STAT3 Mediates Regorafenib-Induced Apoptosis in Hepatocellular Carcinoma Free

Regorafenib (Stivarga), a novel oral multiple kinase inhibitor, potentially inhibits tumor growth through antiangiogenesis (1). Several angiokinases such as VEGFR1/2, PDGFR-β, and FGFR1 are believed to be major targets of regorafenib for cancer treatment (1, 2). Regorafenib was approved for the treatment of patients with metastatic colorectal cancer and advanced gastrointestinal stromal tumors by the FDA in 2012. As the chemical structure of regorafenib is very similar to that of sorafenib, the first and only small-molecule targeted therapy approved for HCC treatment, several in vitro studies and clinical trials have been initiated to evaluate the efficiency of regorafenib for HCC. The safety of regorafenib at a dose of 160 mg once daily has been validated in patients with HCC following first-line sorafenib (3). Participants are currently being recruited for a randomized, double blind, phase III study of regorafenib in patients with HCC who have progressed on sorafenib treatment (NCT01774344).

Protein tyrosine phosphatases (PTP) are also vital regulators involved in many cellular signaling processes, which make PTPs potential prognostic markers and druggable targets. Src homology region 2 (SH2) domain–containing phosphatase 1, SHP-1, acts as a negative regulator of STAT3. The 3D structure of SHP-1 shows that the ligand-free SHP-1 has an autoinhibited conformation through blocking catalytic PTPase via the N-SH2 domain (4–6). However, the underlying molecular mechanism by which SHP-1 is involved in tumorigenesis is still unknown. In this study, we first clarified the molecular mechanism by which regorafenib inhibits STAT3 signaling, and then disclosed the direct effect of regorafenib on SHP-1. We demonstrate that regorafenib has potent anti-HCC potency via targeting STAT3, and provide evidence to suggest that STAT3 may be a useful prognostic marker and druggable target for the treatment of HCC.

The HA59T HCC cell line was obtained from the Bioresources Collection and Research Center (BCRC; Food Industry Research and Development Institute, Hsin-Chu, Taiwan). The PLC/PRF/5 (PLC5), Sk-Hep-1, HepG2, and Hep3B cell lines were obtained from the ATCC. All cells obtained from the BCRC or ATCC were immediately expanded and frozen down such that all cell lines could be restarted every 3 months from a frozen vial of the same batch of cells. No further authentication was conducted in our laboratory. Antibodies for immunoblotting such as cyclin D1 and PARP were purchased from Santa Cruz Biotechnology. Other antibodies such as p-STAT3, STAT3, survivin, Mcl-1, caspase-9, and myc-tag were from Cell Signaling Technology. SHP-1 antibody was purchased from Abcam.

Regorafenib (BAY 73-4506) was purchased from Selleck Chemicals. For cell-based studies, regorafenib at various concentrations was dissolved in DMSO and then added to the cells in serum-free DMEM. The SHP-1 inhibitor (PTP III) was purchased from Calbiochem. The DNA fragmentation was demonstrated in the Cell Death ELISA Assay Kit purchased from Roche Molecular Biochemicals.

Plasmids encoding the human STAT3 and SHP-1 (PTPN6) mutant, in which the SH2/PTP domain was truncated or one aspartic acid at 61 was changed into an alanine residue were cloned into pCMV6-Entry vector with the myc-tag. All of the truncated mutants were confirmed by DNA sequence and their expression level in HCC cells was assayed. SMARTpool siRNAs, including control (D-001810-10) and SHP-1 (PTPN6, L-009778-00-0005), were all purchased from Dharmacon. For transient transfection, plasmids or siRNAs were pretransfected with lipofetamine 2000 (Invitrogene) for 24 hours and then processed with the indicated treatment for another 24 hours as described previously (7).

The human full-length SHP-1–GST fusion protein was inserted into the pGEX-4T-1 vector (Amersham Biosciences). The GST fusion protein of wild-type SHP-1 was expressed in Escherichia coli and affinity purified using glutathione-sepharose 4B beads (Amersham Biosciences).

HCC protein extracts pretreated with regorafenib were incubated with anti–SHP-1 antibody in immunoprecipitation buffer (20 mmol/L Tris-HCl (pH 7.5), 150 mmol/L NaCl, 1 mmol/L EDTA, 1% NP-40, and 1% sodium deoxycholate) overnight. Protein G-Sepharose 4 Fast flow (GE Healthcare Bio-Science) was added to each sample, followed by incubation for 3 hour at 4°C with rotation. A RediPlate 96 EnzChek Tyrosine Phosphatase Assay Kit (R-22067) was used to assay SHP-1 activity (Molecular Probes; Invitrogen). For in vitro assay, SHP-1–containing extract or purified recombinant SHP-1 protein was incubated with regorafenib at the indicated dose for 30 minutes at 4°C and the phosphatase activity of SHP-1 was assessed.

For the subcutaneous xenograft tumor model (n = 10), each mouse was inoculated s.c. in the dorsal flank with 1 × 10⁶ PLC5 cells suspended in 0.1 mL of serum-free medium containing 50% Matrigel (BD Biosciences). When tumors reached 100 to 200 mm³, mice received vehicle or regorafenib (20 mg/kg) by mouth once daily. Tumors were measured twice weekly using calipers and their volumes calculated using the following standard formula: width × length × height × 0.523.

Tumors from patients with HCC who received surgical resection, post-operative treatment and follow-up in Changhua Christian Hospital from June 2012 to June 2013 were enrolled for analysis (CCH IRB no. 120504). Paraffin-embedded HCC tissue sections (4 μm) on poly-1-lysine–coated slides were deparaffinized and rinsed with 10 mmol/L Tris-HCl (pH 7.4) and 150 mmol/L sodium chloride. Paroxidase was quenched with methanol and 3% hydrogen peroxide. Slides were then placed in 10 mmol/L citrate buffer (pH 6.0) at 100°C for 20 minutes in a pressurized heating chamber. After incubation with 1:50 dilution of p-STAT antibody [Rabbit monoclonal [EP2147Y] to STAT3 (phospho Y705; AB76315), Abcam] for 1 hour at room temperature, slides were thoroughly washed three times with PBS. Bound antibodies were detected using the EnVision Detection Systems Peroxidase/DAB, Rabbit/Mouse Kit (Dako). The slides were then counterstained with hematoxylin. Paraffin-embedded sections of mouse kidney tissue and human colon carcinoma were used as positive controls for p-STAT3, respectively, as described in the datasheet from the manufacturer. Negative controls had the primary antibody replaced by PBS. The expression of p-STAT3 was assessed semiquantitatively based on the intensity of staining and the percentage of cell involved by a board certified pathologist. The intensity of staining was scored as negative, low, and high. This study was approved by the ethics committee of the Institutional Review Board of Changhua Christian Hospital. All informed consents from sample donors were in accordance with the Declaration of Helsinki and were obtained at the time of their donation.

Descriptive statistical analysis was used to compare the baseline characteristics of patients and the expression of p-STAT3. A P value of <0.05 was regarded as statistically significant on the two-tailed tests. All statistical analyses were computed using SPSS for Windows software (version 18.0; SPSS, Inc.; ref. 8).

Angiogenesis plays a central role in cancer progression and metastasis, including hepatogenesis. First, we assessed the new antiangiogenesis drug, regorafenib, for treatment of HCC. As shown in Fig. 1A, regorafenib treatment significantly inhibited cell growth in a panel of five HCC cell lines. Regorafenib was also found to induce apoptosis in a dose-dependent manner (Fig. 1B). In addition, regorafenib exhibited significant cell death at clinically relevant doses as measured by DNA fragmentation assay (Fig. 1C). Induction of apoptosis was further confirmed by the activation of caspase-9 and PARP cleavage in regorafenib-treated cells (Fig. 1D). These results suggest that regorafenib has significant anti-HCC activity.

To investigate the potential mechanism of action by which regorafenib induces apoptosis of HCC cells, we first examined the STAT3-related signal pathway, which has been reported to be ubiquitously activated in HCC progression. As shown in Fig. 2A, regorafenib inhibited p-STAT3 and its related proteins, including Mcl-1, cyclin D1, and surviving, in a dose-dependent manner. Regorafenib exhibited significant p-STAT3 inhibition after an extremely short treatment span (Fig. 2B). Regorafenib inhibited p-STAT3 signaling in dose- and time-dependent manners. In addition, regorafenib exhibited strong STAT3 inhibition as determined by p-STAT3 ELISA (Fig. 2C, left). Regorafenib demonstrated more potent inhibition of STAT3 than sorafenib, an approved clinical angiokinase inhibitor, which also shows p-STAT3 inhibitory activity (Fig. 2C, right). To clarify whether STAT3 inhibition induced by regorafenib sensitizes HCC cells to apoptosis, we analyzed the effect of regorafenib in PLC5 cell–overexpressing STAT3. As shown in Fig, 2D, overexpression of STAT3 almost abolished the apoptotic effect of regorafenib, indicating that inhibition of p-STA3–related signaling determines the sensitizing effect of regorafenib on HCC.

To explore the underlying mechanism by which regorafenib inhibits p-STAT3 signaling in HCC, next, we examined potential inhibitors of p-STAT3 such as SHP-1, a known SH2 domain–containing PTP involved in dephosphorylation of STAT3, to clarify its anti-HCC effect. As shown in the results presented in Fig. 3A (left), we found that regorafenib increased SHP-1 activity at the same doses that inhibited p-STAT3 in PLC5 cells. Moreover, regorafenib also activated SHP-1 in other tested HCC cell lines at 10 μmol/L (Fig. 3A, right). Importantly, both specific SHP-1 inhibitor and siRNA against SHP-1 (PTPN6) reversed the regorafenib-induced STAT3 inhibition and apoptosis, suggesting that SHP-1 may be involved in the effect of regorafenib on HCC (Fig. 3B). Furthermore, we found that regorafenib increased SHP-1 activity in purified recombinant SHP-1 proteins and SHP-1–containing immunoprecipitation lysate in vitro (Fig. 3C), indicating that regorafenib activates SHP-1 by direct interaction. As the activity of SHP-1 was strongly regulated by the autoinhibited 3D structure, we further investigated the effect of the autoinhibited N-SH2 domain responsible for autoinhibition with catalytic PTPase, on regorafenib potency. We constructed wild-type, deletion of N-SH2 (dN1), and D61 single mutant (D61A) of SHP-1 to investigate the effect of regorafenib on different SHP-1 statuses (Fig. 4A). As demonstrated in Fig. 4B, regorafenib induced significantly less inhibition of STAT3 and apoptosis in PLC5 cells expressing the dN1 and D61A mutants than in the vector, implying that regorafenib may directly affect the autoinhibition by N-SH2 and PTPase. Furthermore, dN1 decreased regorafenib-induced apoptosis in a dose-dependent manner (Fig. 4C). These results suggest that relieved SHP-1 counteracts regorafenib-induced anti-HCC effect. Regorafenib, therefore, potentially relieves autoinhibition of SHP-1 and further inhibits p-STAT3 signals directly.

To examine whether regorafenib displays anti-HCC potency in vivo, we tested regorafenib orally in an HCC-bearing subcutaneous xenograft mouse model. Regorafenib-treated mice displayed significant tumor inhibition compared with the mice treated with vehicle (Fig. 5A, left). The significant difference between regorafenib-treated (20 mg/kg/d) and vehicle-treated mice was found after 14 days of treatment (**, P < 0.05). In the end of treatment, regorafenib induced 34.72% of tumor volume in this PLC5-bearing animal model. Also, the average tumor weight of mice treated with regorafenib was lower than that of the vehicle-delivered mice in the end of treatment (Fig. 5A, right). Notably, mice treated with 20 mg/kg/d regorafenib did not lose body weight until the end of the treatment (Fig. 5B). To verify the detailed molecular mechanism determining regorafenib-induced anti-HCC potency, we collected the tumor lystate and investigated the SHP-1/p-STAT3–related signaling pathway. As shown in Fig. 5C, regorafenib inhibited p-STAT3 expression and increased the SHP-1 activity in vivo. Moreover, regorafenib showed better antitumor effect at the dose of 40 mg/kg/d treatment. Both the average tumor size and tumor weight of animals were less than 50% of those of control mice at the end of treatment (Fig. 5D). Taken together, we found that regorafenib, a multitarget antiangiokinase inhibitor, exhibited potent anti-HCC effect in vitro and in vivo through direct targeting of SHP-1/STAT3 signaling.

To investigate the clinical relevance of STAT3 signaling, we analyzed the correlation between the expression status of p-STAT3 and important clinical features of the patients, such as clinical stage, tumor differentiation, and cirrhosis severity. There were 142 patients with HCC enrolled for this study and their general characteristics were summarized as in Table 1. Surgical resection was the primary tumor-specific treatment for all our patients. Within a median follow-up of 11.34 months (interquartile range, 8.07–14.7), 16 patients (11.3%) had progression and 14 of them died (9.9%).

Among all the tumor samples examined, highly positive p-STAT3 staining was identified in 92 tumor samples (64.8%). The intensity of p-STAT3 was highly correlated with the aggressiveness of tumors; tumors with high p-STAT3 expressions were more likely to be found in patients with advanced clinical stages (P = 0.009) and poor differentiated tumors (P = 0.035; Fig. 6). The expression of p-STAT3 was not significantly correlated with patient's gender (P = 0.459), severity of fibrosis (P = 0.658) and cirrhosis (P = 0.066), and chronic viral infection (P = 0.085 for hepatitis B virus infection and P = 0.147 for hepatitis C virus infection).

Although one receptor tyrosine kinase inhibitor (TKI) has been approved and several are currently being tested in clinical trials for the treatment of HCC, the unsatisfactory response rate of kinase inhibitors still highlights the urgent need for a precise understanding of which molecule(s) determine HCC progression and may be druggable. In this study, we disclosed that SHP-1 is a potential tumor suppressor, and a direct target of regorafenib for anti-HCC. Previously, SHP-1 was found to be responsible for several immune responses in hematopoietic cells, such as trapping by inhibitory ITAM signaling to form inhibisome clusters (9), negative regulation of T-cell receptor engagement by inactivation of Lck (10), and downregulation of T-cell antigen receptor signaling by direct dephosphorylation of 3BP2 (11).

Recently, several studies have verified the role of SHP-1 in cancer progression. In breast cancer, SHP-1 expression was demonstrated to define a subset of high-grade tumors and correlated with HER-2 (12). SHP-1 also suppressed growth and increased apoptosis in prostate cancer, even under IL6 growth advantage (13). Importantly, promoter hypermethylation of SHP-1 induces ubiquitous activation of the Jak/STAT pathways in human liver cancer (14), suggesting that SHP-1 may be a useful prognostic marker and therapeutic target for cancer treatment. We proved that regorafenib directly targets SHP-1 to inactivate the STAT3-mediated signaling pathway and induces anti-HCC activity in vitro and in vivo, suggesting that SHP-1 may be a druggable protein for the treatment of HCC. Similarly, Nicola and colleagues (15) reported that SHP-1 expression determines the resistance to imatinib for chronic myelogenous leukemia treatment and lower levels of SHP-1 were also found in patients with imatinib treatment failure. In light of these findings and this study, we propose that patients with HCC with lower expression of SHP-1 may be more resistant to regorafenib than those patients with higher SHP-1 expression. Further HCC clinical study is warranted to explore the role of SHP-1 as a prognostic and druggable marker.

Currently, several STAT3 inhibitors are being tested in clinical trials, such as OPB-31121 (NCT00955812) and WP1066 (NCT01904123). Our findings show that P-STAT3 expression is strongly associated with advanced clinical stages and poor differentiation in clinical HCC tissue, further suggesting that STAT3 may be a suitable candidate for prediction of HCC progression (Fig. 6). Notably, STAT3 was believed to be a critical factor involved in microenvironment and cancer development. In our present HCC subcutaneous animal model, it is difficult for us to assay the effect of regorafenib on the tumor microenvironment and angiogenesis. As antiangiogenesis has been considered as the main mechanism of action of TKIs for the treatment of HCC, the further investigation of regorafenib on angiogenesis, tumor microenvironment, and cancer metabolism will be needed to analyze the overall anti-HCC effect of regorafenib.

Here, we reported that regorafenib inhibited p-STAT3–mediated signaling by targeting autoinhibited SHP-1 directly to induce potent anti-HCC activity in vitro and in vivo. The correlation between P-STAT3 expression and HCC prognosis further provided a mechanism-based rationale to target STAT3 by regorafenib. In conclusion, we suggest that STAT3 may be a useful prognostic marker for HCC progression and determine the sensitivity of regorafenib as a druggable target for patients with HCC.

No potential conflicts of interest were disclosed.

Conception and design: W.-T. Tai, C.-W. Shiau, H.-C. Yu, K.-F. Chen

Development of methodology: W.-T. Tai, Y.-S. Li, H.-C. Yu, K.-F. Chen

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): W.-T. Tai, P.-Y. Chu, Y.-L. Chen, Y.-S. Li, L.-J. Chen, P.-Y. Lin, H.-C. Yu

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): W.-T. Tai, M.-H. Hung, H.-C. Yu, K.-F. Chen

Writing, review, and/or revision of the manuscript: W.-T. Tai, C.-W. Shiau, M.-H. Hung, H.-C. Yu, K.-F. Chen

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Y.-L. Chen, P.-L. Chen, J.-C. Su, H.-C. Yu

Study supervision: H.-C. Yu, K.-F. Chen

This research was supported by grants NSC102-2622-B-002-013, NSC102-2325-B-002 -031, and NSC102-2325-B-002 -093 from the National Science Council, Taiwan, and grant NHRI-EX103-10246BI from the National Health Research Institutes.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.