The FXR Agonist, Obeticholic Acid, Suppresses HCC Proliferation & Metastasis: Role of IL-6/STAT3 Signalling Pathway

The nuclear receptor, farnesoid X receptor (FXR), has been recently considered as a tumor suppressor in HCC. IL-6/Janus kinase 2 (Jak-2)/signal transducer and activator of transcription 3 (STAT3) pathway has been implicated as a key player in many cancer types. This study aimed at investigating the potential effect of the FXR agonist, obeticholic acid (OCA), on HCC and the involvement of IL-6/STAT3 pathway. The potential regulation of STAT3 by its main feedback inhibitor target gene, the suppressor of cytokine signaling 3 (SOCS3), triggered by OCA was also explored. Cytotoxicity studies were performed on HepG2, Huh7, and SNU-449 cell lines using OCA alone and combined with the FXR antagonist guggulsterone (Gugg). OCA cytotoxic effect was significantly hampered in presence of Gugg. OCA also caused cell cycle arrest and inhibited invasion and migration of HCC cells. Decrease in STAT3 phosphorylation and SOCS3 upregulation were also observed. Moreover, Jak-2, IL-1β, and IL-6 levels were decreased. These results were correlated with an upregulation of FXR and small heterodimer partner (SHP) levels. Effects of OCA on IL-6/STAT3 main key players were reversed in presence of Gugg. Overall, these findings suggest a potential effect of OCA in HCC via interfering with IL-6/STAT3 signalling pathway in vitro.

Hepatocellular carcinoma (HCC) is the sixth most prevalent type of cancer and the third most common cause of cancer-related death worldwide. It accounts for about 85% of primary liver cancer where its incidence has dramatically increased during the past two decades 1 . Previous studies showed that the molecular pathogenesis of HCC comprises a multi-step process that subsequently leads to the accumulation of various aberrant genetic and epigenetic changes 2 .
The Farnesoid X Receptor (FXR) is a nuclear receptor that has long been considered as a main regulator of metabolism with multifaceted functions in the maintenance of bile acids, lipids, and glucose homeostasis [3][4][5] . Recently, FXR agonists were also found to exert tumor suppression properties in liver tissue through various mechanisms [6][7][8] .
It was clearly stated in previous studies that deficiency of FXR contributes to hepatic carcinogenesis, though the underlying mechanisms haven't yet been totally investigated and could be either FXR-dependent or FXR-independent. FXR deficiency is known to lead to cholestasis 9,10 , which subsequently causes hepatic injury and inflammation 11,12 . It is suggested that inflammation is a pivotal mechanism contributing to hepatic carcinogenesis 13 . One of the main mechanisms which can explain how inflammation leads to HCC is the activation of IL-6/Janus kinase 2 (Jak-2)/signal transducer and activator of transcription 3 (STAT3) signalling pathway, which is considered a potent promoting factor driving tumor formation in the liver 14 . STAT3 activation turns on the expression of a cohort of genes that further induce cell division and inhibit apoptosis 14 . Moreover, STAT3 remarkably induces the gene expression of suppressor of cytokine signalling 3 (SOCS3), which leads to a feedback suppression of STAT3 activity 15,16 . There is a growing body of evidence suggesting that FXR deficiency causes STAT3 activation; however, the underlying mechanism is not yet clear.
Obeticholic acid (OCA), also known as INT-747 or 6α-ethyl-chenodeoxycholic acid, is a first-in-class bile acid analogue, derived from chenodeoxycholic acid, the natural FXR agonist, and shows almost 100-fold greater Migration and invasion assays. To assess cell migration, HepG2 and Huh7 cells (5 × 10 5 cells/well) were seeded on top of the 8.0 μm pore size transwell insert containing 600 μl DMEM medium in the lower chamber. Cells migrated to the other side of membrane were stained with crystal violet and counted after 24 h incubation. To assess cell invasion, cells (1 × 10 5 ) were plated in 8.0 μm pore size Matrigel coated transwell insert. Culture medium was changed twice a day for 3 days. After 48 h, cells that had migrated through the Matrigel to the filter were stained with crystal violet and counted. Experimental design. HepG2 or Huh7 cells were divided into three groups: (1) Blank control group where cells were treated with the vehicle; DMSO, (2) OCA group where cells were treated with IC 50 of OCA; & (3) OCA+ Gugg group where cells were treated with both IC 50 of OCA & 10 µM of Gugg to achieve FXR receptor blockade. All treatments were started 24 hrs after cells were seeded in T-25 flasks. Gene expression levels were assessed at 48 h of treatment exposure, while protein expression levels were assessed at 72 h of exposure. Moreover, the expression of a number of genes (IL-1β, IL-6, STAT3, & FXR) was estimated in THLE-2 normal liver cells for comparison with their corresponding expression in HCC cells.
RNA extraction and real time quantitative polymerase chain reaction (RT-qPCR) assay. Total RNA was isolated using RNeasy Mini Kit (Qiagen, USA). Complementary DNA (cDNA) synthesis and PCR amplification were carried out using iScript TM One-Step RT-PCR Kit with SYBR ® Green (Bio-Rad, CA, USA) according to the manufacturer's instructions using GAPDH as the house-keeping gene, as described previously 26 . The relative mRNA expression levels (fold change from untreated control samples normalized to GAPDH as the housekeeping gene) of caspase-3, FXR, small heterodimer partner (SHP), STAT3, SOCS3, Jak-2, IL-1β, and IL-6 were assessed using the 2 −ΔΔCt analysis method, as previously described by Livak and Schmittgen 27 for all experiments except those performed on THLE-2 cells for estimating the expression of genes in normal cells vs HCC cells. For the latter experiment, absolute quantitation method was used as described by Fronhoffs et al. 28 . The primer sequences used for RT-qPCR are listed in Table 1. Determination of caspase-3 activity. Active caspase-3 level was measured using human active Caspase-3 ELISA kit (Invitrogen, CA, USA) according to the manufacturer's instructions, in order to investigate the effect of OCA on apoptosis. Briefly, cells were washed with PBS, collected and added to extraction buffer containing protease inhibitor (1 ml per 1 × 10 7 cells) then diluted immediately prior to the assay. After completing the steps of the assay, the optical density was measured for each well at 450 nm.
Determination of SOCS3, total STAT3, and phosphorylated STAT3 activity. To determine the effect of OCA on SOCS3, total STAT3 (t-STAT3), and phosphorylated STAT3 (p-STAT3) activity; the human SOCS3 ELISA kit (Cloud-Clone Corp., TX, USA) and the ELISA kit for the determination of t-STAT3 and p-STAT3 (Tyr705) (Abcam, MA, USA) were used according to the manufacturers' instructions. Briefly, cells were washed with PBS, collected and added to the extraction buffer containing protease inhibitors (1 ml per 1 × 10 7 cells) then diluted immediately prior to the assay. After performing all steps of the assay, the optical density of each well was determined within 30 minutes using a microplate reader set at 450 nm.

Statistical analysis.
All values are presented as means ± standard deviation (S.D.) from three independent experiments performed in triplicates. 95% confidence intervals (C.I.) were compared to test significance for the fold changes from control untreated samples in gene expression studies. Statistical analysis was performed by Student's t-test and one-way ANOVA using GraphPad Prism, version 5.00 (GraphPad Software, CA, USA) (v5). Statistical significance was determined at P < 0.05. Data availability. The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Obeticholic acid inhibits HCC cell growth and induces cell cycle arrest. MTT assay showed that
OCA decreased the rate of cell proliferation in HepG2, Huh7, and SNU-449 cell lines, as compared to the corresponding control (Fig. 1). The IC 50 of OCA was determined and found to be 1.067 μM, 1.038 μM, and 0.706 µM on HepG2, Huh7, and SNU-449 cell lines, respectively. However, Gugg caused a 2.24-, 2.95-, and 2.76-fold increase in the IC 50 of OCA in HepG2, Huh7, and SNU-449 cell lines, respectively. Moreover, as shown in Fig. 1, cell cycle analysis showed that OCA caused a 1.06-fold increase in the number of HepG2 cells at G0/G1 phase with a significant decrease in the cells in S phase by 26.13% (P < 0.0001), as compared to control. Whereas in Huh7 cell line, OCA caused cell growth arrest at the S phase where accumulated cells reached 21.18% (P = 0.0003), as compared to untreated cells. Moreover, the cells in M phase were 3.51-fold lower (P = 0.0005) than control in Huh7 cell line. Interestingly, OCA has also caused a 23.68-and 8.32-fold increase (P < 0.0001) in the percentage of cells in the pre-apoptotic phase in HepG2 and Huh-7 cell lines, respectively.

Obeticholic acid inhibits cell migration and invasion of HCC cells. Transwell migration and Matrigel
assays were used to investigate the effect of OCA on both migration and invasion, respectively, in HepG2 and Huh7 cell lines. As shown in Fig. 2, OCA caused 3.19-(P = 0.0031) and 1.65-(P = 0.0181) fold decrease in the number of migrated cells in HepG2 and Huh7 cell lines treated with OCA, as compared to controls, respectively. Moreover, OCA caused 2.43-(P = 0.0075) and 1.48-(P = 0.0245) fold decrease in the invasion percentage in HepG2 and Huh7 cell lines, respectively.

Upregulation of IL-1β, IL-6, & STAT3 and downregulation of FXR in HCC cells vs normal cells.
In order to validate an altered IL-6/STAT3 axis as well as FXR deficiency in HCC cells, the gene expression of IL-1β, IL-6, STAT3, & FXR was assessed in normal liver cells and compared to expression levels in HepG2 and Huh7 cells. As shown in Fig. 2, the expression levels of IL-1β, IL-6, and STAT3 were highly significantly increased (P < 0.0001) in both HepG2 and Huh7 cells, as compared to normal liver cells. Moreover, FXR expression was highly significantly downregulated (P < 0.0001) in HepG2 and Huh7 cells versus normal cells.

Obeticholic acid increases caspase-3 gene expression and protein levels.
To assess the effect of OCA on apoptosis, caspase-3 levels were estimated on both gene and protein levels. OCA caused 6.39-and 5.78-fold increase (P < 0.0001) in active caspase-3 protein levels, as compared to control, in HepG2 and Huh7 cell lines, respectively (Fig. 3a,b). Moreover, it was found that OCA induced upregulation of caspase-3, as determined by the relative mRNA levels, causing 4.80-(95% C.I.: 3.87 to 5.73) and 3.15-(95% C.I.: 2.67 to 3.64) fold increase, in both HepG2 and Huh7 cells, respectively (Fig. 3c,d). These effects were opposed in presence of the FXR antagonist, Gugg, where the fold change expression levels weren't significantly different from control (95% C.I.: 0.91 to 1.38).

Obeticholic acid upregulates FXR and its main target gene SHP in HCC cells.
To investigate the ability of OCA to activate FXR in HCC cells and test whether this activation is related to the antitumour effects observed, mRNA levels of FXR and its main target gene, SHP, were measured using qRT-PCR. As shown in Fig. 3e and 3f, FXR was upregulated in both HepG2 and Huh7 cells treated with OCA reaching 1.57-(95% C.I.: 1.24 to 1.89) and 1.16-(95% C.I.: 1.02 to 1.30) fold increase from control, respectively. Additionally, OCA caused 1.52-(95% C.I.: 1.05 to 1.99) and 1.27-(95% C.I.: 1.11 to 1.43) fold higher relative mRNA SHP expression levels in HepG2 and Huh7 cell lines, respectively.

FXR activation by obeticholic acid induces SOCS3 expression levels.
In order to determine whether FXR activation by OCA resulted in subsequent changes in SOCS3, one of the FXR target genes and also an important feedback repressor of STAT3, we examined the gene and protein expression profiles of SOCS3 in OCA-treated HepG2 and Huh7 cells. As shown in Fig. 5, relative mRNA levels of SOCS3 were 1.55-(95% C.I.: 1.21 to 1.88) and 2.08-(95% C.I.: 1.23 to 2.92) fold higher in HepG2 and Huh7 OCA-treated cells compared to controls, respectively. This was also reflected on the protein level, where SOCS3 showed 1.29-and 1.56-fold higher expression levels in HepG2 and Huh7 cells (P < 0.0001), respectively.

Obeticholic acid inhibits the activation of Jak-2. To further confirm the inhibition of the IL-6/STAT3
signalling pathway by OCA and whether the increase in SOCS3 had subsequently caused an inhibition of Jak-2, the relative mRNA level of Jak-2 in OCA-treated cells was determined. As shown in Fig. 6a and 6b, Jak-2 mRNA level was 1.75-(95% C.I.: 0.30 to 0.84) and 1.28-(95% C.I.: 0.76 to 0.81) fold lower in HepG2 and Huh7 cells treated with OCA, respectively, as compared to control untreated cells.

FXR activation by obeticholic acid inhibits IL-1β and IL-6 expression levels. Since FXR deficiency
reported in HCC was associated with increased levels of IL-1β and IL-6, we were interested in investigating the potential influence of OCA on their expression levels in the present experiment. Interestingly, we found that both IL-1β and IL-6 gene expression levels were  (Fig. 6c-f).

Discussion
The nuclear receptor FXR has long been known for its ability to maintain both glucose and lipid homeostasis in the liver. Recently, the role of FXR in the liver has gone beyond its function as a metabolic regulator and was extended to include liver carcinogenesis. Hence, it is presented as a promising new therapeutic target in HCC  Protein levels were estimated using ELISA. Gene expression levels were estimated using relative qRT-PCR method (fold change from control untreated samples normalized to GAPDH). Values are presented as means ± S.D. from three independent experiments performed in triplicates. *P < 0.05 significant from control untreated cells using Student's t test. # P < 0.05 significant when the 95% C.I. was compared with control untreated cells. @ P < 0.05 significant when the 95% C.I. was compared with OCA+Gugg-treated cells. HepG2 and (f) Huh7 cell lines treated with OCA alone or OCA+Gugg. Protein levels were estimated using ELISA. Gene expression levels were estimated using relative qRT-PCR method (fold change from control untreated samples normalized to GAPDH). Values are presented as means ± S.D. from three independent experiments performed in triplicates. *P < 0.05 significant from control untreated cells using Student's t test. # P < 0.05 significant when the 95% C.I. was compared with control untreated cells. @ P < 0.05 significant when the 95% C.I. was compared with OCA+Gugg-treated cells.
where a tumor suppressor potential of FXR agonists in the liver has been reported both in vivo and in vitro 29 . Multiple mechanisms have been suggested to explain this antitumour effect [30][31][32] . One important mechanism is interfering with the activation of IL-6/Jak-2/STAT3 signalling pathway which was found to have a pivotal role in cancer development and progression 30 . OCA, a potent FXR agonist which has recently been approved by the FDA as an orphan drug for PBC is still undergoing clinical trials for indications closely related to the ability of FXR to maintain bile acid, lipid, and glucose levels in the liver such as NASH 23 . The present study provides the first evidence for a potential anticancer effect of OCA in HCC providing insights into the possible underlying mechanism of action. As shown from the results, activation of FXR by OCA was found to suppress HCC cells proliferation, migration, and invasion. The cytotoxic effect of OCA on HepG2, Huh7, and SNU-449 cell lines was compromised in presence of the FXR antagonist, Gugg. In order to assess the effect of OCA on inducing apoptosis of HCC cells, caspase-3 gene and protein expression levels were determined. It was found that OCA significantly reduced caspase-3 in both HepG2 and Huh7 cell lines. In the present study, OCA was also found to inhibit the activation of IL-6/STAT3 signalling pathway in HCC cell lines (Fig. 7). Our results also showed that OCA was capable of causing a significant repression of important key players in this signalling pathway such as IL-6, IL-1β, and Jak-2 besides STAT3 phosphorylation. As previously reported by Li et al. 30 , FXR deficiency observed in HCC 6 , was found to activate STAT3 pathway by increasing IL-1β expression which subsequently increased IL-6 levels in liver tissue. IL-6 increased levels led to Jak-2 phosphorylation which consequently activated STAT3 Tyr705 and Ser727 phosphorylation. Following this activation step, STAT3 homodimers are formed and translocated to the nucleus leading to the activation of many downstream target genes related to liver cancer development. Among these genes is the SOCS3 which is considered a very important direct FXR downstream target gene in addition to its pivotal role as a feedback inhibitor of the STAT3 pathway 15,16 . Interestingly, in the current study OCA was found to significantly increase the expression of SOCS3 on both gene and protein levels which, therefore, could have suppressed STAT3 phosphorylation. Upregulation of SOCS3 in response to FXR agonists has been previously reported by Xu et al. 33 . In the present experiment, activation of FXR in HCC cell lines by OCA was verified by evaluating the gene expression profile of FXR and its target gene, SHP. As further explained below, OCA caused an upregulation in both FXR and SHP mRNA levels that hence could be correlated to changes observed in STAT3 mRNA and protein levels. FXR was previously reported to directly induce the expression of the HCC suppressor, SHP 34 .
Interestingly, the effects of OCA on IL-1β, IL-6, STAT3, and caspase-3 have been reversed in presence of the FXR antagonist, Gugg, suggesting a causative relationship. Moreover, in order to ensure an altered IL-6/STAT3 axis as well as FXR deficiency in HCC cells, the gene expression levels of IL-1β, IL-6, STAT3, & FXR were assessed Protein levels were estimated using ELISA. Gene expression levels were estimated using relative qRT-PCR method (fold change from control untreated samples normalized to GAPDH). Values are presented as means ± S.D. from three independent experiments performed in triplicates. *P < 0.05 significant from control untreated cells using Student's t test. # P < 0.05 significant when the 95% C.I. was compared with control untreated cells.
in normal liver cells and compared to that in HepG2 and Huh7 cells. An upregulation of IL-1β, IL-6, & STAT3 besides a downregulation of FXR were observed in HCC cells compared to normal cells.
From a mechanistic perspective, there are some pathways that still need to be deciphered. One of these pathways is the activation of the fibroblast growth factor 4 receptor (FGFR4) by fibroblast growth factor (FGF)15 (in mice) or FGF19 (in humans) following FXR stimulation in the intestine. Despite the beneficial effect of FGF15/19 activation on the liver since they were considered crucial for the regenerative process by protecting against bile acid stress and promoting hepatocyte proliferation 35 , it's emerging as an important player in hepatic OCA+Gugg. Gene expression levels were estimated using relative qRT-PCR method (fold change from control untreated samples normalized to GAPDH). Values are presented as means ± S.D. from three independent experiments performed in triplicates. # P < 0.05 significant when the 95% C.I. was compared with control untreated cells. @ P < 0.05 significant when the 95% C.I. was compared with OCA+Gugg-treated cells. carcinogenesis 36 . Hence, exploring the effect of OCA on FGF15/19 in HCC and investigating whether the antiproliferative effect observed could be related to an altered FGF15/19 level need to be addressed in future studies. Moreover, newly synthesized FXR agonists are yet to be investigated in HCC. For instance, Px-102 and Px-104 are non-steroidal FXR agonists that are still in phase II clinical trials, however, they are more potent than OCA and can protect against bile acid overload and hence have broad spectrum of improvements 37,38 .
Taken together, the results of the present study suggest a potential of the FXR agonist, OCA, for inhibiting HCC proliferation, migration, and invasion via interfering with the activation of IL-6/STAT3 signalling pathway. OCA may have decreased STAT3 signalling by abrogating the FXR deficiency and hence the increase in IL-1β & IL-6 as well as increasing the levels of SOCS3, the feedback inhibitor of STAT3. Further studies, however, are required to validate the anticancer effect of OCA in HCC and investigate its potential modulation of the IL-6/ STAT3 signalling pathway in vivo.

Figure 7.
Schematic diagram of the proposed mechanism by which OCA may modulate the IL-6/Jak-2/ STAT3 signalling pathway in HCC cells. FXR deficiency leads to activation of STAT3 pathway by increasing IL-1β which subsequently increases IL-6 levels. IL-6 then binds to IL-6 receptor (IL-6R) leading to Jak-2 phosphorylation via glycoprotein 130 (gp130) consequently activating STAT3 phosphorylation. STAT3 homodimers are then formed and translocated to the nucleus leading to the activation of many downstream target genes responsible for hepatic carcinogenesis such as SOCS3, the FXR target gene and the feedback inhibitor of STAT3.