Abstract
The pregnane X receptor (PXR) is a transcriptional regulator of xenobiotic metabolizing enzymes, including cytochrome P450 3A (CYP3A), and transporters. Pretreatment of mice and rats with inducers of CYP3A increases acetaminophen (APAP) hepatotoxicity. In untreated mice, the amount of hepatic CYP3A11 mRNA is 4-fold greater in PXR(–/–) mice compared to wild-type mice (Guo et al., 2003), a finding anticipated to increase APAP hepatotoxicity in PXR(–/–) mice. We investigated APAP hepatotoxicity in wild-type and PXR(–/–) mice in a C57BL/6 background, with APAP administered by gavage. Despite a 2.5-fold higher level of total hepatic CYP3A protein and a 3.6-fold higher level of CYP3A activity compared to wild-type mice, PXR(–/–) mice were less sensitive to APAP hepatotoxicity. Hepatic levels of CYP2E1 were identical in the two mouse lines, but hepatic CYP1A2 levels were 3-fold greater in wild-type mice compared to PXR(–/–) mice. Caffeine, an inhibitor of CYP1A2 activity and an enhancer of CYP3A activity, decreased APAP hepatotoxicity in wild-type mice. APAP uptake was 1.5-fold greater in wild-type mice compared to PXR(–/–) mice. No significant differences in the formation of APAP glucuronide and sulfate-conjugated metabolites were observed between wild-type and PXR(–/–) mice. Glutathione levels were similar in the two mouse lines and were transiently decreased to similar amounts after APAP administration. Our finding that APAP hepatotoxicity was decreased in PXR(–/–) mice indicates that PXR is an important modulator of APAP hepatotoxicity, through positive modulation of constitutive CYP1A2 expression and possibly through increased APAP absorption.
Acetaminophen (APAP) is an over-the-counter drug commonly used for its analgesic and antipyretic properties. It is also a component of numerous prescription medications. This compound is considered relatively safe at therapeutic doses; however, when taken in overdose, it has been shown to produce potentially fatal centrilobular hepatic necrosis (for review, see James et al., 2003).
At pharmacological doses, APAP is mainly metabolized by phase II xenobiotic-metabolizing enzymes via sulfation and glucuronidation. A small amount of APAP is converted to N-acetyl-p-benzoquinone imine (NAPQI), a reactive metabolite, via cytochromes P450 (P450s) (for review, see James et al., 2003). NAPQI is detoxified by conjugation with glutathione (GSH); however, in APAP overdose, the capacities of phase II pathways are overwhelmed and GSH levels become depleted (for review, see James et al., 2003). The ensuing liver damage may be due to the binding of NAPQI to proteins, the generation of reactive oxygen species, or a combination of both (for review, see James et al., 2003; Jaeschke et al., 2003).
CYP2E1 and CYP3A are the most active human and rat P450 forms that convert APAP to NAPQI (Patten et al., 1993; Thummel et al., 1993). In vitro studies demonstrated that CYP3A has a higher affinity for APAP than does CYP2E1. However, both CYP2E1 and CYP3A would be expected to contribute to APAP hepatotoxicity in vivo since their Vmax/Km ratios for formation of NAPQI are similar (Patten et al., 1993). Despite its relatively low affinity for APAP compared to CYP2E1 and 3A (Patten et al., 1993), at high doses of APAP, CYP1A2 may also contribute to the formation of NAPQI in vivo. Treatment with inducers of CYP1A2 increases APAP hepatotoxicity in rodents (Mitchell et al., 1973; Kalhorn et al., 1990; Jaw and Jeffery, 1993).
CYP2E1 is believed to be the major P450 responsible for APAP bioactivation and toxicity (for review, see James et al., 2003). APAP hepatotoxicity is lower in Cyp2e1(–/–) mice compared to wild-type mice (Lee et al., 1996). However, treatment of animals with inducers of CYP3A, such as dexamethasone (Dex) and phenobarbital, increased APAP hepatotoxicity (Kalhorn et al., 1990; Jaw and Jeffery, 1993). These findings suggest that CYP3A, when elevated in the liver, can play a role in APAP toxicity, in addition to CYP2E1. Alcohol, which has been shown to increase APAP hepatotoxicity, induces both CYP3A and CYP2E1 (for review, see Sinclair et al., 1998). A major role of CYP3A in alcohol-mediated APAP hepatotoxicity was suggested by the finding that triacetyloleandomycin, a specific inhibitor of CYP3A, protected alcohol-pretreated rats from APAP hepatotoxicity (Sinclair et al., 2000a). The above studies were done in animals preinduced for CYP3A before administration of APAP. At this time, it is unclear whether CYP3A contributes to APAP hepatotoxicity in otherwise untreated animals.
The pregnane X receptor (PXR) is a member of the nuclear receptor superfamily that regulates the expression of many drug-metabolizing proteins, including CYP3A (for review, see Maglich et al., 2002). Deletion of PXR leads to a 4-fold elevation of hepatic levels of CYP3A11 mRNA as well as elevation of CYP3A protein (Guo et al., 2003), a finding expected to increase APAP hepatotoxicity in PXR(–/–) mice. Therefore, we compared APAP hepatotoxicity in wild-type and PXR(–/–) mice. Unexpectedly, APAP hepatotoxicity was greater in wild-type mice compared to PXR(–/–) mice. We measured several parameters that may have contributed to this difference in toxicity. Wild-type mice have higher levels of hepatic CYP1A2 and greater absorption of APAP compared to PXR(–/–) mice. These findings suggest that PXR may regulate more than one constitutive function involved in APAP hepatotoxicity.
Materials and Methods
Reagents. Lieber-DeCarli control liquid diet (F1259SP) was obtained from Bio-Serv Inc. (Frenchtown, NJ). Acetaminophen (APAP), p-nitrophenol, HEPES, insulin, Dex, dexamethasone 21-phosphate, dimethyl sulfoxide, the glucuronide metabolite of APAP (APAP-Gluc) used as a standard for HPLC analysis, NADP+, NADPH, Brij 58 (polyoxyethylene 20-cetyl ether), UDP-glucuronic acid (sodium salt), magnesium chloride, dl-isocitric acid, isocitric dehydrogenase, potassium phosphate buffer solutions, 2-acetamidophenol, 1-heptanesulfonic acid, perchloric acid, hydrochloric acid, caffeine, and GSH were purchased from Sigma-Aldrich (St. Louis, MO). Tris, methanol, sodium chloride, trichloroacetic acid, and acetic acid were obtained from Fisher Scientific Co. (Pittsburgh, PA). 7-Methoxyresorufin was obtained from Molecular Probes (Eugene, OR). l-[U-14C]Leucine was purchased from PerkinElmer Life and Analytical Sciences (Boston, MA). The glutathione metabolite of APAP (APAP-SG), used as a standard for HPLC analysis, was a generous gift from Dr. Mark W. Gemborys (ChemDesign Corporation, Fitchburg, MA). The sulfate metabolite of APAP (APAP-SO4), also used as a standard for HPLC analysis, was purchased from Ultrafine (Manchester, UK). The polyclonal rabbit anti-human CYP2E1 antibody, which detected mouse CYP2E1, was purchased from Oxford Biomedical Research (Oxford, MI). The polyclonal goat antibody prepared against rat CYP1A2 detected mouse CYP1A1 and CYP1A2 (Sinclair et al., 2000b). The polyclonal rabbit antibody prepared against human CYP3A4 detected only mouse CYP3A, not CYP1A2 and CYP2E1 (Sinclair et al., 2000b). Electrophoresis and immunoblotting reagents were obtained from Bio-Rad (Hercules, CA). Assay kits for glutathione reductase (GRed) and glutathione peroxidase (GPx) were purchased from Cayman Chemical (Ann Arbor, MI). [3H]Vinblastine sulfate was obtained from Moravek Biochemicals (Brea, CA). TRIzol reagent and SuperScript First-Strand Synthesis Systems were purchased from Invitrogen (Carlsbad, CA). TaqPCR Master Mix, 2X was obtained from USB (Cleveland, OH). Ethidium bromide was purchased from AMRESCO Inc. (Solon, OH). Agarose was obtained from Ambion (Austin, TX). Williams E powdered medium and penicillin G/streptomycin were obtained from Invitrogen. The tk-Renilla plasmid, Lipofectamine 2000, and Dual Luciferase kit were purchased from Promega (Madison, WI). Rat-tail type I collagen was obtained from BD Biosciences Discovery Labware (Bedford, MA). Supersomes expressing human CYP1A2 and 3A4 were obtained from BD Gentest (Woburn, MA). The rabbit polyclonal antibody to 4-hydroxynonenal (4-HNE) was purchased from Alexis Biochemicals (San Diego, CA). The Anti-Rabbit Ig Peroxidase ImmPress kit, hematoxylin QS, and diaminobenzidine were obtained from Vector Laboratories (Burlingame, CA). Matrigel was prepared as previously described (Sinclair et al., 1991).
Animals.PXR(–/–) mice (Xie et al., 2000) were generated in a greater than 98% C57BL/6 background in Dr. Evans' laboratory and maintained at the Veterans Affairs animal facility. Wild-type C57BL/6 mice were purchased from the National Cancer Institute (Frederick, MD). Both mouse lines were maintained in a controlled environment with a 12-h light/dark cycle. Mice were genotyped for the absence of PXR using DNA extracted from either tails or livers with a DNeasy tissue kit (QIAGEN, Valencia, CA), three primers synthesized by Eurogentec North America, Inc. (San Diego, CA), and PCR amplification kits from Takara Mirus Bio (Madison, WI). The primer sequences were as follows: WX48 (sense primer) 5′-agaaacacatagaaacccatccatg-3′; WX47 (antisense primer) 5′-agtccaccaagcctgagcctcctac-3′; and WXNEO (antisense primer) 5′-cttgacgagttcttctgaggggatc-3′ (Dr. J. Sonoda, personal communication). The PCR conditions used were 1 min at 94°C, 2 min at 60°C, and 3 min at 72°C for 35 cycles. The WXNEO primer is specific for the neomycin resistance gene (NEO) cassette that was inserted during deletion of the PXR gene in PXR(–/–) animals. WXNEO and WX48 generated a product of 546 bp in PXR(–/–) mice. WX47 and WX48 generated a product of 497 bp in wild-type mice.
All animal protocols were approved by the Institutional Animal Care and Use Committees at the Department of Veterans Affairs Medical Center and Dartmouth College. Male mice, 3 months of age and weighing 18 to 30 g, were fed liquid Lieber-DeCarli diet for 9 days. The liquid diet was replaced with water 16 h before the administration of APAP. APAP was prepared as previously described (Kostrubsky et al., 1997b). To mimic the route of human ingestion, APAP was administered intragastrically (gavage) at the doses specified in the figure legends. When indicated, caffeine (100 mg/kg) was administered in the same bolus as APAP. Seven hours after APAP administration, mice were anesthetized with carbon dioxide and blood was collected by cardiac puncture. The animals were euthanized by cervical dislocation. For P450 analyses, hepatic microsomes or 20% (w/v) liver homogenates were prepared as previously described (Sinclair et al., 2000b).
Immunochemical Analyses of Cytochromes P450. Hepatic microsomes and 20% (w/v) homogenates were analyzed for CYP3A immunochemically as described previously (Wolf et al., 2004). To reproducibly separate the CYP3A forms, the following modifications were used. The electrophoresis buffer contained 37.5 mM Tris base, 290 mM glycine, and 1.5% (w/v) SDS. The gels were electrophoresed at 160 V for 105 min, with the first 45 min at room temperature and the last 60 min at 4°C. Increasing amounts of human P450 Supersomes were used to determine P450 linearity (see Fig. 4 for CYP1A2 example). Human CYP3A Supersomes were applied to the gel after 45 min of electrophoresis. CYP2E1 was detected immunochemically as described previously (Wolf et al., 2004). After electrophoresis in a mixed anionic-detergent system, CYP1A2 was detected immunochemically (Sinclair et al., 2000b). Human CYP1A2 Supersomes were applied to the gel after 30 min of electrophoresis. An HP Precision Scanner (Hewlett Packard, Palo Alto, CA) was used to scan the immunoblots into Adobe Photoshop (Adobe Systems, Mountain View, CA), and OneDScan software (Scanalytics, Fairfax, VA) was used for quantification. Each band was scanned in triplicate and the intensity expressed as arbitrary units. The values represent the average of three to four mice from each mouse line.
Transport Studies. LLC-PK1 pig kidney epithelial cells were obtained from The American Type Culture Collection (Manassas, VA). Dr. Alfred Schinkel (The Netherlands Cancer Institute, Amsterdam, The Netherlands) generously provided Dr. Erin Schuetz with LLC-PK1 cells expressing human MDR1 (L-MDR1). To investigate whether APAP is a substrate of MDR1, cell lines were cultured and transport assays performed as described (Yasuda et al., 2002). Transportation of [3H]vinblastine sulfate was included as the positive control. APAP was measured in the media by HPLC as described previously (Lindenthal et al., 1993).
Reverse Transcription-PCR Analysis of CYP1A2, CYP3A11, and mdr1a. Total RNA was extracted from either liver or jejunum of wild-type and PXR(–/–) mice using TRIzol reagent. cDNA was prepared from RNA with the SuperScript First-Strand Synthesis System. CYP1A2, CYP3A11, and one set of 18S rRNA primer sequences, previously published (Selvey et al., 2001; Zhang et al., 2003), produced PCR products of 359 bp, 425 bp, and 489 bp, respectively. An additional set of oligonucleotide primers specific for 18S rRNA was designed using the web-based program Primer3. These primers produced a 151-bp product and were as follows: sense primer 5′-gcatatgcttgtctcaaagattaag-3′ and antisense primer 5′-tattagctctagaattaccacagttatcca-3′. Each PCR product was sequenced and integrity verified using BLAST (National Center for Biotechnology Information, National Institutes of Health). Amplification was performed on a Bio-Rad i-Cycler using USB's TaqPCR Master Mix, 2X. CYP1A2/18S rRNA PCR reaction conditions were as follows: an initial denaturation of 94°C for 1 min followed by 25 to 40 cycles of 94°C for 30 s, 50°C for 30 s, and 72°C for 30 s. CYP3A11/18S rRNA PCR reaction conditions were as follows: an initial denaturation of 94°C for 1 min followed by 25 to 35 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 30 s. The PCR products were separated on 2% agarose gels stained with ethidium bromide. The gels were scanned into Adobe Photoshop using an HP Precision Scanner. Each band was scanned in triplicate and quantitated using OneDScan software. The results were expressed as the ratio of either CYP1A2 or CYP3A11 to 18S rRNA. Semiquantitative reverse transcription-PCR conditions for CYP1A2, CYP3A11, and 18S rRNA were established by: 1) determination of the PCR cycle number that generated products in a linear range, 2) performing reactions for CYP1A2 or CYP3A11 in the same tube as the internal control (18S rRNA) to allow for normalization of sample variation, and 3) adjustment of primer concentrations to achieve linearity in PCR product formation with two primer sets in the same reaction. mdr1a mRNA was analyzed by PCR as described (Schuetz et al., 2001).
Transfection Studies in Primary Hepatocyte Cultures. Rat hepatocyte cultures were prepared as previously described (Sinclair et al., 1991), but were maintained on plates coated with rat-tail type I collagen (5 μg/cm2). Forty-eight hours after isolation, rat hepatocytes were transfected with 0.2 μg of pGL3CYP3A23-Luc plasmid (a luciferase reporter construct containing –1360 to +52 base pairs of the 5′ upstream region of the rat CYP3A23 gene) DNA/plate and 0.02 μg of tk-Renilla plasmid (an internal control) DNA/plate using Lipofectamine 2000 in serum-free medium, according to the manufacturer's protocol. After 6 h, the medium was changed and Matrigel added as an overlay (0.2 mg/ml). Twenty-four hours later, the medium was again changed and the rat hepatocytes either remained untreated or were exposed overnight to 10 μM Dex in dimethyl sulfoxide or increasing concentrations of APAP in medium. Cell lysates were harvested using the Dual Luciferase kit and the lysates were analyzed on an Lmax microplate luminometer (Molecular Devices, Sunnyvale, CA). Protein synthesis, serving as an indicator of toxicity (Lindenthal et al., 1993; Kostrubsky et al., 1997a), was measured by the incorporation of [14C]leucine (specific radioactivity, 0.1 mCi/mmol, 0.2 μCi/plate) into protein for 1 h at the end of the overnight incubation with APAP, as previously described (Kostrubsky et al., 1997a).
Immunohistochemical Detection of 4-HNE. 4-HNE was measured immunohistochemically on 4-μm sections of formalin-fixed paraffin-embedded liver sections using the 4-HNE rabbit polyclonal antibody from Alexis Biochemicals. Tissue sections were first deparaffinized and rehydrated, and then incubated in 10 mM sodium citrate buffer (pH 6.0) to unmask antigens. Endogenous peroxidase activity was quenched by incubating sections in 3% hydrogen peroxide. Nonspecific binding was blocked with 2.5% normal horse serum before incubation with the primary antibody. Bound antibody was visualized by Vector Laboratories' ImmPress kit with diaminobenzidine as the substrate, according to the manufacturer's protocol. Sections were counterstained with hematoxylin QS (Vector Laboratories). 4-HNE protein adducts were indicated by areas stained brown.
Additional Assays. The clinical laboratory personnel at the Veterans Affairs Medical Center assayed plasma alanine aminotransferase levels using an AU400e chemistry-immuno analyzer (Olympus America Inc., Melville, NY). Protein concentrations were measured by the procedure of Lowry et al. (1951), using bovine serum albumin as a standard. Hepatic GSH levels were measured in 10% (w/v) liver homogenates as described (Sinclair et al., 2000a). Plasma GPx and GRed activities were measured using assay kits from Cayman Chemical. Changes in levels of GPx and GRed activity in the plasma reflect changes in these enzymes in the liver (Hill et al., 2001; Senthilkumar et al., 2004). p-Nitrophenol O-hydroxylation (PNPH) activity was measured in liver microsomes as described (Wolf et al., 2004). Microsomal methoxyresorufin O-deethylase (MROD) activity was measured as described (Sinclair et al., 2000b). Triazolam hydroxylation activity was measured in liver microsomes as described (Perloff et al., 2000). Microsomal NAPQI formation was measured in the presence of reduced GSH and APAP as the rate of formation of the APAP glutathione adduct, APAP-SG (Sinclair et al., 1991). Glucuronidation of APAP was measured in liver microsomes as described (Court and Greenblatt, 1997). Plasma APAP and its metabolites were extracted by the method of Brunner and Bai (1999) with the following modifications. Ten microliters of plasma were mixed with 20 μl of extraction solution (2% perchloric acid with 0.1 mg/ml theophylline as an internal standard). APAP, APAP-Gluc, APAP-SO4, and APAP-SG in the resulting extract were separated and quantified by HPLC using the methods of Gemborys and Mudge (1981) and Lindenthal et al. (1993) with the following modifications. The HPLC system consisted of a Milton Roy CM4000 Multiple Solvent Delivery System and SM4000 Programmable Wavelength Detector (Milton Roy Company, Rochester, NY), and a Thermo Separations Product Spectra System AS3000 autosampler with a 50-μl loop (Thermo Electron Corporation, Waltham, MA). Twenty-five microliters of extracted sample were injected onto a Dynamax 5-μm C18 Microsorb HPLC column (4.6 mm i.d. x 250 mm; Varian, Inc., Palo Alto, CA). The data output was processed with a Shimadzu Scientific Instruments (Columbia, MD) Chromatopac C-R3A recorder. HPLC buffer A was 3.1 mM heptanesulfonic acid (pH 2.16) and 1% (v/v) acetic acid. Buffer B was 3.1 mM heptanesulfonic acid (pH 2.16), 1% acetic acid (v/v), and 12% acetonitrile (v/v). Initial HPLC conditions were 90% buffer A and 10% buffer B. The metabolites were eluted over a gradient that decreased linearly over 10 min from 90% buffer A to 0% buffer A, held at 0% buffer A for 15 min, and then returned linearly to initial conditions over 10 min. Absorbance was monitored at a wavelength of 254 nm.
Statistical Analyses. Values represent the means and S.E.M. or S.D. as indicated in the figure legends. Analyses were performed either by analysis of variance followed by the Student's Newman-Keuls multiple comparisons test, or by an unpaired t test. Significance was indicated by p values <0.05. Area under the curve calculations and analysis were performed using GraphPad Prism 4 (GraphPad Software Inc., San Diego, CA).
Results
APAP Hepatotoxicity in Wild-Type and PXR(–/–) Mice. Plasma levels of alanine aminotransferase served as a measure of liver damage. Figure 1 shows a dose response of APAP on liver damage in otherwise untreated wild-type and PXR(–/–) mice. Although an increase in APAP hepatotoxicity was noted in wild-type mice at a dose of 200 mg APAP/kg, a dose of 300 mg APAP/kg was significantly more hepatotoxic to wild-type mice than to PXR(–/–) mice (p < 0.01). These results suggest that some PXR-modulated function(s) plays a role in APAP hepatotoxicity.
Comparison of Hepatic Levels of CYP3A, 2E1, and 1A2 in Wild-Type and PXR(–/–) Mice. To identify the P450 forms that may contribute to the increased APAP hepatotoxicity in wild-type mice, we compared hepatic levels of CYP3A, 2E1, and 1A2 in wild-type and PXR(–/–) mice. Six isoforms of mouse CYP3A have been identified. Three of these, CYP3A11, 3A13, and 3A25, are expressed in male mice (for review, see Sakuma et al., 2002). In our study, hepatic levels of total CYP3A protein were 2.5-fold greater in PXR(–/–) mice compared to wild-type mice (p < 0.01; Fig. 2A), similar to the findings of Guo et al. (2003). Hepatic levels of CYP3A activity, measured as the α-hydroxylation and 4-hydroxylation of triazolam (Perloff et al., 2000), were 3.6-fold greater in PXR(–/–) mice compared to wild-type mice (p < 0.01; Fig. 2, B and C). Contrary to our expectations, these findings indicate that elevation of CYP3A in PXR(–/–) mice does not result in increased APAP hepatotoxicity.
CYP2E1 immunoreactive protein levels were similar in wild-type and PXR(–/–) mice (Fig. 3A), consistent with the finding of Guo et al. (2004) that hepatic levels of CYP2E1 mRNA are similar in the two mouse lines. PNPH activity, a specific measure of CYP2E1 in mice (Carlson, 2003; Wolf et al., 2004), was also similar in wild-type and PXR(–/–) mice (Fig. 3B). These findings indicate that hepatic CYP2E1 is not responsible for the differences in APAP hepatotoxicity between the two mouse lines.
Hepatic levels of immunoreactive CYP1A2 protein were 3-fold greater in wild-type mice compared to PXR(–/–) mice (p < 0.01; Fig. 4A). Microsomal MROD activity was approximately 2-fold higher in wild-type animals than in PXR(–/–) animals (p < 0.001; Fig. 4B). CYP1A1 was not detected in hepatic microsomes from untreated mice (Fig. 4A). Since in the absence of CYP1A1, MROD activity is mainly catalyzed by CYP1A2 (Sinclair et al., 2000b), these data suggest that PXR is a positive regulator of the constitutive expression of CYP1A2. Therefore, the higher levels of CYP1A2 in wild-type mice may contribute to the differences in APAP hepatotoxicity.
Formation of NAPQI, the Reactive Metabolite of APAP, by Hepatic Microsomes from Wild-Type and PXR(–/–) Mice. We investigated whether there was a difference between wild-type and PXR(–/–) mice in the ability of hepatic microsomes to generate NAPQI from APAP. We measured formation of NAPQI by measuring the glutathione conjugate of NAPQI, APAP-SG, by HPLC (Sinclair et al., 1991). Two different APAP concentrations were used to analyze APAP activation, based on the kinetic analyses of Patten et al. (1993) and Thummel et al. (1993): 0.5 mM APAP to measure mainly CYP3A activity and 5 mM APAP to measure the composite activities of CYP3A, 2E1, and 1A2. CYP2E1 does not contribute to APAP activation at a concentration of 0.5 mM APAP since the activities are similar in microsomes from wild-type and Cyp2e1(–/–) mice (results not shown). We found no significant difference between wild-type and PXR(–/–) mice in APAP-SG formation at 0.5 mM APAP (Fig. 5). This lack of a difference in APAP-SG formation suggests that the forms of CYP3A that were elevated in PXR(–/–) mice (Fig. 5) do not significantly contribute to NAPQI formation at this concentration of APAP, as measured in vitro.
At the higher concentration of APAP (5 mM), there was also no significant difference in NAPQI formation between wild-type and PXR(–/–) mice (Fig. 5). These findings support the contribution of both CYP1A2 and CYP3A to NAPQI formation in these mice at the high concentration of APAP. The effect of the decrease in CYP1A2 in PXR(–/–) mice may be compensated for by the increase in CYP3A.
Effect of APAP on Hepatic P450 Levels in Wild-Type and PXR(–/–) Mice. APAP has been reported to increase mRNAs for CYP1A2 and 3A11 in wild-type mice, but not in CAR(–/–) (Zhang et al., 2002) or RXRα(–/–) (Wu et al., 2004) mice. Since both these P450s activate APAP, increases in their mRNA were interpreted to play a role in the greater APAP hepatotoxicity observed in wild-type mice. Here, we investigated whether APAP increased hepatic CYP1A2, 2E1, and 3A proteins, as well as CYP1A2 and 3A11 mRNA, in wild-type and PXR(–/–) mice. In PXR(–/–) mice, no changes in hepatic levels of these proteins were detected at either 1 h or 7 h after APAP administration (results not shown). Similar results were seen for mRNA levels of CYP1A2 and 3A11 (results not shown). In wild-type mice, there appears to be an increase in CYP3A protein at 1 and 7 h after APAP and in CYP1A2 protein 7 h after APAP, but these increases were not statistically significant (Fig. 6). No increase in CYP2E1 protein was observed at either time (Fig. 6). CYP3A11 mRNA levels were not increased at 1 h and 7 h after APAP administration (Fig. 6). However, CYP1A2 mRNA was increased 1 h after APAP administration (p < 0.001; Fig. 6) but returned to basal levels by 7 h (Fig. 6).
Effect of APAP on PXR-Mediated Transcription in Primary Hepatocyte Cultures. Treatment with APAP has been reported to increase CYP3A11 mRNA in wild-type, but not in CAR(–/–) mice (Zhang et al., 2002). However, APAP is not a ligand of CAR (Zhang et al., 2002). CYP3A can be regulated by either PXR or CAR (for review, see Maglich et al., 2002). The pGL3CYP3A23-Luc plasmid, a luciferase reporter plasmid containing the upstream regulatory region of CYP3A23 (–1360 to +52 bp), is widely used to identify functional ligands of PXR (Xie et al., 2000). We investigated whether APAP was able to activate the luciferase reporter construct containing a PXR response element transfected into primary rat hepatocytes. Dex, a ligand of PXR (Xie et al., 2000), increased expression of luciferase/renilla activity 9-fold (Fig. 7A), indicating activation of PXR. However, APAP, at concentrations that were nontoxic (2.5 mM and 5 mM; Fig. 7B), caused no increase in luciferase activity (Fig. 7A), indicating that APAP is not a ligand of PXR.
Effect of Caffeine on APAP Hepatotoxicity in Wild-Type and PXR(–/–) Mice. Caffeine has been shown to inhibit CYP1A2 activity, while enhancing CYP3A activity (Lee et al., 1991). We investigated the contribution of CYP1A2 to increased APAP hepatotoxicity in wild-type mice compared to PXR(–/–) mice by administering caffeine. Treatment with 100 mg caffeine/kg alone, administered by gavage, did not cause toxicity (results not shown). As seen in Fig. 1, 300 mg APAP/kg was more hepatotoxic to wild-type mice than to PXR(–/–) mice (p < 0.01; Fig. 8). In wild-type mice, when APAP and caffeine were given in combination, APAP hepatotoxicity was dramatically decreased (p < 0.01; Fig. 8). In PXR(–/–) mice, in contrast, caffeine did not alter APAP hepatotoxicity (Fig. 8). These results suggest that CYP1A2 is an important contributor to APAP hepatotoxicity in wild-type mice.
Intestinal Absorption of APAP by Wild-Type and PXR(–/–) Mice. In our studies, APAP is administered by gavage (i.g.). Therefore, we investigated whether a difference in absorption could contribute to the difference in APAP hepatotoxicity between wild-type and PXR(–/–) mice. In mice administered 300 mg APAP/kg, plasma levels of APAP were measured after 30 and 60 min. The area under the curve (AUC0–60) for wild-type mice was 1.5-fold higher than that in PXR(–/–) mice (p < 0.01; Fig. 9). These findings suggest that APAP absorption was greater in wild-type mice compared to PXR(–/–) mice and that wild-type mice had a greater exposure to APAP.
Role of mdr1 in APAP Transport. Since the uptake of APAP was greater in wild-type mice and mdr1a is involved in the transport of numerous drugs, we investigated whether intestinal mdr1a expression differed between wild-type and PXR(–/–) mice. Consistent with findings in the liver (Guo et al., 2003), mdr1a mRNA (Fig. 10) and protein (results not shown) expressions in the intestine were similar in wild-type and PXR(–/–) mice. In addition, we found no difference in APAP transport (apical to basal and basal to apical) between LLC-PK1 cells (lack MDR1) and L-MDR1 cells (stably overexpress human MDR1), indicating that APAP is not transported by MDR1 (results not shown). Our overall findings indicate that mdr1 is not responsible for the differences in APAP absorption between wild-type and PXR(–/–) mice.
Comparison of APAP Metabolism in Wild-Type and PXR(–/–) Mice. Since phase II metabolism is a major route of APAP elimination, we investigated whether differences in rates of APAP conjugation in vivo could explain differences in APAP hepatotoxicity. Figure 11A shows the plasma levels of the major APAP metabolites 30 and 60 min after administration of 300 mg APAP/kg i.g. In both wild-type and PXR(–/–) mice, the glucuronide conjugate of APAP was the major circulating metabolite, as observed in other mouse lines (Kim and Lee, 1998). No significant differences were seen in the amount of the glucuronide (APAP-Gluc) or the sulfate (APAP-SO4) conjugates in plasma of wild-type and PXR(–/–) mice. Consistent with these findings, hepatic microsomal activities for formation of APAP-Gluc were similar in wild-type and PXR(–/–) mice (Fig. 11B). The glutathione conjugate of APAP (APAP-SG) was detected at very low levels in the plasma at these times (data not shown).
Comparison of Glutathione, Glutathione Reductase, and Glutathione Peroxidase in Wild-Type and PXR(–/–) Mice. Hepatic GSH is a critical component in protection from APAP hepatotoxicity (for review, see Jaeschke et al., 2003; James et al., 2003). Hepatic GSH levels were compared in untreated mice at 1 and 7 h after treatment with 300 mg APAP/kg to determine whether a difference existed that could account for the variation in susceptibility to APAP between the two mouse lines. The hepatic levels of GSH were similar in all untreated mice (Fig. 12). Compared to untreated mice, wild-type and PXR(–/–) mice treated with APAP for 1 h had dramatically decreased hepatic GSH levels (a, p < 0.001; Fig. 12). GSH levels were recovering 7 h after APAP administration in both wild-type (b, p < 0.01; Fig. 12) and PXR(–/–) mice (b, p < 0.001; Fig. 12). However, there were no significant differences between the two mouse lines at any time point. We also measured constitutive activities of two enzymes involved in GSH-mediated protection from oxidative damage, GRed and GPx. Although GRed activity appeared to be greater in PXR(–/–) mice compared to wild-type mice (results not shown), the difference was not significant. GPx activity was similar in the two mouse lines (results not shown). 4-HNE, a product of lipid peroxidation, was measured immunohistochemically and found to be similar in wild-type and PXR(–/–) mice (results not shown). These results suggest that increased APAP hepatotoxicity in wild-type mice is not due to an inability to protect against lipid peroxidation.
Discussion
Using untreated mice in a greater than 98% C57BL/6 background, we found that APAP hepatotoxicity was lower in PXR(–/–) mice compared to wild-type mice (Fig. 1), indicating that the nuclear receptor PXR plays a role in APAP hepatotoxicity. We found two parameters that may account for the increased APAP hepatotoxicity in wild-type mice: 1) greater absorption of APAP and 2) higher hepatic levels of CYP1A2. Although it is widely believed that CYP2E1 is the major P450 mediator of APAP hepatotoxicity (for review, see James et al., 2003), our findings that hepatic CYP2E1 protein (Fig. 3A) and enzyme activity (Fig. 3B) were identical in PXR(–/–) and wild-type mice indicate that CYP2E1 does not account for the differences in APAP hepatotoxicity between the two mouse lines.
CYP1A2 has a low intrinsic activity toward APAP for NAPQI formation (Patten et al., 1993), yet treatment of animals with inducers of CYP1A2 increases APAP hepatotoxicity (Mitchell et al., 1973; Kalhorn et al., 1990; Jaw and Jeffery, 1993). Constitutive levels of CYP1A2 may also contribute to APAP hepatotoxicity. In untreated male mice, which have low levels of CYP3A (for review, see Sakuma et al., 2002), caffeine, a methylxanthine that inhibits CYP1A2 activity and enhances CYP3A activity (Lee et al., 1991), decreases APAP hepatotoxicity (Gale et al., 1987). Due to divergent results, studies with Cyp1a2(–/–) mice have not clarified the role of constitutive CYP1A2 in APAP hepatotoxicity. In one study, there was no difference in APAP hepatotoxicity between wild-type and Cyp1a2(–/–) mice (Tonge et al., 1998), whereas in another, Cyp1a2(–/–) mice were more resistant to APAP hepatotoxicity compared to wild-type mice (Genter et al., 1998). Cyp1a2(–/–)Cyp2e1(–/–) double knockout mice (Zaher et al., 1998) are more resistant to APAP hepatotoxicity than are Cyp2e1(–/–) mice (Lee et al., 1996), suggesting a contribution of CYP1A2 to APAP hepatotoxicity. In our study, the higher hepatic levels of CYP1A2 in wild-type mice (Fig. 4A), associated with greater APAP hepatotoxicity (Fig. 1), support a role for CYP1A2 in APAP hepatotoxicity. Furthermore, a role of CYP1A2 was confirmed by our finding that caffeine protected wild-type mice against APAP hepatotoxicity (Fig. 8). In contrast, our finding that caffeine had no effect on APAP hepatotoxicity in PXR(–/–) mice suggests either that the CYP3A forms elevated in these mice were not activated by caffeine or that enhanced CYP3A activity may have been overcome by CYP1A2 inhibition.
Human and rat CYP3A have a lower Km for APAP than do CYP2E1 and 1A2 (Patten et al., 1993; Thummel et al., 1993). Although untreated PXR(–/–) animals have higher basal levels of CYP3A proteins, and higher CYP3A activity compared to untreated wild-type animals (Fig. 2, A, B, and C), there was no equivalent increase in microsomal formation of NAPQI formation at 0.5 mM APAP (Fig. 5). These findings suggest that the isoforms of CYP3A increased in PXR(–/–) mice do not have the same high affinity and Vmax/Km ratio for APAP as human CYP3A4 and rat CYP3A23. However, our findings that hepatic activities for formation of NAPQI at 5 mM APAP were similar in wild-type and PXR(–/–) mice (Fig. 5) suggest that both CYP1A2 and CYP3A contribute to NAPQI formation at this high APAP concentration. In the microsomal activation of APAP, the lower levels of CYP1A2 in PXR(–/–) mice may have been compensated for by higher levels of CYP3A.
RXRα is a ligand-activated transcription factor in the nuclear receptor superfamily that heterodimerizes with orphan nuclear receptors, including PXR and CAR (Wu et al., 2004). Similar to our findings with PXR(–/–) mice, RXRα(–/–) mice have both lower basal CYP1A2 mRNA levels and lower APAP hepatotoxicity compared to wild-type mice (Wu et al., 2004). The findings of both studies suggest that the PXR/RXRα heterodimer may be a positive regulator of constitutive CYP1A2 expression since deletion of either PXR or its obligatory dimer partner, RXRα, results in decreased hepatic CYP1A2. Zhang et al. (2002) reported that CYP1A2 mRNA levels were increased in CAR(–/–) mice. In contrast to PXR, CAR appears to be a negative regulator of CYP1A2 expression, possibly by binding to RXRα and limiting the amount of RXRα available for association with PXR.
Greater APAP hepatotoxicity in wild-type mice compared to CAR(–/–) and RXRα(–/–) mice has been attributed to increases in CYP1A2 and 3A11, since mRNA for these P450s were increased 2 h after APAP administration in wild-type mice, but not in the knockout mice (Zhang et al., 2002; Wu et al., 2004). In our study, we observed no statistically significant changes in the protein levels of CYP1A2, 2E1, and 3A at either 1 or 7 h after APAP administration in either wild-type (Fig. 6) or PXR(–/–) mice (results not shown). However, since CYP1A2 mRNA was transiently increased by APAP (Fig. 6), a transient increase in the protein may have occurred between 1 and 7 h. In contrast to our findings, CYP3A in rats (Kostrubsky et al., 1997b) and humans (Zhang et al., 2004a), and CYP1A2 and 2E1 in mice (Snawder et al., 1994), are actually decreased after APAP administration, findings attributed to suicidal inactivation of these P450s.
Our findings that APAP hepatotoxicity was lower in PXR(–/–) mice compared to wild-type mice (Fig. 1) are in disagreement with two other recent studies comparing APAP hepatotoxicity in wild-type and PXR(–/–) mice (Guo et al., 2004; Zhang et al., 2004b). A major difference in the experimental design between our study and the other studies is the route of APAP administration. In our studies, the mice were administered APAP by gavage to mimic human exposure, whereas in the other two studies APAP was administered intraperitoneally (Guo et al., 2004; Zhang et al., 2004b). Another difference in the studies is the dose of APAP. We used a lower dose of APAP than the other two studies, which may have allowed us to detect differences in hepatotoxicity. In our study, a dose of 400 mg APAP/kg was equally hepatotoxic in both mouse lines (results not shown). The genetic background of the mice used in these studies may also be a factor in the response to APAP. In our study, the PXR(–/–) mice were in a greater than 98% C57BL/6 genetic background, whereas in the other two studies, mice were in mixed genetic backgrounds (Guo et al., 2004; Zhang et al., 2004b). It is known that heterogeneity in the genetic background of mice can affect their phenotype (for review, see McKinnon and Nebert, 1998).
Hepatic levels of GSH are critical in the protection against APAP toxicity because glutathione conjugation is the major pathway involved in elimination of NAPQI and prevention of oxidative damage (for review, see Jaeschke et al., 2003; James et al., 2003). Levels of GSH, GPx, and GRed were similar in untreated wild-type and PXR(–/–) mice, suggesting that GSH homeostasis cannot account for the different susceptibilities.
Glucuronidation and sulfation are the major routes of APAP metabolism (for review, see James et al., 2003). In our studies, the major APAP metabolite was the glucuronide conjugate (Fig. 11), similar to results from other mouse studies (Kim and Lee, 1998). Despite a higher AUC0–60 for APAP in wild-type mice compared to PXR(–/–) mice (Fig. 9), plasma levels of APAP-Gluc and APAP-SO4 were similar (Fig. 11A), suggesting that these conjugation pathways had become saturated. Consistent with these findings, hepatic microsomal activities for formation of APAP-Gluc were similar in wild-type and PXR(–/–) mice (Fig. 11B). Plasma levels of APAP-SG were too low to enable a comparison of differences in formation of NAPQI in vivo.
PXR may have a role in regulating APAP-transporting proteins in the intestine. We found that the intestinal levels of mdr1a mRNA (Fig. 10) and protein (results not shown) were identical in the two mouse lines, similar to findings in liver (Guo et al., 2003). In addition, APAP was not a substrate for MDR1 (results not shown). Therefore, mdr1a cannot account for the greater exposure to APAP in wild-type mice compared to PXR(–/–) mice. Another transport protein whose constitutive expression is regulated by PXR may be responsible for the differences in APAP absorption between the two mouse lines.
In summary, PXR(–/–) mice were less sensitive to APAP hepatotoxicity than wild-type animals. Our finding that hepatic levels of CYP1A2 in wild-type mice were greater compared to PXR(–/–) mice suggests that PXR acts as a positive regulator of CYP1A2 and that the higher levels of CYP1A2 in wild-type mice have a role in the greater APAP hepatotoxicity in these mice. Caffeine, an inhibitor of CYP1A2 activity, protected against APAP hepatotoxicity in wild-type mice, confirming the role of CYP1A2 in APAP hepatotoxicity. The contribution of a particular P450 to APAP hepatotoxicity may depend on the relative amount of one P450 compared to others that form NAPQI, as well as on the hepatic concentration of APAP. Our findings that wild-type mice had greater intestinal APAP absorption than did PXR(–/–) mice suggest that a transport protein regulated constitutively by PXR may also contribute to the development of APAP toxicity. Our overall findings suggest that PXR constitutively regulates expression of more than one gene involved in APAP hepatotoxicity.
Acknowledgments
We thank Dr. Judith Jacobs for assisting in preparing the manuscript and Heidi Trask for contributing to the transfection studies.
Footnotes
-
This work was supported in part by the Department of Veterans Affairs, National Institutes of Health Research Grants AA12898 (J.F.S.), GM60346 (E.G.S.), and P30 CA21765 Cancer Center Support Grant (E.G.S.), and the American Lebanese Syrian Associated Charities (E.G.S.).
-
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
-
doi:10.1124/dmd.105.005256.
-
ABBREVIATIONS: APAP, acetaminophen; 4-HNE, 4-hydroxynonenal; APAP-Gluc, glucuronide-conjugated metabolite of APAP; APAP-SG, glutathione adduct of APAP; APAP-SO4, sulfate-conjugated metabolite of APAP; AUC, area under the curve; bp, base pair(s); CAR, constitutive androstane receptor; P450, cytochrome P450; Dex, dexamethasone; GPx, glutathione peroxidase; GRed, glutathione reductase; GSH, glutathione; HPLC, high-performance liquid chromatography; i.g., intragastric; MDR/mdr, human and mouse multidrug resistance protein, respectively; MROD, methoxyresorufin O-deethylase; NAPQI, N-acetyl-p-benzoquinone imine; PCR, polymerase chain reaction; PNPH, p-nitrophenol O-hydroxylation; PXR, pregnane X receptor; RXR, retinoid X receptor.
- Received April 20, 2005.
- Accepted September 1, 2005.
- The American Society for Pharmacology and Experimental Therapeutics