Farnesoid X Receptor Antagonizes JNK Signaling Pathway in Liver Carcinogenesis by Activating SOD3

Abstract

The farnesoid X receptor (FXR) is a key metabolic and homeostatic regulator in the liver. In the present work, we identify a novel role of FXR in antagonizing c-Jun N-terminal kinase (JNK) signaling pathway in liver carcinogenesis by activating superoxide dismutase 3 (SOD3) transcription. Compared with wild-type mouse liver, FXR^−/− mouse liver showed elevated JNK phosphorylation. JNK1 deletion suppressed the increase of diethylnitrosamine-induced tumor number in FXR^−/− mice. These results suggest that JNK1 plays a key role in chemical-induced liver carcinogenesis in FXR^−/− mice. We found that ligand-activated FXR was able to alleviate H₂O₂ or tetradecanoylphorbol acetate-induced JNK phosphorylation in human hepatoblastoma (HepG2) cells or mouse primary hepatocytes. FXR ligand decreased H₂O₂-induced reactive oxygen species (ROS) levels in wild-type but not FXR^−/− mouse hepatocytes. FXR knockdown abolished the inhibition of 3-[2-[2-chloro-4-[[3-(2,6-dichlorophenyl)-5-(1-methylethyl)-4-isoxazolyl]methoxy]phenyl]ethenyl]-Benzoic acid (GW4064) on JNK phosphorylation and ROS production induced by H₂O₂ in HepG2 cells. The gene expression of SOD3, an antioxidant defense enzyme, was increased by FXR activation in vitro and in vivo. An FXR-responsive element, inverted repeat separated by 1 nucleotide in SOD3 promoter, was identified by a combination of transcriptional reporter assays, EMSAs, and chromatin immunoprecipitation assays, which indicated that SOD3 could be a direct FXR target gene. SOD3 knockdown abolished the inhibition of GW4064 on JNK phosphorylation induced by H₂O₂ in HepG2 cells. In summary, FXR may regulate SOD3 expression to suppress ROS production, resulting in decreasing JNK activity. These results suggest that FXR, as a novel JNK suppressor, may be an attractive therapeutic target for liver cancer treatment.

c-Jun N-terminal kinase (JNK), a member of the MAPK family, regulates gene expression in response to extracellular stimuli through the phosphorylation of various transcription factors. The JNK pathway is constitutively activated in many types of cancer, including most hepatocellular carcinomas (HCCs) (1). Antisense JNK oligonucleotides inhibit the growth of tumor cells and induce apoptosis (2). Thus, it has been proposed that inhibition of JNK may be beneficial for tumor therapy.

Farnesoid X receptor (FXR) (nuclear receptor subfamily 1, group H, member 4) is a member of the nuclear hormone receptor superfamily of ligand-activated transcription factors. The studies demonstrate that FXR is a central metabolic regulator in bile acid (BA) homeostasis as well as lipid and glucose metabolism (3–6). FXR coordinates the expression of genes involved in BA production, efflux, influx, and detoxification in the liver. Moreover, FXR activation alleviates the hepatotoxicity of hepatotoxins (7–9), protects against cholestatic liver injury in rats (10, 11), suppresses serum deprivation-induced liver cell apoptosis (12), and antagonizes nuclear factor-κB (NF-κB) pathway in hepatic inflammatory response (13). These studies suggest that a key function of FXR is for hepatoprotection, and FXR may suppress chronic inflammation to protect liver from carcinogenesis. FXR also stimulates normal liver regeneration after 70% hepatectomy by regulating Forkhead box protein M1b (9, 14). The hepatoprotective role of FXR is essential for normal liver physiology. For example, FXR null mice spontaneously develop liver tumors due to chronic liver injury when they age (15). Thus, these new findings suggest that FXR is more than a metabolic regulator, and FXR may be a potential drug target for the treatment of liver cancer.

BAs are potent endogenous agents for the induction of liver injury. BAs activate multiple cell signaling cascades, such as the JNK, ERK1/2,and phosphoinositol-3-OH kinase pathways (16, 17), to regulate cell survival and proliferation (18–20). In FXR^−/− mice, BA level is elevated compared with wild-type (WT) mice. FXR^−/− mice spontaneously develop liver tumors beyond 15 months of age (15). It is known that, by activating the JNK1 pathway, BAs can directly induce apoptosis in hepatocytes (21, 22). Because FXR is the primary BA sensor and provides hepatoprotection against BA-induced liver injury (11, 23), we hypothesize that FXR may inhibit JNK signaling and/or other pathways to reduce the toxic effect of BAs and suppress liver cancer development.

In the current study, we identify FXR as a negative regulator of JNK signaling pathway by activating superoxide dismutase 3 (SOD3) transcription and demonstrate that FXR ligands have potential to suppress JNK signaling pathway in liver cells. These findings suggest FXR may be a potential target for therapeutic intervention in human liver carcinogenesis, at least in part by suppressing JNK signaling pathway.

Materials and Methods

Materials

3-[2-[2-chloro-4-[[3-(2,6-dichlorophenyl)-5-(1-methylethyl)-4-isoxazolyl]methoxy]phenyl]ethenyl]-Benzoic acid (GW4064) was purchased from Tocris Bioscience. Phospho-/total-ERK1/2, phospho-/total-p38, phospho-/total-JNK1/2, and nuclear factor-κB p65 (p65) antibodies were purchased from Cell Signaling Technologies. β-Actin antibody was purchased from Santa Cruz Biotechnology, Inc. Chenodeoxycholic acid (CDCA), tetradecanoylphorbol acetate (TPA), H₂O_2, dichlorodihy-drofluorescein diacetate (DCFDA), and diethylnitrosamine (DEN) were purchased from Sigma Chemical.

Cell culture and experimental treatments

Human hepatoblastoma (HepG2) cells were grown in complete culture medium (high-glucose DMEM [with L-glutamine] supplied with 10% [vol/vol] inactivated fetal calf serum, and 1% [vol/vol] antibiotics-antimycotics). Cultures were fed with fresh medium twice weekly. For experiments, 6 × 10⁵ HepG2 cells were seeded in 60-mm culture dishes with complete culture medium. The next day, cells were treated with GW4064 (2μM) for 1 day. Then, cells were collected for quantity real-time PCR analysis. For protein assay, cells were pretreated with GW4064 for 1 day. Then, cells were treated with H₂O₂ for the indicated times. Finally, cells were collected for total protein isolation and Western blot analysis.

Primary hepatocyte culture

Mouse primary hepatocytes from 8-week-old mice were prepared as described previously (13). Hepatocytes were treated with GW4064 (3μM) for 24 hours before RNA isolation. For protein assay, hepatocytes were pretreated with GW4064 or CDCA for 1 day. Then, hepatocytes were treated with TPA or H₂O₂ for the indicated times. Finally, hepatocytes were collected for total protein isolation and Western blot analysis.

RNA isolation and quantitative real-time PCR

Total RNA was extracted from mouse hepatocytes, HepG2 cells and liver using TRI Reagent (Molecular Research Center, Inc). Quantitative real-time PCR was performed using the Power SYBR Green PCR Master Mix protocol (Applied Biosystems). Amplification of β-actin was used as an internal reference. β-Actin primers were obtained from Ambion. Quantitative PCR analysis was conducted using the ABI 7300 Sequence Detection System. Primer sequences are available on request.

PCR array

The Mouse Oxidative Stress and Antioxidant Defense RT² Profiler PCR Array was performed according to the manufacturer's protocol (SABiosciences). WT hepatocytes were isolated and treated with dimethyl sulfoxide (DMSO) (as control) or GW4064 (3μM) for 24 hours. Then, cells were collected for RNA isolation and PCR array.

Animals

The WT and FXR^−/− mice that have been extensively crossed to C57BL/6 background (24) were held in a pathogen-free animal facility under standard a 12-hour light, 12-hour dark cycle. FXR^−/−JNK1^−/− mice were generated by intercrossing FXR^−/− mice with JNK1^−/− mice (The Jackson Laboratory). Mice were fed standard rodent chow and water ad libitum. For DEN-induced liver cancer, 15-day-old WT, FXR^−/−, FXR^−/−JNK1^−/−, and JNK1^−/− mice were injected ip with a single dose of either vehicle or DEN (5 mg/kg body weight). Mice were fed a standard diet and killed 8 months after injection. Tumor number and tumor size were measured. For quantitative real-time PCR assay, WT and FXR^−/− mice were fed 0.2% cholic acid (CA) for 5 days.

Immunoblot analysis

At indicated time points after treatment, HepG2 cells or hepatocytes were lysed for 15 minutes with lysis buffer and centrifuged at 12 000g at 4°C for 15 minutes. Liver tissue extracts were prepared by homogenation in lysis buffer as described previously (25–27). The samples were resolved by 10% SDS-PAGE, transferred to nitrocellulose membranes, and blotted using first antibodies. The membranes were washed with Tris Buffered Saline with 0.1% Tween 20 and then incubated with antirabbit secondary antibody conjugated to horseradish peroxidase (1:5000) (Amersham Biosciences). Bands on blots were visualized using an enhanced chemiluminescence detection system (PerkinElmer Life Sciences) and quantified with a computerized digital imaging system using AlphaImager 2000 software (Alpha Innotech).

Detection of reactive oxygen species (ROS) production

The production of ROS was evaluated using DCFDA. DCFDA is nonfluorescent until hydrolyzed by esterases and oxidized by ROS in cells. After incubation, the cells were rinsed in PBS and consequently loaded by 1μM DCFDA for 45 minutes. Fluorescence values were measured on a CytoFluor 4000 using an excitation and emission wavelength 485 and 535 nm. The intensity value was expressed as the percent of initial value after subtracting out background fluorescence levels (cell fluorescence without the addition of DCFDA).

EMSA assay

Human FXR and retinoid X receptor (RXR) proteins were synthesized in vitro using a TNT quick coupled transcription/translation system according to the manufacturer's protocol (Promega). EMSAs were performed by incubating in vitro translated FXR and RXRα (1 μL each) with ³²P-labeled oligonucleotide duplex under standard binding conditions, and the DNA-protein complex was electrophoretically resolved on a 6% nondenaturing gel. For competition of the EMSA complex with the recombinant FXR and RXR, the unlabeled homologous or heterologous duplex oligonucleotides were added at 100- and 300-fold molar excess during preincubation.

Chromatin immunoprecipitation (ChIP) assay

ChIP assays for HepG2 cells were performed according to the protocol of a kit (Upstate) as reported previously (28). HepG2 cells were treated with DMSO (control group) or GW4064 (2μM) for 24 hours, and then the cells were collected for ChIP assay. The polyclonal anti-FXR antibody (sc-13063X; Santa Cruz Biotechnology, Inc) was used in the immunoprecipitations. IgG was used as a negative control. The final DNA extraction was amplified in a real-time PCR with a primer pair that covered the inverted repeat separated by 1 nucleotide (IR-1) sequence (5′-AAGTTATTGACTT-3′) from −330 to −110 with respect to the transcription start site in the human SOD3 promoter. Sequences of the primers used for ChIP assays are 5′-CTGCTCTGCTCTCATAGGTA-3′ (forward) and 5′-TTGTAGCCAGTGACCTTCAG-3′ (reverse).

FXR and SOD3 small interfering RNA (siRNA)

FXR or SOD3 siRNA and control siRNA were purchased from Santa Cruz Biotechnology, Inc and transfected into HepG2 cells using siRNA transfection reagent (Santa Cruz Biotechnology, Inc). Cell treatment is described in Supplemental Materials and Methods.

Statistics

All data represent at least 3 independent experiments and are expressed as the mean ± SEM. The Student's t test was used to calculate P values. P < .05 is considered significant.

Results

FXR^−/− mouse liver tissue display elevated JNK activity

FXR^−/− mice develop spontaneously liver tumor when they age (15), and FXR^−/− mouse liver shows elevated activity of different signaling pathways (13). BAs can activate the extracellular signal-regulated kinases ERK1/2, the c-Jun NH2-terminal kinases, JNK 1/2, and p38 signaling pathways (16, 17, 22, 29). In this work, firstly, we tested the effect of FXR deficiency on NF-κB signaling and the other 3 well-characterized subfamilies of MAPKs, which include ERK1/2, JNK1/2, and p38. We found elevated p65 protein levels and decreased phosphorylated (p)-ERK1/2 levels in FXR^−/− mouse livers compared with WT livers, which is consistent with our previous report (12, 13). It was observed that, compared with WT controls, livers from FXR^−/− mice showed strong elevated phosphorylation levels of JNK1/2 (Figure 1), but there is no difference for p38 activity between WT and FXR^−/− mouse livers. Furthermore, we measured the expression of factors upstream (phosphorylated stress signaling kinase) and downstream (p-c-JUN) from p-JNK1/2. The expression of phosphorylated stress signaling kinase did not change in these 2 group livers. Compared with WT livers, the expression of p-c-JUN was elevated in FXR^−/− mouse livers (Supplemental Figure 1). These results indicate that FXR may be a negative modulator of JNK signaling pathway.

Figure 1.

FXR^−/− mouse hepatic tissue display elevated JNK activity. Immunoblot analysis for phosphorylated ERK1/2 (P-ERK1/2), total ERK1/2, phosphorylated p38 (P-p38), total p38, phosphorylated JNK1/2 (P-JNK1/2), p65, and β-actin from total protein pools of WT and FXR^−/− individual mouse livers (n = 4).

JNK1 deletion suppresses HCC induced by DEN in FXR^−/− mice

In mouse HCC models, genetic disruption of the JNK1 locus substantially reduced the number and size of HCCs that were induced by DEN. In order to determine the potential link between FXR and JNK in liver carcinogenesis, we compared the HCC incidence in FXR^−/− and FXR^−/−JNK1^−/− littermates 8 months after DEN treatment on day 15 after birth. Both mouse lines developed tumors 8 months after DEN injection (Figure 2). However, the tumor number in FXR^−/−JNK1^−/− mice was significantly decreased compared with that in FXR^−/− mice, although the average tumor sizes were not significantly different between these 2 mouse lines. These results indicate that JNK1 deletion may suppress the initiation of tumors induced by DEN in FXR^−/− mice, suggesting that FXR may suppress JNK pathway to negatively regulate liver carcinogenesis.

Figure 2.

JNK1 deletion suppresses HCC induced by DEN in FXR^−/− mice. A, Livers of 9-month-old male WT, FXR^−/−JNK1^−/−, FXR^−/−, and JNK1^−/− mice receiving a single injection of a low dose of DEN (5 mg/kg) at 15 days of age. B, Numbers of tumors (≥0.5 mm) and maximal tumor sizes (diameters) in livers of male WT, FXR^−/−JNK1^−/−, FXR^−/−, and JNK1^−/− mice at 9 months after DEN (5 mg/kg) injection. *, P < .05.

FXR activation suppresses JNK activation in HepG2 cells and mouse hepatocytes

JNK was discovered almost 20 years ago as the protein kinase responsible for phosphorylating c-Jun at Ser⁶³ and Ser⁷³. These sites had previously been demonstrated to be essential for the c-Jun activity (30). To determine whether FXR is a suppressor of JNK signaling, we measured the effect of an FXR-specific ligand GW4064 on decreasing the levels of phosphorylated JNK. We used H₂O₂ (31), a known activator of JNK signaling pathway, to induce JNK signaling in HepG2 cells. Treatment with H₂O₂ resulted in higher phosphorylated JNK levels than that of controls in HepG2 cells (Figure 3A). JNK phosphorylation induced by H₂O₂ was suppressed by GW4064 treatment. We also used mouse hepatocytes to confirm these results. H₂O₂ and TPA (32, 33) both were used for activating JNK signaling. FXR activation induced by GW4064 or CDCA inhibited the phosphorylation of JNK induced by H₂O₂ or TPA (Figure 3, B and C). These results indicate that FXR activation can suppress JNK activities in both HepG2 cells and mouse primary hepatocytes. Knockdown of FXR abolished the inhibition of GW4064 on JNK activation induced by H₂O₂ in HepG2 cells (Supplemental Figure 2). These results indicate that the effects on JNK phosphorylation observed in HepG2 cells in response to GW4064 require FXR.

Figure 3.

FXR activation suppresses JNK phosphorylation in HepG2 cells and mouse hepatocytes. A, HepG2 cells were pretreated with GW4064 (2μM) for 24 hours. Cells were then treated with H₂O₂ (1mM) for the indicated times. Finally, cells were collected for total protein isolation and immunoblot analysis with antiphosphorylated JNK1/2 or anti-JNK1/2 antibodies. Con, control. B and C, WT hepatocytes were pretreated with 2 FXR ligands GW4064 (3μM) and CDCA (50μM) for 24 hours. Then, cells were treated with H₂O₂ (500μM) (B) or TPA (100nM) (C) for 60 minutes, respectively. Finally, cells were collected for total protein isolation and immunoblot analysis with antiphosphorylated JNK1/2 or anti-JNK1/2 antibodies (n = 3). Bands on blots were quantified with a computerized digital imaging system using AlphaImager 2000 software (Alpha Innotech). The expression levels of phosphorylated (p)-JNK1/2 were normalized by total JNK1/2. n = 3 refers to 3 independent experiments, each performed in triplicate.

FXR activation suppresses ROS production

It has been known that increased ROS levels are required for sustained JNK activation (34). Thus, we next measured ROS production in hepatocytes of WT and FXR^−/− mice. FXR^−/− mice display elevated ROS production compared with WT mice (Figure 4A). FXR activation by its ligand GW4064 decreased ROS production induced by H₂O₂ in WT hepatocytes (Figure 4B), which indicates that FXR activation suppresses JNK activation probably through decreasing ROS production. We also used HepG2 cells to confirm these results. FXR activation by GW4064 suppresses ROS production induced by H₂O₂ in HepG2 cells (Figure 4C). siRNA knockdown of FXR in HepG2 cells did not change ROS levels compared with the control group, whereas FXR knockdown abolished the suppression of FXR agonist GW4064 on ROS levels induced by H₂O₂. This suggests that GW4064 suppresses ROS production induced by H₂O₂ in an FXR-dependent manner.

Figure 4.

FXR activation suppresses ROS production. A, ROS levels in FXR^−/− (FXRKO) mouse hepatocytes compared with WT hepatocytes. B, WT hepatocytes were pretreated with GW4064 (3μM). After 24 hours, cells were treated with H₂O₂ (300μM) for 30 minutes. Cells were then collected for ROS production test (n = 3). *, P < .05. C, HepG2 cells were transfected with siRNA of FXR or control siRNA. Then, cells were treated with vehicle (DMSO) or GW4064 (2μM). After 24 hours, cells were treated with 1mM of H₂O₂ for 30 minutes for detection of ROS production (n = 3). *, P < .05.

PCR array shows that FXR activation up-regulates mRNA levels of genes in oxidant-antioxidant pathways

To better understand the mechanism by which FXR modulates ROS production, we performed a Mouse Oxidative Stress and Antioxidant Defense RT² Profiler PCR Array to determine which genes in oxidative stress were regulated by FXR activation. We compared gene expression in the control and GW4064-treated hepatocytes in WT mice. After GW4064 treatment, it is found that SOD3 gene expression was increased with about 14.31-fold in hepatocytes (Figure 5). SOD3 is an extracellular form of SOD and an antioxidant protein. It has been shown to play an important role in various oxidative stress-dependent pathophysiologies, including hypertension, ischemia-reperfusion injury, and lung injury (35). SOD3 can decrease extracellular O₂•⁻ to protect against oxidant-stress (36). It is proposed that FXR may regulate SOD3 expression to decrease ROS production, resulting in suppression of JNK signaling.

Figure 5.

The mouse oxidative stress and antioxidant defense PCR array results show that FXR activation up-regulated SOD3 expression in mouse hepatocytes. A, GW4064 treatment increased SOD3 expression in hepatocytes. Mouse hepatocytes were treated with GW4064 (3μM) for 24 hours. Then, total RNAs were prepared for PCR array analysis. B, Scatter plot between GW4064-treated and nontreated control groups. Red arrow indicates up-regulated gene area. Red dot indicates SOD3 gene.

FXR activation induces SOD3 gene expression, and FXR suppresses JNK activation and ROS production in an SOD3-dependent manner

We have observed that FXR activation by its ligand GW4064 induced about 2.1-fold increase of SOD3 gene expression in HepG2 cells (Figure 6A). Furthermore, to more clearly determine whether FXR regulates SOD3 gene expression, we isolated the hepatocytes from WT and FXR^−/− mouse liver. GW4064 treatment elevated SOD3 gene expression (∼12-fold) in WT hepatocytes but not FXR^−/− hepatocytes (Figure 6B). In vivo, SOD3 gene expression was down-regulated in FXR^−/− mouse liver compared with WT mice. WT mice, but not FXR^−/− mice, when fed a diet containing 0.2% cholic acid, an endogenous FXR agonist, displayed about 1.6-fold increase in SOD3 gene expression compared with the control group fed a normal diet (Figure 6C). These results indicate that FXR activation can induce SOD3 gene expression both in vitro and in vivo. Knockdown of SOD3 abolished the inhibition of FXR activation on JNK activity and ROS production induced by H₂O₂ (Supplemental Figures 3 and 4). These results indicate that FXR inhibits JNK signaling and ROS production in an SOD3-dependent manner.

Figure 6.

SOD3 is an FXR direct target gene. A, HepG2 cells were treated with GW4064 (2μM) for 24 hours, and then cells were collected for quantitative real-time PCR analysis. *, P < .05. B, WT and FXR^−/− hepatocytes were treated with GW4064 (3μM) for 24 hours, and cells were then collected for quantitative real-time PCR analysis. *, P < .05 vs WT control group. C, CA feeding induced SOD3 gene expression in WT but not FXR^−/− mice. *, P < .05 (n = 4). D, Schematic of human and mouse SOD3 IR-1 elements. E, Schematic of the WT and mutant IR-1 elements. The 3 IR-1 nucleotides that were altered to form the mutant constructs are underlined. F, EMSAs were performed with in vitro-synthesized human FXR/RXR and a ³²P-labeled oligo duplex containing IR-1 motif from the human SOD3 promoter. The positions of the shifted FXR/RXR complex and free probes are indicated. Mut, mutant (changing GAC to CCC). G, Real-time PCR analysis in a ChIP assay for the binding of FXR to the SOD3 promoter in HepG2 cells treated with GW4064 (2μM) for 24 hours (n = 3). *, P < .05 vs the control group. The fraction of immunoprecipitated DNA was normalized to input DNA.

FXR regulates SOD3 transcriptional activation by binding to an IR-1 element in the promoter of SOD3

We then analyzed the SOD3 gene promoter for potential FXR binding sites. We used NUBIScan to identify putative FXR-responsive elements in human SOD3 promoter. This approach identified 1 potential IR-1 in human SOD3 promoter (Figure 6D). The IR-1 element is from 213 to 245 in human SOD3 promoter. The direct interaction of FXR with the human IR-1 element was further substantiated by EMSA (Figure 6, E and F). A mixture of recombinant FXR and RXR proteins, produced by in vitro transcription and translation of the corresponding cDNAs, bound to a ³²P-labeled 213/245 oligo-duplex containing the IR-1 element (Figure 6F, lane 4). The binding of this complex was specifically competed out by the unlabeled homologous element (Figure 6F, lanes 5 and 6) but not by the mutant IR-1 element (Figure 6E, GAC to CCC, and F, lanes 7 and 8). There results suggest that FXR can bind to an IR-1 element (213 to 245) in SOD3 promoter.

To further demonstrate that FXR binds to the promoter of human SOD3 in vivo, we performed ChIP assays on soluble formaldehyde-cross-linked chromatin isolated from untreated and GW4064-treated HepG2 cells with a polyclonal anti-FXR antibody. FXR was bound to the promoter of SOD3 in vivo (Figure 6G), and GW4064 treatment increases the association of FXR with the promoter; this is consistent with the EMSA results shown in Figure 6F. Together, both in vitro and in vivo results demonstrated that SOD3 is up-regulated by FXR activation through binding to the IR-1 element in the promoter of SOD3.

Discussion

Prolonged activation of JNK pathway is frequently detected in many types of cancer, including HCC (1). It has been demonstrated that JNK activity is required for the development of liver cancer in carcinogen-exposed mice, and loss of JNK function in liver results in major defects in cellular proliferation (37). Therefore, identifying the environmental and genetic factors controlling JNK activation remains an important objective with potential pharmacological applications. In the present work, we identified FXR as a negative regulator of JNK signaling pathway and demonstrated that FXR ligands have utility in suppressing JNK signaling pathway in liver cells, thereby highlighting FXR as an attractive target for developing HCC therapy.

ROS, produced in mammalian cells, are mediators of prolonged JNK activation (38). In response to elevated levels of intracellular ROS, mammalian cells produce oxidative stress inducing severe damage to the host, which in turn adapted to face oxidative injury (39). ROS participate in carcinogenesis (40). Disruption of redox balance results in cancer and other diseases. It has been known that the antioxidant defense system, including SODs, protects against the harmful effects of ROS to maintain cellular redox balance (41). In this work, we identified that SOD3 is a direct FXR target gene and that FXR suppresses ROS production by activating SOD3 transcription, which may lead to reducing JNK activities.

High concentrations of BAs are potent endogenous agents for the induction of liver injury (21). FXR as an endogenous BA receptor is a key regulator of BA homeostasis and thus an important sensor of liver function (5). That is, FXR plays an important role in hepatoprotection to prevent the toxic effect of BAs on liver when they reach abnormally high levels. These protective effects for FXR have been attributed to the transcriptional regulation of genes involved in BA metabolism and transportation in hepatocytes (42). The serum BA levels in aged FXR^−/− mice are much higher than that of control WT mice (43). Probably due to the high concentrations of BAs, FXR^−/− mice develop spontaneous liver tumors. We previously have observed that FXR activation may protect liver cells from apoptosis (12). JNK activation causes liver cell apoptosis and then compensatory proliferation, resulting in liver tumor (37). Therefore, FXR may be a potential target for therapeutic intervention in human liver carcinogenesis, at least in part, by suppressing JNK signaling pathway.

It has been reported that FXR could be a potential target for the treatment of diabetes and associated metabolic disorders (4, 8). FXR is essential for normal glucose homeostasis, and treatment of FXR ligands is able to reverse insulin resistance and correct lipid metabolism abnormalities in an obesity animal model (44). These diseases, such as obesity and type 1 and type 2 diabetes, are closely associated with chronic inflammation characterized by abnormal cytokine production, increased acute-phase reactants, and activation of a network of inflammatory signaling pathways (45, 46). Inhibition of NF-κB- or JNK-related inflammation is able to improve glucose metabolism in vivo (27, 47, 48). We previously demonstrated that FXR is a negative modulator of NF-κB-mediated inflammation (13). Here, our data show that FXR activation is able to largely suppress JNK signaling. Therefore, there is a potential link between antiinflammation and treatment of obesity and diabetes through FXR. FXR may be an attractive therapeutic target for metabolic disorders through not only regulation of NF-κB-mediated inflammation but also suppression of JNK pathway.

Additional material

Supplementary data supplied by authors.