Fibroblast growth factor 21 protects against acetaminophen-induced hepatotoxicity by potentiating peroxisome proliferator-activated receptor coactivator protein-1α-mediated antioxidant capacity in mice

Authors

  • Dewei Ye,

    1. State Key Laboratory of Pharmaceutical Biotechnology, University of Hong Kong, Hong Kong, China
    2. Department of Medicine, University of Hong Kong, Hong Kong, China
    3. Department of Pharmacology & Pharmacy, University of Hong Kong, Hong Kong, China
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Yudong Wang,

    1. State Key Laboratory of Pharmaceutical Biotechnology, University of Hong Kong, Hong Kong, China
    2. Department of Medicine, University of Hong Kong, Hong Kong, China
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Huating Li,

    1. Department of Endocrinology and Metabolism, Shanghai Jiao Tong University Affiliated Sixth People's Hospital, Shanghai, China
    Search for more papers by this author
  • Weiping Jia,

    1. Department of Endocrinology and Metabolism, Shanghai Jiao Tong University Affiliated Sixth People's Hospital, Shanghai, China
    Search for more papers by this author
  • Kwan Man,

    1. Department of Surgery, University of Hong Kong, Hong Kong, China
    Search for more papers by this author
  • Chung Mau Lo,

    1. Department of Surgery, University of Hong Kong, Hong Kong, China
    Search for more papers by this author
  • Yu Wang,

    1. State Key Laboratory of Pharmaceutical Biotechnology, University of Hong Kong, Hong Kong, China
    2. Department of Pharmacology & Pharmacy, University of Hong Kong, Hong Kong, China
    Search for more papers by this author
  • Karen S.L. Lam,

    Corresponding author
    1. State Key Laboratory of Pharmaceutical Biotechnology, University of Hong Kong, Hong Kong, China
    2. Department of Medicine, University of Hong Kong, Hong Kong, China
    • Address reprint requests to: Prof. Aimin Xu or Prof. Karen S.L. Lam, State Key Laboratory of Pharmaceutical Biotechnology, and Department of Medicine, University of Hong Kong, Room L8-39, Lab Block, 21 Sassoon Road, Hong Kong. E-mail: amxu@hku.hk or ksllam@hku.hk; fax: +852-28162095.

    Search for more papers by this author
  • Aimin Xu

    Corresponding author
    1. State Key Laboratory of Pharmaceutical Biotechnology, University of Hong Kong, Hong Kong, China
    2. Department of Medicine, University of Hong Kong, Hong Kong, China
    3. Department of Pharmacology & Pharmacy, University of Hong Kong, Hong Kong, China
    • Address reprint requests to: Prof. Aimin Xu or Prof. Karen S.L. Lam, State Key Laboratory of Pharmaceutical Biotechnology, and Department of Medicine, University of Hong Kong, Room L8-39, Lab Block, 21 Sassoon Road, Hong Kong. E-mail: amxu@hku.hk or ksllam@hku.hk; fax: +852-28162095.

    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.

  • Supported by Collaborative Research Fund (HKU2/CRF/12R and HKU3/CRF/11R) from the Research Grant Council of Hong Kong, the National Basic Research Program of China (2011CB504004 and 2010CB945500), Young Scientists Fund of National Natural Science Foundation (81200292), and Theme-Based Research Scheme grant T12-705/11 from the Research Grant Council of Hong Kong.

  • See Editorial on Page 792

Abstract

Acetaminophen (APAP) overdose is a leading cause of drug-induced hepatotoxicity and acute liver failure worldwide, but its pathophysiology remains incompletely understood. Fibroblast growth factor 21 (FGF21) is a hepatocyte-secreted hormone with pleiotropic effects on glucose and lipid metabolism. This study aimed to investigate the pathophysiological role of FGF21 in APAP-induced hepatotoxicity in mice. In response to APAP overdose, both hepatic expression and circulating levels of FGF21 in mice were dramatically increased as early as 3 hours, prior to elevations of the liver injury markers alanine aminotransferase (ALT) and aspartate aminotransferase (AST). APAP overdose-induced liver damage and mortality in FGF21 knockout (KO) mice were markedly aggravated, which was accompanied by increased oxidative stress and impaired antioxidant capacities as compared to wild-type (WT) littermates. By contrast, replenishment of recombinant FGF21 largely reversed APAP-induced hepatic oxidative stress and liver injury in FGF21 KO mice. Mechanistically, FGF21 induced hepatic expression of peroxisome proliferator-activated receptor coactivator protein-1α (PGC-1α), thereby increasing the nuclear abundance of nuclear factor erythroid 2-related factor 2 (Nrf2) and subsequent up-regulation of several antioxidant genes. The beneficial effects of recombinant FGF21 on up-regulation of Nrf2 and antioxidant genes and alleviation of APAP-induced oxidative stress and liver injury were largely abolished by adenovirus-mediated knockdown of hepatic PGC-1α expression, whereas overexpression of PGC-1α was sufficient to counteract the increased susceptibility of FGF21 KO mice to APAP-induced hepatotoxicity. Conclusion: The marked elevation of FGF21 by APAP overdose may represent a compensatory mechanism to protect against the drug-induced hepatotoxicity, by enhancing PGC-1α/Nrf2-mediated antioxidant capacity in the liver. (Hepatology 2014;60:977–989)

Abbreviations
Aco

acyl-coenzyme A oxidase

ALI

acute liver injury

ALT

alanine transaminase

AST

aspartate aminotransferase

APAP

acetaminophen

CYP2E1

cytochrome P450 2E1

CYP4A10

cytochrome P450 4A10

FGF21

fibroblast growth factor 21

G6PDH

glucose-6-phosphate dehydrogenase

GCL-c

γ-glutamylcysteine ligase catalytic subunit

Gpx-1

glutathione peroxidase-1

GSH

glutathione

JNK

c-Jun N-terminal kinase

NAC

N-acetylcysteine

NAPQI

N-acetyl-p-benzoquinone imine

Nox-2

NADPH oxidase-2

Nrf2

nuclear factor erythroid 2-related factor 2

PGC-1α

peroxisome proliferator-activated receptor coactivator protein-1α

Q-PCR

quantitative real-time polymerase chain reaction

ROS

reactive oxidation species

Sod-2

superoxide dismutase-2

TUNEL

terminal deoxynucleotidyl transferase dUTP nick end labeling

Acetaminophen (APAP) is a commonly used over-the-counter analgesic drug with an excellent safety profile when administered in proper therapeutic doses. However, overdose of APAP can cause acute and severe liver injury. In many developed countries, APAP-induced hepatotoxicity has replaced viral hepatitis as the most common cause of acute hepatic failure and is the second most common cause of liver failure requiring transplantation.[1] Mechanistically, APAP is metabolized by the cytochrome P450 system into its reactive intermediate N-acetyl-p-benzoquinone imine (NAPQI), which is then detoxified by conjugation with glutathione (GSH).[2] Therefore, overdose with APAP depletes cellular GSH, consequently leading to the covalent binding of NAPQI to mitochondrial proteins, which in turn triggers mitochondrial dysfunction, adenosine triphosphate (ATP) depletion, oxidative stress, increased phosphorylation, and mitochondrial translocation of c-Jun-N-terminal kinase (JNK) and DNA damage.[2, 3] This is followed by opening of the mitochondrial membrane permeability transition pore with collapse of the membrane potential, consequently leading to massive necrosis of hepatocytes and subsequent release of death-associated molecular pattern molecules such as high-mobility group box 1 protein and heat shock protein 70, which can activate Kupffer cells and neutrophils for production of proinflammatory cytokines to further exacerbate liver damage.[4-6] However, despite extensive studies in the past decade, the pathophysiological responses to APAP overdose remains poorly understood. The therapeutic options for APAP-induced hepatotoxicity are rather limited at this stage.

Fibroblast growth factor 21 (FGF21) is a metabolic hormone secreted predominantly by hepatocytes.[7] Unlike the classical members of the FGF family, FGF21 does not possess heparin-binding properties, enabling it to be released into the circulation.[8] A growing body of evidence from animal studies has demonstrated FGF21 as an important metabolic regulator with pleiotropic effects on glucose and lipid homeostasis.[7, 9] Therapeutic administration of recombinant FGF21 exerts a variety of favorable effects in both obese rodents and nonhuman primates, including reduction of body weight, alleviation of hyperglycemia and insulin resistance, improvement in lipid profiles, and hepatic steatosis.[10, 11] In addition, FGF21 is an important regulator of ketogenesis, gluconeogenesis, and growth hormone resistance in the liver.[11-13]

Despite its multiple metabolic benefits on glucose and lipid homeostasis, circulating levels of FGF21 are elevated in obese subjects and patients with type 2 diabetes.[14, 15] A modest, but significant, elevation of serum FGF21 has also been observed in patients with nonalcoholic fatty liver disease (NAFLD).[16] Several animal studies have demonstrated that FGF21 deficiency increases the susceptibility of mice to cerulein-induced pancreatitis[17] and toxicity of sepsis,[18] suggesting a potentially beneficial effect of FGF21 on acute organ injury. However, the role of FGF21 in APAP-induced acute liver failure has never been explored so far.

In the present study we investigated the dynamic changes in both circulating levels and hepatic expression of FGF21 in response to APAP overdose. Our results demonstrated that APAP overdose induced a rapid and dramatic elevation of both hepatic production and circulating concentrations of FGF21, prior to the occurrence of liver injury. Functionally, we found that FGF21 protected mice from APAP-elicited hepatotoxicity and mortality by induction of peroxisome proliferator-activated receptor coactivator protein-1α (PGC-1α), thereby enhancing the antioxidant capacity in the liver.

Materials and Methods

Animal Studies

Male FGF21 knockout (KO) mice[19] and wild-type (WT) controls in C57BL/6J background were used for this study. To induce APAP hepatotoxicity, freshly prepared APAP solution (Sigma-Aldrich) was administered into mice by intraperitoneal injection at either 500 or 750 mg/kg. All animal experimental procedures were approved by the Committee on the Use of Live Animals for Teaching and Research of the University of Hong Kong and were carried out in accordance with the NIH Guide for the Care and Use of Laboratory Animals.

Histological Analysis, Immunohistochemistry, Immunoassays, and Biochemical Analysis for Serum Samples

Liver tissues were fixed in 10% formalin solution. Hematoxylin-eosin staining and immunohistochemistry were conducted as described elsewhere.[20] Circulating concentrations of FGF21 were determined with an immunoassay (Antibody and Immunoassay Services, University of Hong Kong). Serum ALT and AST levels were measured with their enzymatic assay kits (Sigma-Aldrich, St. Louis, MO).

Measurement of Mitochondrial Number and Mitochondrial Respiratory Chain (MRC) Activities

Mitochondria were isolated from freshly harvested mouse liver tissues as described previously.[21] The mitochondrial DNA copy number and MRC activities were measured as described in the Supporting Information.

Measurement of Reactive Oxygen Species (ROS) in Liver Tissues

ROS levels in liver tissues was determined by lucigenin-enhanced chemiluminescence, as described previously,[20] or by 5-(and-6)-chloromethyl-2′,7′-dichlorodihy-drofluorescein diacetate as a fluorescent probe as described in the Supporting Information. Detailed Materials and Methods are provided in the Supporting Information.

Results

Circulating Levels and Hepatic Expression of FGF21 Are Dramatically Increased at the Very Early Stage During APAP-Induced Hepatotoxicity

The liver is the main contributor to circulating FGF21.[22, 23] To evaluate the effects of APAP on FGF21 production, we measured the dynamic changes of circulating FGF21 levels and its hepatic expression after administration of a single dose of the hepatotoxic drug (500 mg/kg body weight) in C57 BL/6J mice. Serum FGF21 levels increased significantly at 3 hours after APAP injection, reached its peak level at 6 hours, and then declined progressively to its basal levels at 24 hours (Fig. 1A). When calculated as fold changes relative to the baseline, serum FGF21 levels at 6 hours after APAP treatment exhibited an ∼120-fold elevation (Fig. 1B). The marked APAP-induced elevation in serum FGF21 levels was accompanied by a dramatic increase in both messenger RNA (mRNA) and protein expression of FGF21 in the liver, as determined by quantitative real-time polymerase chain reaction (Q-PCR) (Fig. 1C) and immunostaining (Fig. 1D), respectively. By contrast, serum ALT levels started to increase only after serum FGF21 levels declined to its baseline levels after APAP treatment (Fig. 1B). Further tissue profiling analysis demonstrated that APAP-induced marked elevation of FGF21 expression occurred specifically in the liver, but not in pancreas, heart, kidney, epididymal white and brown adipose tissue, muscle, and brain (Supporting Fig. 1), suggesting that increased de novo biosynthesis in the liver is the major contributor to the markedly elevated serum FGF21 levels in APAP-treated mice. When C57 BL/6J mice were intramuscularly injected with a myotoxic dose of bupivacaine to induce severe damage in skeletal muscle but not in the liver, both serum FGF21 levels and FGF21 mRNA expression in skeletal muscle and liver remained unchanged (Supporting Fig. 2A-C). On the other hand, both serum ALT and AST levels in BPVC-treated mice were significantly elevated compared to the vehicle-treated group (Supporting Fig. 2B). Taken together, these findings suggest serum FGF21 as an early and specific biomarker of APAP-induced liver injury.

Figure 1.

Temporal changes in serum levels of FGF21, ALT, and hepatic FGF21 expression in APAP-treated mice. Eight-week-old male C57 BL/6J mice were intraperitoneally injected with either APAP (500 mg/kg body weight) or equal volume of phosphate-buffered saline (PBS) as vehicle control. (A) Serum levels of FGF21 determined at various timepoints after APAP treatment. (B) Fold changes in serum FGF21 and ALT in APAP-treated group relative to their baseline levels. (C) Q-PCR analysis for FGF21 mRNA expression in liver tissues harvested at various timepoints after APAP treatment. (D) Representative images of liver sections stained with anti-FGF21 antibody, harvested from mice at different timepoints after receiving APAP treatment (original magnification: 400×). Quantitative data are expressed as mean ± SEM, n = 6-8. *P < 0.05; **P < 0.01 versus vehicle-treated group.

Figure 2.

FGF21 deficiency exacerbates APAP-induced hepatotoxicity and mortality in mice. (A-D) Eight-week-old male FGF21 WT and age-matched FGF21 (KO) mice were intraperitoneally injected with either APAP (500 mg/kg body weight) or an equal volume of PBS as vehicle control. Serum levels of ALT (A) and AST (B) measured at various timepoints. (C,D) Representative images of H&E-staining and terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining, respectively, in liver sections harvested at 12 hours post-APAP injection. (D, right) Semiquantification of the number of TUNEL-positive cells per field. Data are expressed as mean ± SEM, n = 6-8. *P < 0.05; **P < 0.01 versus WT group. (E) The survival rate of FGF21 KO mice and WT littermates at different timepoints after injection with a lethal dose of APAP (750 mg/kg body weight). n = 9-13 per group, P < 0.05.

APAP-evoked elevation of FGF21 production was accompanied by a significant up-regulation of several key genes involved in gluconeogenesis (glucose-6-phosphatase and phosphoenolpyruvate carboxykinase) and fatty acid metabolism (lipoprotein lipase, pancreatic lipase-related protein 2, carboxyl ester lipase, and β-subunit of ATP synthase) (Supporting Fig. 3A). By contrast, the expression of peroxisome proliferator activated receptor α (PPAR-α) and its downstream target genes angiopoietin-like protein 4 (Angptl4) and medium-chain acyl-CoA dehydrogenase (Mcad) was not altered in response to APAP challenge (Supporting Fig. 3B). The expression of FGF2 and FGF10 in the liver was very low as compared to the brain, and was not induced by APAP overdose (Supporting Fig. 3C). Likewise, APAP overdose had no effects on hepatic expression of FGF15 and FGF23 (Supporting Fig. 3D), or other members of the FGF superfamily (data not shown).

Figure 3.

FGF21 deficiency enhances APAP-induced hepatic ROS accumulation and causes impaired expression of antioxidant genes. Eight-week-old male FGF21 KO mice and WT littermates were intraperitoneally injected with either APAP (500 mg/kg body weight) or equal volume of PBS as vehicle control. Liver tissues were harvested at 6 hours after injection. (A) Hepatic production of ROS was determined by lucigenin-enhanced chemiluminescence and expressed as fold change relative to vehicle-treated WT group. (B) Representative images of dihydroethidium staining in liver sections (left) and semiquantitative analysis of fluorescent intensity per field (right). (C) Malondialdehyde (MDA) contents determined in liver tissues. (D) Protein abundance of phosphorylated JNK (p-JNK) in mitochondrial fraction was determined by western blot analysis (upper) and quantified by densitometry (lower). The mitochondrial protein COX IV was used as a loading control. Hepatic mRNA expression levels of ROS-generating genes (acyl-coenzyme A oxidase [Aco], cytochrome P450 2E1 [CYP2E1], cytochrome P450 4A10 [CYP4A10], NADPH oxidase-2 [Nox2], E), and antioxidant genes (γ-glutamylcysteine ligase catalytic subunit [GCL-c], glutathione peroxidase-1 [Gpx-1], superoxide dismutase-2 [Sod2], glucose-6-phosphate dehydrogenase [G6pdh], F) were determined by Q-PCR. Data are expressed as mean ± SEM, n = 6-8. *P < 0.05 versus WT group.

FGF21 Deficiency Enhances APAP-Evoked Hepatotoxicity and Mortality in Mice

To investigate the roles of FGF21 in APAP-induced acute liver injury, we performed serological and histological comparison between FGF21 KO mice and WT littermates treated with the same dose of APAP (500 mg/kg). As expected, serum levels of both ALT and AST in WT mice were elevated progressively after APAP administration (Fig 2A,B). Notably, the amplitude of APAP-induced elevation of these two markers of liver injury in FGF21 KO mice was substantially higher than in WT littermates. Histological analysis of liver sections revealed that the extent of APAP-induced necrosis in FGF21 KO mice was much more severe than that in WT controls (Fig. 2C). Likewise, DNA fragmentation, a characteristic feature of APAP-induced hepatocyte death,[4] was further exacerbated in APAP-treated FGF21 KO mice as compared to the WT controls (Fig. 2D).

To test the impact of FGF21 deficiency on APAP-induced mortality, both FGF21 WT and KO mice were intraperitoneally injected with a lethal dose of APAP (750 mg/kg) and the survival of these mice was monitored every 3 hours until 48 hours postinjection. The mortality rate in APAP-treated FGF21 KO mice was much higher than APAP-treated WT controls throughout the observation period (Fig 2E). At 48 hours after APAP injection, over 60% WT mice remained alive, whereas the survival rate of FGF21 KO mice was less than 30%, suggesting that FGF21 plays a protective role in APAP-induced hepatotoxicity and mortality.

FGF21 Deficiency Markedly Increases APAP-Induced Hepatic ROS Accumulation Due to Impaired Antioxidant System

To decipher the molecular basis by which FGF21 deficiency exacerbates APAP-induced hepatotoxicity, we measured GSH and ROS levels in the liver tissues of FGF21 KO mice and WT littermates treated with or without APAP. In both groups of mice, hepatic GSH was almost completely depleted within 2 hours after injection with APAP. FGF21 deficiency had no obvious effect in the magnitude of APAP-induced decline in cellular GSH levels (Supporting Fig. 4). On the other hand, APAP-induced elevations of hepatic ROS production and lipid peroxidation were obviously augmented in FGF21 KO mice when compared to those in WT littermates (Fig. 3A-C). These changes in FGF21 KO mice were accompanied by increased mitochondrial abundance of phosphorylated c-Jun N-terminal kinase (JNK) (Fig. 3D), which has been documented to amplify APAP-induced mitochondrial oxidative stress.[24]

Figure 4.

Replenishment of recombinant FGF21 reversed the increased susceptibility of FGF21 KO mice to APAP-induced hepatotoxicity. Eight-week-old male FGF21 KO mice were intraperitoneally injected with APAP (500 mg/kg body weight) or PBS as vehicle control, followed by intraperitoneal administration of recombinant mouse FGF21 (rmFGF21, 2 mg/kg body weight) at 2 hours after APAP treatment. (A) Serum levels of ALT (left) and AST (right) measured at various timepoints as indicated. n = 6-8, *P < 0.05 versus APAP group. (B) Representative images of H&E staining in liver tissues collected at 12 hours post-APAP injection (original magnification: 100×). (C) Survival rate of FGF21 KO mice receiving a lethal dose of APAP treatment (750 mg/kg), followed by intraperitoneal injection with either rmFGF21 (2 mg/kg) or PBS as vehicle control at 2 hours post-APAP injection. Data are expressed as percentage of the survival mice in each timepoint. n = 12-13 per group. P < 0.05. (D) The levels of ROS in liver tissues collected at 6 hours post-APAP injection were determined by lucigenin-enhanced chemiluminescence (left) and dihydroethidium staining (right), respectively. (E) Western blot analysis of phosphorylated JNK in the mitochondrial fraction isolated from liver tissues harvested at 6 hours post-APAP injection. (F) Hepatic mRNA expression levels of antioxidant genes were determined by Q-PCR. Data are expressed as mean ± SEM, n = 6-8. *P < 0.05; **P < 0.01.

As excessive ROS accumulation occurs as a result of the imbalance between ROS generating and scavenging systems, we next compared the hepatic expression levels of several key genes involved in both processes between FGF21 KO mice and WT littermates with Q-PCR analysis. Consistent with a previous report,[25] the expressions of both prooxidant genes (acyl-coenzyme A oxidase, cytochrome P450 2E1, cytochrome P450 4A10, NADPH oxidase-2), and antioxidant genes (γ-glutamylcysteine ligase catalytic subunit, glutathione peroxidase-1, superoxide dismutase-2, glucose-6-phosphate dehydrogenase) in WT littermates were significantly up-regulated in response to APAP challenge (Fig. 3E,F). FGF21 deficiency had no obvious impact on the mRNA abundance of the prooxidant genes (Fig. 3E), but led to a marked impairment in APAP-induced expression of the aforementioned antioxidant genes (Fig. 3F), suggesting that increased hepatic oxidative stress may be attributed to the defective ROS scavenging system in APAP-treated FGF21 KO mice.

Replenishment of FGF21 Reverses the Increased Susceptibility of FGF21 KO Mice to APAP-Induced Hepatotoxicity by Restoring the Antioxidant Activity

To further ascertain the roles of FGF21 in APAP-induced hepatotoxicity, we next investigated whether supplementation of FGF21 can rescue the defective antioxidant system and severe liver injury observed in APAP-treated FGF21 KO mice. To this end, FGF21 KO mice were treated with APAP, followed by intravenous injection with recombinant mouse FGF21 (rmFGF21) (2 mg/kg body weight). Circulating FGF21 started to rise immediately after administration with rmFGF21, peaked at 1.5 hour, and was sustained at high levels until 6 hours (Supporting Fig. 5). Treatment with rmFGF21 significantly protected FGF21 KO mice from APAP-induced liver injury, as evidenced by a markedly attenuated elevation in serum ALT and AST levels and an obvious reduction in liver necrosis (Fig. 4A,B). Furthermore, the mortality rate of FGF21 KO mice induced by a lethal dose of APAP was significantly reduced by administration of rmFGF21 (Fig. 4C). At the molecular level, replenishment of rmFGF21 dramatically decreased APAP-induced hepatic ROS accumulation (Fig. 4D) and mitochondrial abundance of activated JNK (Fig. 4E), but significantly increased the expression of several ROS-scavenging genes (Fig. 4F), suggesting that FGF21 treatment prevents APAP-induced hepatotoxicity by restoring the antioxidant activities in the liver.

Figure 5.

FGF21 induces PGC-1α expression and nuclear abundance of Nrf2 in the liver of APAP-treated mice. (A,B) Eight-week-old male FGF21 KO mice or WT littermates were intraperitoneally injected with either APAP (500 mg/kg body weight) or PBS as vehicle control. Protein abundance of Nrf2 in nuclear fraction and PGC-1α in total lysate of the livers collected at 6 hours after APAP injection was determined by western blot analysis. The bar charts in the lower panels represent the densitometric quantification of Nrf2 (A) and PGC-1α (B) relative to histone and β-actin, respectively. (C,D) Western blot analysis for protein levels of PGC-1α in total lysate and Nrf2 in nuclear fractions in the livers of FGF21 KO mice receiving APAP treatment, followed by intraperitoneal injection of rmFGF21 (2 mg/kg) as in Fig. 4. The bar charts in the lower panel are densitometric analysis as in (A,B). Data are expressed as mean ± SEM, n = 6-8. *P < 0.05.

APAP overdose reduced both mitochondrial DNA copy number and activities of MRC complexes (I, II+III, IV, and V) in WT mice, and these changes were further aggravated in FGF21 KO mice. Conversely, treatment with rmFGF21 significantly reversed APAP-induced reduction of mitochondrial DNA copy number and MRC complex activities in both FGF21 KO and WT mice (Supporting 6A-E).

APAP-Induced Hepatic Expression of PGC-1α and Nuclear Accumulation of Nrf2 Is Mediated by FGF21 in Mice

Nuclear factor erythroid 2-related factor 2 (Nrf2) is a master regulator for transcriptional activation of a wide spectrum of genes related to the antioxidant defense system.[26] Notably, Nrf2 is under the direct control of PGC-1α,[27, 28] a transcriptional coactivator that has been shown to be induced by FGF21.11,29 We therefore compared the hepatic levels of Nrf2 and PGC-1α between FGF21 KO mice and WT littermates. Both mRNA and protein levels of Nrf2 and PGC-1α were progressively increased by APAP treatment (Fig. 5A,B; Supporting Fig. 7). The maximum induction of both PGC-1α and Nrf2 occurred at 6 hours, when circulating FGF21 reached its peak levels after APAP treatment (Fig. 1A). Notably, APAP-induced hepatic expression of both PGC-1α and Nrf2 were markedly attenuated in FGF21 KO mice (Fig. 5A,B; Supporting Fig. 7). In contrast, replenishment of rmFGF21 into APAP-treated FGF21 KO mice resulted in a significant elevation in hepatic expression of PGC-1α and Nrf2 (Fig. 5C,D), suggesting that FGF21 acts as an upstream inducer of these two transcriptional regulators in response to APAP challenge.

PGC-1α Confers Protective Effects of FGF21 Against APAP-Induced Hepatotoxicity

To investigate whether induction of PGC-1α is required for the hepatoprotective effects of FGF21 against APAP-induced liver injury, we next treated FGF21 KO mice with adenovirus expressing small interfering RNA (siRNA) against PGC-1α (siPGC-1α) or scramble control by tail vein injection (5 × 108 plaque-forming units [p.f.u.]/mouse) for 7 days, followed by APAP treatment and subsequent administration of rmFGF21. The expression level of PGC-1α in the liver tissue was decreased by 80% as compared to scramble control in FGF21 KO mice after intraperitoneal injection of rmFGF21 for 4 hours (Supporting Fig. 8A). Notably, APAP-induced elevations in serum levels of the liver injury enzymes (ALT and AST, Fig. 6A) and necrosis area in hematoxylin and eosin (H&E)-stained liver sections (Fig. 6B) in siPGC-1α-treated mice were significantly higher than those in the scramble control group. The antioxidant effect of rmFGF21 was also dramatically abolished in siPGC-1α-treated KO mice, as demonstrated by the markedly increased ROS accumulation in the liver (Fig. 6C,D) and enhanced phosphorylation of mitochondrial JNK (Fig. 6E). Furthermore, rmFGF21-induced expression of Nrf2 (Fig. 6E, right) and several antioxidant genes (Fig. 6F) were all abrogated in siPGC-1α-treated mice. Similarly, adenovirus-mediated knockdown of PGC-1α in WT mice significantly augmented APAP-induced liver injury, oxidative stress, and JNK activation, which was accompanied by impaired compensatory up-regulation of the antioxidant genes (Fig. 6A-F), further supporting that induction of the Nrf2/antioxidant pathway is the downstream event of PGC-1α in response to FGF21 stimulation.

Figure 6.

siRNA-mediated knockdown of PGC-1α in the liver markedly blunts the protective effects of FGF21 on APAP-induced hepatotoxicity. Eight-week-old male FGF21 KO mice and age- and sex-matched WT littermates were intravenously injected with recombinant adenovirus expressing siRNA against PGC-1α (siPGC-1α) or scramble (scr) (5 × 108 p.f.u./mouse). At day 7 after adenoviral injection, mice were intraperitoneally injected with APAP (500 mg/kg body weight) and subsequently treated with rmFGF21 administration (2 mg/kg body weight) as in Fig. 4. (A) Serum levels of ALT and AST were measured at different timepoints after APAP injection. *P < 0.01 versus KO scramble groups; #P < 0.01 versus WT scramble groups. The liver tissues harvested at 6 hours after APAP injection were subjected to H&E staining to visualize necrosis areas (B, original magnification of 100×); measurement of ROS accumulation with luminescence-based assays (C), dihydroethidium staining (D), western blot analysis for mitochondrial abundance of phosphorylated JNK (E, left), nuclear enrichment of Nrf2 (E, right), and Q-PCR analysis for expression levels of the antioxidant genes (F), respectively. Data are expressed as mean ± SEM, n = 6-8. *P < 0.05, **P < 0.01.

Figure 7.

Adenovirus-mediated hepatic overexpression of PGC-1α reverses the increased susceptibility of FGF21 KO mice to APAP-induced hepatotoxicity. Recombinant adenovirus encoding PGC-1α or luciferase (luc) were infused into 8-week-old male FGF21 KO mice or WT littermates by tail vein injection (5 × 108 p.f.u./mouse). At day 7 postadenovirus injection, mice were intraperitoneally injected with APAP (500 mg/kg body weight). (A) Serum levels of ALT and AST measured at different timepoints. *P < 0.05 versus luciferase-treated KO group. (B) Representative images of H&E-stained liver sections collected at 12 hours after APAP injection (original magnification: 100×). Hepatic levels of ROS determined by lucigenin-enhanced chemiluminescence (C) and dihydroethidium staining (D, original magnification: 400×) in liver tissues collected at 6 hours post-APAP injection. (E) Representative western blots of p-JNK in mitochondrial extracts (left) and Nrf2 in nuclear fraction (right) from liver tissues harvested at 6 hours post-APAP injection. Histograms under each blot represent the densitometric quantification of relative abundance of each target protein. (F) The mRNA expression levels of the four antioxidant genes determined by Q-PCR. Data are expressed as mean ± SEM, n = 6-8. *P < 0.05, **P < 0.01.

We next investigated whether overexpression of PGC-1α is sufficient to counteract the exacerbation of APAP-induced liver injury by FGF21 deficiency in mice. To this end, FGF21 KO mice or WT littermates were administered recombinant adenovirus expressing PGC-1α or luciferase control by tail-vein injection (5 × 108 p.f.u./mouse), followed by APAP treatment at 7 days after adenoviral infection. The hepatic expression of PGC-1α in FGF21 KO mice infected with PGC-1α-expressing adenovirus was elevated by ∼4-fold compared to those mice infected with luciferase-expressing adenovirus (Supporting Fig. 8B). The adenovirus-mediated ectopic expression of PGC-1α led to a marked attenuation of APAP-induced elevations of serum ALT and AST levels and a substantial reduction of necrosis to a level comparable to WT littermates (Fig. 7A,B). Likewise, APAP-induced increase in hepatic ROS levels (Fig. 7C,D) and mitochondrial accumulation of phosphorylated JNK (Fig. 7E, left) in PGC-1α-expressing adenovirus-treated FGF21 KO mice were largely abrogated when compared with those in FGF21 KO mice receiving luciferase-expressing adenoviruses. Furthermore, these changes caused by adenovirus-medicated PGC-1α overexpression were accompanied by markedly increased nuclear abundance of Nrf2 (Fig. 7E, right), as well as the up-regulation of the antioxidant genes (Fig. 7F), suggesting that overexpression of PGC-1α is sufficient to reverse the increased susceptibility of FGF21 KO mice to APAP-induced hepatotoxicity. Adenovirus-mediated overexpression of PGC-1α in WT mice also led to a mild alleviation in APAP-induced liver injury and oxidative stress as well as a higher expression level of the antioxidant genes as compared to those WT mice with adenoviral expression of luciferase control (Fig. 7A-F).

Discussion

In the present study we found that APAP overdose causes a drastic elevation in both circulating levels and hepatic expression of FGF21, which in turn acts as a feedback signal to protect mice from APAP-induced hepatotoxicity by inducing the antioxidant capacity in the liver (Supporting Fig. 9). We obtained several lines of evidence demonstrating that APAP-evoked elevation of FGF21 represents a compensatory response for the liver to defend against APAP-induced liver damage. On the one hand, the obvious liver damage (as determined by serum ALT levels) occurs only after the dramatic elevation of serum FGF21 is markedly attenuated in APAP-treated mice, suggesting that the massive onset of cell death is attributed in part to the decompensated elevation of serum FGF21. On the other hand, mice lacking FGF21 exhibit a markedly exacerbated liver injury and increased mortality in response to APAP overdose. Conversely, the increased susceptibility of FGF21 KO mice to APAP-induced hepatotoxicity is reversed by the replenishment with rmFGF21. In line with our findings, up-regulated FGF21 expression in acinar cells has been shown to protect pancreatic acini from overt damage in mice with cerulein-induced pancreatitis.[17] Furthermore, therapeutic administration of exogenous FGF21 also protects ob/ob obese mice from endotoxin- and cecal ligation-induced toxicity of sepsis.[18]

Hepatic FGF21 expression under the nutrient-deficient conditions is under the control of nuclear receptor PPAR-α.[12] However, our results showed that the hepatic expression of PPARα and its well-established downstream target genes Angptl4 and Mcad were not significantly altered in C57 mice in response to APAP treatment (Supporting Fig. 3B), suggesting that PPARα is unlikely to be a key mediator of APAP-induced expression of FGF21. The precise nature of the transcriptional factor(s) conferring APAP-induced transactivation of the FGF21 gene remains to be identified.

Hepatic oxidative stress triggered by the binding of the APAP metabolite NAPQI to mitochondrial proteins is a central mediator of APAP-induced acute liver injury.[6] Excessive accumulation of both ROS and peroxynitrite leads to structural alterations of mitochondrial proteins and DNA and the cessation of ATP production, thereby inducing necrosis.[30, 31] Furthermore, ROS can induce activation and mitochondrial translocation of JNK, which in turn further exacerbates mitochondrial oxidative stress and permeability transition.[32] Treatment with either antioxidants or pharmacological inhibitors of JNK can alleviate APAP-induced liver injury in animals.[31, 33] Likewise, our present study demonstrated that the protective effects of FGF21 against APAP-induced hepatotoxicity is largely attributed to its ability to prevent hepatic oxidative stress, as evidenced by the fact that APAP-induced ROS accumulation, lipid peroxidation, and mitochondrial translocation of activated JNK in the liver tissues are augmented by FGF21 deficiency, but are blocked by replenishment of recombinant FGF21. Furthermore, our data suggest that the alleviation of hepatic oxidative stress by FGF21 is unrelated to the prooxidant system, but due to a marked elevation in the expression of a cluster of antioxidant enzymes.

In response to oxidative stress, the antioxidant defense system is often activated as a compensatory response to protect cell damage by maintaining cellular redox homeostasis.[26] The nuclear factor Nrf2, a transcription factor essential for activation of the antioxidant defense system, has been shown to be actively involved in scavenging of APAP overdose-induced ROS in mice.[26] Upon activation by oxidants, Nrf2 translocates from cytosol to nuclei, where it is recruited to the antioxidant responsive element sequence (TGACnnnGC) in the regulatory region of its target genes as a heterodimer with small Maf proteins.[34] Mice lacking Nrf2 are more susceptible to APAP-induced hepatotoxicity compared to WT littermates.[35] In contrast, constitutive activation of hepatic Nrf2 by genetic disruption of the Nrf2 endogenous inhibitor keap1 gene confers mice with significant resistance to APAP-induced liver damage by inducing the expression of antioxidant genes.[36] In line with these reports, our present study showed that APAP overdose-induced nuclear accumulation of Nrf2 and expression of a cluster of antioxidant enzymes is impaired in FGF21 KO mice, whereas replenishment of recombinant FGF21 alone is sufficient to reverse these changes, suggesting that the compensatory up-regulation of Nrf2-mediated expression of antioxidant genes is induced at least in part by FGF21 in response to APAP overdose.

PGC-1α is a transcriptional coactivator that regulates gene expression by interacting with other transcription factors, especially peroxisome proliferator-activated receptor γ.[37] Notably, PGC-1α can coordinate the expression of many antioxidant genes,[28] possibly by inducing the expression of Nrf2 as well as by coactivating with Nrf2 for the transcriptional activation of its target genes through an unknown pathway.[38] In this study, we provided several lines of evidence demonstrating that induction of PGC-1α by FGF21 is an early event obligatory for FGF21-mediated antioxidant capacity and subsequent protection against APAP-induced liver injury. First, the time course of APAP-induced PGC-1α expression matches well with those of FGF21, both of which reach the maximum at 6 hours followed by a progressive decline (Supporting Fig. 7). Second, APAP-induced expression of PGC-1α is abrogated in FGF21 KO mice, whereas treatment with recombinant FGF21 alone is sufficient to restore hepatic expression of PGC-1α. Third, the therapeutic benefits of recombinant FGF21 on elevation of nuclear Nrf2 abundance, induction of antioxidant genes, and consequent alleviation of oxidative stress and hepatotoxicity are largely abrogated by siRNA-mediated knockdown of PGC-1α expression in the liver of APAP-treated mice. Conversely, adenovirus-mediated hepatic overexpression of PGC-1α is able to prevent the impairments in APAP-induced nuclear accumulation of Nrf2 and expression of the antioxidant genes, and increased susceptibility to APAP-induced oxidative stress and liver damage in FGF21 KO mice. In support of our conclusion, induction of PGC-1α by FGF21 has been shown to play an indispensable role in mediating FGF21-induced gluconeogenesis and fatty acid oxidation in hepatocytes,[11] and mitochondrial biogenesis in adipocytes.[29] Collectively, these findings support the role of PGC-1α as an early effector that mediates the multiple biological actions of FGF21. However, both a previous study[11] and our results (Supporting Fig. 10) showed that rmFGF21 was unable to induce the expression of PGC-1α or its downstream target genes in isolated primary hepatocytes, suggesting that induction of hepatic PGC-1α by FGF21 may require an unknown factor(s) that are lost after separation of hepatocytes from the liver. It is also possible that FGF21 exerts its hepatic actions indirectly, by way of induction of other hepatoprotective factors. Further studies are warranted to delineate the detailed mechanisms underlying FGF21 actions in hepatocytes.

In summary, this study demonstrated that the drastic elevation of hepatic FGF21 production in response to APAP overdose is an early and obligatory step for the body to activate the antioxidant defense system for alleviation of the drug-induced oxidative stress and hepatotoxicity, by inducing PGC-1α expression, which in turn triggers Nrf2-mediated expression of a cluster of antioxidant genes (Supporting Fig. 9). FGF21 may represent not only a potential early diagnostic biomarker, but also a promising therapeutic agent for APAP-induced acute liver injury.

Acknowledgment

We thank Dr. Morichika Konishi and Dr. Nobuyuki Itoh (Kyoto University) for providing FGF21 KO mice, Dr. Jiandie Lin (University of Michigan) for providing the adenoviral vector for PGC-1α expression, and Dr. Macr Montiny (Salk Institute for Biological Studies) for providing adenoviral vectors encoding siRNA against PGC-1α.

Ancillary

Advertisement