Introduction
Liver cancer is a significant health issue, particularly in developing countries where it is inevitably fatal (1,2), and worldwide it is the third most common cause of death from cancer. In 2008, an estimated 749,000 newly diagnosed liver cancer cases and 695,000 deaths were reported (1). The 5-year survival rate of these patients is between 6.5 and 8.3% (2).
The most common form of liver cancer is hepatocellular carcinoma (HCC). It is well accepted that the primary causes of HCC are nonspecific cirrhosis, steatohepatitis and viral hepatitis, and according to clinical data the incidence of HCC and the associated mortality continue to increase (3,4). However, due to the lack of specific early symptoms and symptoms that overlap with other liver diseases, HCC is often diagnosed at an advanced stage when curative surgical resection is no longer viable. At this advanced stage of disease, liver transplantation may be an option, but it is difficult to carry out due to limited donor livers, fierce immune rejection and huge expenditure. Chemotherapy or radiation treatment may be helpful for some patients, but are usually palliative. Many HCC patients have concomitant liver cirrhosis or other liver diseases, and the above therapies have only limited applications. Thus, novel approaches for HCC patients are urgently needed to provide both preventive and curative strategies.
Towards this end, the drug metformin, commonly used to suppress glucose production by the liver (5), has recently been targeted as a chemoprophylatic agent and for treatment of various human cancers. Belonging to the biguanide class of drugs, metformin is used as a first-line therapy for diabetes mellitus type 2 (adult-onset diabetes), but epidemiological studies have shown that when used against type 2 diabetes it also reduced the risk of cancer (6,7). Moreover, according to several clinical trials, type 2 diabetes is an independent risk factor for HCC development, and use of metformin appeared specifically to be associated with a lower risk of liver cancer (8,9). Features of diabetes such as insulin resistance, glycometabolic disorder, and hepatic lipid accumulation may also occur in HCC, precancerous nonalcoholic fatty liver disease and liver cirrhosis (10–12). Metformin can increase insulin sensitivity and reduce serum levels of glucose, lipotoxicity and inflammatory cytokines, which may in turn help to suppress oncogenesis (13,14).
Metformin inhibits glucose production in the liver and improves hyperglycemia via activation of AMP-activated protein kinase (AMPK), an enzyme with an important role in insulin signaling, whole body energy balance, and the metabolism of glucose and fat. AMPK is a serine/threonine kinase, which for its activation requires phosphorylation by upstream kinases at a specific threonine residue (Thr-172) (15–19). AMPK has been considered a target for cancer therapy or prophylaxis.
Previous studies have reported that metformin is able to inhibit the viability of ovarian cancer cell lines (8) and to suppress renal carcinoma in vivo(20). In experiments with hamsters, the drug also showed preventive activity for pancreatic cancer (21). The multiple effects of metformin in cancers such as breast, ovarian, prostate and colon cancers, appear to act through cell cycle arrest, apoptosis, prevention of angiogenesis and enhancement of chemosensitivity (22–24) through AMPK-dependent or -independent pathways (25,26).
In the present study, we investigated the effects of metformin on HCC in vitro and in vivo, and the underlying molecular events.
Materials and methods
Cell culture and cell viability assay
The human HCC cell lines HepG2 and PLC/PRF/5 were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA) and were cultured in RPMI-1640 medium (Gibco-BRL, Grand Island, NY, USA) supplemented with 10% fetal calf serum (Sijiqing, Hangzhou, China), 100 U/ml penicillin and 100 U/ml streptomycin, in a humidified incubator at 37°C with 5% CO2 and 95% air.
To assess changes in cell viability, we performed a methyl thiazolyl tetrazolium (MTT) assay. Briefly, HepG2 and PLC/PRF/5 cells were seeded at a density of 3×103 cells/well in 96-well plates and cultured for 24 h. The medium was then replaced with RPMI-1640 medium supplemented with 2.5, 5.0, 10 or 20 mM metformin (American Biomol, Farmingdale, NY, USA) or 100, 250, 500 or 1,000 μM 5-aminoimidazole-4-carboxamide ribonucleoside (AICAR; Cell Signaling Technology, Inc., Danvers, MA, USA). Following the above treatments, MTT (Sigma Chemicals, St. Louis, MO, USA) at 5 mg/ml in phosphate-buffered saline (PBS) was added to each well, and the cells were incubated for an additional 4 h. The supernatant was discarded, and 150 μl of dimethyl sulfoxide was added to each well for 10 min to dissolve the formazan crystals. The optical density levels of the cell cultures were measured spectrophotometrically using a dual beam microplate reader at 490 nm.
Flow cytometric assessment of cell cycle distribution
Cells (5×105) were grown and treated with metformin (10 mM) or AICAR (500 μM) for 72 h. The cells were trypsinized and fixed in 70% ethanol overnight. The following day, the cells were stained with propidium iodide (50 μg/ml; Sigma) for 30 min at 4°C and then analyzed with a flow cytometer. The data were further analyzed using Windows Multiple Document Interface for Flow Cytometry (WinMDI) software.
DAPI staining to visualize apoptosis
To assess apoptosis after drug treatment, cells were stained fluorescently with 4′,6-diamidino-2-phenylindole (DAPI) to detect nuclear condensation and fragmentation. Cells (1×105) were cultured in medium with or without metformin (10 mM) or AICAR (500 μM) for 72 h. The cell culture medium was then discarded, and 1 μg/ml of DAPI solution (Roche, Shanghai, China) was added to the cell culture and incubated for 15 min in the dark. After that, the cells were washed with methanol and examined under a fluorescence microscope.
Nude mouse tumor model
The Ethics Committee for Animal Experimentation of the Shantou University Medical College approved the animal experiments. BALB/c-nu mice were obtained from SLC (Guangzhou, China). Freshly cultured (human hepatoma) PLC/PRF/5 cells (1×106) were injected subcutaneously into 6- to 8-week-old male nu/nu mice at 4 sites in both flanks (x2). After 1 week, metformin dissolved in PBS was intragastrically administered at 250 mg/kg body weight/day. The control group received PBS only. Tumor formation and growth were measured every 2 days when xenografts were visible. Mouse body weight was measured the day before intragastric administration and was repeatedly measured at week 7. Three hours after intragastric drug administration, serum glucose was measured with a blood glucose monitor.
Immunohistochemistry
Immunohistochemistry was performed on tumor xenograft tissues for assessment of gene expression. After animals were euthanized, tumor xenograft tissues were resected for tissue processing (fixed in 10% buffered formalin, embedded in paraffin, and cut into 4-mm sections). For immunohistochemistry, the sections were deparaffinized, rehydrated, and then incubated in 3% (v/v) hydrogen peroxide for 10 min at room temperature. The sections were subjected to antigen retrieval by heating in 10 mM (pH 6.0) citrate buffer for 20 min in a microwave oven, incubation in 20% normal serum for 50 min at room temperature, and further incubation with a rabbit anti-cyclin D1, anti-cyclin E or anti-p27KIP antibody (Maixin Bio, Fuzhou, China) overnight at 4°C. The following day, the sections were washed with PBS thrice and then incubated with a secondary antibody, followed by incubation with streptavidin-peroxidase solution (both from Vector Laboratories, Burlingame, CA, USA). The color reaction was performed using 3,39-diaminobenzidine (DAB; Sigma) and counterstaining with hematoxylin. The negative control was carried out omitting the primary antibody in the tissue sections.
The stained sections were reviewed and scored under a light microscope independently by 2 investigators. The images of the immunostained tissue sections were scanned with an Olympus charged coupled device (CCD) camera and analyzed by ImageJ (National Institutes of Health). The positive expression rate was calculated as the positive cell number/total cell number in a field. The mean of 5 low-power fields was used.
Protein extraction and western blotting
Cultured tumor cells and HCC tissues were lysed in RIPA buffer (Bocai, Shanghai, China) supplemented with a protease and phosphatase inhibition mixture (10 μl), NaF (10 μl of 100 mM), sodium orthovanadate (10 μl of 100 mM) and phenylmethylsulfonyl fluoride (PMSF; 10 μl of 100 mM).
For western blotting, equal amounts of the protein samples were resolved via SDS-PAGE and transferred to nitrocellulose membranes (Millipore, Billerica, MA, USA). The membranes were then blocked with 10% non-fat dried milk in Tris-buffered saline with Tween-20 (TBST) for 1 h at room temperature. The membranes were incubated with primary antibodies overnight at 4°C. The next day, the membranes were washed three times with TBST and incubated with horseradish peroxidase-conjugated secondary antibodies (Sigma) for 2 h at room temperature.
The peroxidase activity was detected using enhanced chemiluminescence (ECL; Pierce Biotechnology, Inc., Rockford, IL, USA) and exposed to X-ray film (Kodak, Guangzhou, China). The positive protein bands were quantitated using laser densitometry.
Statistical analyses
Data are expressed as means ± standard error and were analyzed using SPSS 17.0 software (SPSS, Inc., Chicago, IL, USA). The statistical difference between two measurements was calculated using the Student’s t-test. Results were considered statistically significant at P<0.05.
Results
Effects of metformin on HCC cell viability
To determine the effects of metformin on HCC cell viability, the HCC cell lines PLC/PRF/5 and HepG2 were treated with different concentrations of metformin or the AMPK activator AICAR for 72 h in vitro. The data showed that metformin significantly inhibited the viability of HCC cells in a dose-dependent manner (Fig. 1A), and AICAR was also able to reduce HCC cell viability (Fig. 1B).
Metformin induces cell cycle arrest at the G0/G1 phase in HCC cells
Treatment of HCC cells with 10 mM metformin for 72 h induced HCC cell cycle arrest at the G1/G0 phase (58.99% compared with 69.28% in HepG2 cells and 50.14% compared with 65.96% in PLC/PRF/5 cells). Moreover, treatment with 500 μM AICAR for 72 h replicated the metformin-induced cell cycle arrest phenomenon in these 2 HCC cell lines (Fig. 2).
Molecularly, treatment of HepG2 cell lines with metformin or AICAR induced expression of p21CIP and p27KIP, but downregulated cyclin D1 expression (Fig. 2). These data suggest that metformin and AICAR modulated expression of cyclin D1, p21CIP and p27KIP to arrest the cell cycle at the G0/G1 phase, and that AMPK activation may be responsible for this effect.
Metformin induces apoptosis of HCC cells in vitro
We next assessed the effect of metformin on induction of apoptosis of tumor cells. Treatment of HepG2 cell lines with 10 mM metformin for 24 h induced apoptosis. Apoptotic cells with condensed and fragmented nuclei were visualized by staining with DAPI (Fig. 3).
Association of metformin with AMPK activation in vitro
We further investigated expression of phosphorylated AMPK (p-AMPK), an active form of AMPK, in HCC cells. Metformin was able to activate AMPK, similar to AICAR (Fig. 4). This result indicates that AMPK activation may be involved in the antiproliferative effects of metformin on HCC cells.
Metformin inhibites HCC cell xenograft formation and growth in nude mice
To assess the in vivo effects of metformin, we produced an HCC cell xenograft model in nude mice and then treated the mice with metformin through intragastric administration (250 mg/kg body weight/day for 7 weeks) 1 week after subcutaneous injection of PLC/PRF/5 cells. Our data showed that metformin markedly suppressed the growth of tumor xenografts when compared to the isovolumic PBS-treated controls (Fig. 5A and B). In addition, the morbidity due to HCC in the mouse model was reduced from 41 to 16%, after metformin treatment (compared with the control, n>10). However, metformin did not have any effect on mouse body weight and serum glucose level (Fig. 5C and D). These data may support the view that metformin is relatively safe and contributed to the prophylaxis and treatment of HCC.
Effects of metformin on expression of cell cycle regulators and p-AMPK in vivo
To explore the molecular mechanisms responsible for the metformin-induced antiproliferative effects in vivo, expression of different cell cycle regulator genes was analyzed in the PLC/PRF/5 cell tumor xenograft tissues. Western blotting data showed that the expression of p21CIP and p27KIP proteins was significantly increased by metformin treatment when compared with that of the controls (Fig. 6A). Immunohistochemical data showed that metformin inhibited expression of cyclin D1 and cyclin E proteins, but upregulated the level of p27KIP in the metformin-treated tumor xenografts (Fig. 6B–E). Furthermore, we found that metformin treatment upregulated expression of phosphorylated AMPK protein (Fig. 7).
Discussion
In the present study, we investigated the effects of metformin in vitro and in vivo, and the underlying molecular events. Our data showed that treatment with metformin in vitro reduced HCC cell viability in a dose-dependent manner. In vitro, metformin treatment induced HCC cell cycle arrest at the G1/G0 phase and apoptosis. In the mouse HCC cell xenograft model, metformin not only blocked tumor progression, but also reduced tumor morbidity. Molecularly, metformin upregulated p21CIP and p27KIP expression, but downregulated cyclin D1 levels both in vitro and in vivo. We also found evidence to suggest that these effects of metformin may be through the upregulation of p-AMPK protein. Thus, metformin warrants further evaluation as a novel strategy for the treatment and prevention of HCC.
The conventional view of HCC risk factors focused on viral hepatitis and cirrhosis. Yet recent epidemiological data have shown an association between diabetes, obesity, and insulin resistance and the increased incidence of HCC (13). The HCC risk in type 2 diabetics was found to be as high as 7.1-fold higher than in non-diabetic patients, depending on the duration of diabetes and the treatment protocol used (27). Thus, metformin could be useful for the prevention of HCC, particularly in type 2 diabetic patients. Our current data that metformin induces cell cycle arrest at the G1/G0 phase supports this notion.
Previous studies, including in vitro experiments, animal models, and epidemiological analyses have shown that metformin use is associated with a lower rate of cancer development and that metformin inhibits tumor cell proliferation (6,20). Previous studies have revealed that metformin is able to inhibit HCC growth by arresting the tumor cell cycle at G1 phase, through regulation of the AMPK-dependent pathway (28). However, Xiong et al(31) showed that the induction of HCC cell cycle arrest and apoptosis by metformin was through an AMPK-independent pathway. In the present study, we confirmed that metformin treatment upregulated p-AMPK levels in HCC cells in vitro and in nude mouse xenografts in vivo, but further study is required to determine whether AMPK activation is essential for the antitumor effects of metformin.
Several lines of evidence indicate that AMPK activation may be involved in metformin-induced cell cycle arrest of tumor cells (22,28). Metformin was found to exert its antitumor activity through the AMPK-dependent pathway (29). HeLa cells are deficient in the liver kinase B1 (LKB1; the upstream activator of AMPK) allele and are insensitive to metformin-induced antitumor effects (25). Our current data showed that metformin treatment increased the level of AMPK phosphorylation in vivo and in vitro. Furthermore, AICAR, an AMPK activator, was able to replicate the other antiproliferative effects of metformin in HCC cells in vitro. These data indicate that metformin-activated AMPK contributes to the inhibitory effects of metformin in HCC cells. However, data that contradicts this finding has been observed (26,30), suggesting that the effects of metformin on tumor cells are AMPK-independent. The reason for this may be because different experimental conditions were used to test the effects of metformin (28,31).
We recognize that metformin is not only an AMPK activator, but also possesses antiproliferative effects, which may be involved in other gene pathways and multiple regulators. Therefore, we propose that AMPK-dependent and AMPK-independent pathways may coexist. In any case, the most destructive feature of tumor cells is uncontrolled cell cycle progression, and as a central metabolic sensor AMPK can influence cell proliferation via the cell cycle, in addition to mediating the metabolism of fatty acids, glycogen, and proteins (16,28). From this point of view, AMPK may be a target for tumor therapy.
The regulation of cell proliferation is dependent on a balance between cell division and death. In cancer cells, this balance swings toward proliferation, and cancer cells manifest dysregulated cell cycle and immortality. In the present study, metformin arrested HCC cells at the G1 phase of the cell cycle and induced apoptosis. This is consistent with the findings of Qu et al(32), and reveals the diversity of the effects of metformin. Therefore, metformin may be useful in the control of HCC progression.
In regular cell cycle progression, the transition from G1 to S phase depends on the regulation of specific cyclins and cyclin-dependent kinase inhibitors (CDKIs), including cyclin D1, cyclin E, p21CIP and p27KIP. Cyclin D1 and cyclin E promote cell DNA synthesis and cell growth. Overexpression of cyclin D1 and cyclin E promotes cancer progression (33,34). In contrast, downregulation of cyclin D1 and cyclin E expression restricts the cell cycle progression to the G0/G1 phase and inhibits tumor cell proliferation (22,35). Our current data showed that metformin treatment decreased the levels of cyclin D1 and cyclin E both in HCC cells in vitro and in nude mouse xenografts.
Moreover, p21CIP downregulates the level of phosphorylated retinoblastoma (Rb) to induce cell cycle arrest (36) and combines with proliferating cell nuclear antigen (PCNA) to reduce DNA replication (37). The cell cycle-negative regulator of p27KIP can suppress activation of cyclin-dependent kinase (CDK) (38), inactivate activated cyclin/CDK complexes (35,39), and reduce Rb phosphorylation and E2F release to inhibit gene transcription and control cell cycle progression (40). Our present in vivo and in vitro data collectively showed that metformin inhibited the expression of cyclin D1 protein, but upregulated p21CIP and p27KIP expression. These data suggest that metformin can aid in the control of HCC.
From a clinical perspective, evaluation of chemotherapeutics must include, not only their antitumor effects, but also the financial cost and side-effects. As a traditional antidiabetic agent, metformin is popular since it is relatively inexpensive and safe. In our in vivo experiment, we found that metformin did not influence the weight and serum glucose level of animals. This may be additional support for the use of metformin in antitumor therapy. However, further investigation is needed to determine the therapeutic dose range of metformin for antitumor therapy, and whether metformin treatment has adverse effects within this range. Recently, there have been several clinical chemoprevention trials using metformin and we await publication of the positive data.