1521-009X/42/11/1834–1842$25.00 DRUG METABOLISM AND DISPOSITION Copyright ª 2014 by The American Society for Pharmacology and Experimental Therapeutics

http://dx.doi.org/10.1124/dmd.114.059352 Drug Metab Dispos 42:1834–1842, November 2014

Evaluation of In Situ Generated Valproyl 1-O-b-Acyl Glucuronide in Valproic Acid Toxicity in Sandwich-Cultured Rat Hepatocytes Jayakumar Surendradoss, Thomas K. H. Chang, and Frank S. Abbott Faculty of Pharmaceutical Sciences, University of British Columbia, Vancouver, British Columbia, Canada Received June 2, 2014; accepted August 21, 2014

ABSTRACT inducers evaluated, b-naphthoflavone produced the greatest increase in VPA-G formation. This was accompanied by an attenuation of the increase in BODIPY 558/568 C12 accumulation, but did not affect the change in LDH release or total GSH content in VPA-treated hepatocytes. Inhibition of in situ formation of VPA-G by borneol was not accompanied by substantive changes in the effects of VPA on any of the toxicity markers. In a comparative study, in situ generated diclofenac glucuronide was not toxic to rat hepatocytes, as assessed using the same chemical modulators, thereby demonstrating the utility of the sandwich-cultured rat hepatocyte model. Overall, in situ generated VPA-G was not toxic to sandwich-cultured rat hepatocytes, suggesting that VPA glucuronidation per se is not expected to be a contributing mechanism for VPA hepatotoxicity.

Introduction

chemical reactivity, which is attributed to the chemical structure of the aglycone moiety (Stachulski et al., 2006). Postulated mechanisms for the toxicity of acyl glucuronides include the following: 1) a direct impairment of the function of a key protein that is covalently modified; 2) an indirect immune reaction to the antigenic drug-protein adducts; and 3) formation of more reactive acyl-glutathione thioester conjugates with intracellular glutathione (GSH), resulting in GSH depletion and possibly covalent binding to proteins (Shipkova et al., 2003; Skonberg et al., 2008). Although acyl glucuronides are hypothesized to be involved in various toxicities, including hepatotoxicity, of carboxylic acid-containing drugs (Regan et al., 2010; Boelsterli, 2011), it remains to be established whether there is a causal role for these reactive species. Valproic acid (VPA) is a commonly used antiepileptic drug that is effective against various types of seizures and epileptic syndromes (Loscher, 2002). This drug undergoes microsomal glucuronidation, mitochondrial b-oxidation, and microsomal cytochrome P450–mediated oxidation (Abbott and Anari, 1999). Glucuronidation is a major metabolic pathway for VPA and contributes to the biotransformation of about 30–50% of the administered dose of VPA in humans (Silva et al., 2008). Valproyl 1-O-b-acyl glucuronide (VPA-G) is one of the least reactive acyl glucuronides investigated to date (Stachulski et al., 2006). Yet, it undergoes intramolecular acyl migration to form positional isomers of VPA-G (Dickinson et al., 1984) and appears to be responsible, at least partly, for the formation of VPA-protein adducts in vitro in rat hepatocytes (Porubek et al., 1989). In rats (Tong et al., 2003) and pediatric patients (Michoulas et al., 2006), VPA increases the in vivo levels of 15-F2tisoprostane, which is a marker of lipid peroxidation (Halliwell and

Drug-induced hepatotoxicity is a common cause of acute liver failure (Tujios and Fontana, 2011). Evidence for drug-induced hepatotoxicity leads to attrition during drug development, refusal of drug approval, and black box warning or postmarketing withdrawal (Regev, 2013). Druginduced hepatotoxicity can be intrinsic, the occurrence of which is relatively common, predictable, and dose-dependent (Russmann et al., 2009), or idiosyncratic, which occurs in a rare, unpredictable, and often dose-independent fashion in a few susceptible patients (Regev, 2013). Various mechanisms of drug-induced hepatotoxicity have been proposed, including formation of reactive electrophilic metabolites (Srivastava et al., 2010; Leung et al., 2012). One such class of reactive metabolites is the acyl glucuronides (Kalgutkar et al., 2005). Acyl glucuronides, which are enzymatic products formed by glucuronidation of carboxylic acids (Stachulski et al., 2006), are capable of undergoing the following: 1) hydrolysis to parent aglycone mediated by b-glucuronidases, nonspecific esterases, hydroxyl ion, or serum albumin; 2) intramolecular acyl migration to form positional isomers that are resistant to b-glucuronidase–mediated hydrolysis; and 3) covalent binding to proteins via transacylation or glycation mechanisms (Regan et al., 2010). Acyl glucuronides, however, differ widely in their This work was supported by the Canadian Institutes of Health Research [Grant MOP-13744 to F.S.A. and T.K.H.C.]. J.S. received a Four Year Doctoral Fellowship from University of British Columbia. T.K.H.C. received a Senior Scholar Award from the Michael Smith Foundation for Health Research. dx.doi.org/10.1124/dmd.114.059352.

ABBREVIATIONS: DFN-G, diclofenac 1-O-b-acyl glucuronide; DMSO, dimethylsulfoxide; GSH, glutathione; LDH, lactate dehydrogenase; 49-OH-DFN, 49-hydroxy diclofenac; 5-OH-DFN, 5-hydroxy diclofenac; PBS, phosphate-buffered saline; PCN, pregnenolone 16a-carbonitrile; UGT, uridine 59-diphospho-glucuronosyltransferase; UHPLC-MS/MS, ultra-high performance liquid chromatography–tandem mass spectrometry; VPA, valproic acid; VPA-G, valproyl 1-O-b-acyl glucuronide. 1834

Downloaded from dmd.aspetjournals.org at ASPET Journals on March 22, 2015

Acyl glucuronides are reactive electrophilic metabolites implicated in the toxicity of carboxylic acid drugs. Valproyl 1-O-b-acyl glucuronide (VPA-G), which is a major metabolite of valproic acid (VPA), has been linked to the development of oxidative stress in VPA-treated rats. However, relatively little is known about the toxicity of in situ generated VPA-G and its contribution to VPA hepatotoxicity. Therefore, we investigated the effects of modulating the in situ formation of VPA-G on lactate dehydrogenase (LDH) release (a marker of necrosis), BODIPY 558/568 C12 accumulation (a marker of steatosis), and cellular glutathione (GSH) content in VPA-treated sandwich-cultured rat hepatocytes. VPA increased LDH release and BODIPY 558/568 C12 accumulation, whereas it had little or no effect on total GSH content. Among the various uridine 59-diphospho-glucuronosyltransferase

Evaluating Toxicity of VPA Glucuronide in Rat Hepatocytes

Materials and Methods Chemicals, Reagents, and Solvents. Sodium VPA (CAS 1069-66-5), sodium diclofenac (CAS 15307-79-6), sodium phenobarbital (CAS 57-30-7), b-naphthoflavone (CAS 6051-87-2), 3-methylcholanthrene (CAS 56-49-5), L-sulforaphane (CAS 142825-10-3), quercetin (CAS 117-39-5), dexamethasone (CAS 50-02-2), clofibrate (CAS 637-07-0), pregnenolone 16a-carbonitrile (PCN; CAS 1434-54-4), trans-stilbene oxide (CAS 1439-07-2), (2)-borneol (CAS 464-45-9), and dimethylsulfoxide (DMSO; CAS 67-68-5) were obtained from Sigma-Aldrich (St. Louis, MO). DFN-G (CAS 64118-81-6), 49-hydroxydiclofenac (49-OH-DFN; CAS 64118-84-9), and 5-hydroxy diclofenac (5-OH-DFN; CAS 69002-84-2) were purchased from Toronto Research Chemicals (North York, ON, Canada). The LDH Cytotoxicity Detection Kit was obtained from Roche Diagnostics (Indianapolis, IN), and the Glutathione Assay Kit was from Cayman Chemical (Ann Arbor, MI). Williams’ medium E, liver perfusion medium, hepatocyte wash medium, heat-inactivated fetal bovine serum, phosphate-buffered saline (PBS; pH 7.4), 10 Hank’s balanced salt solution, 10 Dulbecco’s PBS, penicillin-streptomycin, L-glutamine, and BODIPY 558/568 C12 were obtained from Invitrogen (Burlington, ON, Canada). Matrigel basement membrane matrix was obtained from BD Biosciences (Mississauga, ON, Canada). Percoll was purchased from GE Healthcare (Baie d’Urfe, QC, Canada). Ammonium acetate, ethyl acetate, diethyl ether, acetonitrile, methanol, n-hexanes, glacial acetic acid, and sodium hydroxide were obtained from Fisher Scientific (Ottawa, ON, Canada). Animals. Adult male Sprague-Dawley rats (175–200 g) were obtained from Charles River Laboratories (Senneville, QC, Canada) and were housed and cared for in our animal care facility, as described previously (Surendradoss et al., 2012). All animal experiments were approved by the University of British Columbia Animal Care Committee and were conducted in accordance with the guidelines of the Canadian Council on Animal Care. Isolation, Culture, and Treatment of Rat Hepatocytes. Rat hepatocytes were isolated by a two-step collagenase perfusion method (Seglen, 1993), as described previously (Tong et al., 2005a). Plating of hepatocytes (0.7  106 cells/well), preparing the sandwich-culture configuration, and culturing of hepatocytes were performed, as reported earlier (Surendradoss et al., 2012). At 120 hours after plating, sandwich-cultured rat hepatocytes were treated with

VPA, culture medium (vehicle for VPA), diclofenac, or DMSO (vehicle for diclofenac) for the next 24 hours at the concentrations indicated in each figure legend. In other experiments, at 48 hours after plating, cultured hepatocytes were pretreated with an inducer of uridine 59-diphospho-glucuronosyltransferase (UGT) or an inhibitor of glucuronidation and then treated with VPA, diclofenac, or vehicle, as described in each figure legend. Quantification of VPA-G Concentration. At the end of the drug treatment period, culture supernatant was collected and hepatocytes were lysed with 2% Triton X-100 in PBS (pH 7.4) containing 20 mM EDTA. Each sample was transferred into a microfuge tube and stored at 280C until analysis. VPA-G concentrations in culture supernatant and cell lysate were quantified using a validated ultra-high performance liquid chromatography–tandem mass spectrometry (UHPLC-MS/MS) assay with [2H6]VPA-G as the internal standard (Surendradoss et al., 2013). The UHPLC-MS/MS system consisted of an Agilent 1290 Infinity Binary Pump, a 1290 Infinity Autosampler, and a 1290 Infinity Thermostatted Column Compartment (Agilent Technologies, Mississauga, Ontario, Canada), which was connected to an AB Sciex QTRAP 5500 hybrid linear ion-trap triplequadrupole mass spectrometer equipped with a Turbo Spray source (AB Sciex, Concord, Ontario, Canada). The mass spectrometer was operated in negative ionization mode. Quantification of DFN-G Concentration. As a precaution to maintain the stability of DFN-G, the culture supernatant samples and the stock solutions of calibration standards were thawed and maintained on wet ice during sample preparation and were acidified with 2 M acetic acid solution (4% v/v final concentration) to reduce intramolecular acyl migration and/or hydrolysis of DFN-G (Sparidans et al., 2008). To quantify DFN-G concentrations in culture supernatant from diclofenac-treated hepatocytes, 10 ml of sample and 10 ml of 50 mg/ml solution of [2H6]VPA-G (internal standard) were added to 480 ml of assay diluent (85% of 2 mM ammonium acetate in water and 15% of 2 mM ammonium acetate in a 9:1 mixture of acetonitrile and water), vortex-mixed for 10 seconds, and centrifuged at 10,600g for 5 minutes at 4C. A 15-ml volume was injected onto the UHPLC-MS/MS system. The calibration curve of DFN-G ranged from 2.1 to 2120 nM. The mobile phases, chromatographic gradient, and mass-spectrometric conditions were the same as those described previously for VPA-G assay (Surendradoss et al., 2013). The declustering potential and collision energy settings were 260 V and 212 V for DFN-G, and 240 V and 220 V for [2H6]VPA-G, respectively. DFN-G was analyzed using the total ion current of the multiple reaction monitoring transition m/z 470.2→192.8 (Koga et al., 2011), with the internal standard [2H6]VPA-G transition pairs being m/z 325.1→149.3 and 325.1→174.9. LDH Assay. LDH release was used as a marker of cell necrosis (Jauregui et al., 1981). LDH activity in the culture supernatants and cell lysate samples was determined using the Cytotoxicity Detection Kit (Roche Diagnostics) (Surendradoss et al., 2012). LDH released into the culture supernatant was expressed as a percentage of the total cellular LDH activity (i.e., sum of the LDH activity in the culture supernatant and cell lysate). BODIPY 558/568 C12 Assay. Cellular accumulation of BODIPY 558/568 C12 was used as an index of steatosis (Fujimura et al., 2009) and determined, as described previously (Surendradoss et al., 2012). Fluorescence was measured at a lex of 484 nm and lem of 618 nm in a Biotek Synergy Mx microplate reader (Biotek Instruments, Winooski, VT). Each blank well had culture medium containing BODIPY 558/568 C12 and a test compound, but without cells. BODIPY 558/568 C12 accumulation was expressed as fold increase in fluorescence in drug-treated wells over that in vehicle-treated control wells. Total GSH Assay. Cellular content of total GSH was quantified using the Glutathione Assay Kit (Cayman Chemical), as described previously (Surendradoss et al., 2012). The rate of formation of the reaction product, 5-thio-2-nitrobenzoic acid, was determined spectrophotometrically in a kinetic mode at a wavelength of 405 nm in a Labsystems Multiskan Ascent multiwell plate reader (Thermo Electron, Burlington, ON, Canada). The blank sample consisted of equal volumes of PBS (pH 7.4; supplemented with 1 mM EDTA) and metaphosphoric acid, but without cell homogenate. Quantification of Oxidative Metabolites of VPA. Concentrations of the oxidative metabolites of VPA in culture supernatants from VPA-treated hepatocytes were quantified using a gas chromatography–mass spectrometry assay (Surendradoss et al., 2012). Determination of 49-OH-DFN and 5-OH-DFN. Concentrations of 49-OH-DFN and 5-OH-DFN in the culture supernatants of diclofenac-treated

Downloaded from dmd.aspetjournals.org at ASPET Journals on March 22, 2015

Whiteman, 2004). The increase in plasma and hepatic levels of 15-F 2t -isoprostane in rats administered VPA was accompanied by an increase in the levels of VPA-G (Tong et al., 2005b). In another in vivo study in rats, hepatic and urinary concentrations of VPA-G did not correlate with serum levels of a-glutathione-S-transferase (Lee et al., 2009), which is a marker of hepatotoxicity (Bailey et al., 2012). The therapeutic use of VPA by humans is associated with a rare, but potentially fatal, idiosyncratic hepatotoxicity (Nanau and Neuman, 2013). Although the mechanism of VPA hepatotoxicity is not understood, VPA-G has been proposed to play a role (Tong et al., 2005b). However, there is no direct evidence as to whether in situ generated VPA-G is hepatotoxic. Therefore, to increase our understanding of the toxicological significance of VPA-G, the present study was conducted to determine whether in situ generated VPA-G is toxic and whether it contributes to the toxicity of VPA in sandwich-cultured rat hepatocytes. It is now increasingly recognized that the sandwich-cultured hepatocyte model is appropriate for studying hepatic biotransformation and toxicity of drugs and other chemicals (Swift et al., 2010). Diclofenac is another well-known carboxylic acid drug that is associated with idiosyncratic hepatotoxicity (Tang, 2003) Although diclofenac undergoes glucuronidation to form an unstable acyl glucuronide, the hepatotoxicity of this drug has been attributed to the cytochrome P450–mediated oxidative metabolites of diclofenac (Tang, 2003). As a comparison, the present study also determined the effect of modulating the in situ formation of a more reactive acyl glucuronide, diclofenac 1-O-b-acyl glucuronide (DFN-G), on the release of lactate dehydrogenase (LDH) in cultured hepatocytes treated with diclofenac. The findings of the present study provide insight to the question as to whether VPA-G plays a role in VPA hepatotoxicity.

1835

1836

Surendradoss et al.

Results Concentration of In Situ Generated VPA-G in Culture Supernatant and Cell Lysate of Sandwich-Cultured Rat Hepatocytes Treated with VPA. As assessed in hepatocytes treated with VPA (1 mM), the concentration of VPA-G was 251 6 12 and 6.8 6 1.5 mM (mean 6 S.E.M.; n = 4 rats per treatment group) in the culture supernatant and cell lysate, respectively. As more than 97% of the in situ generated VPA-G was localized in the culture supernatant, VPA-G concentration was quantified in culture supernatant in all the subsequent experiments, unless indicated otherwise. Time Course and Concentration-Response Relationship in the In Situ Formation of VPA-G in Sandwich-Cultured Rat Hepatocytes Treated with VPA. As shown in Fig. 1A, the in situ concentration of VPA-G continued to increase in the culture supernatants in a linear fashion over the 1- to 24-hour post-treatment period. Concentration-response experiments indicated that the in situ formation of VPA-G increased from 20 to 348 mM in response to increases in VPA concentration from 0.03 to 3 mM, reached a peak at 3–10 mM VPA, and decreased at $20 mM VPA (Fig. 1B). Based

on these results, VPA concentrations of 10 and 15 mM were chosen to investigate the toxicity of in situ generated VPA-G, as these concentrations of VPA resulted in maximal or near-maximal formation of VPA-G (Fig. 1B) and elicited a measurable response in the toxicity markers (Surendradoss et al., 2012). Effect of Various Known UGT Inducers on In Situ Formation of VPA-G in Sandwich-Cultured Rat Hepatocytes Treated with VPA. Except for rat UGT2B1 (Pritchard et al., 1994), the identity of specific rat UGT enzymes catalyzing VPA glucuronidation is not known. Therefore, prior to investigating the toxicity of the in situ generated VPA-G in sandwich-cultured rat hepatocytes treated with VPA, initial experiments were performed to identify chemical modulators of VPA glucuronidation. Cultured hepatocytes were pretreated with a known UGT enzyme inducer, such as b-naphthoflavone (Viollon-Abadie et al., 2000), L-sulforaphane (Kohle and Bock, 2006), phenobarbital (Soars et al., 2004), 3-methylcholanthrene (Jemnitz et al., 2000), quercetin (Soars et al., 2004), clofibrate (Jemnitz et al., 2000), dexamethasone (Jemnitz et al., 2000), PCN (Shelby and Klaassen, 2006), or trans-stilbene oxide (Shelby and Klaassen, 2006) once every 24 hours for 72 hours, followed by VPA treatment for the next 24 hours. As shown in Fig. 2A, b-naphthoflavone, L-sulforaphane, and phenobarbital increased the in situ formation of VPA-G by 2.3-, 1.7-, and 1.9-fold, respectively, whereas none of the other chemicals had an effect (data not shown). As a comparison, b-naphthoflavone, L-sulforaphane, and phenobarbital increased the in situ formation of DFN-G by 3.4-, 4.6-, and 2.9-fold, respectively (Fig. 2A). Among the UGT inducers investigated, b-naphthoflavone yielded the greatest increase in in situ VPA-G formation (Fig. 2A). As evident from the concentration-response data, pretreatment with b-naphthoflavone at 20 mM concentration produced the maximal increase in in situ concentrations of VPA-G in culture supernatants and cell lysates of VPA-treated cells (Table 1). Therefore, this concentration (20 mM) was used in subsequent modulation experiments involving b-naphthoflavone. Effect of Increasing In Situ VPA-G Formation by b-Naphthoflavone on Markers of Toxicity in Sandwich-Cultured Rat Hepatocytes Treated with VPA. To investigate the effects of increasing the in situ formation of VPA-G on VPA toxicity in sandwich-cultured rat hepatocytes, we assessed LDH release (a marker of necrosis), BODIPY 558/568 C12 accumulation (marker of steatosis), and cellular content of total GSH, all of which are relevant to VPA hepatotoxicity (Jurima-Romet et al., 1996; Silva et al., 2008). Cultured hepatocytes were pretreated with b-naphthoflavone (20 mM) or DMSO (0.1% v/v; vehicle) once every 24 hours for 72 hours, followed by treatment with VPA (10 or 15 mM) or culture medium (vehicle) for the next 24 hours. At the end of the treatment period, LDH release, BODIPY 558/568 C12 accumulation, and cellular concentration of total GSH were measured. As shown in

Fig. 1. Time course and concentration-response relationship in the in situ formation of VPA-G in sandwich-cultured rat hepatocytes treated with VPA. (A) Hepatocytes were cultured for 120 hours and then treated with VPA (1 mM) or culture medium (vehicle). At 1, 2, 4, 8, 12, and 24 hours after drug treatment, an aliquot of culture supernatant was collected and VPA-G concentrations were quantified by UHPLC-MS/MS. (B) Hepatocytes were cultured for 120 hours and then treated with VPA (0.03–100 mM) or culture medium (vehicle) for 24 hours. Data are expressed as mean 6 S.E.M. for three or four rats.

Downloaded from dmd.aspetjournals.org at ASPET Journals on March 22, 2015

hepatocytes were determined semiquantitatively using a liquid chromatography– tandem mass spectrometry assay adapted from Sparidans et al. (2008). Briefly, 50 ml of the culture supernatant sample was added to 150 ml of assay diluent (60% of solvent A, 8.5 mM ammonium acetate in water containing 0.0075% formic acid, and 40% of solvent B, methanol), vortex-mixed for 10 seconds, and centrifuged at 10,600g for 5 minutes at 4C. A 5 ml volume of the supernatant was injected onto the UHPLC-MS/MS system. The calibration curves of 49-OH-DFN and 5-OH-DFN ranged from 0.03 to 16 mM. The mobile phases and the chromatographic gradient conditions used in this assay were the same as those described previously (Sparidans et al., 2008). Analysis was performed under multiple reaction monitoring mode on a QTRAP 5500 linear ion trap mass spectrometer operated in positive electrospray ionization, using the following instrument settings: curtain gas, 30 U; ion source gas 1, 60 U; ion source gas 2, 40 U; collision-activated dissociation gas level, high; ion source temperature, 400C; ion spray voltage, 5500 V; collision cell exit potential, 18 V; entrance potential, 10 V; and dwell time, 150 ms. The 49-OH-DFN and 5-OH-DFN were analyzed using the sum of the total ion currents of the multiple reaction monitoring transitions m/z 312.0→230.9, m/z 312.0→266.0, and m/z 312.0→294.0. Whereas the declustering potential was 66 V, the collision energy settings for the three MRM transitions were 27, 19, and 15 V for the three transitions, respectively. Under the assay conditions employed, the retention times of 49-OH-DFN and 5-OH-DFN were 4.39 and 5.03 minutes, respectively. Statistical Analysis. Data were analyzed by one-way or two-way analysis of variance, as appropriate, and when there were significant differences, the data were further analyzed by the Student-Newman-Keuls multiple comparison test (Sigmaplot for Windows, version 11.0; Systat Software, Chicago, IL). The level of statistical significance was set a priori at P , 0.05.

1837

Evaluating Toxicity of VPA Glucuronide in Rat Hepatocytes

Fig. 2B, b-naphthoflavone alone did not affect LDH release, whereas 10 and 15 mM VPA increased it. b-Naphthoflavone pretreatment did not further increase the LDH release by VPA (Fig. 2B). In a comparative experiment, b-naphthoflavone increased DFN-G concentrations by 3.4-fold (Fig. 2A), and this was accompanied by an attenuation of LDH release in diclofenac-treated hepatocytes (Fig. 2B). Pretreatment with b-naphthoflavone alone did not affect BODIPY 558/568 C12 accumulation, whereas 10 and 15 mM VPA increased BODIPY 558/568 C12 accumulation (Fig. 3A). This increase in BODIPY 558/568 C12 accumulation by VPA was attenuated by b-naphthoflavone pretreatment (Fig. 3A). In contrast to the increases observed for the LDH and BODIPY markers, VPA (10 and 15 mM) had no effect on total GSH content (Fig. 3B). As shown in Fig. 3B, b-naphthoflavone pretreatment alone led to an increase in the concentration of total GSH. However, treatment with 15 mM VPA, but not 10 mM VPA, decreased the concentration of total GSH (by ;25%) in the b-naphthoflavone–pretreated hepatocytes. Effect of Borneol on In Situ Formation of VPA-G in SandwichCultured Rat Hepatocytes Treated with VPA. Another approach to investigate the toxicity of in situ formed VPA-G is to determine the toxicological consequence of inhibiting its metabolic formation. Initial experiments were performed to determine the effect of borneol, which is a known inhibitor of glucuronidation (Watkins and Klaassen, 1983; Dong and Smith, 2009), on in situ formation of VPA-G in sandwichcultured rat hepatocytes treated with VPA. Cultured hepatocytes were pretreated with borneol (0.25, 0.5, 0.75, or 1 mM) or DMSO (0.1% v/v; vehicle) for 0.5 hour, followed by cotreatment with VPA (10 or 15 mM) or culture medium (vehicle) for the next 24 hours. Borneol, over the concentration range of 0.25–1 mM, decreased both extracellular and intracellular concentrations of VPA-G in VPA-treated hepatocytes (Table 1). As 1 mM borneol produced the greatest inhibition of in situ VPA-G formation (Table 1), this concentration was selected for the subsequent experiments. As shown in Fig. 4A, whereas 1 mM borneol decreased the extracellular concentration of VPA-G, the same concentration of borneol did not affect the in situ formation of DFN-G in hepatocytes treated with diclofenac for 24 hours (Fig. 4A). Effect of Decreasing In Situ VPA-G Formation by Borneol on Markers of Toxicity in Sandwich-Cultured Rat Hepatocytes Treated with VPA. The next experiment was to investigate the effects of decreasing in situ formation of VPA-G by borneol on

markers of toxicity in sandwich-cultured rat hepatocytes treated with VPA. Cultured hepatocytes were pretreated with borneol (1 mM) or DMSO (0.1% v/v; vehicle) for 0.5 hour, followed by a 24-hour cotreatment with VPA (10 or 15 mM) or culture medium (vehicle). As shown in Fig. 4B, borneol alone did not affect LDH release, whereas 10 and 15 mM VPA treatment increased LDH release by 2- and 4.5-fold, respectively. In hepatocytes treated with 10 or 15 mM VPA, cotreatment with borneol (1 mM) decreased in situ concentration of VPA-G, but it had no effect on LDH release (Fig. 4B). By comparison, borneol (1 mM) attenuated LDH release in hepatocytes treated with 400 mM diclofenac (Fig. 4B), but this was not accompanied by a change in the in situ concentration of DFN-G (Fig. 4A). As shown in Fig. 5A, treatment of cultured hepatocytes with 10 and 15 mM VPA increased BODIPY

TABLE 1 Concentration-response effect of b-naphthoflavone and borneol on in situ formation of valproyl 1-O-b-acyl glucuronide in sandwich-cultured rat hepatocytes treated with valproic acid Experiment 1: At 48 hours after plating, hepatocytes were pretreated with b-naphthoflavone (BNF; 5, 10, 15, or 20 mM) or 0.1% dimethylsulfoxide (DMSO; vehicle) every 24 hours for 72 hours and then treated with valproic acid (VPA; 10 mM) or culture medium (vehicle) for the next 24 hours. Experiment 2: At 120 hours after plating, hepatocytes were pretreated with borneol (0.25, 0.5, 0.75, or 1 mM) or 0.1% DMSO (vehicle) for 0.5 hour. Subsequently, hepatocytes were treated with VPA (10 mM) or culture medium (vehicle) in the presence of borneol or DMSO at concentrations indicated above for 24 hours. In both experiments, culture supernatants and cell lysates were collected at the end of the 24-hour treatment period and subjected to ultra-high performance liquid chromatography–tandem mass spectrometry assay for valproyl 1-O-b-acyl glucuronide (VPA-G) concentrations. Data are expressed as mean 6 S.E.M. for three or four rats. In Situ Concentration of VPA-G (fold increase over control) Pretreatment/Treatment

Experiment 1 DMSO 0.1% v/v BNF 5 mM BNF 10 mM BNF 15 mM BNF 20 mM Experiment 2 DMSO 0.1% v/v Borneol 0.25 mM Borneol 0.5 mM Borneol 0.75 mM Borneol 1 mM

Culture Supernatant

Cell Lysate

1.0 1.7 2.2 2.2 2.3

6 6 6 6 6

0.0 0.06* 0.23* 0.17* 0.20*

1.0 6 0.0 Not determined Not determined Not determined 2.1 6 0.09*

1.0 0.90 0.83 0.72 0.64

6 6 6 6 6

0.0 0.03* 0.02* 0.04* 0.04*

1.0 0.85 0.78 0.72 0.65

*Significantly different from the DMSO vehicle control group, P , 0.05.

6 6 6 6 6

0.0 0.05* 0.04* 0.04* 0.03*

Downloaded from dmd.aspetjournals.org at ASPET Journals on March 22, 2015

Fig. 2. Effect of UGT enzyme inducers on in situ formation of VPA-G and LDH release in sandwich-cultured rat hepatocytes treated with VPA. (A) Hepatocytes were cultured for 48 hours and then pretreated with b-naphthoflavone (BNF; 20 mM), L-sulforaphane (L-SFN; 5 mM), sodium phenobarbital (PB; 2 mM), or vehicle (0.1% DMSO for BNF and L-SFN; culture medium for PB) once every 24 hours for 72 hours. Subsequently, the hepatocytes were treated with VPA (10 mM), culture medium (vehicle for VPA), diclofenac (DFN; 400 mM), or DMSO (0.1% v/v; vehicle for DFN) for the next 24 hours. (B) Hepatocytes were pretreated with BNF and then treated with VPA (10 or 15 mM), diclofenac (400 mM), or their respective vehicle, as described above in (A). Data are expressed as mean 6 S.E.M. for three rats. *Significantly different from the respective vehicle-pretreated control group, P , 0.05. #Significantly different from the respective culture medium-treated group, P , 0.05. †Significantly different from the DMSO + DMSO group, P , 0.05. ‡Significantly different from the BNF + DMSO and the DMSO + diclofenac groups, P , 0.05. VPA-G concentrations were 256 6 45 and 213 6 35 mM per 0.7 million cells in the culture medium- and DMSO-pretreated groups, respectively. DFN-G concentrations were 23 6 2.0 and 20 6 1.8 mM per 0.7 million cells in the culture medium- and DMSO-pretreated groups, respectively.

1838

Surendradoss et al.

Fig. 3. Effect of b-naphthoflavone on (A) BODIPY 558/568 C12 accumulation and (B) total GSH content in sandwich-cultured rat hepatocytes treated with VPA. Hepatocytes were cultured for 48 hours and then pretreated with b-naphthoflavone (BNF; 20 mM) or DMSO (0.1% v/v; vehicle) once every 24 hours for 72 hours. Subsequently, the hepatocytes were treated with VPA (10 or 15 mM) or culture medium (vehicle) for the next 24 hours. Data are expressed as mean 6 S.E.M. for three to five rats. *Significantly different from the DMSO + culture medium group, P , 0.05; #significantly different from the corresponding DMSO + culture medium or DMSO + VPA groups, P , 0.05; ‡significantly different from the BNF + culture medium and DMSO + 15 mM VPA groups, P , 0.05. The total GSH content was 7.4 6 1.6 mM per 0.7 million cells in the DMSO + culture medium group.

P450–mediated oxidation (Abbott and Anari, 1999). Therefore, we determined whether b-naphthoflavone and borneol are capable of modulating the oxidative biotransformation of VPA, as a means to gain insight into the observed lack of toxicity of VPA-G in cultured hepatocytes treated with VPA (e.g., Fig. 2B and 4B). As shown in Table 2, b-naphthoflavone increased the concentrations of 4-keto-VPA, 4-OH-VPA, and 3-OH-VPA (two- to threefold), but not the other oxidative metabolites of VPA in cultured hepatocytes treated with VPA. By comparison, borneol decreased the concentrations of only 4-OH-VPA (60%) and 5-OH-VPA (40%) (Table 3). Effect of b-Naphthoflavone and Borneol on In Situ Generated 49-OH-DFN and 5-OH-DFN in Sandwich-Cultured Rat Hepatocytes Treated with Diclofenac. Diclofenac undergoes biotransformation not only by glucuronidation, but also by cytochrome P450–mediated oxidation to produce 49-OH-DFN and 5-OH-DFN, which undergo subsequent oxidation to form highly electrophilic benzoquinone imine intermediates (Tang, 2003). Therefore, we determined the effects of b-naphthoflavone and borneol on the in situ concentrations of 49-OH-DFN and 5-OH-DFN in diclofenac-treated hepatocytes to rationalize the attenuation in LDH release by b-naphthoflavone (Fig. 2B) and borneol (Fig. 4B). Interestingly, b-naphthoflavone pretreatment differentially altered the concentrations of the two hydroxy metabolites of diclofenac; b-naphthoflavone increased the concentration of 49-OH-DFN by over twofold, whereas it attenuated the concentration of 5-OH-DFN by 30% (Fig. 7). Borneol did not affect the concentration of 49-OH-DFN, but it decreased the concentration of 5-OH-DFN by 80% (Fig. 7).

Fig. 4. Effect of borneol on in situ formation of VPA-G and LDH release in sandwich-cultured rat hepatocytes treated with VPA. Hepatocytes were cultured for 120 hours and then pretreated with borneol (1 mM) or DMSO (0.1% v/v; vehicle) for 0.5 hour. Subsequently, the hepatocytes were treated with 10 mM VPA (A and B), 15 mM VPA (B), culture medium (vehicle for VPA; A and B), 400 mM diclofenac (DFN; A and B), or DMSO (0.1% v/v; vehicle for DFN; A and B) for the next 24 hours in the presence of borneol (1 mM) or DMSO (0.1% v/v). Data are expressed as mean 6 S.E.M. for three rats. #Significantly different from the DMSO + VPA group, P , 0.05; *significantly different from the DMSO + culture medium group or borneol + culture medium group, P , 0.05; †significantly different from the DMSO + DMSO group, P , 0.05; ‡significantly different from the borneol + DMSO and DMSO + diclofenac groups, P , 0.05. VPA-G concentration was 283 6 41 mM per 0.7 million cells in the DMSO + VPA group. DFN-G concentration was 20 6 1.8 mM per 0.7 million cells in the DMSO + diclofenac group.

Downloaded from dmd.aspetjournals.org at ASPET Journals on March 22, 2015

558/568 C12 accumulation by 1.9- and 2.5-fold, respectively, and this increase was not affected by borneol. Treatment of cultured hepatocytes with 15 mM VPA decreased the total GSH content by 35%, which was further decreased by borneol (Fig. 5B). In contrast, neither 10 mM VPA nor cotreatment with borneol affected total GSH content (Fig. 5B). Time Course of the Effect of Borneol on the In Situ Formation of VPA-G and LDH Release in Sandwich-Cultured Rat Hepatocytes Treated with VPA. A time course experiment was performed to further characterize the effect of borneol on in situ formation of VPA-G and LDH release in VPA-treated hepatocytes. VPA increased VPA-G concentrations in the culture supernatants over a 2- to 24-hour period, and this increase was attenuated by borneol at 4, 8, and 24 hours after VPA treatment (Fig. 6A). VPA increased LDH release, but this did not occur until 24 hours post-treatment, and the effect was further enhanced by borneol (Fig. 6B). By comparison, diclofenac increased DFN-G concentrations over the 2- to 24-hour time period (Fig. 6C), whereas borneol decreased the in situ formation of DFN-G only at 2 and 4 hours after drug treatment, but it had no effect at 8 hours, and in fact modestly increased DFN-G formation at 24 hours after drug treatment (Fig. 6C). Treatment with diclofenac resulted in an increase in LDH release at 8 and 24 hours post-treatment, and this increase was attenuated by cotreatment with borneol (Fig. 6D). Effect of b-Naphthoflavone and Borneol on In Situ Generated Oxidative Metabolites of VPA in Sandwich-Cultured Rat Hepatocytes Treated with VPA. In addition to glucuronidation, VPA is also biotransformed by mitochondrial b-oxidation and cytochrome

Evaluating Toxicity of VPA Glucuronide in Rat Hepatocytes

1839

Fig. 5. Effect of borneol on (A) BODIPY accumulation and (B) total GSH content in sandwich-cultured rat hepatocytes treated with VPA. Hepatocytes were cultured for 120 hours and then pretreated with borneol (1 mM) or DMSO (0.1% v/v; vehicle) for 0.5 hour. Subsequently, the hepatocytes were treated with VPA (10 or 15 mM) or culture medium (vehicle) for 24 hours in the presence of borneol (1 mM) or DMSO (0.1% v/v). Data are expressed as mean 6 S.E.M. for three or four rats. *Significantly different from the DMSO + culture medium group, P , 0.05; #significantly different from the borneol + culture medium, P , 0.05; †significantly different from the borneol + culture medium and DMSO + 15 mM VPA groups. The total GSH content was 13 6 1.1 mM per 0.7 million cells in the DMSO + culture medium group.

Discussion

Fig. 6. Time course on the effect of borneol on (A) VPA-G concentration, (B) LDH release after VPA treatment, (C) DFN-G concentration, and (D) LDH release following diclofenac treatment in sandwich-cultured rat hepatocytes. Hepatocytes were cultured for 120 hours and then pretreated with borneol (1 mM) or DMSO (0.1% v/v; vehicle) for 0.5 hour. Subsequently, the hepatocytes were treated with VPA (15 mM), culture medium (vehicle for VPA), diclofenac (DFN; 400 mM), or DMSO (0.1% v/v; vehicle for DFN) for the next 2, 4, 8, and 24 hours in the presence of borneol (1 mM) or DMSO (0.1% v/v). Data are expressed as mean 6 S.E.M. for four rats. *Significantly different from either DMSO + VPA (A) or DMSO + diclofenac (C) treatment group, P , 0.05; †significantly different from either DMSO + culture medium (B) or DMSO + DMSO treatment group (D), P , 0.05; ‡significantly different from either the borneol + culture medium and DMSO + VPA treatment groups (B) or the borneol + DMSO and DMSO + diclofenac treatment groups (D), P , 0.05.

Downloaded from dmd.aspetjournals.org at ASPET Journals on March 22, 2015

Acyl glucuronides have been implicated in the hepatotoxicity of carboxylic acid drugs; however, there is no direct experimental evidence linking an acyl glucuronide to hepatotoxicity (Stachulski et al., 2013). VPA-G, which is an acyl glucuronide (Dickinson et al., 1984), has been linked to the formation of covalent VPA adducts with hepatocellular proteins (Porubek et al., 1989) and the development of oxidative stress in rats by VPA (Tong et al., 2005b). Yet, the

hepatotoxic potential of VPA-G remains to be investigated. As shown in the present study, a major finding is that in situ generated VPA-G did not appear to be toxic to sandwich-cultured rat hepatocytes. This conclusion is based on the following experimental evidence: 1) an increase in the in situ formation of VPA-G by b-naphthoflavone was not accompanied by an increase in VPA toxicity (in fact, it led to an attenuation of BODIPY 558/568 C12 accumulation); and 2) inhibition of in situ VPA-G formation by borneol did not result in an attenuation

1840

Surendradoss et al. TABLE 2 Effect of b-naphthoflavone on in situ formation of oxidative metabolites of valproic acid in sandwich-cultured rat hepatocytes treated with valproic acid At 48 hours after plating, hepatocytes were pretreated with 20 mM b-naphthoflavone (BNF) or 0.1% dimethylsulfoxide (DMSO; vehicle) once every 24 hours for 72 hours and then treated with valproic acid (VPA; 10 or 15 mM) or culture medium (vehicle) for the next 24 hours. Culture supernatants were collected at the end of the 24-hour treatment period and subjected to gas chromatography–mass spectrometry assay for the quantification of the concentrations of oxidative metabolites of VPA. Data are expressed as mean 6 S.E.M. for three rats. In Situ Concentration of VPA Metabolites (mM/0.7  106 cells) Metabolite

10 mM VPA

0.06 0.15 2.5 1.4 None 0.16 1.7 1.2 1.7 5.7 2.0

6 0.01 6 0.01 6 0.30 6 0.27 detected 6 0.01 6 0.27 6 0.20 6 0.13 6 1.2 6 0.63

0.09 6 0.01 0.50 6 0.06* 4.9 6 0.54* 1.2 6 0.17* None detected 0.19 6 0.01 1.7 6 0.21 1.3 6 0.14 1.5 6 0.06 11 6 4.2 1.7 6 0.32

20 mM BNF

0.1% DMSO

0.05 0.22 2.1 1.3 None 0.15 1.7 1.0 1.8 5.5 2.1

6 0.01 6 0.02 6 0.25 6 0.25 detected 6 0.01 6 0.34 6 0.15 6 0.21 6 1.5 6 0.99

0.08 6 0.01 0.58 6 0.08* 5.1 6 0.46* 1.2 6 0.24 None detected 0.19 6 0.02 1.8 6 0.34 1.1 6 0.13 1.8 6 0.20 18 6 5.0* 2.2 6 0.79

*Significantly different from the DMSO-pretreated vehicle control group, P , 0.05.

of VPA toxicity. The lack of an effect of b-naphthoflavone on VPA toxicity cannot be due to the increased cellular GSH content by b-naphthoflavone because depletion of GSH has been reported as a consequence rather than a cause of VPA toxicity in cultured rat hepatocytes (Kiang et al., 2011). In support of our finding that in situ generated VPA-G is nontoxic to rat hepatocytes, previous studies have shown that VPA-G is one of the least reactive and the most stable acyl glucuronides (Stachulski et al., 2006). Furthermore, as demonstrated under the cell culture conditions employed in this study, acyl migration of VPA-G was not significant (Surendradoss et al., 2013), suggesting that VPA-G did not form more reactive positional isomers. Overall, VPA glucuronidation appears to be a typical detoxification pathway, as determined in the current study in sandwich-cultured rat hepatocytes. Another novel finding of the present study was obtained from evaluating the effects of various known inducers of UGTs on increasing in situ formation of VPA-G in sandwich-cultured rat hepatocytes treated

with VPA. Among the various inducers investigated in this study, only b-naphthoflavone, phenobarbital, and L-sulforaphane, but not 3-methylcholanthrene, quercetin, clofibrate, dexamethasone, PCN, or trans-stilbene oxide, increased in situ formation of VPA-G in sandwich-cultured rat hepatocytes. The observed magnitude of the increase (;twofold) in VPA glucuronidation by b-naphthoflavone, phenobarbital, and L-sulforaphane is consistent with the notion that the inducibility of UGT enzymes is less than that of cytochrome P450 enzymes (Soars et al., 2004). The increase in the extent of drug glucuronidation in response to prototypical UGT inducers is usually ;twofold (Lin and Wong, 2002; Soars et al., 2004). As shown in a previous ex vivo study (Shelby and Klaassen, 2006), b-naphthoflavone induces the hepatic gene expression of rat UGT1A3, UGT1A6, and UGT1A7, whereas PB induces UGT2B1. Therefore, these UGT enzymes are likely catalysts of VPA glucuronidation. In fact, rat recombinant UGT2B1 has been shown to glucuronidate VPA (Pritchard et al., 1994).

TABLE 3 Effect of borneol on in situ formation of oxidative metabolites of valproic acid in sandwich-cultured rat hepatocytes treated with valproic acid At 120 hours after plating, hepatocytes were pretreated with 1 mM borneol or 0.1% dimethylsulfoxide (DMSO; vehicle) for 0.5 hour. Subsequently, hepatocytes were treated with valproic acid (VPA) (10 or 15 mM) or culture medium (vehicle) in the presence of 1 mM borneol or 0.1% DMSO for 24 hours. Culture supernatants were collected at the end of the 24-hour treatment period and subjected to gas chromatography–mass spectrometry assay for the quantification of the concentrations of oxidative metabolites of VPA. Data are expressed as mean 6 S.E.M. for four rats. In Situ Concentration of VPA Metabolites (mM/0.7  106 cells) Metabolite

10 mM VPA 0.1% DMSO

4-ene-VPA 4-keto-VPA 4-OH-VPA 5-OH-VPA (E)-2,4-diene-VPA (E,Z)-2,39-diene-VPA (E,E)-2,39-diene-VPA 3-ene-VPA (E)-2-ene-VPA 3-OH-VPA 3-keto-VPA

0.08 0.09 3.4 1.6 None 0.19 1.5 1.3 1.8 7.1 1.4

6 0.01 6 0.01 6 0.56 6 0.10 detected 6 0.02 6 0.09 6 0.13 6 0.16 6 0.36 6 0.33

1 mM Borneol

0.05 6 0.01* 0.09 6 0.02 1.4 6 0.45* 0.94 6 0.05* None detected 0.19 6 0.02 1.4 6 0.20 1.2 6 0.16 1.9 6 0.20 6.2 6 1.5 1.1 6 0.35

*Significantly different from the DMSO-cotreated vehicle control group, P , 0.05.

15 mM VPA 0.1% DMSO

0.04 0.13 2.2 1.7 None 0.18 1.6 1.1 2.2 11 2.8

6 0.01 6 0.01 6 0.22 6 0.23 detected 6 0.01 6 0.22 6 0.13 6 0.29 6 3.1 6 0.75

1 mM Borneol

0.04 0.16 1.9 1.3 None 0.20 1.5 1.0 2.2 7.2 2.1

6 0.01 6 0.02 6 0.21 6 0.19 detected 6 0.01 6 0.14 6 0.10 6 0.23 6 1.1 6 0.97

Downloaded from dmd.aspetjournals.org at ASPET Journals on March 22, 2015

4-ene-VPA 4-keto-VPA 4-OH-VPA 5-OH-VPA (E)-2,4-diene-VPA (E,Z)-2,39-diene-VPA (E,E)-2,39-diene-VPA 3-ene-VPA (E)-2-ene-VPA 3-OH-VPA 3-keto-VPA

15 mM VPA 20 mM BNF

0.1% DMSO

Evaluating Toxicity of VPA Glucuronide in Rat Hepatocytes

As a comparison with the results obtained with VPA-G, the present study also investigated the effects of modulating in situ formation of DFN-G on hepatocyte toxicity of diclofenac. The increase in VPA-G formation by b-naphthoflavone did not affect LDH release in VPAtreated sandwich-cultured rat hepatocytes, whereas the increase in DFN-G formation by b-naphthoflavone resulted in an attenuation of LDH release in diclofenac-treated hepatocytes. The reason for the differential effect of b-naphthoflavone on the toxicity of VPA and diclofenac is not known, but it may involve distinct modulation of other biotransformation pathways of these two drugs. Other than glucuronidation, pretreatment of rat hepatocytes with b-naphthoflavone also increased cytochrome P450–mediated biotransformation of VPA, as shown in the present study. However, the in situ formation of these VPA oxidative metabolites has been reported not to influence VPA toxicity in sandwich-cultured rat hepatocytes (Kiang et al., 2010; Surendradoss et al., 2012). In the case of diclofenac, the hepatocyte toxicity of this drug has been attributed to the formation of cytochrome P450–mediated oxidative metabolites of diclofenac (Kretz-Rommel and Boelsterli, 1993). In the present study, b-naphthoflavone attenuated the hepatocyte toxicity of diclofenac, even though it increased in situ formation of 49-OH-DFN. The reason for the attenuation of diclofenac toxicity is not known, but it may relate to the increase in cellular GSH content by b-naphthoflavone. GSH is involved in the conjugation and detoxification of cytochrome P450–mediated reactive metabolites of diclofenac (Tang et al., 1999). Overall, the increase in the in situ formation of two acyl glucuronides (VPA-G and DFN-G), which differ greatly (;150-fold) in chemical stability (Stachulski et al., 2006), did not enhance the hepatocyte toxicity of the parent drugs VPA and diclofenac. As shown in the present study, cotreatment of sandwich-cultured rat hepatocytes with borneol and VPA decreased VPA-G formation over a period of 4–24 hours after VPA treatment, but this was not associated with a decrease in LDH release over the same time period. By comparison, cotreatment of hepatocytes with borneol and diclofenac was seen to decrease DFN-G levels for only up to 4 hours after diclofenac treatment. The apparent recovery of the in situ levels of DFN-G over the 4- to 24-hour period was accompanied by a marked decrease in the extent of LDH release in the 8- and 24-hour time points after the cotreatment of hepatocytes with borneol and

diclofenac. A possible explanation for the observed decrease in LDH release by borneol in diclofenac-treated hepatocytes is the substantive attenuation of the in situ levels of the oxidative metabolite 5-OH-DFN. Previously, borneol was reported to increase rather than decrease LDH release in diclofenac-treated rat hepatocytes (Kretz-Rommel and Boelsterli, 1993). The reason for the differences in the effect of borneol on LDH release in diclofenac-treated hepatocytes shown in the present study and in the earlier study is not known; however, it may relate to the differences in the duration of borneol treatment [24 hours in the present study and 2 hours in the study by Kretz-Rommel and Boelsterli (1993)]. Borneol inhibits the glucuronidation of drugs and other chemicals by depleting the hepatocellular levels of the cofactor uridine 59-diphospho-glucuronic acid and inhibiting the catalytic activity of specific UGT enzymes that catalyze glucuronidation (Watkins and Klaassen, 1982). As evident by our time course data, the inhibitory effect of borneol on in situ VPA-G formation was apparent from 4 to 24 hours after VPA treatment, whereas its inhibition of in situ DFN-G formation no longer occurred at 8 hours after diclofenac treatment. The differences in the temporal profile on the effect of borneol on in situ formation of VPA-G and DFN-G suggest that borneol does not inhibit the glucuronidation of VPA and diclofenac by the same mechanism. Sandwich-cultured hepatocytes have been used to study the toxicity of acyl glucuronides (Dong and Smith, 2009). However, a limitation of this model is the lack of enterohepatic recycling. As evident from a preliminary experiment, hydrolysis of VPA-G was not apparent in our sandwich-cultured rat hepatocyte model. Pretreatment or cotreatment with an inhibitor of VPA-G hydrolysis (Suzuki et al., 2010), D-saccharolactone or a carbapenem (meropenem or imipenem), did not enhance in situ concentrations of VPA-G, as assessed at 24 hours after treatment with VPA (data not shown). However, VPA-G is subject to efficient hydrolysis and enterohepatic recycling in rats in vivo (Dickinson et al., 1979; Liu and Smith, 2006). Thus, the conjugationdeconjugation cycling may increase the hepatocellular exposure to VPA-G in vivo. It remains to be investigated whether such an increase in exposure to VPA-G in vivo contributes to the hepatotoxicity of VPA. In conclusion, similar to the findings on DFN-G in the same cell culture model, in situ generated VPA-G was not toxic to sandwichcultured rat hepatocytes, as evident from the experiments using chemical modulators of VPA glucuronidation. The findings of this study, together with the available information regarding the chemical reactivity of VPA-G (Stachulski et al., 2006), lead us to conclude that in situ generated VPA-G does not contribute to the hepatocyte toxicity of VPA. As reported previously, the in situ formation of other reactive metabolites of VPA, such as 4-ene-VPA (Kiang et al., 2010) and (E)-2,4-diene-VPA (Surendradoss et al., 2012), also does not play a role in VPA toxicity. Valproyl-S-acyl CoA is another metabolite of VPA formed by the b-oxidation pathway (Silva et al., 2008). It has been proposed that bioactivation of carboxylic acids may also involve acyl-CoA thioesters and acyl-glutathione thioesters, which are even more reactive than acyl glucuronides (Skonberg et al., 2008). The toxicological significance of acyl-CoA thioester metabolites has been relatively less studied in comparison with acyl glucuronides, although the intracellular localization of acyl-CoA thioesters makes them more probable mediators of hepatotoxicity associated with carboxylic acid drugs (Darnell and Weidolf, 2013). In a previous study, it was speculated that the formation of valproyl-S-acyl CoA thioester may contribute to the hepatotoxicity of VPA (Grillo et al., 2001). It would be of interest in the future to investigate the role of valproyl-S-acyl-CoA and other downstream b-oxidation metabolites

Downloaded from dmd.aspetjournals.org at ASPET Journals on March 22, 2015

Fig. 7. Effect of b-naphthoflavone and borneol on in situ formation of 49-OH-DFN and 5-OH-DFN in sandwich-cultured rat hepatocytes treated with diclofenac. At 48 hours after plating, hepatocytes were pretreated with b-naphthoflavone (BNF; 20 mM) or DMSO (vehicle; 0.1% v/v) once every 24 hours for 72 hours, followed by treatment with diclofenac (400 mM) or 0.1% DMSO (vehicle) for the next 24 hours. In another experiment, hepatocytes were cultured for 120 hours and pretreated with borneol (1 mM) or DMSO (0.1% v/v; vehicle) for 0.5 hour, followed by treatment with diclofenac (400 mM) or 0.1% DMSO (vehicle) in the presence of borneol or DMSO for the next 24 hours. Data are expressed as mean 6 S.E.M. for three rats. *Significantly different from the DMSO group, P , 0.05. The 49-OH-DFN and 5-OH-DFN concentrations were 1.0 6 0.20 and 0.46 6 0.20 mM per 0.7 million cells, respectively, in the DMSO group.

1841

1842

Surendradoss et al.

of VPA in the hepatotoxicity of VPA using the sandwich-cultured hepatocyte model. Acknowledgments The authors thank Andras Szeitz (Faculty of Pharmaceutical Sciences, University of British Columbia) for technical assistance with the UHPLC-MS/MS assays for VPA-G and DFN-G. Authorship Contributions Participated in research design: Surendradoss, Chang, Abbott. Conducted experiments: Surendradoss. Performed data analysis: Surendradoss. Wrote or contributed to the writing of the manuscript: Surendradoss, Chang, Abbott. References

Address correspondence to: Dr. Frank S. Abbott, Faculty of Pharmaceutical Sciences, University of British Columbia, 2405 Wesbrook Mall, Vancouver, BC, V6T 1Z3, Canada. E-mail: [email protected]

Downloaded from dmd.aspetjournals.org at ASPET Journals on March 22, 2015

Abbott FS and Anari MR (1999) Chemistry and biotransformation, in Valproate (Loscher W, ed) pp 47–75, Birkhauser Verlag, Basel, Switzerland. Bailey WJ, Holder D, Patel H, Devlin P, Gonzalez RJ, Hamilton V, Muniappa N, Hamlin DM, Thomas CE, and Sistare FD,, et al. (2012) A performance evaluation of three drug-induced liver injury biomarkers in the rat: alpha-glutathione S-transferase, arginase 1, and 4-hydroxyphenylpyruvate dioxygenase. Toxicol Sci 130:229–244. Boelsterli UA (2011) Acyl glucuronides: mechanistic role in drug toxicity? Curr Drug Metab 12: 213–214. Darnell M and Weidolf L (2013) Metabolism of xenobiotic carboxylic acids: focus on coenzyme A conjugation, reactivity, and interference with lipid metabolism. Chem Res Toxicol 26: 1139–1155. Dickinson RG, Harland RC, Ilias AM, Rodgers RM, Kaufman SN, Lynn RK, and Gerber N (1979) Disposition of valproic acid in the rat: dose-dependent metabolism, distribution, enterohepatic recirculation and choleretic effect. J Pharmacol Exp Ther 211:583–595. Dickinson RG, Hooper WD, and Eadie MJ (1984) pH-dependent rearrangement of the biosynthetic ester glucuronide of valproic acid to beta-glucuronidase-resistant forms. Drug Metab Dispos 12:247–252. Dong JQ and Smith PC (2009) Glucuronidation and covalent protein binding of benoxaprofen and flunoxaprofen in sandwich-cultured rat and human hepatocytes. Drug Metab Dispos 37: 2314–2322. Fujimura H, Murakami N, Kurabe M, and Toriumi W (2009) In vitro assay for drug-induced hepatosteatosis using rat primary hepatocytes, a fluorescent lipid analog and gene expression analysis. J Appl Toxicol 29:356–363. Grillo MP, Chiellini G, Tonelli M, and Benet LZ (2001) Effect of alpha-fluorination of valproic acid on valproyl-S-acyl-CoA formation in vivo in rats. Drug Metab Dispos 29:1210–1215. Halliwell B and Whiteman M (2004) Measuring reactive species and oxidative damage in vivo and in cell culture: how should you do it and what do the results mean? Br J Pharmacol 142: 231–255. Jauregui HO, Hayner NT, Driscoll JL, Williams-Holland R, Lipsky MH, and Galletti PM (1981) Trypan blue dye uptake and lactate dehydrogenase in adult rat hepatocytes—freshly isolated cells, cell suspensions, and primary monolayer cultures. In Vitro 17:1100–1110. Jemnitz K, Veres Z, Monostory K, and Vereczkey L (2000) Glucuronidation of thyroxine in primary monolayer cultures of rat hepatocytes: in vitro induction of UDP-glucuronosyltranferases by methylcholanthrene, clofibrate, and dexamethasone alone and in combination. Drug Metab Dispos 28:34–37. Jurima-Romet M, Abbott FS, Tang W, Huang HS, and Whitehouse LW (1996) Cytotoxicity of unsaturated metabolites of valproic acid and protection by vitamins C and E in glutathionedepleted rat hepatocytes. Toxicology 112:69–85. Kalgutkar AS, Gardner I, Obach RS, Shaffer CL, Callegari E, Henne KR, Mutlib AE, Dalvie DK, Lee JS, and Nakai Y,, et al. (2005) A comprehensive listing of bioactivation pathways of organic functional groups. Curr Drug Metab 6:161–225. Kiang TK, Teng XW, Karagiozov S, Surendradoss J, Chang TK, and Abbott FS (2010) Role of oxidative metabolism in the effect of valproic acid on markers of cell viability, necrosis, and oxidative stress in sandwich-cultured rat hepatocytes. Toxicol Sci 118:501–509. Kiang TK, Teng XW, Surendradoss J, Karagiozov S, Abbott FS, and Chang TK (2011) Glutathione depletion by valproic acid in sandwich-cultured rat hepatocytes: role of biotransformation and temporal relationship with onset of toxicity. Toxicol Appl Pharmacol 252:318–324. Koga T, Fujiwara R, Nakajima M, and Yokoi T (2011) Toxicological evaluation of acyl glucuronides of nonsteroidal anti-inflammatory drugs using human embryonic kidney 293 cells stably expressing human UDP-glucuronosyltransferase and human hepatocytes. Drug Metab Dispos 39:54–60. Köhle C and Bock KW (2006) Activation of coupled Ah receptor and Nrf2 gene batteries by dietary phytochemicals in relation to chemoprevention. Biochem Pharmacol 72:795–805. Kretz-Rommel A and Boelsterli UA (1993) Diclofenac covalent protein binding is dependent on acyl glucuronide formation and is inversely related to P450-mediated acute cell injury in cultured rat hepatocytes. Toxicol Appl Pharmacol 120:155–161. Lee MS, Lee YJ, Kim BJ, Shin KJ, Chung BC, Baek DJ, and Jung BH (2009) The relationship between glucuronide conjugate levels and hepatotoxicity after oral administration of valproic acid. Arch Pharm Res 32:1029–1035. Leung L, Kalgutkar AS, and Obach RS (2012) Metabolic activation in drug-induced liver injury. Drug Metab Rev 44:18–33. Lin JH and Wong BK (2002) Complexities of glucuronidation affecting in vitro in vivo extrapolation. Curr Drug Metab 3:623–646. Liu JH and Smith PC (2006) Predicting the pharmacokinetics of acyl glucuronides and their parent compounds in disease states. Curr Drug Metab 7:147–163.

Löscher W (2002) Basic pharmacology of valproate: a review after 35 years of clinical use for the treatment of epilepsy. CNS Drugs 16:669–694. Michoulas A, Tong V, Teng XW, Chang TK, Abbott FS, and Farrell K (2006) Oxidative stress in children receiving valproic acid. J Pediatr 149:692–696. Nanau RM and Neuman MG (2013) Adverse drug reactions induced by valproic acid. Clin Biochem 46:1323–1338. Porubek DJ, Grillo MP, and Baillie TA (1989) The covalent binding to protein of valproic acid and its hepatotoxic metabolite, 2-n-propyl-4-pentenoic acid, in rats and in isolated rat hepatocytes. Drug Metab Dispos 17:123–130. Pritchard M, Fournel-Gigleux S, Siest G, Mackenzie P, and Magdalou J (1994) A recombinant phenobarbital-inducible rat liver UDP-glucuronosyltransferase (UDP-glucuronosyltransferase 2B1) stably expressed in V79 cells catalyzes the glucuronidation of morphine, phenols, and carboxylic acids. Mol Pharmacol 45:42–50. Regan SL, Maggs JL, Hammond TG, Lambert C, Williams DP, and Park BK (2010) Acyl glucuronides: the good, the bad and the ugly. Biopharm Drug Dispos 31:367–395. Regev A (2013) How to avoid being surprised by hepatotoxicity at the final stages of drug development and approval. Clin Liver Dis 17:749–767, xi. Russmann S, Kullak-Ublick GA, and Grattagliano I (2009) Current concepts of mechanisms in drug-induced hepatotoxicity. Curr Med Chem 16:3041–3053. Seglen PO (1993) Isolation of hepatocytes by collagenase perfusion, in In Vitro Biological Systems (Tyson CA and Frazier JM, eds) pp 231–243, Academic Press, San Diego, California. Shelby MK and Klaassen CD (2006) Induction of rat UDP-glucuronosyltransferases in liver and duodenum by microsomal enzyme inducers that activate various transcriptional pathways. Drug Metab Dispos 34:1772–1778. Shipkova M, Armstrong VW, Oellerich M, and Wieland E (2003) Acyl glucuronide drug metabolites: toxicological and analytical implications. Ther Drug Monit 25:1–16. Silva MF, Aires CC, Luis PB, Ruiter JP, Ijlst L, Duran M, Wanders RJ, and Tavares de Almeida I (2008) Valproic acid metabolism and its effects on mitochondrial fatty acid oxidation: a review. J Inherit Metab Dis 31:205–216. Skonberg C, Olsen J, Madsen KG, Hansen SH, and Grillo MP (2008) Metabolic activation of carboxylic acids. Expert Opin Drug Metab Toxicol 4:425–438. Soars MG, Petullo DM, Eckstein JA, Kasper SC, and Wrighton SA (2004) An assessment of UDP-glucuronosyltransferase induction using primary human hepatocytes. Drug Metab Dispos 32:140–148. Sparidans RW, Lagas JS, Schinkel AH, Schellens JH, and Beijnen JH (2008) Liquid chromatography-tandem mass spectrometric assay for diclofenac and three primary metabolites in mouse plasma. J Chromatogr B Analyt Technol Biomed Life Sci 872:77–82. Srivastava A, Maggs JL, Antoine DJ, Williams DP, Smith DA, and Park BK (2010) Role of reactive metabolites in drug-induced hepatotoxicity. Handbook Exp Pharmacol 196:165–194. Stachulski AV, Baillie TA, Park BK, Obach RS, Dalvie DK, Williams DP, Srivastava A, Regan SL, Antoine DJ, and Goldring CE,, et al. (2013) The generation, detection, and effects of reactive drug metabolites. Med Res Rev 33:985–1080. Stachulski AV, Harding JR, Lindon JC, Maggs JL, Park BK, and Wilson ID (2006) Acyl glucuronides: biological activity, chemical reactivity, and chemical synthesis. J Med Chem 49: 6931–6945. Surendradoss J, Chang TK, and Abbott FS (2012) Assessment of the role of in situ generated (E)-2,4-diene-valproic acid in the toxicity of valproic acid and (E)-2-ene-valproic acid in sandwich-cultured rat hepatocytes. Toxicol Appl Pharmacol 264:413–422. Surendradoss J, Szeitz A, Teng XW, Chang TK, and Abbott FS (2013) A rapid and sensitive assay to quantify valproyl 1-O-acyl glucuronide in supernatants of sandwich-cultured rat hepatocytes using ultra-high performance liquid chromatography-tandem mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci 932:40–49. Suzuki E, Yamamura N, Ogura Y, Nakai D, Kubota K, Kobayashi N, Miura S, and Okazaki O (2010) Identification of valproic acid glucuronide hydrolase as a key enzyme for the interaction of valproic acid with carbapenem antibiotics. Drug Metab Dispos 38:1538–1544. Swift B, Pfeifer ND, and Brouwer KL (2010) Sandwich-cultured hepatocytes: an in vitro model to evaluate hepatobiliary transporter-based drug interactions and hepatotoxicity. Drug Metab Rev 42:446–471. Tang W (2003) The metabolism of diclofenac—enzymology and toxicology perspectives. Curr Drug Metab 4:319–329. Tang W, Stearns RA, Bandiera SM, Zhang Y, Raab C, Braun MP, Dean DC, Pang J, Leung KH, and Doss GA,, et al. (1999) Studies on cytochrome P-450-mediated bioactivation of diclofenac in rats and in human hepatocytes: identification of glutathione conjugated metabolites. Drug Metab Dispos 27:365–372. Tong V, Chang TK, Chen J, and Abbott FS (2003) The effect of valproic acid on hepatic and plasma levels of 15-F2t-isoprostane in rats. Free Radic Biol Med 34:1435–1446. Tong V, Teng XW, Chang TK, and Abbott FS (2005a) Valproic acid II: effects on oxidative stress, mitochondrial membrane potential, and cytotoxicity in glutathione-depleted rat hepatocytes. Toxicol Sci 86:436–443. Tong V, Teng XW, Karagiozov S, Chang TK, and Abbott FS (2005b) Valproic acid glucuronidation is associated with increases in 15-F2t-isoprostane in rats. Free Radic Biol Med 38: 1471–1483. Tujios S and Fontana RJ (2011) Mechanisms of drug-induced liver injury: from bedside to bench. Nat Rev Gastroenterol Hepatol 8:202–211. Viollon-Abadie C, Bigot-Lasserre D, Nicod L, Carmichael N, and Richert L (2000) Effects of model inducers on thyroxine UDP-glucuronosyl-transferase activity in vitro in rat and mouse hepatocyte cultures. Toxicol In Vitro 14:505–512. Watkins JB and Klaassen CD (1982) Effect of inducers and inhibitors of glucuronidation on the biliary excretion and choleretic action of valproic acid in the rat. J Pharmacol Exp Ther 220:305–310. Watkins JB and Klaassen CD (1983) Chemically-induced alteration of UDP-glucuronic acid concentration in rat liver. Drug Metab Dispos 11:37–40.