Introduction
Osteosarcoma (OS), the most common tumor of the bone, is a rare malignant neoplasm affecting mostly children and adolescents. Although long-term survival in high-grade OS has markedly improved in the last decades, owing to neoadjuvant chemotherapy (1), data emerging from clinical studies show that 35–45% of OS patients have a natural or acquired drug-resistance (2).
The possibility of identifying tumor molecular background and signaling pathway key end-points may provide new targets for planning tailored therapies combined with conventional therapeutic modalities (3).
Metformin (1,1-dimethylbiguanide hydrochloride) belongs to the biguanide class of oral hypoglycemic agents and is widely used as an antidiabetic drug (4,5) by regulating glucose homeostasis and reducing insulin resistance.
Recent evidence indicates that metformin may reduce the risk of cancer and improve prognosis and that in patients with type 2 diabetes it reduces the risk of cancer (6–8).
In vitro and in vivo data (2,9–11) emphasized the role of 5′-monophosphate-activated protein kinase (AMPK) in action mechanism of metformin and demonstrated that the reduction of tumor cell proliferation and survival is mediated by inactivation of mTOR in both insulin-dependent and -independent pathways (12).
AMPK is a heterotrimeric serine/threonine kinase composed of a catalytic α subunit, and two regulatory subunits, β and γ (13,14). Activation of AMPK requires an allosteric change induced by AMP, as well as phosphorylation at Thr172, that inhibits the downstream target mTOR implicated in protein synthesis and proliferation (15) and promotes vascular endothelial growth factor expression and angiogenesis (16–20).
In vitro and in vivo studies demonstrated that metformin inhibits tumor cell growth and survival in numerous tumors (8,21–23), emphasizing its role as an antineoplastic agent through a variety of responses including inhibition of growth factor signaling pathway, and/or cell arrest in G1 phase (8,24,25).
The present study investigated the antitumor effects of metformin on OS cell lines alone and in combination with cisplatin (CDDP), a DNA-damaging chemotherapeutic drug frequently used in OS patients.
Findings of the present study indicated that metformin may sensitize OS cells to CDDP through inactivation of critical intracellular end-points and lengthening of cell cycle phases.
Materials and methods
Reagents
Anti-cyclin D1 (HD11) and anti-p-p53 (hSer20) were obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). Anti-phospho AMPKα (Thr172), anti-AMPKα, anti-phospho-p70S6K (S6K1) (Thr389), anti-IGF-1Rβ, anti-phospho Chk1 (Ser345) were purchased from Cell Signaling Technology (Beverly, MA, USA). Anti-cyclin A and anti-cyclin E were obtained from Calbiochem, Merck KGaA, (Darmstadt, Germany). Anti-actin was from Sigma Chemical Co., (St. Louis, MO, USA). Horseradish peroxidase-conjugated anti-rabbit IgG, anti-mouse IgG were purchased from GE Healthcare. Enhanced chemiluminescent substrate LiteAblot Plus was obtained from EuroClone S.p.A (Pero, Milan, Italy).
Metformin was obtained from Sigma-Aldrich Biotechnology (St. Louis, MO, USA), and diluted in PBS 1X to make a 1 M stock solution that was stored at −20°C. It was used across a range of concentrations at 0, 5, 10, 20 and 40 mM diluted in media.
Cisplatin was purchased by Teva Pharmaceuticals B.V. (Utrecht, The Netherlands) and was stored at 4°C; it was used across a range of concentrations at 0.01, 0.1, 1.0, 10 and 100 ng/ml diluted in media.
Cell lines and culture conditions
Human OS cell lines U2OS (pRB+/+, p53+/+), 143B (pRB+/+, p53+/+) and MG63 (pRB+/+, p53−/−) were obtained from the American Type Culture Collection (ATCC, Rockville, MD, USA) and cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), L-glutamine (2 mM), 100 U/ml penicillin and 100 μg/ml streptomycin (Invitrogen) at 37°C in a 5% CO2 humidified incubator. Cells were routinely passed when they reached ~80% confluence.
Cell growth and sensitivity study
The number of adherent, viable cells was assessed microscopically using an improved Neubauer haemocytometer and proliferation was assessed as the percentage of cells that excluded 0.2% trypan blue. Cells were seeded at 100,000/well in 6-well plates and incubated in medium containing 10% FBS. Twenty-four hours after seeding, cells were treated either with or without increasing doses of metformin (0, 5, 10, 20 and 40 mM) for 24, 48 and 72 h. After 24, 48 and 72 h, cells were washed once with Dulbecco’s phosphate-buffered saline (PBS) 1X, harvested by trypsinization and cell number was determined using trypan blue.
IC50 and IC30 values, defined as the concentration of drugs inhibiting cell growth by 50 and 30%, respectively, were calculated for experiments with 72 h of treatment.
Cells were also treated with increasing doses of CDDP (0.01, 0.1, 1.0, 10 and 100 ng/ml) and cytotoxicity was evaluated as cell viability up to 72 h.
For combined treatment, cells were treated at the same time in combination with metformin IC30 and CDDP at different concentrations for 72 h; cells were also treated in sequential manner with metformin IC30 for 72 h, followed by 24 h of CDDP treatment at different concentrations.
Flow cytometry for apoptosis
OS cells were seeded at 100,000/well in 6-well plates, allowed to attach overnight, and incubated with or without an IC50 dose of metformin for 48 and 72 h. According to the protocol kit (MEBCYTO Apoptosis kit; MBL International, Woburn, MA, USA), the adherent cells were trypsinized, detached, and combined with floating cells from the original growth medium, centrifuged and washed twice with PBS 1X. Cells were re-suspended in 500 μl of staining solution containing FITC-conjugated Annexin V antibody and propidium iodide (PI) for 30 min and analyzed by flow cytometry.
The number of viable (Annexin−/PI−), apoptotic (Annexin+/PI−) and necrotic (Annexin+/PI+) cells were determined with the CellQuest Software (BD Biosciences, San Jose, CA, USA), using a peak fluorescence gate to exclude cell aggregates during cell cycle analysis in a FACSCalibur flow cytometer (Becton-Dickinson, San Jose, CA, USA).
Cell cycle analysis by FACS
OS cells were plated in 6-well plates (100,000 cells/well), allowed to attach overnight, and incubated with IC50 doses of metformin. After 48 and 72 h they were harvested by trypsinization, fixed with 70% ethanol and washed with appropriate buffer (PAT) several times. After α-bromodeoxyuridine incorporation and α-mouse FITC incubation as secondary antibody, cells were stained for total DNA content with a solution containing PI (1:5 in PAT). Cell cycle distribution was then analyzed with a FACScan flow cytometer (Becton-Dickinson).
Protein extraction and western blot analysis
Expression levels of proteins were determined by western blot analysis. After 48 h of IC50 metformin incubation, cells were washed three times with PBS and lysed in 100–400 μl lysis buffer [20 mM Tris-HCl (pH 7.5)], 150 mM NaCl, 2.5 mM sodium pyrophosphate, 1 mM β-glycerol phosphate, 1 mM Na3VO4, 1 mM EGTA, 1% Triton and complete protease inhibitor mixture inhibitors from Roche Diagnostics (Laval, QC, Canada). Cellular debris was removed by centrifugation at 14,000 × g for 20 min at 4°C. Following assay for total protein (Bio-Rad Laboratories, Mississauga, ON, Canada), clarified protein lysates (50 μg) were boiled for 5 min and analyzed by 8.0–15% SDS-polyacrylamide gel, followed by blotting at 40 V for 1 h and 100 V for 2 h. Blots were probed with anti-p-p53 (Ser20) (1:200), anti-IGF-IRβ (1:200), anti-phospho-AMPKα (Thr172) (1:1,000), anti-AMPKα (1:1,000), anti-phospho-p70S6K (S6K1) (Thr389) (1:1,000), anti-phospho Chk1 (Ser345) (1:1,000), anti-cyclin A (1:300), anti-cyclin E (1:200), and anti-cyclin D1 (1:200). Horseradish peroxidase-conjugated anti-rabbit IgG or anti-mouse IgG were used as secondary antibodies. The signal was visualized by enhanced chemiluminescent substrate LiteAblot Plus (EuroClone S.p.A.) and quantified using GS-800 imaging densitometer (Bio-Rad Laboratories, Hercules, CA, USA). A rabbit anti-actin antibody was used as control.
Statistical analysis
All experiments were performed three times and results are expressed as means ± SD. Significance was analyzed by the Student’s t-test and a probability value of P≤0.05 was considered to indicate a statistically significant difference.
Results
Susceptibility of OS cell lines to metformin
When OS cell lines were exposed to increasing doses of metformin (0–40 mM), a progressive loss of proliferation up to 72 h was observed when cell growth decreased by 75% for U2OS, 89% for MG63 and 82% for 143B (Fig. 1).
Cell sensitivity evaluation indicated that U2OS, MG63 and 143B were sensitive to metformin with IC50 mean values at 72 h of 9.13±0.3, 8.72±0.4 and 7.29±0.7, respectively, and IC30 mean values of 4.11±0.7, 6.2±1.1 and 3.2±0.4, respectively.
Effect of metformin on cell cycle and apoptosis
Following exposure of U2OS to IC50 dose of metformin, cell cycle analysis revealed a transient arrest in G2 phase at 48 h, while a longer exposure (72 h) caused accumulation of cells in S phase (Fig. 2) with a significant time-dependent induction of apoptosis (from 4.6% in non-stimulated cells to 17.2 and 21.7%, respectively, in stimulated cells) (Fig. 3).
Conversely, 143B responded to the IC50 doses of metformin with relevant arrest of cells in G1 at 48 h associated with a decrease of number of cells in S and G2 phase. The following 72 h treatment resulted in lengthening of S phase, concomitant with a significant decrease of G2 phase (Fig. 2) and a moderate induction of apoptosis when compared to non-treated cells (8.10% non-treated cells, 8.86% at 48 h and 11.20% at 72 h) (Fig. 3).
In MG63, metformin treatment was effective only at the 72 h with accumulation of cells in G1 and G2 phases concomitant with strong decrease in S phase (Fig. 2). No cases showed apoptotic induction by Annexin V-FITC assay (7.6% in non-treated cells, 6.79% at 48 h, 8% at 72 h) (Fig. 3), suggesting a predominant cytostatic effect of metformin exposure.
Protein analysis
All OS cell lines were positive to IGF-IRβ and total AMPKα without showing changes in expression levels after metformin exposure. However, at 48 h of IC50 treatment, phosphorylation level of p-AMPKα Thr172 increased in all cell lines and accumulation of p53 (Ser20) was seen in wild-type-p53 U2OS and 143B.
p70S6K phosphorylated at Thr389, substrate of mTOR activity, markedly decreased after treatment (Fig. 4).
When proteins involved in cell cycle control were analyzed, both wild-type U2OS and 143B cells showed increased expression of Chk1 (Ser345) associated with downregulation of active cyclin A and cyclin E. No significant changes in the volume of electrophoretic bands were seen for MG63. In parallel, we observed a loss of cyclin D1 expression in 143B and MG63 and to a lesser extent in U2OS (Fig. 4).
Susceptibility of OS cell lines to CDDP
All OS cell lines were exposed to increasing doses of CDDP up to 100 ng/ml for 72 h; MG63 and 143B did not show cell growth inhibition, while U2OS had a slight reduction of 30% with the maximum dose of CDDP (Fig. 5).
Metformin sensitizes OS cells to CDDP
First U2OS, 143B and MG63 were exposed to increasing concentrations of cisplatin (0.01–100 ng/ml) combined with sub-toxic doses of metformin (IC30) for 72 h.
Data demonstrate that U2OS and 143B responded to simultaneous treatment with reduction of cell proliferation of 33% (P<0.01) and 60% (P<0.001), respectively, when compared with CDDP alone showing a synergistic effect up to 1.0 ng/ml of CDDP for U2OS and 100 ng/ml for 143B. MG63 responded to a lesser extent by reduction of cell proliferation of 27% (P<0.05 at maximum dose of CDDP). An antagonistic effect was observed between the two drugs at any dose.
Subsequently, we evaluated whether pre-treatment with metformin better sensitizes OS cells to CDDP treatment by administering the drugs in sequence. OS cells were exposed to IC30 metformin for 72 h, followed by increasing doses of CDDP for 24 h.
In U2OS and MG63, cell proliferation dropped by 78% (P<0.001) and 44% (P<0.01), respectively, with respect to CDDP alone, while 143B responded with a percentage of decrease equal to that of simultaneous treatment (60%) (P<0.01) (Fig. 5).
When CDDP was administered after metformin, a synergistic interaction was seen in all cell lines.
Discussion
The first choice in OS treatment consists of combined chemotherapeutic treatments often associated with serious problems, such as frequent acquisition of drug-resistant phenotypes and toxic side-effects that impair the quality and expectancy of life in sarcoma patients.
Identification of critical end-points implicated in the control of tumor cell survival (26) may provide the rationale for new combined regimens able to overcome conventional treatment failure.
Several experimental approaches have demonstrated the therapeutic potential of mTOR inhibitors (27) and the strengthening of cell response to anticancer agents through checkpoint activation and arrest of cell cycle (28,29).
Evidence shows that metformin may inhibit tumor cell growth (30) and enhance the effect of chemotherapy through different anticancer mechanisms including insulin-dependent and/or -independent activity (12,31).
Our data show that OS cell lines differing in proliferation, transmigration and genetic background (32) respond to metformin by decreasing cell proliferation through cell cycle lengthening associated or not with apoptosis induction.
This effect appears to be correlated with increased expression of AMPKα phosphorylated at Tyr172 and inhibition of mTOR downstream signaling pathway measured by dephosphorylation of p70S6K, resulting in an inhibition of protein synthesis and cell growth (33).
Some reports support the hypothesis that inhibition of cell proliferation by AMPKα activation is determined other than by mTOR signaling inhibition, by arresting cell cycle through activation of phospho-p53 and downregulation of cyclin-dependent kinase (CDK) activity (34,35).
It is well known that cell cycle is regulated by phosphorylation and dephosphorylation events controlled by CDK/cyclin complexes and CDK inhibitors that arrest cell growth at G1/S and/or G2/M checkpoints (28,36).
Our data showed that in wt-p53 U2OS and 143B cell lines, metformin treatment induced accumulation of p53 (Ser20) associated with apoptosis induction and prevalent lengthening of S phase after long-term exposure (72 h). This delay in cell cycle progression resulted from activation of phospho-Chk1 at Ser345 that activates S and G2 checkpoints through downregulation of cyclin A and cyclin E. Evidence that Chk1 contributes to cell cycle checkpoints in human cells comes from studies showing that Chk1 is an important regulator of S phase arrest and its disruption abrogates S and G2 checkpoints (37,38). These events may contribute to sensitize our wt-p53 OS cell lines to CDDP showing a synergistic effect with metformin both in combined and sequence treatments. Null-p53 MG63 where no activation of phospho-Chk1 was seen, responded to long-term exposure of metformin with prevalent accumulation of cells in G1 associated with downregulation of cyclin D1 without apoptosis induction, suggesting cytostatic rather than cytotoxic effect. Ben Sahra et al(8) demonstrated that in prostate cancer, the block of cell cycle in G1 by metformin is not mediated by the AMPK pathway. By contrast, in breast cancer, inhibitors of AMPK induced downregulation of D1 and G1 arrest even in mut-p53 cells (35).
Moreover, CDDP in sequence with metformin was more effective in decreasing MG63 cell proliferation than in simultaneous treatment, where the two agents presented antagonistic effects.
These results show that treatment with metformin induces significant growth inhibition of OS cell lines through arrest of cell cycle and decrease of S6K activity mediated by AMPKα phosphorylation. In addition, metformin may sensitize OS cells otherwise resistant to CDDP in a p53-independent manner through synergistic drug-drug interaction.
Our data may have clinical relevance for novel therapeutic strategies for the treatment of OS.