Clinical investigations suggest that hepatotoxicity after acetaminophen (APAP) overdose could be more severe in the context of obesity and nonalcoholic fatty liver disease. The pre-existence of fat accumulation and CYP2E1 induction could be major mechanisms accounting for such hepatic susceptibility. To explore this issue, experiments were performed in obese diabetic ob/ob and db/db mice. Preliminary investigations performed in male and female wild-type, ob/ob, and db/db mice showed a selective increase in hepatic CYP2E1 activity in female db/db mice. However, liver triglycerides in these animals were significantly lower compared with ob/ob mice. Next, APAP (500 mg/kg) was administered in female wild-type, ob/ob, and db/db mice, and investigations were carried out 0.5, 2, 4, and 8 h after APAP intoxication. Liver injury 8 h after APAP intoxication was higher in db/db mice, as assessed by plasma transaminases, liver histology, and terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling assay. In db/db mice, however, the extent of hepatic glutathione depletion, levels of APAP-protein adducts, c-Jun N-terminal kinase activation, changes in gene expression, and mitochondrial DNA levels were not greater compared with the other genotypes. Furthermore, in the db/db genotype plasma lactate and β-hydroxybutyrate were not specifically altered, whereas the plasma levels of APAP-glucuronide were intermediary between wild-type and ob/ob mice. Thus, early APAP-induced hepatotoxicity was greater in db/db than ob/ob mice, despite less severe fatty liver and similar basal levels of transaminases. Hepatic CYP2E1 induction could have an important pathogenic role when APAP-induced liver injury occurs in the context of obesity and related metabolic disorders.
Acetaminophen (N-acetyl-p-aminophenol; APAP) is one of the most widely prescribed drugs for the management of pain and hyperthermia. This drug is metabolized mainly in the liver by phase II-conjugating enzymes into the harmless glucuronide and sulfate conjugates, which represent in the urine approximately 55 and 30% of the initial APAP dose, respectively (Prescott, 1980). In addition, a small amount of APAP is oxidized to the highly reactive metabolite N-acetyl-p-benzoquinone imine (NAPQI) by several cytochromes P450, more specifically by CYP2E1 (Dahlin et al., 1984; Gonzalez, 2007). Although NAPQI is safely detoxified by hepatic reduced glutathione (GSH) when APAP is taken at the recommended dosage, high levels of NAPQI after overdose can induce several detrimental effects. Indeed, hepatocellular GSH stocks are quickly depleted, and as soon as GSH is no longer available for NAPQI detoxication, this reactive metabolite binds to different macromolecules, including proteins (Tirmenstein and Nelson, 1989; Muldrew et al., 2002). It is noteworthy that these early cellular events are followed by c-Jun N-terminal kinase (JNK) activation, mitochondrial dysfunction, overproduction of reactive oxygen species and peroxynitrite, massive hepatocellular necrosis, and acute liver failure (Cover et al., 2005; Jaeschke and Bajt, 2006; Hinson et al., 2010).
Although APAP is usually deemed a safe drug, APAP intoxication is relatively common, in particular in the context of unintentional overdosage (Larson et al., 2005; Craig et al., 2011). Several predisposing factors could enhance the risk and the severity of APAP-induced acute liver failure, including malnutrition, alcoholic liver disease, hepatitis C virus infection, and nonalcoholic fatty liver disease (NAFLD), which is often associated with obesity and type 2 diabetes (Nguyen et al., 2008; Myers and Shaheen, 2009). NAFLD encompasses a large spectrum of liver lesions such as fatty liver, nonalcoholic steatohepatitis (NASH), and cirrhosis (Diehl, 2002; Trak-Smayra et al., 2011). It is noteworthy that hepatic CYP2E1 activity is frequently enhanced in obese individuals with NAFLD (Chalasani et al., 2003; Aubert et al., 2011) and in type 2 diabetics (Wang et al., 2003).
Different investigations in rodents have dealt with APAP-induced hepatotoxicity in the context of obesity, type 2 diabetes, and NAFLD. However, these investigations have shown conflicting results. Indeed, whereas some studies showed increased hepatotoxicity (Corcoran and Wong, 1987; Kon et al., 2010; Kučera et al., 2012), others demonstrated no difference or an obvious protection (Blouin et al., 1987; Ito et al., 2006; Sawant et al., 2006). Although the exact reasons for this discrepancy are currently unknown, several hypotheses can be put forward. First, hepatic CYP2E1 activity can significantly vary between the different rodent models of obesity (Aubert et al., 2011). Second, obesity and NAFLD seem to be associated with increased APAP glucuronidation and clearance (Xu et al., 2012), and this could have afforded protection against APAP-induced liver injury in some studies. Third, basal levels of oxidative stress and mitochondrial dysfunction as well as lipid accumulation in liver can also greatly fluctuate between these models. The extent of lipid accretion could be a critical factor because it can sensitize cells to cell death induced by different types of insults (Feldstein et al., 2003; Reinartz et al., 2010). Thus, the main goal of the present study was to obtain information concerning the factors that could significantly modulate APAP-induced hepatotoxicity in the context of obesity and related metabolic disorders. To this end, we performed investigations in obese diabetic ob/ob and db/db mice presenting fatty liver with or without high hepatic CYP2E1 activity, respectively.
Materials and Methods
All mice were purchased from Janvier (Le-Genest-St-Isle, France). After their arrival at the animal facility, the mice were acclimatized for at least 1 week in an environmentally controlled room with a 12-h light/dark cycle and free access to food and water. In this study, wild-type C57BL/6J(+/+) (henceforth referred to as wild-type mice), C57BL/6J-ob/ob mice (ob/ob mice), and C57BL/KsJ-db/db mice (db/db mice) were used for all investigations. It is noteworthy that a previous study showed that wild-type C57BL/6J and C57BL/KsJ mice were similarly susceptible to APAP-induced liver injury (Harrill et al., 2009). In a first series of preliminary investigations, 10-week-old male and female wild-type, ob/ob, and db/db mice were used to determine hepatic CYP2E1 activity. Thereafter, 10- to 12-week-old female wild-type, ob/ob, and db/db mice were used for all experiments dealing with APAP-induced hepatotoxicity. In these experiments, mice were treated with APAP (Sigma-Aldrich, St. Quentin-Fallavier, France) or the vehicle control after an overnight fast. To this end, APAP was dissolved in warm saline and injected intraperitoneally at the dose of 500 mg/kg body weight, whereas saline was administered to control animals. After 0.5, 2, 4, or 8 h, blood was drawn from the retro-orbital sinus with heparinized capillary Pasteur pipettes for biochemistry analyses. Mice were then sacrificed by cervical dislocation, and livers were quickly removed. Although a majority of the liver fragments were immediately frozen in liquid nitrogen, some of them were rapidly processed for appropriate histological staining. Collected liver fragments frozen in liquid nitrogen were subsequently stored at −80°C until use. All experiments were performed according to national guidelines for the use of animals in biomedical research and approved by the local Ethics Committee in Animal Experiment of the Université de Rennes 1.
Immediately after collection, blood was centrifuged for 10 min at 1000g, and plasma was stored at −20°C until assay. Plasma activity of alanine aminotransferase (ALT) and aspartate aminotransferase (AST) and glucose levels were measured on an automatic analyzer AU2700 (Olympus Diagnostics, Rungis, France) with Olympus commercial kits (OSR6107, OSR6109, and OSR6121, respectively). Plasma lactate and β-hydroxybutyrate were measured with commercial kits purchased from Biovision Inc. (Mountain View, CA) and Cayman Chemical (Ann Arbor, MI), respectively. Plasma levels of APAP, APAP-glucuronide, and APAP-sulfate were measured by using a liquid chromatography/tandem mass spectrometry system from Thermo Fisher Scientific (Waltham, MA). In brief, separation of the analytes was carried out on a Thermo Hypersil Gold C18 column (3.0 μm; 2.1 × 100 mm; Thermo Fisher Scientific). The mobile phase consisted of aqueous 1% formic acid and 95% methanol (80:20, v/v). Calibration or mouse plasma (25 μl) samples were supplemented with deuterated APAP (Promochem, Molsheim, France) as internal standard and treated with methanol. The retention times for APAP-glucuronide, APAP-sulfate, APAP, and internal standard were 2.3, 2.9, 3.5 and 3.5 min, respectively. Data acquisition, peak integration, and calibration were performed by using Xcalibur 2.1 software (Thermo Fisher Scientific).
Liver Histology and Terminal Deoxynucleotidyl Transferase-Mediated dUTP Nick-End Labeling Assay.
To evaluate the different hepatic lesions, liver fragments were fixed in 10% neutral formalin and embedded in paraffin. Then, 4-μm-thick sections were cut, stained with hematoxylin, eosin, and safran (HES), and blindly examined by an experienced pathologist (B. Turlin). After the examination of 10 different liver lobules per mouse, a score of necrosis from 0 to 3 was given based on the surface of the necrotic areas and the nature of the cytoplasmic and nuclear alterations. HES staining was also used to evaluate the presence of macrovacuolar and microvesicular steatosis. Cell death was assessed with the TUNEL assay by using the In Situ Cell Death Detection Kit purchased from Roche Diagnostics (Indianapolis, IN), according to the manufacturer's recommendations.
Reduced GSH and APAP-Protein Adducts in Liver.
Reduced GSH was determined by an enzymatic reaction in which it reduces 5,5′-dithio-bis(2-nitrobenzoic acid) to generate 2-nitro-5-thiobenzoic acid that is spectrophotometrically measured at 412 nm (Robin et al., 2005a). For the assessment of APAP-protein adducts, APAP-cysteine was measured by using high-pressure liquid chromatography separation and electrochemical detection as described previously (Muldrew et al., 2002).
CYP2E1 and Caspase 3 Activities.
CYP2E1 activity was assessed on hepatic microsomes as described previously (Robin et al., 2005b). To prepare microsomes, liver fragments (approximately 500 mg) were homogenized in 5 ml of 50 mM Tris buffer supplemented with 0.25 M sucrose, 1 mM EDTA, 25 μM phenylmethylsulfonyl fluoride, and 1 mM dithiothreitol, pH 7.4, and the homogenate was centrifuged at 1000g for 10 min at 4°C. The supernatant (liver tissue homogenate) was subsequently centrifuged at 9000g for 20 min at 4°C. The new supernatant was finally centrifuged at 100,000g for 1 h at 4°C. The pellet (microsomal fraction) was resuspended in 200 μl of 0.1 M sodium phosphate buffer, pH 7.4, containing 10% glycerol, and stored at −80°C until use. Microsomal proteins were determined by the method of Lowry et al. (1951) based on the Folin phenol reagent. CYP2E1 activity was subsequently assessed by measuring hydroxylation of aniline into p-aminophenol (Robin et al., 2005b) and hydroxylation of chlorzoxazone into 6-hydroxychlorzoxazone (Guillouzo and Chesne, 1996). Hepatic caspase 3 activity was assessed in liver homogenates as described previously (Dumont et al., 2010).
Western Blot Analysis.
To assess liver expression of total and phosphorylated JNK1/2 (JNK1/2 and P-JNK1/2, respectively), hepatic proteins underwent SDS-polyacrylamide electrophoresis. In brief, liver fragments were homogenized in a lysis buffer (Cell Signaling Technology, Danvers, MA) supplemented with 1 mM phenylmethylsulfonyl fluoride. Homogenates were then centrifuged at 4500g at 4°C to remove tissue debris. Forty micrograms of protein were then separated by electrophoresis on 4 to 12% gradient Bis-Tris gels (Invitrogen, Cergy-Pontoise, France), transferred to Hybond ECL nitrocellulose membranes (GE Healthcare, Chalfont St. Giles, Buckinghamshire, UK) and immunoblotted with antibodies against JNK1/2 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA), P-JNK1/2 (Thr183/Tyr185) (Calbiochem, San Diego, CA), and heat shock cognate protein 70 (HSC70) (Tebu-bio, Le Perray en Yvelines, France). Blots were then incubated with appropriate secondary antibodies, and protein bands were revealed by enhanced chemiluminescence in a Chemi-smart imager (Thermo Fisher Scientific, Illkirch, France). HSC70 was used to normalize protein loadings, and quantification was performed with BIO-1D software (Vilber Lourmat, Marne la Vallée, France). To correlate JNK phosphorylation status to its activity, the nonradioactive stress-activated protein kinase/JNK assay kit (Cell Signaling Technology) was used according to the manufacturer's recommendations.
Real-Time Quantitative PCR Analysis.
To study gene expression in liver, total RNA was extracted with the SV Total RNA Isolation System from Promega (Madison, WI), which included a direct DNase treatment step. RNA quantity and purity were assessed with a Nanodrop ND-1000 spectrophotometer (Nyxor Biotech, Paris, France), and RNA integrity was checked after migration on a 1% agarose gel. cDNAs were prepared by reverse transcription of 1 μg of total RNA by using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Courtaboeuf, France). cDNAs were thus amplified with specific primers by using the Power SYBR Green PCR Master Mix (Applied Biosystems), in an ABI Prism 7900 instrument (Applied Biosystems). The PCR conditions were one cycle at 50°C for 2 min and one cycle at 95°C for 10 min followed by 40 cycles at 95°C for 15 s and 60°C for 1 min. Amplification of specific transcripts was confirmed by melting curve profiles generated at the end of each run. Moreover, PCR specificity was further ascertained with a 1.75% agarose gel electrophoresis by checking the length of the PCR products. Expression of the mouse ribosomal protein S6 was used as reference, and the 2−ΔΔCt method was used to express the relative expression of each selected gene (Begriche et al., 2008a; Massart et al., 2012). Quantification of the mitochondrial DNA (mtDNA) was also performed by RT-qPCR after extraction of total genomic DNA by using the QIAGEN DNeasy tissue kit (QIAGEN, Courtaboeuf, France). In brief, 20-mg liver samples were digested in a lysis buffer supplemented with proteinase K. After complete tissue digestion, genomic DNA was then extracted and purified according to the manufacturer's recommendations. mtDNA levels were then assessed by RT-qPCR using primers specific to the mitochondrial cytochrome c oxidase II gene, whereas the nuclear DNA-encoded S6 gene was used for normalization. Sequences of the primers used in this study are available on request.
Data are presented as means ± S.E.M. Analysis of variance (ANOVA) was performed to assess statistical significance. For the preliminary investigations carried out with male and female wild-type, ob/ob, and db/db mice, a two-way ANOVA was used with the factors of sex and genotype. For the subsequent experiments performed with female wild-type, ob/ob, and db/db mice treated with APAP or vehicle control, a two-way ANOVA was performed with the factors of genotype and APAP treatment. The one-way ANOVA and Mann-Whitney tests were used in some experiments. When ANOVA provided significant differences, individual means were compared with the Tukey's Honestly Significant Difference test.
Preliminary Investigations in Lean and Obese Mice.
The present study was performed to obtain information concerning factors that could significantly modulate APAP-induced hepatotoxicity in the context of obesity and NAFLD. In particular, we wanted to determine whether hepatic CYP2E1 induction and/or the degree of fatty liver could have a significant effect on early APAP-induced liver injury.
CYP2E1 plays a key role in APAP hepatotoxicity, because this enzyme generates the toxic metabolite NAPQI (Jaeschke and Bajt, 2006; Gonzalez, 2007). Moreover, hepatic CYP2E1 activity is frequently enhanced in obese individuals with associated metabolic diseases such as NAFLD and type 2 diabetes (Chalasani et al., 2003; Aubert et al., 2011). However, to the best of our knowledge, hepatic CYP2E1 activities in male and female wild-type, ob/ob, and db/db mice have never been compared in the same study. In a first set of preliminary investigations, we found that microsomal CYP2E1-mediated aniline and chlorzoxazone hydroxylase activities were significantly increased in female db/db mice compared with the other groups of animals (Table 1). Moreover, there was a positive correlation between plasma glucose (Table 1) and CYP2E1-mediated aniline hydroxylase activity in liver (r2 = 0.315; P < 0.01), supporting previous data about a possible role of type 2 diabetes in hepatic CYP2E1 induction (Aubert et al., 2011). Thus, female db/db mice were selected in the present study as a murine model of obesity associated with hepatic CYP2E1 induction. It is noteworthy that several investigations in mice of different genetic background (including C57BL/6) showed that APAP biotransformation and toxicity significantly varied between males and females (Dai et al., 2006; Lee et al., 2009; Masubuchi et al., 2011). Hence, only females of the different mouse genotypes were selected for further investigations.
The degree of fatty liver is another factor that could play a significant role in APAP-induced liver injury in the context of obesity. Indeed, excess lipid can sensitize cells to cell death induced by different types of insults (Feldstein et al., 2003; Reinartz et al., 2010). In this study, hepatic fat deposition was higher in female ob/ob mice with a combination of macrovacuolar and microvesicular steatosis, whereas steatosis was mostly microvesicular in female db/db mice (Fig. 1A). Moreover, liver triglycerides were significantly enhanced by 134% in female ob/ob mice compared with female db/db mice (Fig. 1B). These data are in keeping with previous results also obtained in female ob/ob and db/db mice (Sahai et al., 2004). Greater lipid deposition in ob/ob mice also explained higher liver weight and liver-to-body-weight ratio in these animals (Table 1). Finally, liver lipids in female wild-type mice (n = 6) were much lower compared with the obese animals because total lipids and triglycerides were 18 ± 3 and 4 ± 1 mg·g−1 liver, respectively.
Early Hepatotoxicity after APAP Intoxication in Female Wild-Type, ob/ob, and db/db Mice.
To evaluate APAP-induced liver injury, 500 mg/kg APAP was administered to female wild-type, ob/ob, and db/db mice, and plasma transaminases were determined 0.5, 2, 4, and 8 h after APAP intoxication. It is noteworthy that basal activity of transaminases was higher in obese mice (Fig. 2A), reflecting the presence of moderate hepatic cytolysis associated to NASH. Whereas plasma transaminases were unchanged in all groups of mice 0.5 h after APAP administration (data not shown), the activity of transaminases was significantly enhanced after 2 h, but only in wild-type mice (609 ± 144 and 2542 ± 614 UI/liter, respectively, for ALT and AST; n = 6–7 mice). After 4 h, plasma transaminases were significantly enhanced in all groups of treated mice, although there was no difference between the different genotypes. Indeed, ALT activity was 1022 ± 328, 783 ± 176, and 619 ± 136 UI/liter, whereas AST values were 1495 ± 412, 1603 ± 368, and 1291 ± 239 UI/liter, respectively, in wild-type, ob/ob, and db/db mice (n = 9–12 animals per group).
After 8 h, plasma transaminases were further enhanced in the different groups of treated mice, but particularly in db/db mice (Fig. 2B). Indeed, AST and ALT activity in db/db mice was increased by 2- or 3-fold compared with wild-type and ob/ob mice. In all mice, the areas of cytolysis were generally located around the central veins (i.e., where CYP2E1 is mainly expressed), although in some obese mice these areas extended to the mediolobular zones (Fig. 2C). When a score of necrosis was attributed as described under Materials and Methods, all treated wild-type (n = 9) mice were scored 2, whereas one, six, and two ob/ob mice were scored 0, 1, and 3, respectively. For the db/db genotype, one, five, and three mice were scored 1, 2 and 3, respectively. Thus, necrosis was in general less severe in ob/ob mice, whereas the highest frequency of score 3 was observed in db/db mice. It is noteworthy that APAP intoxication in wild-type mice induced macrovacuolar steatosis in some hepatocytes (Fig. 2C), as described previously in humans and mice (Yohe et al., 2006; Begriche et al., 2011). Moreover, APAP intoxication aggravated steatosis in obese mice, in particular in ob/ob mice (Fig. 2C). Finally, a TUNEL assay was performed, and the percentage of TUNEL-positive nuclei was measured in the different groups of mice. As shown in Fig. 3, the TUNEL-positive nuclei were located around the central vein, and the percentage of these nuclei was significantly higher in db/db mice compared with ob/ob mice. However, measurement of caspase 3 activity showed virtually no activity of this enzyme, whatever the group of treated mice (data not shown). This result is in keeping with previous investigations showing that hepatic necrosis, rather than apoptosis, is the major pathway of cell death after APAP intoxication (Jaeschke and Bajt, 2006; Han et al., 2010; Hinson et al., 2010).
GSH and APAP-Protein Adducts in Liver.
Early APAP hepatotoxicity is associated with a rapid depletion of liver GSH reserves, as well as covalent binding of NAPQI to macromolecules, including proteins (Tirmenstein and Nelson, 1989; Muldrew et al., 2002; Jaeschke and Bajt, 2006). In this study, decreased GSH and increased APAP-protein adducts in liver were observed 0.5, 2, 4, and 8 h after APAP intoxication in all groups of mice, although there were some differences between wild-type and obese mice (Fig. 4). Indeed, GSH depletion and recovery occurred earlier in wild-type mice compared with ob/ob and db/db mice (Fig. 4A). Levels of APAP-protein adducts in liver reached a plateau as soon as 2 h in wild-type mice, but only after 4 h in ob/ob and db/db mice. In addition, levels of APAP-protein adducts in obese mice after 0.5, 2, and 8 h were significantly lower compared with wild-type mice (Fig. 4B).
Total and Phosphorylated JNK in Liver.
APAP hepatotoxicity is associated with a rapid phosphorylation of JNK, inducing its activation, which then enhances mitochondrial dysfunction and other detrimental cellular events (Han et al., 2010; Hinson et al., 2010). In this study, JNK phosphorylation was observed 2, 4, and 8 h after APAP intoxication (Fig. 5), but not after 0.5 h (data not shown). After 2 h, JNK was phosphorylated in all wild-type mice, but only in a few obese mice (one of three animals in each genotype). After 4 h, JNK was phosphorylated in all mice, but there was no significant difference between genotypes. After 8 h, JNK phosphorylation was still high in ob/ob mice, but was reduced in wild-type and db/db mice (Fig. 5). Measurement of hepatic JNK activity was in keeping with these results, with a trend toward lower JNK activation in ob/ob and db/db mice after 2 h and higher JNK activity in ob/ob mice after 8 h (data not shown).
Hepatic mRNA Expression of Genes after APAP Intoxication.
After APAP intoxication, there is a rapid alteration in the hepatic expression of numerous genes involved in oxidative stress and antioxidant defense, glutathione synthesis, APAP metabolism, cell repair, and lipid homeostasis, including mitochondrial and peroxisomal fatty acid oxidation (Heinloth et al., 2004; Beyer et al., 2007). In this study, hepatic expression of the following genes were determined after APAP administration: NF-E2-related factor-2 (Nrf2), heme oxygenase-1 (HO-1), heat shock protein 70 (Hsp70), insulin-like growth factor binding protein-1 (IGFBP-1), γ-glutamylcysteine synthetase (γGCS), Tribbles homolog 3 (Trib3), glutathione transferases (GSTs) π-1 (GSTP1) and GST κ (GSTK), peroxisome proliferator-activated receptor-α (PPARα), l-carnitine palmitoyltransferase 1 (l-CPT1), acyl-coenzyme A oxidase (ACO), and pyruvate dehydrogenase kinase 4 (PDK4). Recent investigations showed that increased Trib3 expression was a good marker to detect GSH depletion secondary to its conjugation (Gao et al., 2010), and high basal PDK4 expression could increase sensitivity to APAP-induced hepatotoxicity (Liu et al., 2010). As shown in Fig. 6A, the expression of several of these genes was up-regulated as early as 0.5 h after APAP administration, especially IGFBP-1, γGCS, Trib3, and PDK4. Expression of these four genes was still increased after 2, 4, and 8 h (Fig. 6, B–D). It is noteworthy that γGCS expression was particularly enhanced in wild-type mice 0.5 and 2 h after APAP intoxication (Fig. 6, A and B), whereas γGCS up-regulation was principally observed in obese mice after 8 h (Fig. 6D). Because γGCS is a key enzyme of the GSH biosynthetic pathway, these results could explain, at least in part, why the recovery of liver GSH was faster in wild-type mice compared with ob/ob and db/db mice (Fig. 4A). HO-1 expression was highly up-regulated 2 h after APAP intoxication in wild-type mice and afterward in all genotypes (Fig. 6, B-D). A reduction of PPARα expression was observed 2, 4, and 8 h after APAP administration. It is noteworthy that 8 h after APAP administration expression of PPARα and its targets l-CPT1 and ACO was particularly decreased in obese mice (Fig. 6D). Moreover, at this time point, expression of the mitochondrial fatty acid oxidation enzyme medium-chain acyl-CoA dehydrogenase (another PPARα target) was significantly reduced in ob/ob and db/db mice by 32 and 37%, respectively, whereas it was unchanged in wild-type mice (data not shown). Significant reduction of the PPARα signaling pathway in ob/ob and db/db mice could be one mechanism explaining the aggravation of fatty liver in these obese animals (Fig. 2).
Hepatic mtDNA Levels.
APAP intoxication in mice can induce early mtDNA depletion in liver (Cover et al., 2005). In the present study, hepatic mtDNA levels were significantly decreased by 32% in ob/ob mice 2 h after APAP intoxication, but mtDNA depletion was not observed in the other groups of treated mice (Table 2). In addition, whereas mtDNA was virtually unchanged 4 h after APAP treatment in all groups of animals, mtDNA levels were significantly enhanced after 8 h in all treated mice (Table 2). It was also noteworthy that basal hepatic mtDNA levels were increased in ob/ob mice as reported previously (Robin et al., 2005a). Increased mtDNA levels in liver could be an adaptive response to oxidative stress (Robin et al., 2005a).
Plasma Lactate and β-Hydroxybutyrate.
Drug-induced mitochondrial dysfunction can induce hyperlactatemia (or lactic acidosis) and complex changes in the plasma ketone bodies acetoacetate and β-hydroxybutyrate (Fromenty and Pessayre, 1995; Labbe et al., 2008). Indeed, although plasma ketone bodies are classically reduced when hepatic fatty acid oxidation is impaired, plasma levels of these oxidative products can be paradoxically enhanced when the Krebs cycle is inhibited in extra-hepatic tissues (Fromenty and Pessayre, 1995; Labbe et al., 2008). In this study, plasma lactate was decreased in mice 2, 4, and 8 h after APAP administration, in particular in ob/ob mice (Table 2). It is noteworthy that plasma glucose in ob/ob mice was significantly reduced by 47 and 57%, respectively, 4 and 8 h after APAP administration (data not shown). Thus, reduced glucose availability might have reduced lactate production in ob/ob mice. Plasma β-hydroxybutyrate levels were significantly decreased 4 h after APAP administration in all genotypes, but especially in obese mice (Table 2). After 2 and 8 h, however, plasma β-hydroxybutyrate was either decreased or increased in the different groups of animals (Table 2).
Plasma Levels of APAP, APAP-Sulfate, and APAP-Glucuronide.
Plasma levels of APAP, APAP-sulfate, and APAP-glucuronide were determined in treated mice 0.5, 2, 4, and 8 h after APAP intoxication (Table 3). After 0.5 and 2 h, APAP levels were higher in obese mice compared with wild type, whereas the converse was observed later on. Levels of APAP-sulfate were significantly higher in ob/ob and db/db obese mice compared with wild-type mice, but only 2 h after APAP administration. Finally, APAP-glucuronide levels after 0.5 and 2 h were significantly increased in obese mice compared with wild-type mice (Table 3). It is noteworthy that whatever the time points the levels of APAP-glucuronide were always the highest in ob/ob mice. This is in keeping with recent investigations showing higher APAP glucuronidation in hepatic microsomes from ob/ob mice compared with wild-type mice (Xu et al., 2012).
The present study was carried out to explore the underlying mechanisms that could potentially explain why NAFLD enhances the severity of APAP-induced hepatotoxicity. Our data clearly indicated that the degree of fatty liver in obese mice was not a key pathogenic factor in early hepatotoxicity induced by 500 mg/kg APAP. Indeed, whereas basal levels of hepatic triglycerides was lower in db/db mice compared with ob/ob mice, liver injury 8 h after APAP intoxication was significantly higher in the former group. Moreover, APAP-induced hepatic cytolysis was approximately similar between wild-type and ob/ob mice.
Leptin-deficient ob/ob mice and leptin-resistant db/db mice are widely used as models of obesity, type 2 diabetes, and NAFLD. One of the major advantages of these genetic murine models of obesity is the rapid and consistent occurrence of insulin resistance, hyperglycemia, and NAFLD (Lindström, 2007; Trak-Smayra et al., 2011; Massart et al., 2012), whereas high-fat diet-induced obesity is not always associated with these secondary metabolic disorders, even after several months of feeding (Burcelin et al., 2002; Begriche et al., 2008a). It is noteworthy that both ob/ob and db/db mice are characterized by “borderline NASH,” because fatty liver is associated with moderate hepatic necroinflammation and mild fibrosis (Li et al., 2003; Begriche et al., 2008b; Trak-Smayra et al., 2011). Moreover, this study and previous investigations (Trak-Smayra et al., 2011) showed comparable basal levels of plasma transaminases between ob/ob and db/db mice, indicating a similar degree of hepatic cytolysis. However, db/db mice have less lipid deposition in liver compared with ob/ob mice (Sahai et al., 2004; Trak-Smayra et al., 2011), a feature that has been confirmed in this study.
Several experimental and clinical investigations reported that NAFLD could sensitize to drug-induced acute liver injury (Tarantino et al., 2007; Begriche et al., 2011). However, the present study indicates that fatty liver per se and pre-existent hepatic cytolysis are not systematically involved in this sensitization. This strongly suggests the presence of other pathogenic factors. Although further investigations will be required to determine the exact causes of higher sensitivity of db/db mice to APAP hepatotoxicity, increased basal hepatic CYP2E1 activity could have played a role. Enhanced hepatic CYP2E1 activity has been associated with more severe APAP-induced hepatotoxicity in the context of chronic ethanol intoxication (Zhao and Slattery, 2002; Gonzalez, 2007). This major role of CYP2E1 is caused by the generation of NAPQI, a highly reactive metabolite that depletes GSH (Jaeschke and Bajt, 2006) and binds to different cellular proteins (Tirmenstein and Nelson, 1989; Muldrew et al., 2002).
In our investigations, however, depletion of GSH and levels of APAP-protein adducts in liver were not higher in db/db mice. Regarding these adducts, it is conceivable that a small subset of critical proteins could have been specifically damaged by NAPQI in db/db mice. It is noteworthy that NAPQI could bind to, and thus inhibit, different enzymes of the plasma membrane, cytosol, and mitochondria (Tsokos-Kuhn et al., 1988; Landin et al., 1996; Gupta et al., 1997). However, it is still uncertain whether APAP-induced hepatotoxicity is caused primarily by APAP binding to these enzymes or other yet unidentified critical proteins. Moreover, the cellular targets of NAPQI could be different in female db/db mice. Nevertheless, our data strongly suggested that subtle events downstream of APAP metabolic activation were responsible for the higher susceptibility of db/db mice. In this respect, it is noteworthy that recent investigations in interleukin-4(−/−) mice showed higher susceptibility to APAP-induced liver injury despite lower levels of APAP-protein adducts (Ryan et al., 2012).
JNK activation and mitochondrial dysfunction play a significant role in APAP liver injury (Han et al., 2010; Shinohara et al., 2010; McGill et al., 2011). However, our investigations did not reveal greater hepatic JNK activation in db/db mice after APAP intoxication. Early hepatic mtDNA depletion was detected in ob/ob mice 2 h after APAP intoxication but was not observed in db/db mice. Later on, 8 h after APAP administration, mtDNA levels were increased in treated mice whatever the genotype. Plasma levels of lactate and β-hydroxybutyrate were also measured as surrogate markers of mitochondrial dysfunction (Fromenty and Pessayre, 1995; Labbe et al., 2008). However, after APAP treatment, plasma concentrations of lactate were decreased in all groups of mice, especially ob/ob mice. In addition, plasma levels of β-hydroxybutyrate were either increased or decreased in db/db mice, but the extent of these changes were also greater in ob/ob mice. Finally, hepatic expression of the mitochondrial enzymes CPT1, PDK4, and GSTK was similarly modified after APAP treatment in the different groups of animals. Further investigations will be needed to determine whether specific mitochondrial alterations could have been responsible for increased APAP hepatotoxicity in db/db mice.
Obesity is a major issue for health because it is associated with several metabolic disorders, including insulin resistance, type 2 diabetes, and NAFLD. Moreover, there is increasing evidence that obesity and NAFLD can increase the risk and the severity of liver injury induced by several drugs. Indeed, in addition to APAP, this is suspected for halothane, tamoxifen, irinotecan, methotrexate, and some antibiotics (Tarantino et al., 2007; Aubert et al., 2011; Begriche et al., 2011). Obesity and NAFLD also favor hepatotoxicity of toxic compounds such as ethanol, carbon tetrachloride, and thioacetamide (Robin et al., 2005a; Donthamsetty et al., 2007; Kučera et al., 2011). It is noteworthy that CYP2E1 metabolizes several of these drugs and toxic derivatives, and hepatic CYP2E1 activity is frequently enhanced in obese individuals with NAFLD (Chalasani et al., 2003; Aubert et al., 2011). Thus, along with different rodent models of diet-induced obesity and NAFLD (Aubert et al., 2011), we believe that female db/db mice could be a useful murine model for exploring liver injury induced by these compounds.
Participated in research design: Aubert, Robin, and Fromenty.
Conducted experiments: Aubert, Begriche, Delannoy, Pajaud, Ribault, Lepage, McGill, and Lucas-Clerc.
Performed data analysis: Aubert, Morel, Turlin, Robin, and Fromenty.
Wrote or contributed to the writing of the manuscript: Aubert, Robin, Jaeschke, and Fromenty.
We thank Alain Fautrel and Pascale Bellaud from the Plate-forme d'Histopathologie for excellent technical support.
This work was supported by the Institut Nationale de la Santé et de la Recherche Médicale. J.A. and K.B. were recipients of grants from the President of Université de Rennes 1 and the Région Bretagne, respectively.
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
- acyl-coenzyme A oxidase
- alanine aminotransferase
- analysis of variance
- aspartate aminotransferase
- γ-glutamylcysteine synthetase
- glutathione transferase
- GST κ
- GST π-1
- hematoxylin, eosin, and safran
- heme oxygenase-1
- heat shock cognate protein 70
- Honestly Significant Difference
- heat shock protein 70
- insulin-like growth factor binding protein-1
- c-Jun N-terminal kinase
- phosphorylated JNK
- l-carnitine palmitoyltransferase 1
- mitochondrial DNA
- nonalcoholic fatty liver disease
- N-acetyl-p-benzoquinone imine
- nonalcoholic steatohepatitis
- NF-E2-related factor-2
- polymerase chain reaction
- real-time quantitative PCR
- pyruvate dehydrogenase kinase 4
- peroxisome proliferator-activated receptor-α
- Tribbles homolog 3
- terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling
- Received February 28, 2012.
- Accepted May 29, 2012.
- Copyright © 2012 by The American Society for Pharmacology and Experimental Therapeutics