Farnesoid X receptor activation induces antitumour activity in colorectal cancer by suppressing JAK2/STAT3 signalling via transactivation of SOCS3 gene

Abstract Farnesoid X receptor (FXR, encoded by NR1H4), a bile acid‐activated nuclear receptor, is widely implicated in human tumorigenesis. The FXR agonist obeticholic acid (OCA) has preliminarily displayed tumour suppressor potential. However, the anticancer effects of this agent on colorectal cancer (CRC) remain unclear. In this study, the treatment of colon cancer cells with OCA inhibited cell proliferation and invasion in vitro, retarded tumour growth in vivo and prevented the G0/G1 to S phase transition. Moreover, the expression of active caspase‐3, p21 and E‐cadherin was up‐regulated and the expression of cyclin D1, c‐Myc, vimentin, N‐cadherin and MMP9 was down‐regulated in OCA‐treated colon cancer cells. Mechanistic studies indicated that OCA treatment suppressed the activity of JAK2/STAT3 pathway by up‐regulating SOCS3 expression. Colivelin, an agonist of JAK2/STAT3 pathway, antagonized the tumour‐suppressive effect of OCA on colon cancer cells. Dual‐luciferase reporter and quantitative chromatin immunoprecipitation (qChIP) assays further confirmed that OCA promoted SOCS3 transcription by enhancing the binding of FXR to the FXRE/IR9 of the SOCS3 promoter. In conclusion, our study demonstrates that targeting FXR and improving its function might be a promising strategy for CRC treatment.

Epidemiological studies and animal studies have suggested that prolonged exposure to faecal bile acids contributes to the occurrence of inflammatory bowel disease (IBD) and gastrointestinal cancer. 6 Increased amounts of bile acids generate an excess of detrimental injury on colonic mucosa including ROS-induced DNA damage, 7 inflammation 8 and hyperproliferation. 9 Farnesoid X receptor (FXR, encoded by NR1H4), a bile acid-activated nuclear receptor, regulates metabolism homeostasis of lipid, cholesterol and glucose. 10 Mounting evidence supports a pivotal role for FXR in human tumorigenesis. FXR exerts tumour-promoting effect in oesophageal 11 and pancreatic carcinomas. 12 However, some studies conclude the opposite. FXR-deficient mice are susceptible to developing hepatocellular carcinoma (HCC). 13 Diminished FXR correlates with late tumour stage and often predicts a poor prognosis. 14 The deficiency of FXR promotes intestinal inflammation and colon tumorigenesis. 15 Conversely, activation of intestinal FXR can suppress abnormal cell growth and curtail CRC progression. 16 Thus, targeting FXR and restoring its function might be an attractive tactics for CRC treatment.
Obeticholic acid (OCA) is a novel FXR agonist, a derivative of chenodeoxycholic acid (CDCA), and shows almost 10-fold greater potency than CDCA. 17 Importantly, OCA has been approved by the US Food and Drug Administration (FDA) for the treatment of primary biliary cholangitis. 18 Recent studies indicate that OCA inhibits the malignant potential of cholangiocarcinoma 19 and HCC. 20 However, the anticancer effect of this agent on colon cancer cells remains to be elucidated. Herein, we evaluated the impact of OCA on the aggressive phenotype of colon cancer cells and further investigated the underlying mechanisms.

| Lentiviral vectors and transfection
The phU6-EGFP-shRNA-SOCS3 lentiviral vectors and their control vectors were constructed and prepared by GeneChem Co., Ltd. All transfections were performed according to the manufacturer's instructions.

| Cell growth curve, CCK8, colony formation, cell cycle and apoptosis assays
Cell growth curve and CCK8 assays were performed as described previously. 21 For colony formation assay, three hundred cells were seeded and cultured for 14 days. Colonies (≥50 cells/colony) were counted. Cell cycle distributions were evaluated by flow cytometry as previously described. 21 For apoptosis assay, cells were labelled with Annexin V PE/7-AAD (BD Biosciences) according to the manufacturer's protocol as previously described. 22 Each experiment was performed in triplicate.

| Wound-healing assays
Cells were cultured in 6-well plates until confluent. Then, 3 artificial vertical lines were created with pipette tips (10 µL) in each well.
The wells were washed with phosphate-buffered saline (PBS) to remove cell debris. The cells were then cultured for an additional 48 hours. The scratch lines were imaged under a microscope, and the scratch distances were measured. Each experiment was performed in triplicate.

| Transwell assays
Cell migration and invasion were measured by using Transwell plates (Corning) with or without Matrigel (BD). In both assays, the lower chamber was filled with 600 μL of RPMI 1640 medium containing 20% FBS. For the migration assay, the cells were trypsinized and The migratory or invading cells on the underside of the membrane were fixed with 4% paraformaldehyde and stained with 1% crystal violet. The number of cells was counted in three randomly selected fields of fixed cells under an inverted microscope. Each experiment was performed in triplicate.

| RNA isolation and real-time PCR
RNA isolation, complementary DNA (cDNA) synthesized and realtime PCR were performed as described previously. 21 The sequences of primers were summarized in Supplementary Table S1. Each experiment was performed in triplicate.

| Total Protein extraction and Western blot
The detailed protocol was performed as described previously. 21 The antibody information was presented in Supplementary Table S2.
Each experiment was performed in triplicate.

| Immunofluorescence and immunocytochemistry
For Immunofluorescence (IF), the cells were fixed with 4% paraformaldehyde for 20 minutes and permeabilized with 0.2% Triton X-100 for 10 minutes. After blocking with 5% bovine serum albumin (BSA) for 30 minutes at room temperature, the cells were incubated at 4°C overnight with primary antibodies against E-cadherin, vimentin and STAT3 (1:100 dilution). The dishes were washed three times with PBS for 10 minutes each and then incubated with Alexa Fluor 594-con- For immunocytochemistry (ICC), the cells were fixed with 4% paraformaldehyde for 20 minutes and permeabilized with 0.2% Triton X-100 for 10 minutes and then incubated with the primary antibodies against E-cadherin and vimentin. The following protocol is the same as the Immunohistochemistry (IHC) assay.

| Luciferase reporter assay
For promoter analyses, a fragment of the SOCS3 5′-flanking sequence (from −3182 bp to + 425 bp) and other truncated fragments were cloned into the luciferase reporter vector pGL3.0 Basic Vector (Promega) to generate a SOCS3 full promoter reporter construct and the truncated ones. The primers for plasmid constructs were presented in Supplementary Table S1. The detailed protocol was performed as described previously. 21 Each experiment was performed in triplicate.

| Quantitative chromatin immunoprecipitation
Cells were subjected to ChIP using the EZ-ChIP Kit (Millipore).
The detailed protocol was performed as described previously. 21 Real-time PCR was conducted to amplify the regions of DNA fragments by using special primers (Supplementary Table S1). Each experiment was performed in triplicate.

| Statistical analysis
The differences among the groups were compared by Student's ttest or one-way ANOVA. All statistical analyses were performed using the SPSS statistical package (SPSS Inc, Chicago, IL, USA). P < .05 was considered statistically significant.

| FXR activation inhibits the proliferation of colon cancer cells in vitro
The effect of FXR activation on tumour cell viability and growth was as-

| FXR activation retards the tumour growth in nude mice
Next, xenograft mouse model was conducted to assess the effect of FXR activation on tumour growth in vivo. The xenograft tumours in the OCA-treated group showed a decline in growth rate in relative to the DMSO-treated group (Figure 2A Collectively, our data indicated that FXR activation retards the tumour growth in nude mice.

| FXR activation inhibits the invasion and migration of colon cancer cells by arresting EMT
Considering that tumour metastasis is the leading cause of cancer-related death in CRC, we thus aimed to evaluate the impact of FXR activation on the invasive and migratory abilities of colon cancer cells. The wound-healing assay showed that OCA treatment resulted in a decreased rate of wound healing ( Figure 3A,B).
The Transwell assays showed that the number of invasive and migrating cells in the OCA-treated group was less than that in the DMSO-treated group ( Figure 3C,D). Epithelial-mesenchymal transition (EMT) has been linked to the mobility and dissemination of CRC by conferring increased invasiveness and metastatic potential to cells. 23,24 Our study further assessed the levels of

| FXR activation suppresses JAK2/STAT3 pathway by up-regulating SOCS3 expression in colon cancer cells
The Janus tyrosine kinases/signal transducer and activator of transcription (JAK/STAT) pathway, which is negatively regulated by the SOCS family, plays a crucial role in cellular responses by regulating survival, proliferation, invasion and differentiation. 25,26 Our study attempted to detect the impact of FXR activation on the activity of JAK/STAT signalling. OCA treatment inhibited the protein expression of p-STAT3 and p-JAK2 ( Figure 4A,B). However, OCA treatment had no significant impact on phosphorylated JAK1 expression ( Figure 4A,B). Furthermore, we found that OCA treatment resulted in an up-regulation of SOCS3 expression at both mRNA and protein levels ( Figure 4A,B). To further validate the involvement of SOCS3 in OCA-mediated inhibition of JAK/STAT signalling, we knocked down SOCS3 in colon cancer cells.
Depleting SOCS3 expression restored the activity of JAK/STAT signalling in OCA-treated cells ( Figure 4D,E). Phosphorylated STATs dimerize and translate into the nucleus to regulate target gene transcription. 27 Thus, we attempted to assess the localization of STAT3 protein by using IF staining. Exposure to OCA resulted in increased STAT3 protein accumulation in the cell cytoplasm ( Figure 4F,G). Taken together, these results suggest that FXR activation might suppress JAK2/STAT3 pathway in colon cancer cells by up-regulating SOCS3 expression.

F I G U R E 3 Farnesoid X receptor (FXR) activation inhibits the invasion and migration of colon cancer cells in vitro. A,
It has been demonstrated that FXR usually modulates target genes transcription by specifically binding to FXR response element (FXRE). 28 We employed the dual-luciferase reporter assay to assess the impact of FXR activation on SOCS3 transcription. All of these results indicate that FXR activation inhibits JAK2/ STAT3 pathway by regulating SOCS3 transcription in colon cancer cells.

| Activation of JAK2/STAT3 signalling rescues the tumour-suppressive effect of FXR activation
In order to determine that JAK2/STAT3 signalling is indispensable for the tumour-suppressive effect of FXR activation on colon cancer cells, colivelin and IL-6, the agonists of JAK/STAT3 pathway, 29,30 were adopted to treat FXR-activated cells. IL-6 is a classic STAT3 activator, while colivelin, a potent humanin (HN) derivative, has recently been proved to activate JAK2/STAT3 signalling via binding to cell surface receptors involving CNTFRα, WSX-1 and gp130. 30 In the current study, colivelin and IL-6 rescued the proliferative and invasive abilities inhibited by FXR activation (Figure 6A-F, Figure 7A,B).
Consistent with the observations above, colivelin and IL-6 up-regulated the protein levels of p-STAT3, p-JAK2, vimentin, c-Myc and cyclin D1 but reduced the level of p21, active CASP-3 and E-cadherin ( Figure 7C-F). Taken together, these results demonstrate that JAK2/ STAT3 signalling participates in OCA-mediated tumour inhibition.

| D ISCUSS I ON
Recently, due to the limited effectiveness of conventional therapeutic strategies, there has been a strong interest in the field of novel antitumour drugs. These novel drugs characterized by the good markers available are the major goals of molecular oncology.
The tumour-suppressive role of FXR in colorectal tumorigenesis has inspired us to restore FXR activity as a novel therapeutical strategy. Our study indicated that FXR activation by OCA inhibited the proliferative potential of colon cancer cells by arresting G1/S transition and inducing apoptosis. In support of these data, treatment with OCA significantly altered the levels of growth-and F I G U R E 5 Farnesoid X receptor (FXR) transactivates the expression of SOCS3 by binding to the FXRE/IR9 in the SOCS3 promoter in colon cancer cells. A and B, The activities of the full SOCS3 promoter reporter construct and the truncated ones in Caco-2 (A) and HT-29 (B) cells after obeticholic acid (OCA) exposure using the dual-luciferase assay. C, Schematic representation of a putative FXRE in the human SOCS3 promoter region (SOCS3/ IR9). Enrichment level of FXR binding to the putative FXRE in the SOCS3 promoter region in Caco-2 (D) and HT-29 (E) cells determined by the qChIP assay. *P < .05 apoptosis-related proteins. FXR agonist GW4064 has been reported to trigger apoptosis and retard tumour growth by activating its target SHP. 31  SOCS3, an important member of the SOCS family, negatively regulates JAK/STAT pathway. 38 SOCS3 expression is reduced in CRC tissues due to promoter methylation. 39 Lack of SOCS3 contributes to accelerative intestinal crypt growth and facilitates tumour growth. 40 In the current study, OCA treatment increased the mRNA and protein levels of SOCS3. Similarly, restoration of SOCS3 by FXR agonist GW4064 is observed in hepatocellular inflammation and HCC, 41,42 indicating that FXR-SOCS3 axis may serve as a new potential target for the prevention and treatment of HCC.
Furthermore, our study revealed that OCA treatment could promote F I G U R E 7 Activation of the JAK2/ STAT3 pathway rescues the inhibitory effect of Farnesoid X receptor (FXR) activation on the growth-, apoptosis-and EMT-related proteins of colon cancer cells. A, The effect of colivelin and IL-6 on the invasion of FXR-activated Caco-2 and HT-29 detected by invasion and migration assays. B, The number of invasive and migrating cells in Caco-2 and HT-29 cells. C and D, Western blot bands of the growth-, apoptosis-and EMT-related proteins in FXR-activated Caco-2 (C) and HT-29 (D) cells after colivelin and IL-6 exposure. E and F, Quantitative analysis of the growth-, apoptosis-and EMT-related proteins in Caco-2 (E) and HT-29 (F) cells. *P < .05 SOCS3 transcription by binding to the −3182 bp to −2834 bp of the SOCS3 promoter, as determined by the dual-luciferase reporter assay. Nuclear hormone receptors regulate gene transcription by recognizing the motif consisting of repeats with the core element AGGTCA aparted by one or more nucleotides. 43 We predicted a potential FXRE/IR9 (−2884 bp to −2864 bp) in the −3182 bp to −2834 bp of the SOCS3 promoter by using the JASPAR database.
The qChIP assay demonstrated that FXR specifically and directly binds to the FXRE/IR9 of SOCS3 promoter. Some pharmacological compounds have been shown to repress STAT3 activity by restoring SOCS3 expression. [44][45][46] Manipulation of SOCS3 expression via the regulatory role of FXR might be expected to be a promising therapeutic option for cancer chemotherapy. Intriguingly, the activation of FXR by GW4064 decreased the expression of HER2, the upstream of JAK/STAT3 pathway. 47 Besides JAK/STAT signalling, OCA might exert tumour suppressor role by directly regulating FXR target genes SHP 31 and MMP7 48 or suppressing Wnt/β-catenin 49 and EGFR/ERK signalling. 50 Hence, OCA exerts its tumour suppressor functions via multiple molecular mechanisms.
In the present study, the FXR agonist OCA inhibited colon cancer cell proliferation and invasion by repressing JAK2/STAT3 pathway via regulating SOCS3 transcription. Our study provided the first evidence that FXR activation by OCA may represent a potential novel therapeutic strategy for CRC treatment.

CO N FLI C T O F I NTE R E S T
All authors declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data used to support the findings of this study are included within the article.