Disruption of peroxisome proliferator-activated receptor α in hepatocytes protects against acetaminophen-induced liver injury by activating the IL-6/STAT3 pathway

Background & Aims: Peroxisome proliferator-activated receptor α (PPARα) is a ligand-activated transcription factor abundantly expressed in liver. PPARα activator has been previously reported to protect against acetaminophen-induced hepatotoxicity, but fenofibrate, a lipid-lowering drug that activates PPARα, has a common side-effect causing liver injury. Thus, the exact effect of liver PPARα on drug-induced liver injury remains obscure. Methods: Hepatocyte-specific Ppara knockout mice and littermate wild-type control mice were intraperitoneally injected with acetaminophen (400 mg/kg body weight). Blood and liver samples were collected at different time points. We measured phase I and II cytochrome P450 enzymes, glutathione, reactive oxygen species, cytokines including Il6, and pSTAT3 by reverse transcriptase quantitative PCR, colorimetric, immunohistochemistry analyses and Western blotting. Results: Hepatic expression of PPARα was significantly decreased in DILI patients. Disruption of the Ppara gene in hepatocytes significantly reduced acetaminophen-induced liver injury in mice. ROS production rather than the expression levels of phase I and II cytochrome P450 enzymes was reduced in hepatocyte-specific Ppara knockout mice compared to control mice after acetaminophen administration. Mechanistically, hepatocyte-specific Ppara knockout mice had upregulated activation of the hepatoprotective pathway IL-6/STAT3 compared to wild-type mice, as evidenced by hepatic Il6 mRNA levels, hepatic protein levels of STAT3 and phosphorylated STAT3 were much higher in hepatocyte-specific Ppara knockout mice than in wild-type mice post acetaminophen injection. Conclusions: Hepatocyte-specific disruption of the Ppara gene protects against acetaminophen-induced liver injury by reducing oxidative stress and upregulating the hepatoprotective IL-6/STAT3 signaling pathway.


Introduction
Acetaminophen (APAP) is the most common over the-counter antipyretic and analgesic medicine, but overdose has become the headmost cause of acute liver failure (ALF), with high mortality and economic burden [1,2]. At therapeutic doses, most APAP is metabolized into nontoxic metabolites by UDP-glucuronosyltransferases (UGT) and sulfotransferases (SULT) [3,4]. A small percentage of it is Ivyspring International Publisher converted by phase I cytochrome P450 (CYP) enzymes to a chemically reactive metabolite (N-acetyl-pbenzoquinone imine, NAPQI), which can be detoxified by directly binding with glutathione (GSH) or by GSH-s-transferases (GST) [4]. However, with APAP overdose, GSH is exhausted, resulting in NAPQI accumulation and covalent binding to mitochondrial proteins, ultimately aggravating mitochondrial dysfunction, which is the key mechanism of APAP hepatotoxicity [5,6]. There are two hit theories, the first hit for mitochondrial GSH depletion and ROS production and the second hit for JNK activation and translocation into mitochondria, aggravating ROS generation and mitochondrial membrane permeability transition (MPT), leading to hepatocyte death [5]. However, the treatment is limited to N-acetylcysteine (NAC), which works only by restoring the levels of GSH at the early phase [7]. Those who progress to acute liver failure often need liver transplantation [8]. There is still a lack of effective targeted drugs for this disease.
Precise control of APAP metabolism into toxic metabolites is essential. Nuclear receptors are transcription factors that serve as xenobiotic-activated receptors and help regulate genes involved in drug metabolism and transporters [9]. Peroxisome proliferator-activated receptor (PPAR) α is a member of the nuclear receptor superfamily. Activation of PPARα by its ligands (fatty acid/eicosanoid/fibrates) controls fatty acid metabolism and regulates various biological transformation processes in different tissues, especially in the liver [10]. Earlier results demonstrated that a PPARα activator (fibrates) could inhibit APAP hepatotoxicity in WT mice [11,12] but lacked protection against APAP hepatotoxicity in Ppara -/mice [13]. However, the mechanism of this protection is controversial. Some studies revealed that pretreatment with clofibrate produced an increase in GSH levels, which promoted the exclusion of the toxic metabolite NAPQI [11], but others evidenced that the protection was not due to altered glutathione homeostasis [14]. Using liquid chromatography-mass spectrometry (LC-MS) and multivariate data analysis, it was reported that treatment with APAP induced liver injury and suppressed PPARα-regulated pathways, followed by inhibition of fatty acid oxidation and accumulation of acylcarnitines in the serum, thus promoting continuous oxidative stress in the liver [15]. A study by Patterson et al suggests that the PPARα target gene uncoupling protein 2 (UCP2) inhibited APAP-induced elevated reactive oxygen species (ROS) generation [16]. However, PPARα is enriched in various tissues with abundant fatty acid oxidation, including the liver, heart, brown adipose tissue, skeletal muscles, intestines and kidneys [10]. Therefore, it remains unknown whether the lack of protection against APAP-induced liver injury in PPARα-null mice relies on the contribution of PPARα activity in the liver or in other organs. Furthermore, it is reported that liver biochemistry abnormalities develop during fenofibrate therapy in 5-10% of patients. The clinical pattern of fenofibrate associated liver injury has ranged from an acute, self-limited hepatitis to a more persistent and even chronic liver injury [17]. So PPARα activity and its effect in the different cell types needs to be investigated.
In the present study, we aimed to investigate the unexplored role of PPARα in hepatocytes in models of APAP-induced liver injury in mice. We generated Ppara ΔHep mice and found that disruption of PPARα in hepatocytes protects against APAP-induced liver injury by decreasing APAP-induced oxidative stress and upregulating the IL-6/STAT3 signaling pathway.

Human Liver Histology and Immunohistochemistry
Eleven drug-induced liver injury (DILI) samples from patients who underwent a liver biopsy at Beijing Youan Hospital and three healthy control (HC) samples from patients who underwent hepatic cyst resection at Peking University First Hospital were analyzed. Human liver sections were immunostained for PPARα using an anti-PPARα antibody (Invitrogen, PA1-822A, US). The number of PPARα + cells in the liver sections was counted in 10 randomly selected fields (×400), and the average of the 10 fields was calculated. The study protocol and the use of bio-samples for research were approved by the Ethics Committee of the Peking University First Hospital.

Mice
Eight-to ten-week-old male C57BL/6 background mice were used in this study. WT mice were purchased from Vital River Laboratory (Beijing, China).
Hepatocyte-specific PPARα knockout (Ppara ΔHep ) mice and control floxed (Ppara fl/fl ) mice were constructed as described previously [18]. Briefly, mice with a floxed allele of exon 5 Ppara were built up by using homologous recombination in ES cells and back-crossed with transgenic mice expressing Cre recombinase regulated by the albumin (Alb-Cre) promoters for at least eight generations to create hepatocyte-specific (Ppara ΔHep ) knockout mice. Genomic DNA was extracted from the mouse tail using PCR to confirm the genotype. All animal care and experiments were approved by the Capital Medical University Animal Care and Use Committee. All institutional and national guidelines for the care and use of laboratory animals were followed.

APAP-induced liver injury
After fasting overnight, all the mice were intraperitoneally injected with 400 mg/kg fresh solutions of APAP (Sigma-Aldrich, St. Louis, MO) dissolving in warmed saline. Mice were sacrificed at different time points after treatment. Liver and blood samples were collected.

Biochemical Analysis for Serum Samples
Prior to sacrifice, blood was collected from the inner canthus vein with a lancet into heparin-coated tubes. Plasma was collected by centrifugation (8000 rpm, 5 min, 4 °C). Serum levels of alanine transaminase (ALT) and aspartate aminotransferase (AST) were measured using enzyme dynamic methods with the ALT/GPT Reagent Kit and AST/GOT Reagent Kit (Zhongsheng Technologies, Beijing, China), according to the manufacturer's instructions.

Liver Histology and Immunohistochemistry
Following euthanasia by overdosage, mouse livers were removed, weighed, and dissected into three parts for histological or biological analysis. Formalin-fixed liver samples were embedded in paraffin, sectioned at 4 µm, and subjected to routine hematoxylin and eosin (H&E) staining following standard methods. Immunohistochemical staining for PPARα and CYP2E1 performed using antibodies against 5-bromo-2-deoxyuridine (BD Bioscience, San Jose, CA), according to the manufacturer's protocol. The following primary antibodies were used: anti-PPARα (Abcam, Cambridge, UK) and anti-Cyp2e1 (Millipore, US). The number of PPARα positive cells in the liver sections was counted in 10 random fields (×200), and the average of the 10 fields was calculated. The average area of CYP2E1 in 10 random fields (×200) were analyzed.

Real-Time Quantitative Polymerase Chain Reaction (qRT-PCR)
Total RNA was extracted from liver tissues using TRIzol reagent (Life Technologies, USA) according to the manufacturer's instructions. Two micrograms of total RNA were reverse-transcribed into cDNA using the GoScript TM Reverse Transcription System (Promega, USA). qRT-PCR was performed in duplicate for each sample using a CFX Connect Real-Time System (Bio-Rad, USA) with SYBR Green Mix (TaKaRa, Japan). The amplification of specificity was confirmed by melting curve profiles. The mRNA level of β-actin was used as an internal standard. Primer sequences were listed in Table 1.

Measurement of GSH in Liver Tissues
Hepatic levels of glutathione (GSH) were measured in whole-liver homogenates (50-100 mg each of frozen liver tissue) using a glutathione assay kit (Nanjing Jiancheng Bioengineering Institute, China) according to the manufacturer's protocols.

Measurement of Reactive Oxygen Stress (ROS) in Liver Tissues
Hepatic levels of malondialdehyde (MDA) were analyzed in freshly collected liver tissue lysate using a Lipid Peroxidation MDA Assay Kit (Beyotime, China) according to the manufacturer's protocols. Superoxide in liver tissues was measured by dihydroethidium (DHE) staining after incubation at 37 °C for 30 minutes (Santa Cruz, CA). 4′,6-Diamidino-2-phenylinodole (DAPI; 7 × 10 −6 mol/L) was added for 5 min to visualize the nuclei and images captured the following day using an Eclipse 55i microscope (Nikon GmbH, Düsseldorf, Germany) and a Fluoro Pro MP 5000 camera (Intas Science Imaging Instruments GmbH, Göttingen, Germany) at 200-fold magnification. Detection wavelength settings were 406-458 nm for DAPI (blue), and 575-700 nm for DHE (red) imaging.

Statistical analysis
Data were expressed as means ± SEM. Student's t test was performed to compare two groups in the experiment. The four groups, Pparα fl/fl + Saline, Pparα fl/fl +APAP, Pparα ΔHep +Saline, and Pparα ΔHep +AP AP, were analyzed by one-way analysis of variance (ANOVA), followed by Tukey's post-hoc test. All data were analyzed using GraphPad Prism 7 software for Windows (La Jolla, CA). P values less than 0.05 were considered significant.

Decreased PPARα expression in DILI patients than HC
To test the expression level of PPARα in DILI patients and HCs, human liver sections were immunostained for PPARα. As indicated in Supplementary Fig. 1, the mean percentage of PPARα + cells in DILI patients and HCs was 45.1% and 59.2% (P=0.04), respectively, and the expression of PPARα in liver sections was significantly decreased in DILI patients compared with HCs ( Supplementary  Fig. 1A, B). Consistent with the PPARα expression level, histological analysis showed typical decreased PPARα expression in DILI patients compared with HCs ( Supplementary Fig. 1A). The clinical characteristics of the patients and healthy controls were showed in Supplementary Table 1. This finding suggests that PPARα signaling pathway plays a partial role in modulating drug induced liver injury.

Hepatic Ppara is downregulated in WT mice after APAP treatment
To examine the effects of APAP treatment on PPARα expression, we first studied the expression of PPARα after administration of toxic doses of APAP (400 mg/kg body weight) to WT mice. As indicated in Fig. 1, the serum levels of ALT and AST were significantly elevated at 6 hours and 12 hours after APAP treatment (Fig. 1A, B). Consistent with the ALT and AST levels, histological analysis established typical necrosis of the area located around the central hepatic veins at 12 hours after APAP treatment (Fig.  1C). Additionally, the relative mRNA expression of Ppara as well as Ppara-target genes (Cyp4a10, Cyp4a14) was downregulated in WT mice after treatment with toxic doses of APAP (Fig. 1D), which was consistent with previous studies suggesting that the mRNA expression levels of Ppara were inhibited after APAP treatment in WT mice [15,16].

Disruption of PPARα in hepatocytes significantly reduces APAP-induced liver injury
To examine the role of PPARα in hepatocytes in APAP-induced liver injury, Ppara ΔHep and littermate control Ppara fl/fl mice were administered an overdose of APAP. Hepatocyte-specific PPARα deficiency was confirmed by genotyping, qRT-PCR and Western blot analysis, as previous study. After APAP injection, Ppara fl/fl mice showed gradually elevated serum ALT and AST levels as well as extensive liver necrosis, peaking at 9 hours, and then declined progressively but remained at much higher levels at 24 hours after APAP injection ( Fig. 2A, B). Interestingly, hepatocyte-specific PPARα deficiency reduced ALT and AST levels at different time points after APAP injection ( Fig. 2A, B). In addition, H&E staining revealed that Ppara fl/fl mice had a large area of liver necrosis over time, peaking at 9 hours, whereas Ppara ΔHep mice strikingly decreased the area of necrosis after APAP injection (Fig. 3C, D). The liver body weight ratio was unchanged in Ppara fl/fl and Ppara ΔHep mice after APAP injection, as shown in Supplementary Fig. 2. Taken together, these findings suggested that hepatocyte-specific PPARα deficiency significantly reduced APAP-induced liver injury.

Disruption of PPARα in hepatocytes does not affect hepatic expression of APAP metabolizing enzymes
Previous studies have shown that whether PPARα affects APAP-metabolizing enzymes is controversial [11,14]. To test whether the protection of Ppara ΔHep mice against APAP-induced liver injury relied on the alteration of APAP metabolism, the expression of phase I, II, and III hepatic metabolizing enzymes was examined. After APAP injection, the mRNA levels of the metabolic enzymes Cyp2e1, Ugt1a1, Sult2a1, GSTM1, GSTM2, drug transporter Mrp3 as well as the protein levels of CYP2E1, a major P450 in the conversion of APAP into NAPQI, were not significantly different from those in Ppara fl/fl mice. Although the expression of the Cyp1a2, Gstm1 and Gstm2 genes was higher in Ppara ΔHep mice than in Ppara fl/fl mice, this tendency was also observed under basal conditions without APAP treatment (Fig. 3).
Furthermore, hepatic levels of GSH were almost completely depleted and comparable at 3 h and 6 h after APAP injection in both groups of mice (Fig. 4A). This was consistent with previous studies showing that the liver content of GSH was exhausted after APAP treatment [19][20][21]. These results suggested that the associated protective role of Ppara ΔHep mice in APAP-induced liver injury was not due to APAP bioactivation and detoxification.

Disruption of PPARα in hepatocytes decreases APAP-induced oxidative stress
Previous studies revealed that hepatic oxidative stress was the key mechanism in APAP hepatotoxicity [5,6] and that up-regulation of PPARα activation by Wy14,643 inhibited fatty acid oxidation after APAP treatment, resulting in suppression of lipid peroxidation and the production of superoxide, the major component of reactive oxygen species (ROS) [16,22]. To further explore the mechanisms by which Ppara ΔHep mice protected against APAP-induced liver injury, the hepatic content of malondialdehyde (MDA, lipid peroxidation markers) and dihydroethidium (DHE) staining for superoxide in liver tissue were examined. As illustrated in Fig. 4B, the hepatic level of MDA was lower in the APAP-treated Ppara ΔHep mice than in the Ppara fl/fl mice. Notably, DHE staining further confirmed that Ppara ΔHep mice greatly diminished ROS production after APAP treatment compared to that of Ppara fl/fl mice 3 hours post-APAP injection (Fig. 4C). These findings suggested that the protection of Ppara ΔHep mice in APAP-induced liver injury relied on amelioration of APAP-induced oxidative stress.

Disruption of PPARα in hepatocytes increases the IL-6/STAT3 signaling pathway after APAP treatment
PPARα is known to regulate anti-inflammatory effects by trans-repression mechanisms, frequently targeting NF-kB and AP-1 signaling [23]. We next investigated inflammatory cytokines in the liver. As illustrated in Fig. 5, qRT-PCR indicated that the gene expression of the inflammatory cytokine Il6 was elevated over time and peaked at 9 hours in Ppara ΔHep mice compared to Ppara fl/fl mice after APAP treatment (Fig. 5A, B), and the mRNA levels of Tnfa and Il1b were unchanged (Fig. 5A). Activation of the IL-6/STAT3 pathway plays an important role in promoting liver regeneration after APAP treatment [24], and hepatocyte-specific STAT3 knockout inhibits hepatocyte proliferation and liver regeneration, aggravating APAP-induced liver injury [25]. We next detected the levels of STAT3 phosphorylation at 9 hours after APAP treatment. As shown in Fig. 5C, STAT3 phosphorylation was upregulated in Ppara ΔHep mice compared to Ppara fl/fl mice after APAP treatment, indicating that STAT3 may be involved in the protection of APAP-induced liver injury in Ppara ΔHep mice.

Discussion
APAP-induced liver injury is a major public health problem that is strongly linked with acute liver failure. Most cases come from an unconscious ingestion of cold medicine containing APAP [1,2]. The majority of APAP is detoxified by the phase II enzymes UGT and SULT, while a small part is converted by P450 enzymes to the reactive metabolite NAPQI [3,4]. At therapeutic doses, the formation of NAPQI was detoxified by GSTM enzymes or directly bound to GSH to yield a GSH conjugate. After taking toxic doses of APAP, however, GSH is exhausted by conjugation. The highly reactive intermediate metabolite NAPQI accumulates, resulting in acute hepatic necrosis [4]. Thus, it is essential to control the formation of toxic metabolites produced by APAP metabolism, which is known to be regulated by Figure 3. Disruption of Ppara in hepatocytes does not affect APAP metabolism or detoxification. Eight-to ten-week-old male Ppara Δ Hep mice and Ppara fl/fl mice were treated with APAP or saline. Liver tissues were collected at 9 hours after APAP treatment. qPCR analyses of the expression of phase I metabolizing enzymes (A). Phase II metabolizing enzymes (B). Drug transporter, GST (phase II metabolizing enzymes) (C). IHC of CYP2E1 protein expression (D) and the area of CYP2E1 in 5 randomly selected fields (10X) were analyzed (E). Data were represented as means ± SEM (n=5-7/group). *P<0.05, **P<0.01, ***P<0.001. nuclear receptors [9]. PPARα is a member of the nuclear receptor superfamily. Whole-body Ppara -/mice show impaired fatty acid oxidation and accumulation of liver and serum levels of acylcarnitines, resulting in deterioration of APAP hepatotoxicity [15]. However, PPARα is also expressed in other organs to function in fatty acid β-oxidation [10]. Whether the role of PPARα in hepatocytes in APAP-induced liver injury is still unknown. Here, we explored the impact of hepatocyte-specific deletion of PPARα on APAPinduced liver injury in vivo. The present study demonstrates that hepatocyte-specific disruption of PPARα plays an important protective role against APAP-induced liver injury by reducing APAPinduced oxidative stress and upregulating the IL-6/ STAT3 signaling pathway to promote liver repair.  Although it is not clear whether drugs exert hepatotoxic effects, PPARα may be involved in the hepatotoxic mechanism. In this paper, we found that the expression of PPARα was significantly decreased in DILI patients compared to HCs. This result suggested that PPRAα plays a major role in the process of drug-induced liver injury. APAP bioactivation was the foremost process associated with APAP-induced liver injury. Our data revealed that the protective effect of Ppara ΔHep against APAP-induced liver injury was not due to APAP bioactivation and detoxification. This supports the concept that PPARα does not affect APAP metabolism [26]. Additionally, we found that GSH was exhausting, which agreed with a previous study showing that GSH was exhausting following APAP overdosing [27]. However, no differences in hepatic GSH content after APAP treatment were found between Ppara ΔHep mice and Ppara fl/fl mice. This result suggested that the protection of Ppara ΔHep mice against APAP toxicity is not due to APAP metabolism.
Hepatic oxidative stress triggered by inhibition of mitochondrial respiration through the covalent binding of NAPQI to mitochondrial proteins is a central element of APAP-induced liver injury [5,6]. Excessive accumulation of ROS induces mitochondrial permeability transition (MPT) and the cessation of ATP generation, ultimately leading to hepatocyte necrosis [5]. Our present study revealed that APAP induced significantly elevated ROS production in Ppara fl/fl mice but diminished ROS production in almost all Ppara ΔHep mice, as evidenced by the markedly lower levels of hepatic MDA content and decreased number of DHE-positive cells. Thus, hepatocyte-specific PPARα deficiency protects against APAP-induced liver injury by suppressing liver oxidative stress. However, previous studies revealed that the PPARα activators fenofibrate and Wy-14643 blocked APAP-induced oxidative stress, while whole-body Ppara -/mice did not [15,16]. The induction of the PPARα target gene UCP2 might be associated with a protective role, as confirmed by the sensitivity of Ucp2 -/mice to APAP-induced toxicity [16].
It is well-known that PPARα signaling prevents the induction of mitochondrial and ER stress, and mitochondrial β-oxidation activity was significantly down-regulated in the livers of Ppara-null mice, PPARα agonist improves hyperglycemia-induced oxidative stress [28][29][30][31]. There is some evidence that PPARα also has anti-inflammatory effects by counteracting nuclear factor kappa B (NF-κB) pathway [32].The results from the study of human primary hepatocytes treated with PPARα agonists also validate the anti-inflammatory role of PPARα [33]. A recently study on autoimmune myocarditis reported PPARα may play an important role by regulating IL-6/STAT3 pathway on Th17 cell differentiation which contributes to inflammatory response [34]. And more studies have also showed that PPARα agonists inhibits the expression of IL-6, while PPARα knockout mice displays an increase in IL-6 secretion when stimulated with LPS [35][36][37][38]. Besides that, PPARα was proved anti-tumor effects in hepatoma cells by regulating NF-κB signaling [39]. More importantly, macrophage-specific Ppara deletion did not impact Wy-14,643 induced up-regulation of lipid metabolism, cell proliferation, or DNA damage and repair-related gene expression. It has been proved that hepatomegaly and hepatocyte proliferation were observed in wild-type and Ppara ΔMac mice, but not Ppara ΔHep and wholebody Ppara-null mice. Furthermore, down-regulation of interleukin (IL)-15 and IL-18 after PPARα-agonist treatment was relied on the presence of macrophage PPARα. Growing evidence indicates PPARα may play the different roles in the different cell types, tissues, and underlying conditions [18,40]. So, the reasons may be that the effect of PPARα on APAP-induced liver injury is likely dependent on PPARα activity in different cells. Other cells, such as hepatic nonparenchymal cells, might play an important role in APAP-induced liver injury. This needs to be further investigated.
PPARα is known to interact with proinflammatory transcription factors, such as NF-κB, cJun and p65, to negatively regulate their target genes, Il1b, Il6 and Tnfa [23]. In this paper, we found that the mRNA levels of Il6 were significantly increased in Ppara ΔHep mice after APAP treatment, while Il1b and Tnfa were not. In the liver, STAT3, which is mainly activated by IL-6, IL-10 and IL-22, has been shown to play an important role in promoting liver proliferation and protecting against a variety of toxin-induced liver injuries [24,41,42]. Furthermore, a previous study reported that activation of the IL-6/STAT3 pathway played an important role in promoting liver regeneration after APAP treatment [24], and hepatocyte-specific STAT3 knockout inhibited hepatocyte proliferation and liver regeneration, aggravating APAP-induced liver injury in the early phase [25]. In this study, we revealed that APAP induced upregulation of STAT3 phosphorylation in Ppara ΔHep mice compared to Ppara fl/fl mice, indicating that STAT3 may be involved in promoting hepatocyte proliferation and liver regeneration in Ppara ΔHep mice, which may be attributed to the protection of APAP-induced liver injury in Ppara ΔHep mice compared to Ppara fl/fl mice.
Altogether, the present study performed in Ppara ΔHep mice after APAP overdose demonstrated that hepatocyte-specific PPARα deficiency plays an important protective role against APAP-induced oxidative stress and increases the IL-6/STAT3 signaling pathway to promote liver repair.