Background
Hepatocellular carcinoma (HCC) is the fifth most frequently diagnosed cancer in men and the seventh in women worldwide [
1]. Surgical resection has been regarded as the main treatment for HCC [
2,
3]. However, postoperative recurrence and metastasis are still the main obstacles for long-term survival of HCC [
4,
5].
Sorafenib is the first-line oral multi-kinase inhibitor acting on advanced liver cancer. Some evidence suggest that sorafenib can block Raf⁄MEK⁄ERK signaling pathway to inhibit tumor cell proliferation and it can also target the tyrosine kinase receptor vascular endothelial growth factor receptor-2 (VEGFR-2) or platelet-derived growth factor receptor (PDGFR) to produce inhibition of angiogenesis [
6‐
8]. However, sorafenib has been proved to have limited survival benefits with very low response rates because of drug resistance [
9]. Hypoxic environment in solid tumor is one of the vital factors for the treatment of resistance [
10]. Many evidence indicate that hypoxia-inducible factors (HIFs) are essential for tumor cells to adapt to low oxygen environments [
11]. Recent research findings revealed that sorafenib could reduce the expression of HIF-1α, promoting the hypoxic response switch from HIF-1α- to HIF-2α-dependent pathways, resulting in upregulation of HIF-2α [
12]. The high expression of HIF-2α induced by sorafenib is the main cause for HCC cells resistant to therapy in hypoxia [
13].
It has been reported that, under certain circumstances, anti-angiogenic drugs could heighten invasiveness and increase lymphatic or distant metastasis [
14]. Our previous results demonstrated that relatively low dosages of sorafenib promoted HCC invasion and metastasis by downregulating tumor suppressor gene HTATIP2, also known as TIP30, which is a 30-kDa human cellular protein that was purified as a HIV-1 Tat-interacting protein [
15]. Until now, the relationship between HIF-2α and TIP30 has yet not been reported.
Metformin is a widely used drug for the treatment of type 2 diabetes [
16]. Recently, some articles pointed out that metformin may be associated with reduced cancer incidence in diabetic patients [
17,
18], including prostate, breast, pancreas, and liver cancer [
19‐
22]. In addition, it was also reported that metformin could target NLK (Nemo-like kinase) to inhibit non-small cell lung cancer (NSCLC) cell proliferation and stemness [
23]. Here, we showed that metformin could increase the sensitivity of HCC cells to sorafenib and inhibit HCC recurrence and metastasis after surgical resection.
Discussion
Surgical operation is the primary treatment for clinical HCC patients. However, recurrence and metastasis are the major factors leading to poor prognosis [
4,
5]. Sorafenib is the first-line treatment for advanced hepatocellular carcinoma. However, resistance of tumor cells to sorafenib is one of the main reasons for drug utilization [
9]. It seems that finding an effective way to inhibit the resistance may improve the efficiency of sorafenib.
Hypoxia is a prevalent phenomenon in solid tumor in which HIF-2α plays an important role. Previous results indicated that HIF-2α was connected with tumorigenesis, invasion, and metastasis [
28,
29]. Here, we revealed that metformin could sensitize hypoxic HCC cells to sorafenib and synergistically suppress the expression of HIF-2α. It has been reported that decreased TIP30 was associated with EMT [
26], and our previous results pointed out that sorafenib promoted EMT process by inhibiting TIP30 [
15]. In the present study, we found that overexpression of HIF-2α led to downregulation of TIP30 and subsequently promoted the process of EMT, and knocking-down of HIF-2α had the opposite effects. In vitro CHIP assays showed a 214-bp band was amplified immunoprecipitated with anti-HIF-2α antibodies under hypoxic conditions. More broadly, our data support a role for TIP30 as a unique HIF-2α target gene involved in the regulation of cancer recurrence and metastasis.
The previous article reported that exosome-mediated transfer of miR-122 via microRNA (miRNA)-modified adipose tissue-derived mesenchymal stem cells (AMSCs) can enhance the chemosensitivity of HCC cells [
30]. Our results showed that combination of metformin and sorafenib inhibited HIF-2α expression to a large degree, indicating that metformin could regain hypoxic HCC cells to be sensitive to sorafenib treatment in vitro. Subsequently, an orthotopic xenograft mouse model was established to explore the influence of metformin combined with sorafenib on recurrence and metastasis after surgical resection of HCC. The subcutaneous tumor was removed and cut into about 2 mm
3 slice for in situ tumor implantation, and we tried to ensure the uniformity of the tumor as far as possible before implantation. But unavoidably, there may be some disadvantages because the variations may exist in cancer cells. Zhang et al. indicated that arsenic trioxide (As
2O
3) induced HCC cancer stem cells (CSCs) differentiation, inhibited recurrence, and prolonged survival after hepatectomy by targeting GLI1 expression [
31]. Here, we found that metformin in combination with sorafenib could significantly inhibit the recurrence and metastasis of primary liver cancer in mice after surgical resection. Lower dosage of sorafenib have been found to promote invasion and metastasis of HCC cells [
15], and we showed that combined therapy could obviously reduce this effect.
In general, the above features of metformin and the encouraging results presented herein warrant future investigation of the use of metformin for combating HCC, especially in combination with sorafenib.
Methods
Cell culture and drugs
The highly metastatic human hepatocellular carcinoma cell line MHCC97H obtained from at the Liver Cancer Institute of Fudan University was maintained in Dulbecco’s modified eagle’s medium (DMEM, Gibco, UK), supplemented with 10 % fetal bovine serum, 100 units/ml penicillin, and 100 mg/ml streptomycin and cultured in a 37 °C incubator with 5 % CO2 in the air. Sorafenib (Bayer Healthcare, Leverkusen, Germany) was dissolved in dimethyl sulfoxide at a final concentration of 20 mM, and metformin (Bristol-Myers Squibb, China) was dissolved in PBS at a final concentration of 1 M for in vitro assay. Small hairpin RNA (ShRNA) construct against HIF2α (Cat. No. HSH004903-LVRH1GP) and control (Cat. No. CSHCTR001-LVRH1GP), lentiviral plasmid overexpress HIF-2α (Cat. No. EX-M0910-Lv105-5), overexpress control (Cat. No. EX-eGFP-Lv-105), and Lenti-Pac™HIV Expression Packaging Kit (Cat. No. HPK-LvTR-20) were all purchased from GeneCopoeia (America).
Western blot analysis
Cells were lysed with the lysate containing 25 mM Tris-HCl (pH 7.6), 150 mM NaCl, 1 mM EDTA, 1 % Na-deoxycholate, 0.1 % sodium dodecyl sulfate (SDS), and 1 % Triton X-100, as well as protease and phosphatase inhibitors. Protein concentration was measured with Bradford reagent (Sangon Biotech, Cat. No. B724DB0009). At least 20 μg of sample was used for detecting the protein expression. The antibodies used were as follows: anti-HIF-2 alpha/EPAS1 rabbit polyclonal antibody (1:1000, Novus, Cat. No. NB100-122), anti-TIP30 rabbit monoclonal antibody (1:1000, abcam, Cat. No. ab177961), anti-E-Cadherin rabbit polyclonal antibody (1:1000, abcam, Cat. No. ab15148), and anti-N-Cadherin rabbit polyclonal antibody (1:1000, abcam, Cat. No. ab18203).
Animals and in vivo experiment
Male BALB/c nude mice weighed 18–20 g and aged 4–6 weeks were purchased from the Animal Research Center, Beijing, China. To construct a nude subcutaneous tumor model, 1 × 10
7 MHCC97H cells were resuspended in 0.2 ml PBS and injected into the left flank of the mouse. When the tumor volume reached about 1 cm in diameter, the subcutaneous tumor was removed and cut into about 2 mm
3 slice, and the pieces from one tumor of one mouse were reimplanted into all the subsequent mice. The recipient mice were anesthetized with pentobarbital sodium salt (1 %, 25 mg/kg), and a small piece of tumor was implanted into the left lobe of the liver. Fourteen days after the orthotopic implantation, a second operation was carried out to remove the lobe where the tumor was implanted. Then, the mice were randomly divided into four groups. On the third day, after tumor resection, the mice were orally treated either with 0.9 % sodium chloride (control), 30 mg/kg sorafenib (sorafenib) [
15], 200 mg/kg metformin (metformin) [
32], or 30 mg/kg sorafenib in combination with 200 mg/kg metformin (sorafenib + metformin) once daily. All the drugs were dissolved in 0.9 % sodium chloride. The mice were treated for 37 days and killed 48 h after the last treatment. The tumor volume (
V) was measured with vernier caliper and calculated with
V = 1/2(length × width
2). The lung tissues were fixed with 10 % formaldehyde, and the HCC tumor tissues were cut into two parts for subsequent fixation with 10 % formaldehyde or frozen resection.
Immunohistochemistry
The tumor tissue samples and lung tissues were fixed by 10 % formalin, embedded with paraffin, and cut into 5-mm-thick sections. In order to observe the metastatic node in the lung, the lung tissues were stained with hematoxylin and eosin. For immunohistochemistry, the tumor tissue sections were deparaffinized, rehydrated, subsequently subjected to antigen retrieval with 121 °C for 5 min, and incubated with 3 % H2O2 for 10 min to inactivate endogenous peroxidase. The specimens were blocked with 10 % goat serum for 1 h, and then incubated with primary antibody of Ki67 or CD31 overnight in 4 °C. The next day, the slice was incubated with secondary antibody for 1 h at room temperature. Then, the slice was colored by DAB Substrate-Chromogen System. TUNEL was detected using the TUNEL Apoptosis Detection Kit (KeyGENBioTECH, China, Cat. No. KGA7022).
Cell viability and apoptosis assay
MHCC97H cells were plated in 96-well plates at 4000 cells/well (n = 6) containing 100 μl of DMEM +10 % FBS treated with 400 μM CoCl2 and cultured for 24 h, then incubated with sorafenib or metformin for another 48 h. DMEM (100 μl) and CCK8 (10 μl) were added to each well and incubated for 2 h. Then, the absorbance was detected with a microplate reader at a test wavelength of 450 nm.
Annexin V/PI was applied to investigate the impact of sorafenib or metformin on cell apoptosis. After treatments, the cells were added with 5 μl Annexin V and 10 μl PI staining supplied by the Annexin V-FITC Apoptosis Detection Kit (SANGON, Cat. No. BS6336), then the results were measured by flow cytometry using a FACS flow cytometer (Becton Dickinson verse, San Jose, CA).
Chromatin immunoprecipitation assay
MHCC97H cells were fixed at 37 °C for 10 min with 1 % formaldehyde. The cells were then collected and lysed on ice for 30 min in cell lysis buffer (50 mM EDTA, 1 % SDS, 50 mM Tris-HCl) containing protease inhibitors and 1 mM PMSF. Nuclear chromatin was broken into small fragments by sonicating the nuclear lysate on ice using a Misonix Sonicator 3000 equipped with a microtip. One percent of the volume of samples (input) was saved for the subsequent PCR analysis. The samples were precleared with protein A Sepharose (GE Healthcare, Cat. No. 10043746). Equal aliquots of precleared chromatin samples were incubated overnight at 4 °C with either specific rabbit HIF-2α or nonspecific rabbit antiserum IgG. Immune complexes were collected by incubation with 20 ml protein A Sepharose at 4 °C for 4 h. The complexes were washed once with buffer I (1 % Triton X-100, 2 mM EDTA, 150 mM NaCl, 20 mM Tris-HCl, 0.1 % SDS), buffer II (1 % Triton X-100,2 mM EDTA, 50 mM NaCl, 20 mM Tris-HCl, 0.1 % SDS), buffer III (0.23 mM LiCl, 1 % NP40, 1 % deoxycholate, 1 mM EDTA, 10 mM Tris-HCl), and TE buffer (10 mM Tris-HCl (pH = 8.0),1 mM EDTA), respectively. Immune complexes were then eluted from the beads by incubation twice under agitation at 37 °C with 100 μl of elution buffer (0.1 M NaHCO3, 1 % SDS). The eluted material and input was cross-linked at 65 °C for 6 h. Immunoprecipitated DNA was purified by PCR purification kits. Primers used for PCR correspond to the TIP30 promoter region, primers: 5′ primer: 5′-CAAACTTAGGAAGGG TCGCG-3′; 3′ primer: 5′-ATCAGAGCATCCCACCTTCC-3′.
Statistical analysis
All data were expressed as mean ± SD. The Student t test was used for the comparison of measurable variants of the two groups. P < 0.05 was defined as statistically significant. All statistical analyses were done using statistical software (SPSS 13.0 for Windows; SPSS, Inc., Chicago, IL).
Acknowledgments
This work was supported by the National Natural Science Foundation of China (No. 81372635, No. 81101871, No. 81572434, and No. 81201644) and the Major Program of Natural Science Foundation of Tianjin (No. 11JCZDJC18800).
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
ABY, MQC, ZGG, BFZ, JRG, HYZ, HKL, YLC, FF, WZ, TQS, QL, and XLZ carried out all the experiments, prepared figures and drafted the manuscript. ABY, HFY, AND HCS participated in data analysis and interpretation of results. TZ designed the study, participated in data analysis and interpretation of results. All authors read and approved the manuscript.