One adduct with MH+ at m/z ) 500 [(MH + 16 + GSH)+] eluted at 4.97 min and gave rise to fragments m/z ) 407.0 (loss of Gly + H2O), m/z ) 353.0 (loss of Glu + H2O), and m/z ) 335.0 (loss of Gly + 2H2O) (Figure 2A). The second adduct with MH+ at m/z ) 516 [(MH + 32 + GSH)+] eluted at 18.6 min and gave rise to fragment ions m/z ) 497.9 (loss of H2O), m/z ) 441.0 (loss of Gly), and m/z ) 385.9 (loss of Gly + H2O) (Figure 2B). When NAC was used as the trapping agent, two similar conjugates were also observed. The two adducts were identified at 11.1 [(MH + 16 + NAC)+] and 26.4 min [(MH + 32 + NAC)+], respectively. The M1 was independently synthesized by chemical methods (Table 1). Coelution experiments with the HLM samples further confirmed that the M + 16 metabolite was the M1. Incubations of M1 with GSH or NAC followed by LC/MS analysis revealed the formation of conjugates that exactly matched the singly oxidized GSH and NAC conjugates (Rt ) 4.9 and 11.2 min, respectively) observed in HLM studies. When M1 was incubated with NAC, a conjugate was formed, which eluted at 11.2 min (Figure 3). This is in agreement with the conjugate observed in HLM samples. The parent MH+ at m/z ) 356 of the peak at 11.1 min yielded a daughter ion at m/z ) 193 (loss of NAC). Efforts to isolate and characterize the singly oxidized adduct by NMR spectroscopy were unsuccessful owing to its instability. Adducts formed from incubations with
the synthetic standard, M1, and GSH or NAC were found to be acid sensitive and readily undergo water elimination to give a new species 2-ABT-NAC, C1, which was chemically synthesized and thoroughly characterized (Table 1). C1 was demonstrated to coelute with the new peak (Rt ) 42.4 min) identified in HLM studies containing NAC (Figure 4). The elution was monitored by LCESI/MS-MS yielding the parent MH+ at m/z ) 338 with fragment ions at m/z ) 319.8 (loss of H2O), m/z ) 295.8 (loss of COCH3), m/z ) 277.9 (loss of NHCOCH3), and m/z ) 235.9. Additional characteristic features of C1 from 1 H NMR show disappearance of the C3 proton (7.8 ppm) and appearance of two characteristic NAC signals corresponding to the CH2 adjacent to the S (3.2 ppm) and to the NAC acetyl CH3 (1.7 ppm) (Table 1). Structural identification of the second M + 32 GSHNAC adduct has been more elusive. Synthetic measures were pursued to isolate the next compound, 2-ABTsulfone, 7. Conjugation experiments of 2-ABT-sulfone, 7, with GSH-NAC gave rise to an adduct in each case; however, this product did not coelute with the M + 32 GSH-NAC adducts observed in microsomal incubations. Studies are currently ongoing to determine the precise structure of the M + 32 GSH/NAC adduct species. Mass spectral data for the doubly oxidized adduct do not show an extended loss of water as seen with the singly oxidized species. On the basis of this limited evidence, we suspect that this structure could be a ring-hydroxylated S-oxide analogue. Half-Life of M1 with GSH. To determine the nonenzymatic reactivity of M1 with GSH, we measured the half-life of M1 in the presence and absence of nucleophile (GSH or NAC) with varying pH values. Samples were analyzed by LC/MS to monitor for the loss of starting material. The results are summarized in Table 2 and show that the conjugation occurs almost instantaneously at low pH values, while slowing down at increasing pH values. In all cases, it was observed that a decrease in
142
Chem. Res. Toxicol., Vol. 17, No. 2, 2004
Joshi et al.
Figure 4. LC chromatograms with mass filter m/z ) 338 showing peak associated with 2-ABT-NAC (C1) at Rt ) 42.4 min from coelution experiments. LC chromatogram resulting from C1 analysis (A). LC chromatogram resulting from Zileuton-dosed SpragueDawley rat urine (B). Table 2. Half-Life (t1/2) of M1 in the Presence of 2 mM GSH at Various pH Values pH conditions
t1/2 (h)
2.0 4.0 6.5 7.4 8.0 10.0
24
chromatographic analysis by reversed phase HPLC. The elution was monitored by ESI-MS/MS and both the synthetic standard C1 and a component of the urine sample were shown to coelute at 42.4 min (Figure 4) yielding the same daughter ions. The parent MH+ at m/z ) 338.0 of the peak at 42.4 min yielded daughter ions at m/z ) 319.8 (loss of H2O), m/z ) 295.8 (loss of COCH3), m/z ) 277.9 (loss of NHCOCH3) and m/z ) 235.9.
Discussion
Figure 5. P450 metabolism of 2-ABT.
the amount of M1 corresponded to an increase in the amount of conjugate formed. Metabolism of 2-ABT by cDNA-Expressed P450s. To identify specific P450 enzymes responsible for the oxidation of 2-ABT, incubations were carried out with a number of commercially available cDNA expressed human P450 enzymes (1A2, 2A6, 3A4, 4A11, 2B6, 2C8, 2C9, and 2E1). Data obtained from the incubations of 2-ABT with each P450 gave rise to only the M1-GSH-NAC adduct. No adduct formation was detected with FMO3. The data demonstrate that a number of P450s are involved in the formation of M1. From the enzymes studied, P450 2E1 and 1A2 are the primary enzymes responsible for the production of M1. All other recombinant P450s were limited in their ability to generate M1 (Figure 5). Identification of in Vivo Metabolism of Zileuton. Urine samples from rats treated twice daily with 275 mg/ kg/day zileuton were collected, purified by solid phase extraction, and studied by LC/MS. The analysis identified the formation of the proposed conjugate, C1, in urine. This was further confirmed by adding the synthetic C1 to the solid phase-extracted mixture and subjected to
The mechanism and proximate chemical species responsible for the hepatotoxicity of zileuton remain unresolved. To address this question, we have investigated the biotransformation of zileuton and its derivatives in several subcelluar, cellular, and in vivo rat model systems. The data accumulated by studies reported here and elsewhere (8) are consistent with the following overall pathway for induction of zileuton toxicity: initial zileuton metabolism leading to 2-ABT formation, followed by subsequent bioactivation to a reactive intermediate(s). The initial transformation of this proposed pathway has been reported to occur in vitro as reported by Alvarez and Slade (8). In addition, we have identified a mercapturic acid (C1), a conjugated derivative of 2-ABT, in rat urine after zileuton administration. Although C1 could be generated through alternative pathways not involving initial formation of 2-ABT, we have shown that C1 is a metabolic product of 2-ABT in microsomal samples. These data are consistent with the hypothesis of initial 2-ABT formation. Unfortunately, we were unable to detect 2-ABT as a direct metabolite in rats after zileuton administration. Because rats have not been dosed with 2-ABT, the detailed mechanism of 2-ABT metabolism is yet to be established. Consequently, HLM incubations demonstrated a substantial capacity to metabolize 2-ABT within a 30 min period. Therefore, the steady state levels of 2-ABT in the rat might have been below our current level of detection. The oxidative bioactivation of 2-ABT to both singly and doubly oxidized species could hold relevance for its toxicity. For example, cell culture experiments demonstrate a significant difference in the GI50 values for toxicity of 2-ABT found between MCL5 cells. These cells
Oxidative Metabolism of 2-Acetylbenzothiophene
overexpress a panel of P450s capable of bioactivating 2-ABT. This difference was not seen in the control, K562 cells, which do not overexpress exogenous P450 enzymes (Figure 1). Data from in vitro experiments with HLM incubated with 2-ABT show the formation of two oxidized metabolites, a singly oxidized species along with a doubly oxidized species. Each of these two metabolites underwent conjugation with GSH and NAC, suggesting the formation of two potential reactive metabolites. Formation of both the oxidized metabolites and the conjugates was dependent upon the presence of NADPH implicating the involvement of one or more cytochrome P450 enzymes. The singly oxidized metabolite was identified as the S-oxide, M1, by comparison with an authentic standard. It was also demonstrated that M1 underwent nucleophilic addition by GSH or NAC in chemical reactions to generate a M + 16 + GSH/NAC adduct, which then eliminates a water molecule to give 2-ABT-GSH (data not shown) or 2-ABT-NAC (C1) adducts (Scheme 1). Subsequent processing of an initial GSH adduct in vivo would yield the NAC conjugate C1. Kinetic studies of the nonenzymatic reaction between M1 and GSH in buffer gives a t1/2 ∼ 8 h at physiological pH (Table 2). Conversely, microsomal incubations show almost complete loss of 2-ABT after 30 min with little M1 detected. This discrepancy suggests two possible roles: (i) microsomal protein binding between M1 and surrounding proteins or (ii) involvement of GSH S-transferase enzymes in mediating the nucleophilic addition of GSH to M1. Metabolism of 2-ABT using a variety of recombinant human P450 enzymes illustrated that the specific activities of two P450s, 2E1 and 1A2, were prominent for the oxidation of the benzothiophene to the corresponding S-oxide (Figure 5). Although the activity of 2E1 and 1A2 for oxidation of 2-ABT is significantly higher than the other P450s, the contribution of metabolism from 3A4 and possibly other enzymes may be important in vivo due to their abundance. In humans, 3A4 is the most abundant P450 enzyme expressed in the liver, representing close to 30% of total P450 present (13). Studies are currently ongoing to determine the structure of the doubly oxidized metabolite and adducts (M + 32 + GSH-NAC) observed in HLM studies. In an effort to identify the structure of the doubly oxidized metabolite and adducts, 2-ABT-sulfone (7) was synthesized. Studies demonstrated that 7 was not identical to the doubly oxidized metabolite observed in HLM. The sulfone, 7, did undergo facile conjugation with both GSH and NAC, but the resulting conjugates did not coelute with the observed doubly oxidized M + 32 + GSH-NAC adducts found in HLM samples (data not shown). Comparisons between the fragmentation data obtained from LC/ESI-MS/MS analysis of the M + 16 and M + 32 conjugates revealed no extended loss of water molecules in the doubly oxidized conjugate (Figure 2B). On the basis of this limited evidence, we propose that the doubly oxidized structure is a ring-hydroxylated S-oxide moiety. This assignment is based solely on MS data and is not enough to determine its structure. Synthetic efforts to isolate a ring-hydroxylated derivative of 2-ABT, which would give access to the proposed doubly oxidized metabolite, have been unsuccessful. The work presented here provides a plausible mechanistic rationale for the hepatotoxicity associated with
Chem. Res. Toxicol., Vol. 17, No. 2, 2004 143
zileuton therapy. We propose that zileuton undergoes a biotransformation giving 2-ABT, which is then metabolized through oxidation to a reactive intermediate(s). The ability of a drug to generate a reactive intermediate capable of alkylating proteins is generally considered necessary, but not always sufficient, criteria for a drug to cause an idiosyncratic drug reaction. We have observed C1, a mercapturate of 2-ABT, in rats and observed the bioactivation of 2-ABT to its S-oxide, M1, in a variety of in vitro assays. In addition, we have identified the presence of an additional reactive metabolite derived by two oxidative transformations of 2-ABT, to M + 32. These reactive species undergo conjugation with GSH and NAC and therefore could be capable of forming covalent adducts with surrounding proteins. Such liver bioactivation and covalent adduct formation are thought to be common mechanisms for many known hepatotoxins (14). The toxicity of zileuton emphasizes the importance of understanding the metabolism and bioreactivity of the benzothiophene substructure in future drug development.
References (1) Pirmohamed, M., and Park, B. K. (2001) Genetic susceptibility to adverse drug reactions. Trends Pharmacol Sci. 22, 298-305. (2) White, T. J., Arakelian, A., and Rho, J. P. (1999) Counting the costs of drug-related adverse events. Pharmacoeconomics 15, 445-458. (3) Dieckhaus, C. M., Thompson, C. D., Roller, S. G., and Macdonald, T. L. (2002) Mechanisms of idiosyncratic drug reactions: the case of felbamate. Chem.-Biol. Interact. 142, 99-117. (4) Uetrecht, J. P. (1999) New concepts in immunology relevant to idiosyncratic drug reactions: the “danger hypothesis” and innate immune system. Chem. Res. Toxicol. 12, 387-395. (5) Drazen, J. M., Israel, E., and O’Byrne, P. M. (1999) Treatment of asthma with drugs modifying the leukotriene pathway. N. Engl. J. Med. 340, 197-206. (6) Sweeny, D. J., and Nellans, H. N. (1995) Stereoselective glucuronidation of zileuton isomers by human hepatic microsomes. Drug Metab. Dispos. 23, 149-153. (7) Machinist, J. M., Mayer, M. D., Shet, M. S., Ferrero, J. L., and Rodrigues, A. D. (1995) Identification of the human liver cytochrome P450 enzymes involved in the metabolism of zileuton (ABT-077) and its N-dehydroxylated metabolite, Abbott-66193. Drug Metab. Dispos. 23, 1163-1174. (8) Alvarez, F. J., and Slade, R. T. (1992) Kinetics and mechanism of degradation of Zileuton, a potent 5-lipoxygenase inhibitor. Pharm. Res. 9, 1465-1473. (9) Dansette, P. M., Thang, D. C., el Amri, H., and Mansuy, D. (1992) Evidence for thiophene-S-oxide as a primary reactive metabolite of thiophene in vivo: formation of a dihydrothiophene sulfoxide mercapturic acid. Biochem. Biophys. Res. Commun. 186, 16241630. (10) Pouzet, P., Erdelmeier, I., Dansette, P. M., and Mansuy, D. (1998) Synthesis of (4-Chlorophenyl)-(1-oxo-1λ(4)-benzo[b]thien-2-yl)methanone and study of its reactivity towards sulfur- and oxygencontaining nucleophiles. Tetrahedron 54, 14811-14824. (11) Valadon, P., Dansette, P. M., Girault, J. P., Amar, C., and Mansuy, D. (1996) Thiophene sulfoxides as reactive metabolites: formation upon microsomal oxidation of a 3-aroylthiophene and fate in the presence of nucleophiles in vitro and in vivo. Chem. Res. Toxicol. 9, 1403-1413. (12) Mansuy, D. (1997) Molecular structure and hepatotoxicity: compared data about two closely related thiophene compounds. J. Hepatol. 26 (Suppl. 2), 22-25. (13) Ortiz de Montellano, P. R. (1999) The Cytochrome P450 Oxidative System. Handbook of Drug Metabolism, Marcel Dekker, New York. (14) Holme, J. A., Dahlin, D. C., Nelson, S. D., and Dybing, E. (1984) Cytotoxic effects of N-acetyl-p-benzoquinone imine, a common arylating intermediate of paracetamol and N-hydroxyparacetamol. Biochem. Pharmacol. 33, 401-406.
TX0341409
