Prevention of acetaminophen (APAP)-induced hepatotoxicity by leflunomide via inhibition of APAP biotransformation to N-acetyl-p-benzoquinone imine

Toxicology Letters 180 (2008) 174–181

Contents lists available at ScienceDirect

Toxicology Letters journal homepage: www.elsevier.com/locate/toxlet

Prevention of acetaminophen (APAP)-induced hepatotoxicity by leflunomide via inhibition of APAP biotransformation to N-acetyl-p-benzoquinone imine Su Ching Tan, Lee Sun New, Eric C.Y. Chan ∗ Department of Pharmacy, Faculty of Science, National University of Singapore, 18 Science Drive 4, Singapore 117543, Singapore

a r t i c l e

i n f o

Article history: Received 20 April 2008 Received in revised form 2 June 2008 Accepted 3 June 2008 Available online 8 June 2008 Keywords: Leflunomide A77–1726 Acetaminophen Hepatotoxicity N-Acetyl-p-benzoquinone imine Microsomes

a b s t r a c t Acetaminophen (APAP) is safe at therapeutic levels but causes liver injury via N-acetyl-p-benzoquinone imine (NAPQI)-induced oxidative stress when overdose. Recent studies indicated that mitochondrial permeability transition (mPT) plays a key role in APAP-induced toxicity and leflunomide (LEF) protects against the toxicity through inhibition of c-jun NH2 -terminal protein kinase (JNK)-mediated pathway of mPT. It is not clearly understood if LEF also exerts its protective effect through inhibition of APAP bioactivation to the toxic NAPQI. The present work was undertaken to study the effect of LEF on the bioactivation of APAP to NAPQI. Mechanism-based inhibition incubations performed in mouse and human liver microsomes (MLM and HLM) indicated that inhibition of APAP bioactivation to NAPQI was observed in MLM but not in HLM. Furthermore, LEF but not its active metabolite, A77–1726, was shown to be the main inhibitor. When APAP and LEF were incubated with human recombinant P450 enzymes, CYP1A2 was found to be the isozyme responsible for the inhibition of APAP bioactivation. Species variation in CYP1A2 enzymes probably accounted for the different observations in our MLM and HLM studies. We concluded that inhibition of NAPQI formation is not a probable pathway that LEF protects APAP-induced hepatotoxicity in human. © 2008 Elsevier Ireland Ltd. All rights reserved.

1. Background Acetaminophen (APAP) is a widely used analgesic and antipyretic drug that is safe and effective at therapeutic levels. However, when excessive doses are ingested, APAP can cause severe liver injury that can be fatal (Jaeschke and Bajt, 2006). Due to its ready availability, APAP overdose is associated with the most common cause of drug-induced hepatotoxicity in the United States and the United Kingdom (Lee, 2004). For many years, P450-dependent biotransformation of APAP to its reactive metabolite, N-acetyl-pbenzoquinone imine (NAPQI) has been clearly identified as the mechanism mediating APAP-induced hepatotoxicity. While bioactivation of APAP to NAPQI is mediated by P450s 1A2, 2E1, and 3A4 in both mouse and human, P4502E1 has been implicated as the most

Abbreviations: A77–1726, teriflunomide, N-(4-trifluoromethylphenyl)-2-cyano3-hydroxycrotoamide; APAP, acetaminophen; DDTC, sodium diethyldithiocarbamate trihydrate; DMSO, dimethyl sulfoxide; ESI +ve, electrospray positive ionization mode; ESI −ve, electrospray negative ionization mode; GSH, glutathione; HLM, human liver microsomes; JNK, c-jun NH2 -terminal protein kinase; LC/MS/MS, liquid chromatography-tandem mass spectrometry; LEF, leflunomide; MLM, mouse liver microsomes; mPT, mitochondrial permeability transition; NAC, N-acetylcysteine; NAPQI, N-acetyl-p-benzoquinone imine; P450, cytochrome P450. ∗ Corresponding author. Tel.: +65 65166137; fax: +65 67791554. E-mail address: [email protected] (E.C.Y. Chan). 0378-4274/$ – see front matter © 2008 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.toxlet.2008.06.001

relevant isozyme in human (Snawder et al., 1994; Manyike et al., 2000; Jaeschke and Bajt, 2006). NAPQI requires glutathione (GSH) for detoxification by forming its GSH-adduct. Once the intracellular stores of GSH are depleted, excess NAPQI may react with cellular proteins, including mitochondrial proteins, thus causing liver cell necrosis (Kon et al., 2004; Jaeschke and Bajt, 2006; Terneus et al., 2007). Mitochondrial permeability transition (mPT) is recently elucidated as the principal mechanism underlying APAP-induced liver injury. The mPT is characterized by an abrupt increase in permeability of the inner mitochondrial membrane, resulting in osmotic matrix expansion and ultimately the rupture of outer mitochondrial membrane. Cytochrome c and other proapoptotic factors including endonuclease G and apoptosis-inducing factor (AIF) are then released into the cytosol (Masubuchi et al., 2005; Latchoumycandane et al., 2006, 2007). NAPQI-induced oxidative stress and increase in cellular Ca2+ are responsible for the opening of the transition pore. The mPT then leads to oncotic necrosis of the hepatocytes (Masubuchi et al., 2005; Jaeschke and Bajt, 2006). It has also been demonstrated that c-Jun NH2 -terminal protein kinase (JNK) plays a critical role in mediating the mPT (Schwabe et al., 2004). N-Acetylcysteine (NAC) is currently the drug of choice for the management of APAP overdose. NAC, a precursor of GSH, diminishes APAP toxicity by increasing GSH levels and acts as an antioxidant to antagonize NAPQI-induced oxidative stress

S.C. Tan et al. / Toxicology Letters 180 (2008) 174–181

(Smilkstein et al., 1991; Terneus et al., 2007). However, in order for NAC to be effective, this antidote should be administered relatively early (i.e., within 8 h) after APAP overdose (Smilkstein et al., 1988; Latchoumycandane et al., 2007). Leflunomide (LEF) is an immunomodulatory and disease-modifying anti-inflammatory agent used for the treatment of rheumatoid arthritis, allograft and xenograft rejection, and systemic lupus erythematous (Rozman, 2002; Yao et al., 2003). Following oral administration, LEF is almost completely converted to its pharmacological active metabolite, A77–1726 (teriflunomide) (Rozman, 2002). Recent data have shown that LEF protects immortalized human hepatocytes from APAPinduced toxicity. LEF exerts its cytoprotective effect via inhibition of the phosphorylation of JNK1/2 and prevention of the mPT pore opening in human hepatocytes (Latchoumycandane et al., 2006). In vivo studies in mice have shown that LEF suppresses APAP-induced phosphorylation of JNK1/2, and hence inhibits Bcl-2 and Bcl-XL inactivation (Latchoumycandane et al., 2007). This in turn prevents the mPT and subsequent steps that lead to oncotic necrosis of the liver. LEF may also alleviate APAP-induced hepatocyte injury via the prevention of peroxynitrite formation, which is an activator of the JNK pathway (Latchoumycandane et al., 2007). The combined results obtained from in vitro data in hepatocytes and in vivo data in mice indicated indirectly that the mechanism of protection may not involve the inhibition of the P450-mediated bioactivation of APAP to its toxic metabolite, NAPQI. It is suggested that LEF could be administered at a relatively late point, as its hepatoprotective effect is via the inhibition of the signaling pathway distal to the bioactivation of APAP (Latchoumycandane et al., 2006, 2007). The metabolism of LEF to A77–1726 is mediated by the cytochrome P450 enzyme system. While P450s 1A2, 2C19, and 3A4 are the isozymes involved in the biotransformation of LEF to A77–1726, P4501A2 is the major enzyme responsible for the metabolism in human (Kalgutkar et al., 2003). On the other hand, the bioactivation of APAP to NAPQI is also mediated by P450s 1A2, 2E1, and 3A4 in both mouse and human. In view of the common metabolic pathways, we hypothesized that LEF and APAP may interact at the hepatic metabolism level and the level of toxic NAPQI may be reduced in the presence of LEF. Based on its clinical pharmacokinetics, we understand that LEF is rapidly metabolized to A77–1726 in human after oral dosing and the parent drug is usually not detected in the plasma (Lucian et al., 1995; Rozman, 2002). In view of this clinical disposition, it becomes important to determine whether LEF and/or A77–1726 are capable in the inhibition of NAPQI formation. It is also important to investigate and elucidate such drug–drug interaction using human liver microsomes (HLM) as studies conducted so far has been limited to animals. In this present work, we aimed to study the effects of LEF and A77–1726 on the bioactivation of APAP to NAPQI in vitro and to determine the specific isozyme that is responsible in the interaction between the drugs. Three in vitro models, namely MLM, HLM, and human recombinant P450 enzymes are used in our investigations. The knowledge gained from our study will help to further clarify the mechanism of action of LEF when it is experimented as an antidote for the treatment of APAP-induced hepatotoxicity. 2. Materials and methods 2.1. Chemicals HPLC grade acetonitrile was purchased from Fisher Scientific (Leicestershire, UK). Ammonium acetate of 99% purity was obtained from VWR International Ltd. (Leicestershire, UK). Water was purified using a Milli-Q water purification system (Millipore, Bedford, MA, USA). Potassium phosphate monobasic (ACS grade) and potassium phosphate dibasic (ACS grade) were purchased from Mallinckrodt Baker (Phillipsburg, NJ, USA) and Sigma–Aldrich (St. Louis, MO, USA), respectively. APAP, reduced l-glutathione (GSH), furafylline, ketoconazole, and sodium diethyldithiocarbamate trihydrate (DDTC) were purchased from Sigma–Aldrich (St. Louis, MO,

175

USA). LEF (Sigma–Aldrich) and A77–1726 (Merck Pte. Ltd., Singapore) were generous gifts from Professor Urs Alex Boelsterli from the School of Pharmacy at the University of Connecticut. MLM (male CD-1), pooled HLM, human recombinant P450s 1A2, 2E1, and 3A4 and cofactors were purchased from BD Gentest (Woburn, MA, USA). All other chemicals and reagents used for the experiments were of analytical grades. 2.2. UPLC/QTOF/MS/MS analysis of NAPQI–GSH For the accurate mass measurement, a sample consisting of 800 ␮M APAP and 2 ␮M LEF incubated in HLM for 60 min at 37 ◦ C was used. The sample was analysed using an ACQUITY UPLC system (Waters, Milford, MA, USA) interfaced with a quadrupole, orthogonal acceleration time-of-flight tandem mass spectrometer (QTOFMS) equipped with ESI source (Q-Tof PremierTM , Waters, Manchester, UK). The UPLC/QTOFMS system was controlled by MassLynx 4.1 software (Waters). Chromatographic separations were performed on an AQUITY UPLC BEH C18 1.7 ␮m 50 mm × 2.1 mm i.d. column (Waters). The column heater and autosampler were kept 60 and 4 ◦ C, respectively. The mobile phases consisted of solvent A (0.1% formic acid in water) and solvent B (0.1% formic acid in acetonitrile). The optimized elution conditions were as follows: gradient of 0.1–10.0% solvent B (0–1.80 min), 10.0–95.0% solvent B (1.80–1.81 min), isocratic at 95.0% solvent B (1.81–2.14 min) and isocratic at 0.1% solvent B (2.15–2.50 min). The flow rate was 0.7 mL/min. The eluent was splitted post-column (ratio of 2:1 between waste and MS) into the ESI source using a T-splitter so as to enhance the MS sensitivity. The QTOFMS system was tuned for optimum sensitivity and resolution in electrospray positive ionization mode (ESI +ve) mode using leucine enkephalin (50 pg/␮L infused at 5 ␮L/min). The QTOF/MS/MS analysis was operated in “V” mode and the optimized conditions were as follows: capillary voltage, 3400 V; sampling cone, 24 V; source temperature, 100 ◦ C; desolvation temperature, 350 ◦ C; cone gas flow, 0 L/h; desolvation gas flow, 500 L/h; collision energy, 15 eV; MCP detector voltage, 1800 V; pusher voltage, 915 V; pusher voltage offset, −0.70 V; puller voltage, 640 V. The precursor mass was set at 457.1 Da and continuum data were acquired for each sample from 50 to 600 Da with a 0.2 s scan time and a 0.02 s interscan delay. Prior to analysis, the system was calibrated in ESI +ve mode using 0.5 M sodium formate solution infused at a flow rate of 4 ␮L/min. All analyses were acquired using an independent reference spray via the LockSprayTM interface to ensure high mass accuracy and reproducibility; the [M+H]+ ion of leucine enkephalin (2 ng/␮L infused at 3 ␮L/min) was used as the reference lock mass (m/z 556.2771). The LockSprayTM was operated at a reference scan frequency, reference cone voltage and collision energy of 10 s, 30 and 18 V, respectively. 2.3. Mechanism-based inhibition in MLM LEF and APAP stock solutions were prepared in methanol and water, respectively. The final organic content in the incubation media was less than 0.003% (v/v). To investigate the effect of LEF and A77–1726 on the inhibition of NAPQI–GSH formation in MLM, a two-step incubation scheme was used, in which the metabolism of LEF to A77–1726 was allowed to take place during the preincubation period. LEF (0, 0.5, 2, and 15 ␮M) was preincubated in MLM (protein concentration, 1 mg/mL) with cofactors (1.55 mM NADP+ , 3.3 mM glucose-6-phosphate, 3.3 mM magnesium chloride, 0.4 U/mL glucose-6-phosphate dehydrogenase and 0.05 mM sodium citrate) and 5 mM of GSH in 50 mM potassium phosphate buffer (pH 7.4) at 37 ◦ C. The reaction mixture was warmed at 37 ◦ C for 5 min before adding LEF. The total incubation volume was 200 ␮L. The preincubation period was set to be 0, 10, and 20 min. At the end of each stipulated preincubation period, one portion (55 ␮L) of the preincubation reaction mixture was withdrawn and added to cold acetonitrile (110 ␮L) to quench the reaction. Five microlitres of APAP (final concentration, 800 ␮M) was then added to the remaining reaction mixture and it was further incubated for 60 min at 37 ◦ C. Each reaction was terminated by the addition of each aliquot to two volumes of cold acetonitrile, and the sample was centrifuged at 13,000 rpm for 10 min. The supernatant obtained was subjected to liquid chromatography/tandem mass spectrometry (LC/MS/MS) analysis. All incubations were performed in triplicates. 2.4. Incubation of A77–1726 in MLM A77–1726 stock solution was prepared in DMSO. The final organic content in the incubation media was less than 0.05% (v/v). To investigate the effect of A77–1726 on the inhibition of formation of NAPQI–GSH, A77–1726 (0, 0.01, 0.1, 0.5, 1, 2, and 5 ␮M) and APAP (800 ␮M) were incubated for 60 min at 37 ◦ C using similar reagents as stated in the previous section. The reaction mixture was warmed at 37 ◦ C for 5 min before adding A77–1726 and APAP. The total incubation volume was 200 ␮L. Reactions were terminated by addition of two volumes of cold acetonitrile, and the samples were centrifuged at 13,000 rpm for 10 min and analysed using LC/MS/MS for NAPQI–GSH formation. All incubations were performed in triplicates. 2.5. Mechanism-based inhibition in HLM To investigate the effect of LEF and A77–1726 on the inhibition of formation of NAPQI–GSH, mixtures containing LEF (0, 0.5, 2, and 15 ␮M) and HLM (protein

176

S.C. Tan et al. / Toxicology Letters 180 (2008) 174–181

concentration, 1 mg/mL) were preincubated for 0, 10, and 20 min before the addition of APAP (800 ␮M), and then allowed to incubate for additional 60 min at 37 ◦ C. The conditions and reagents adopted were similar to the MLM mechanism-based inhibition study. Reactions were terminated by addition of two volumes of cold acetonitrile, and the samples were centrifuged at 13,000 rpm for 10 min and analysed subsequently for NAPQI–GSH formation. All incubations were performed in triplicates. 2.6. Human recombinant P450 enzyme experiments To identify the human P450 isozyme that is responsible for the inhibition of NAPQI–GSH formation by LEF, reaction mixtures each consisted of human recombinant P450s 1A2, 2E1, and 3A4, respectively (P450 concentration, 20 pmol/mL), cofactors, 5 mM of GSH, and 50 mM potassium phosphate buffer were prepared. These reaction mixtures were warmed at 37 ◦ C for 5 min. LEF (0, 2, and 15 ␮M) was then added and incubations were carried out for 60 min at 37 ◦ C. Reactions were terminated by the addition of two volumes of cold acetonitrile, and the samples were centrifuged at 13,000 rpm for 10 min and analysed. All incubations were performed in triplicates. For positive control experiments, specific P450 inhibitors were added to the respective incubation mixtures (furafylline, DDTC and ketoconazole at a final concentration of 5, 5, and 2 ␮M for P4501A2, P4502E1, and P4503A4, respectively). Furafylline and ketoconazole stock solutions were prepared in DMSO while DDTC was prepared in water. The final organic concentration of the incubation media was not more than 0.25% (v/v). 2.7. LC/MS/MS analysis The microsomal assay samples were analysed using an Agilent 1100 HPLC (Agilent Technologies, Santa Clara, CA, USA) interfaced with a hybrid triple quadrupole linear ion trap mass spectrometer (QTRAP MS) equipped with TurboIonSpray ESI source (2000 QTRAP, Applied Biosystems, Foster City, CA, USA). The HPLC and QTRAP MS systems were both controlled by Analyst 1.4.1 software (Applied Biosystems). Chromatographic separations were performed on a Phenomenex Luna C18 (2), 50 mm × 2.0 mm i.d. column. The column heater and autosampler were kept 60 and 4 ◦ C, respectively. For the analysis of APAP, NAPQI–GSH and O-deethylated phenacetin, the flow rate was 0.5 mL/min and the mobile phase consisted of solvent A (0.1% formic acid in water) and solvent B (0.1% formic acid in acetonitrile). The optimized elution conditions were as follows: gradient of 2.0–20.0% solvent B (0–3.00 min), gradient of 20.0–100.0 solvent B (3.00–3.01 min), isocratic at 100.0% solvent B (3.01–4.00 min) and isocratic at 2.0% solvent B (4.01–7.00 min). For the analysis of LEF and A77–1726, the flow rate was 0.5 mL/min and the mobile phase consisted of solvent A (10 mM ammonium acetate in 90% water and 10% acetonitrile) and solvent B (10 mM ammonium acetate in 10% water and 90% acetonitrile). The optimized elution conditions were as follows: 32.0–100.0% solvent B (0–3.00 min), isocratic at 100.0% solvent B (3.00–4.00 min) and isocratic at 32.0% solvent B (4.01–7.00 min). The human recombinant P450 assay samples were assayed using an ACQUITY UPLC system (Waters, Milford, MA, USA) interfaced with a hybrid triple quadrupole linear ion trap mass spectrometer (QTRAP MS) equipped with TurboIonSpray ESI source (2000 QTRAP, Applied Biosystems, Foster City, CA, USA). The UPLC and QTRAP MS systems were controlled by separate processes of MassLynx 4.1 (Waters) and Analyst 1.4.1 software (Applied Biosystems), respectively. Chromatographic separations were performed on an AQUITY UPLC BEH C18 1.7 ␮m 50 mm × 2.1 mm i.d. column (Waters). For the analysis of APAP and NAPQI–GSH, the column heater and autosampler were kept at 40 and 4 ◦ C, respectively. The flow rate was 0.5 mL/min and the same mobile phases as described above were used. The optimized elution conditions were as follows: gradient of 0.1–10.0% solvent B (0–2.70 min), 10.0–95.0% solvent B (2.70–2.71 min), isocratic at 95.0% solvent B (2.71–3.09 min), and isocratic at 0.1% solvent B (3.10–3.50 min). For the analysis of LEF and A77–1726, the column heater and autosampler were kept at 60 and 4 ◦ C, respectively. The flow rate was 0.5 mL/min and the same mobile phases as described above were used. The optimized elution conditions were as follows: gradient of 15.0–65.0% solvent B (0–2.70 min), 65.0–95.0% solvent B (2.70–2.71 min), isocratic at 95% solvent B (2.71–3.10 min) and isocratic at 15.0% solvent B (3.11–3.50 min). Electrospray positive ionization mode (ESI +ve) was used for APAP and NAPQI–GSH analysis. Multiple reaction monitoring (MRM) experiments using m/z 152.10–110.10 and 457.00–328.00 were performed to profile and quantitate APAP and NAPQI–GSH, respectively. For the analysis of LEF and A77–1726, electrospray negative ionization mode (ESI −ve) was used. Multiple reaction monitoring experiments using m/z 269.10–82.00 were performed to profile and quantitate LEF and A77–1726. The MS conditions for all the MRM experiments are summarized in Table 1. 2.8. Data analysis For the mechanism-based inhibition studies, repeated measures ANOVA test was used for the comparison of the analyte levels between the various experimental groups, followed by the Dunnett’s test to determine significant differences between the experimental and control group means. For the remaining studies, one-

Table 1 Optimized MS parameters for the detection of APAP, O-deethylated phenacetin, NAPQI–GSH, LEF and A77–1726 Parameter

Curtain gas (psi) Ion spray voltage (V) Temperature (◦ C) Gas 1 (psi) Gas 2 (psi) Interface heater CAD gas Declustering potential (V) Entrance potential (V) Collision energy (V) Collision cell entrance potential (V) Collision cell exit potential (V) Dwell time (ms)

Values APAP

NAPQI–GSH

LEF and A77–1726

25 4500 550 60 60 On Medium 53 10 23 19.84 2.50 500

25 4500 550 60 60 On Medium 80 10 25 29.90 4.00 500

25 −1750 550 60 60 On Medium −65 −5.50 −25 −23.43 −1.00 300

way ANOVA test was used for comparison of the analyte levels between the various experimental groups, followed by Tukey’s multiple comparison test to determine significant differences between the group means. P < 0.05 was considered statistically different in all studies.

3. Results 3.1. Accurate mass analysis of NAPQI–GSH Using MassLynx, the potential calculated masses, mass accuracy (mDa and ppm), i-FIT (the likelihood that the isotopic pattern of the elemental composition matches a cluster of peaks in the spectrum) values and elemental compositions associated with the measured mass of NAPQI–GSH and its fragments were generated and studied. Using a mass tolerance of 10 ppm, the experimentally determined masses/mDa/ppm/i-FIT/elemental composition of NAPQI–GSH ([M+H]+ ) was 457.1383/−1.0/−2.2/0.0/C18 H25 N4 O8 S and its fragments were as follows—(i) [M−75]+ : 382.1070/−0.3/−0.8/5.7/ C16 H20 N3 O6 S; (ii) [M−129]+ : 328.0961/−0.6/−1.8/0.2/C13 H18 N3 O5 S; (iii) [M−146]+ : 311.0707/0.5/1.6/3.6/C13 H15 N2 O5 S; (iv) [M−188]+ : 269.0606/1.0/3.7/7.9/C11 H13 N2 O4 S; (v) [M−232]+ : 225.0704/0.6/ 2.7/0.7/C10 H13 N2 O2 S; (vi) [M−249]+ : 208.0428/−0.4/−1.9/0.9/C10 H10 NO2 S; and (vii) [M−275]+ : 182.0279/0.3/1.6/0.2/C8 H8 NO2 S, respectively (see Fig. 1). 3.2. MLM experiments To validate whether LEF and/or A77–1726 (the major metabolite of LEF) inhibit the bioactivation of APAP to NAPQI in MLM, we performed the mechanism-based inhibition study where A77–1726, the active metabolite of LEF, was allowed to be generated prior to the addition of APAP, the test substrate. Briefly, LEF (0.5, 2, and 15 ␮M) was preincubated for 0, 10, and 20 min before adding APAP. Our experimental concentration of APAP was estimated at 800 ␮M based on its reported concentration in human plasma (1 mM), and extent of protein binding (20%) when APAP was overdosed (Drugdex® Evaluations, 2008). In control experiment, LEF was omitted from the APAP incubations. With reference to Fig. 2A, it was clear that for all the three investigated concentrations of LEF, the remaining levels of the drug declined as the preincubation time was increased from 0 to 20 min, indicating the concentration- and time-dependent biotransformation of LEF. As shown in Fig. 2B, an increasing level of A77–1726 was observed as the concentration and preincubation time of LEF were increased. This result further confirmed that LEF was metabolized to its major metabolite, A77–1726, in a concentration- and time-dependent manner. While it is well established that LEF is metabolized to A77–1726, these

S.C. Tan et al. / Toxicology Letters 180 (2008) 174–181

177

Fig. 1. Extracted UPLC/QTOF/MS/MS ion spectrum of NAPQI–GSH (m/z 457). The CID pathways of NAPQI–GSH and its respective fragments are shown.

Fig. 2. Time- and concentration-dependent effects of LEF on the bioactivation of APAP to NAPQI–GSH in MLM at 37 ◦ C. Levels of (A) LEF, (B) A77–1726 and (C) NAPQI–GSH (expressed as a percentage of the control) measured with and without preincubation of LEF in MLM. *P < 0.01 vs. control. Data points are mean ± S.D. (n = 3).

in vitro experiments were important to allow correlation between the dynamic levels of these two compounds with the observed inhibition of NAPQI–GSH formation. The results of the mechanism-based inhibition study in MLM were presented in Fig. 2C. The amount of NAPQI–GSH formed was expressed as a percentage of the control where APAP was incubated without LEF. This index was used to indicate the percent marker activity of the biotransformation of APAP to NAPQI. When compared to the control, the inhibition of NAPQI formation observed in each test sample (with and without preincubation) was found to be significant (P < 0.01). When there was no preincubation of LEF (0 min), the percent marker activity decreased expectedly as the amount of LEF was increased. It was also observed that for the samples incubated with LEF at 2 and 15 ␮M, increasing the preincubation time from 0 to 20 min did not result in a decrease in marker activity. These higher levels of NAPQI formation with declining levels of LEF due to preincubation suggested that the biotransformation of APAP to its toxic metabolite, NAPQI, was inhibited mainly by LEF and not A77–1726. In contrary, at lower concentrations of LEF (0.5 ␮M), preincubation was observed to enhance the inhibition of APAP bioactivation. We repeated this experiment a few times and similar inhibition profiles were consistently observed. We suspected that at a lower concentration of LEF (0.5 ␮M), the degree of inhibition by LEF was smaller and the effect of the mechanismbased inhibition became less apparent. To further ascertain that A77–1726 does not have an inhibitory effect on the bioactivation of APAP in MLM, we conducted a concentration–response experiment, by adding A77–1726 directly to the microsomal incubations. A77–1726 at 0.01, 0.1, 0.5, 1, 2, and 5 ␮M were incubated with 800 ␮M APAP in MLM for 60 min at 37 ◦ C. In control experiment, A77–1726 was omitted from the APAP incubations. As shown in Fig. 3, A77–1726 at very low concentrations (0.01 and 0.1 ␮M) resulted in moderate but not statistically significant decline in NAPQI formation as compared to the control. It was only at higher concentrations (≥ 0.5 ␮M) that A77–1726 resulted in significant inhibition (P < 0.05) in the formation of NAPQI. In our mechanism-based inhibition study using MLM, the concentrations of A77–1726 after preincubation of LEF were determined using standard calibration in our study and its levels were found

178

S.C. Tan et al. / Toxicology Letters 180 (2008) 174–181

Fig. 4. Time- and concentration-dependent effects of LEF on the bioactivation of APAP to NAPQI–GSH in HLM after incubation for 60 min at 37 ◦ C. Data points are mean ± S.D. (n = 3). Fig. 3. Concentration-dependent effect of A77–1726 on the formation of NAPQI–GSH in MLM after incubation for 60 min at 37 ◦ C. *P < 0.05 vs. control. Data points are mean ± S.D. (n = 3).

to range from 0 to 0.2 ␮M. Collectively, these results confirmed that LEF was the main inhibitor in our mechanism-based inhibition study of APAP bioactivation, while its major metabolite, A77–1726, only played a very minor role in the inhibition process.

nificantly as compared to the control where APAP was incubated without LEF (see Fig. 4). Preincubation of LEF also did not influence the levels of NAPQI significantly. Similar metabolism of LEF to A77–1726 was observed in HLM after preincubation. Our results confirmed that both LEF and A77–1726 did not inhibit the bioactivation of APAP to NAPQI in HLM. Representative MRM chromatograms of NAPQI–GSH, LEF and A77–1726 detected in HLM and CYP2E1 are illustrated in Fig. 5.

3.3. HLM experiments 3.4. Human recombinant P450 enzymes experiments To ascertain whether LEF also exhibits inhibitory effect on the bioactivation of APAP, a similar mechanism-based inhibition study was performed in HLM incubations containing LEF (0.5, 2, and 15 ␮M) and 800 ␮M APAP. In control experiments, LEF was omitted from the APAP incubations. It was clear that LEF at the three investigated concentrations did not alter the levels of NAPQI sig-

To identify the P450 isozyme that is responsible for the inhibition of APAP bioactivation to NAPQI, LEF (2 and 15 ␮M) was incubated with human recombinant P450 enzymes 1A2, 2E1 and 3A4 at 37 ◦ C, in the presence of 800 ␮M APAP. As a control experiment, APAP was incubated in the absence of LEF. Specific inhibitors

Fig. 5. MRM chromatograms of NAPQI–GSH, LEF and A77–1726 detected in (A) HLM (incubation time = 60 min) analysed on the Agilent 1100 HPLC coupled with the 2000 QTRAP MS and (B) CYP2E1 (incubation time = 60 min) ACQUITY UPLC coupled with the 2000 QTRAP MS.

S.C. Tan et al. / Toxicology Letters 180 (2008) 174–181

179

4. Discussion

Fig. 6. Effect of LEF on the bioactivation of APAP to NAPQI–GSH in human recombinant P450s (A) 1A2, (B) 2E1 and (C) 3A4 enzymes after incubation for 60 min at 37 ◦ C. P < 0.05 vs. control. Data points are mean ± S.D. (n = 3).

of the isozymes were incubated with APAP as positive controls to validate the performance of the assays. As shown in Fig. 6A and C, the formation of NAPQI was significantly inhibited in the presence of furafylline (P4501A2 inhibitor) and ketoconazole (P4503A4 inhibitor) when APAP was incubated with P450s 1A2 and 3A4, respectively. Although not statistically significant, the formation of NAPQI was also markedly inhibited in the presence of DDTC (P4502E1 inhibitor) when APAP was incubated with the P4502E1 isozyme (Fig. 6B). As shown in Fig. 6A, LEF at both 2 and 15 ␮M inhibited the formation of NAPQI significantly in a concentrationdependent manner. In contrast, incubations of LEF and APAP in both P450s 2E1 and 3A4 did not show any significant reduction in the levels of NAPQI as compared to APAP alone (see Fig. 6B and C). Our data confirmed that LEF inhibited P4501A2 in the bioactivation of APAP to NAPQI while it did not exhibit any inhibitory effect on P450s 2E1 and 3A4, which are also the enzymes capable of catalyzing the bioactivation of APAP to its toxic metabolite.

The aim of our work was to investigate whether LEF and/or A77–1726 inhibit the bioactivation of APAP to NAPQI in vitro, and the P450 isozyme that is involved in the inhibition process. This was based on our hypothesis that LEF and APAP may interact pharmacokinetically as the metabolism of both drugs is mediated by common P450 isozymes. In order to establish the formation of NAPQI in all the experiments, accurate mass measurement of the NAPQI–GSH conjugate was conducted using QTOFMS. In our QTOFMS experiment, accurate mass measurement was rendered possible by the simultaneous but independent acquisition of reference ions of leucine enkephalin via the LockSprayTM interface. Our results demonstrated that the mass accuracy measurements were consistently high for NAPQI–GSH and all of its seven fragments (less than 5 RMS ppm). The identity of NAPQI–GSH was further confirmed based on its fragmentation pattern (see Fig. 1) which is related to that of a GSH conjugate. Some of the observed mass peaks of NAPQI–GSH corresponds to the hallmark loss of glycine (m/z = 75) and pyroglutamic acid (m/z 129) that is typical of the CID fragmentation of GSH. The QTOF/MS/MS spectral data were also consistently observed in our QTRAP/MS/MS experiments and other studies (Lee, 2004; Jaeschke and Bajt, 2006). Finally, it was observed that NAPQI–GSH was not formed in APAP-deprived microsomal incubations. Collectively, we confirmed that NAPQI–GSH was formed in all microsomal experiments in our study. In this study, quantitative method validations were performed for LEF and A77–1726. On the other hand, for NAPQI–GSH (where the standard compound is not available), its relative quantitation was based on the direct measurement of its peak ion intensity in the MRM chromatogram. Our validation results for LEF and A77–1726 have been reported earlier (Seah et al., in press). The current work was inspired by the papers published by Latchoumycandane et al. (2006, 2007). Their papers presented clear and strong evidence that the JNK pathway is implicated in the protection of APAP-induced liver toxicity by LEF. Importantly, they demonstrated the potential of LEF in rescuing patients who are overdosed with APAP but are diagnosed several hours after the over dosage. This increases our curiosity if LEF or A77–1726 also protects the liver by inhibiting NAPQI formation when LEF is administered early. If LEF truly inhibits both the bioactivation of APAP to NAPQI and the JNK pathway, administering LEF early (before 4 h) may prevent APAP-induced liver toxicity effectively via dual pathways. This was the point that we investigated in this paper. Based on Latchoumycandane et al. (2006), LEF was shown not to inhibit the metabolism of APAP to NAPQI using human hepatocytes by measuring the in vitro level of GSH. However, it has been reported that the GSH tripeptide is unstable and prone to auto-oxidation to GSSG, especially at elevated temperature (Rossi et al., 2002). During the incubation of the cells at 37 ◦ C, auto-oxidation of GSH to GSSG may occur. Further, the GSH assay, using monochlorobimane (MCB) as the fluorescence substrate, requires heating the sample at 100 ◦ C for 3 min to remove glutathione S-transferase (Latchoumycandane et al., 2006). This step would also promote the auto-oxidation of GSH to GSSG and rendered the measurement inaccurate. On the other hand, using direct measurement of the reactive metabolite GSH conjugate, our MLM incubation study demonstrated clearly the concentration-dependent inhibition of the bioactivation of APAP to NAPQI by LEF. Our results suggested that the mode of protection by LEF against APAP-induced hepatotoxicity in mice might be the outcome of dual effects. Firstly, LEF may exert its hepatoprotective effect in the early phase of APAP-induced toxicity, via the prevention of oxidative bioactivation of APAP to its toxic metabolite, NAPQI. The second protection mode of LEF is based on the findings from previous study that LEF inhibits JNK1/2-mediated mPT

180

S.C. Tan et al. / Toxicology Letters 180 (2008) 174–181

and prevents the formation of peroxynitrite (Latchoumycandane et al., 2007), a pivotal mediator of APAP-induced toxicity (Hinson et al., 1998; Jaeschke and Bajt, 2006). However, one may challenge that LEF is usually not detected in plasma after oral administration. Hence, its inhibition of the bioactivation of APAP to NAPQI and the associated hepatoprotective action due to this pathway may be insignificant in an in vivo experiment. As LEF is a prodrug which is rapidly and almost completely converted to its active metabolite, A77–1726, we sought to determine if A77–1726 is responsible in the inhibition process in MLM. Our results revealed that at very low concentrations (<0.5 ␮M), A77–1726 did not inhibit the formation of NAPQI significantly. It was only at higher concentrations of A77–1726 (≥0.5 ␮M) that concentration-dependent inhibition was observed. These results supported our earlier observation that inhibition of NAPQI formation by A77–1726 was not significant in our mechanism-based inhibition experiment as the amount of A77–1726 formed during preincubation fell within the low concentration range. Nonetheless, we have to be mindful that such an absence of in vitro inhibition on NAPQI formation by A77–1726 cannot be extrapolated directly to an in vivo experiment as the pharmacokinetic parameters of this metabolite is unknown. Assuming the free drug concentration of A77–1726 is greater than 0.5 ␮M in the liver, after considering its steady state plasma concentration and plasma/tissue binding, our data suggested that it may be possible for A77–1726 to inhibit NAPQI formation in vivo. As such, future animal study should include the pharmacokinetic profiling of the levels of LEF and A77–1726 in both blood and liver tissue samples. As LEF has been shown to interfere with the bioactivation of APAP in MLM, we are interested to determine whether such drug–drug interaction is also present in the HLM. No study on the investigation of the effect of LEF on the bioactivation of APAP has been performed to date in HLM. In this study, it was found interestingly that LEF did not inhibit the formation of NAPQI in HLM. Our findings were consistent with the results reported by the previous in vitro study in immortalized human hepatocytes (Latchoumycandane et al., 2006), in which LEF did not interfere with the bioactivation of APAP to its reactive metabolite, but rather the liver protection effect was achieved via the inhibition of the more distal pathways. Taken together, our results confirmed that LEF does not exert its hepatoprotective effect due to the involvement of metabolic interactions in human liver cells. For the human recombinant P450 enzyme incubations, the enzymes used were P450s 1A2, 2E1, and 3A4. These enzymes were chosen based on the understanding that they are all involved in the metabolism of APAP to NAPQI, while P450s 1A2 and 3A4 are responsible for the metabolism of both LEF and APAP to their other respective metabolites. Our results identified P4501A2 as the sole enzyme involved in the inhibition of APAP bioactivation by LEF, while P450s 2E1 and 3A4 were not inhibited. Collectively, it was deduced that the effects of LEF in MLM resulted from inhibition of CYP1A2, and the absence of effects in HLM was likely to be due to species differences in APAP bioactivation and CYP1A2 content/activity. While various studies had reported species differences in the P450 enzymes which accounted for the differences in enzyme catalytic activities (Guengerich, 1997; Lewis et al., 1998), using the specific metabolism of phenacetin to its O-deethylated metabolite, it was further confirmed in our study that the CYP1A2 activity in MLM was five to six times greater than that in HLM (data not presented). Clinically, other P450 isozymes including CYP2E1 and CYP3A4 are involved in the bioactivation of APAP to NAPQI, and CYP2E1 has been implicated as the most relevant isozyme in human. Since CYP2E1 activity was found not to be inhibited by LEF in vitro, we speculated that LEF would not contribute significantly to the inhibition of NAPQI formation in human.

In conclusion, we demonstrated that the mode of protection by LEF in mice against APAP-induced toxicity via the inhibition of NAPQI formation is probable and future animal study is needed to confirm it. In contrast, such inhibition was not observed in HLM in vitro. Our study supported the previous findings (Latchoumycandane et al., 2006) that the hepatoprotective effect of LEF in human liver cells was not due to direct inhibition of APAP bioactivation but via the inhibition of JNK1/2-mediated mPT and prevention of peroxynitrite formation. Conflict of interest statement The authors declare that there are no conflicts of interest. Acknowledgements This work was supported by the National University of Singapore grant (R-148-050-088-101/133 to E.C.Y. Chan) and the Department of Pharmacy Final Year Project grant (R-148-000-003-001). The authors are also grateful to Professor F. Peter Guengerich at the Vanderbilt-Ingram Cancer Center for providing constructive suggestions to improve our paper. References Drugdex® Evaluations, 2008. Acetaminophen. In: Micromedex® Healthcare Series. Thomson Healthcare, North Carolina. Guengerich, F.P., 1997. Comparisons of catalytic selectivity of cytochrome P450 subfamily enzymes from different species. Chem. Biol. Interact. 106, 161–182. Hinson, J.A., Pike, S.L., Pumford, N.R., Mayeux, P.R., 1998. Nitrotyrosine–protein adducts in hepatic centrilobular areas following toxic doses of acetaminophen in mice. Chem. Res. Toxicol. 11, 604–607. Jaeschke, H., Bajt, M.L., 2006. Intracellular signaling mechanisms of acetaminopheninduced liver cell death. Toxicol. Sci. 89, 31–41. Kalgutkar, A.S., Nguyen, H.T., Vaz, A.D.N., Doan, A., Dalvie, D.K., McLeod, D.G., Murray, J.C., 2003. In vitro metabolism studies on the isoxazole ring scission in the antiinflammatory agent leflunomide to its active ␣-cyanoenol metabolite A77 1726: mechanistic similarities with the cytochrome P450-catalyzed dehydration of aldoximes. Drug Metab. Dispos. 31, 1240–1250. Kon, K., Kim, J.S., Jaeschke, H., Lemasters, J.J., 2004. Mitochondrial permeability transition in acetaminophen-induced necrosis and apoptosis of cultured mouse hepatocytes. Hepatology 40, 1170–1179. Latchoumycandane, C., Seah, Q.M., Tan, R.C.H., Sattabongkot, J., Beerheide, W., Boelsterli, U.A., 2006. Leflunomide or A77 1726 protect from acetaminophen-induced cell injury through inhibition of JNK-mediated mitochondrial permeability transition in immortalized human hepatocytes. Toxicol. Appl. Pharmacol. 217, 125–133. Latchoumycandane, C., Goh, C.W., Ong, M.M.K., Boelsterli, U.A., 2007. Mitochondrial protection by the JNK inhibitor leflunomide rescues mice from acetaminopheninduced liver injury. Hepatology 45, 412–421. Lee, W.M., 2004. Acetaminophen and the U.S Acute Liver Failure Study Group: lowering the risks of hepatic failure. Hepatology 40, 6–9. Lewis, D.F.V., Ioannides, C., Parke, D.V., 1998. Cytochrome P450 and species differences in xenobiotic metabolism and activation of carcinogen. Environ. Health Perspect. 106, 633–641. Lucian, J., Dias, V.C., LeGatt, D.F., Yatscoff, R.W., 1995. Blood distribution and singledose pharmacokinetics of leflunomide. Ther. Drug Monit. 17, 454–459. Manyike, P.T., Kharasch, E.D., Kalhorn, T.F., Slattery, J.T., 2000. Contribution of CYP2E1 and CYP3A to acetaminophen reactive metabolite formation. Clin. Pharmacol. Ther. 67, 275–282. Masubuchi, Y., Suda, C., Horie, T., 2005. Involvement of mitochondrial permeability transition in acetaminophen-induced liver injury in mice. J. Hepatol. 42, 110–116. Rossi, R., Milzani, A., Dalle-Donne, I., Giustarini, D., Lusini, L., Colombo, R., Di Simplicio, P., 2002. Blood glutathione disulfide: in vivo factor or in vitro artifact? Clin. Chem. 48, 742–753. Rozman, B., 2002. Clinical pharmacokinetics of leflunomide. Clin. Pharmacokinet. 41, 421–430. Schwabe, R.F., Uchinami, H., Qian, T., Bennett, B.L., Lemasters, J.J., Brenner, D.A., 2004. Differential requirement for c-Jun NH2 -terminal kinase in TNF␣- and Fas-mediated apoptosis in hepatocytes. FASEB J. 18, 720–722. Seah, Q.M., New, L.S., Chan, E.C.Y., Boelsterli, U.A., in press. Oxidative bioactivation and toxicity of leflunomide in immortalized human hepatocytes and kinetics of the non-enzymatic conversion to its major metabolite, A77 1726. Drug Metab. Lett. Smilkstein, M.J., Knapp, G.L., Kulig, K.W., Rumack, B.H., 1988. Efficacy of oral Nacetylcysteine in the treatment of acetaminophen overdose. Analysis of the national multicenter study (1976–1985). N. Engl. J. Med. 319, 1557–1562.

S.C. Tan et al. / Toxicology Letters 180 (2008) 174–181 Smilkstein, M.J., Bronstein, A.C., Linden, C., Augenstein, W.L., Kulig, K.W., Rumack, B.H., 1991. Acetaminophen overdose: a 48-hour intravenous N-acetylcysteine treatment protocol. Ann. Emerg. Med. 20, 1058–1063. Snawder, J.E., Roe, A.L., Benson, R.W., Roberts, D.W., 1994. Loss of CYP2E1 and CYP1A2 activity as a function of acetaminophen dose: relation to toxicity. Biochem. Biophys. Res. Commun. 203, 532–539.

181

Terneus, M.V., Kiningham, K.K., Carpenter, A.B., Sullivan, S.B., Valentovic, M.A., 2007. Comparison of S-adenosyl-l-methionine and N-acetylcysteine protective effects on acetaminophen hepatic toxicity. J. Pharmacol. Exp. Ther. 320, 99–107. Yao, H.W., Li, J., Jin, Y., Zhang, Y.F., Li, C.Y., Xu, S.Y., 2003. Effect of leflunomide on immunological liver injury in mice. World J. Gastroenterol. 9, 320–323.