MET was provided free of charge by Sumitomo Dainippon Pharma Co., Ltd. (Osaka, Japan). GEM was purchased from Eli Lilly Japan (Hyogo, Japan).

The human pancreatic cancer cell line, BxPC-3, was purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA) and cultured in RPMI-1640 medium (Wako Pure Chemical Industries Ltd., Osaka, Japan) supplemented with 10% (v/v) heat-inactivated fetal bovine serum (FBS; Invitrogen/Thermo Fisher Scientific, Waltham, MA, USA) in a humidified 5% CO2 incubator at 37°C. BxG30, a moderately GEM-resistant cell line, was established as previously described (18). Briefly, the BxPC-3 cells were allowed to acclimatize by stepwise exposure to increasing concentrations of GEM beginning at 1.5 ng/ml and increasing by 1-10 ng/ml increments at every passage for 6 months. A BxPC-3-derived cell line cultured in the presence of a final concentration of 30 ng/ml GEM was named BxG30.

Animals were maintained according to institutional regulations in facilities approved by the Animal Care Committee of Kitasato University (Tokyo, Japan) and in accordance with Japanese government guidelines for animal experiments. The protocol of the animal experiments was reviewed and approved by The Laboratory Animal care and use Committees of Kitasato University on April 1st, 2013 (approval no. 13023). Animal experiments were performed in accordance with the ethical guidelines of the Kitasato Institute. Isoflurane was used as an inhaled anesthetic. Humanitarian endpoints, such as performing euthanasia treatment at an appropriate time, when unbearable pain was involved, were considered. The Institutional Animal Care and Use Committee Guidebook (20) was referred to as regards the humanitarian endpoints. In this study, no mice had to be euthanized due to poor conditions before the end of the experiment.

During a preliminary toxicity test of MET monotherapy, certain adverse events, such as appetite loss and body weight loss, occurred at doses of 800 mg/kg and increased significantly from the dose of 1,500 mg/kg. Finally, we decided on a 600 mg/kg dose of MET, as this induced a consistent anti-tumor effect without any adverse events. For the purposes of this study, 96 male BALB/c Slc-nu/nu mice (6 weeks old) were obtained from CLEA Japan Inc. (Tokyo, Japan). The body weights of the mice ranged from 15 to 19 g on arrival. The mice were maintained in a specific-pathogen-free condition (temperature, 23±2°C; humidity, 55±10%) and were provided with autoclaved food and water. After being allowed to acclimatize for 1 week, 2×10⁶ BxPC-3 and BxG30 cells in 0.1 ml of 1% phosphate-buffered saline (PBS) were injected into both the flanks of the mice. The mice were randomly divided into 4 groups as follows: i) The control group [no treatment, final tumor number (n)=10]; ii) the GEM-treated group (80 mg/kg, n=11); iii) the MET-treated group (600 mg/kg, n=11); and iv) the combination treatment group (GEM + MET, n=12). The BxG30 cells were injected into the other mice, and the mice were divided into 4 groups according to the treatments (control group, n=14; GEM group, n=13; MET group, n=10; GEM + MET group, n=14). Treatment was initiated 2 weeks following implantation. MET was diluted in PBS to 600 mg/kg and administered orally every day for 4 weeks. GEM was dissolved in PBS and injected intraperitoneally at a dose of 80 mg/kg every week (on days 15, 22, 29 and 36). Estimated tumor volumes and body weights were measured each week. The estimated tumor volume was calculated using the Battelle Columbus Laboratories Protocol according to the following formula (21): TV = (A×B²)/2, where TV = tumor volume, A = major axis and B = minor axis.

Following sacrifice on day 42, the tumors were dissected and weighed. The estimated tumor volumes of the control group (C) and treatment groups (T) were calculated as described above. The relative tumor volumes (TRW and CRW) were calculated at defined time points by evaluating the ratio of (T) and (C) to the estimated tumor volumes at the beginning of treatment, and the treatment to the control ratio (T/C%) was calculated as TRW divided by CRW. The minimal T/C% over the treatment period was assessed and used to calculate the therapeutic effect. A value of T/C% <50% was required for the treatment to be considered effective.

Although some tumor ulceration was observed, all mice exhibiting ulceration, bleeding or necrosis also exhibited good general conditions without any signs of infection, anemia or abnormal behavior. Thus, no treatment was administered for the ulcerations as it was not deemed necessary.

The BxG30 cells were seeded at a density of 2.2×10³ cells per well in 96-well plates containing culture medium supplemented with 10% FBS. After 24 h, the cultures were washed and incubated in medium alone (control) or medium containing MET or GEM at various concentrations (final concentrations of MET: 0, 1.65, 2.475, 3.3 and 4.125 µg/ml; final concentrations of GEM: 0, 2, 3, 4, 5 and 10 ng/ml) in quadruplicate. The number of viable cells was counted after 72 h using a Cell Counting Kit-8 (Nacalai Tesque, Inc., Kyoto, Japan) according to the manufacturer’s instructions. The assay reagent was a tetrazolium compound (WST-8) that is reduced by live cells to produce a colored formazan product whose absorbance can be measured at 450 nm. The quantity of the formazan product is directly proportional to the number of live cells in the culture. The inhibition index (I.I., %) was calculated using the following formula: I.I. (%) = (b−c)/(b−a) ×100, where a = the optical density (OD) of the cells, b = the OD of the cells + WST-8 reagent, and c = the OD of the cells + WST-8 reagent + anti-tumor agent) and IC₅₀ values were calculated. A classical isobologram was used to evaluate the synergistic effects of GEM and MET. All experiments were repeated at least 3 times.

The BxG30 cells were pre-cultured at a density of 1.5×10⁵ cells per well in RPMI-1640 medium containing 10% FBS for 24 h. After removing the culture medium and washing with PBS, the cells were treated with MET and GEM (final concentrations of MET: 0, 0.825, 1.65 and 3.3 µg/ml; final concentrations of GEM: 0, 1, 2, 4 and 10 ng/ml). The cells were cultured for 24 h and lysed in lysis buffer [Cell Signaling Technology (CST) Japan, Tokyo, Japan] containing Protease Inhibitor Cocktail Set III (Wako Pure Chemical Industries, Osaka, Japan) and Phosphatase Inhibitor Cocktail Solution I (Wako Pure Chemical Industries). The lysates were centrifuged at 23,000 × g for 20 min at 4°C and the supernatants were collected. The samples were stored at -80°C and used for western blot analysis.

For protein extraction, cells were plated into 6-well cell culture plates (Corning, Inc. Corning, NY, USA) at a density of 1.5×10⁵ cells in serum-containing medium and then incubated for 24 h at 37°C. The cells were rinsed twice with PBS and scraped into Cell Lysis Buffer (CST) and Phosphatase Inhibitor Cocktail Solution I (Wako, Osaka, Japan) was added. Following incubation for 20 min on ice, cell lysates were cleared by 20 min of centrifugation at 15,000 × g at 4°C. Protein concentration was determined using the BCA Protein Assay Reagent (Pierce, Rockford, IL, USA), and conditioned to 15 µg per lane. Samples were boiled for 3 min at 100°C and 4X LDS Sample Buffer (Life Technologies Japan Ltd., Tokyo, Japan), and 5% 2-mercaptoethanol (Nacalai Tesque Inc.). Equal amounts of total protein were separated on a 12.5% SDS polyacrylamide gel at a constant current of 20 mA. Separated proteins were transferred to PVDF membranes (Life Technologies Japan Ltd.) at 30V by using iBlot^® Gel Transfer Device (Life Technologies Japan Ltd.). The membrane was blocked with 5% dry-milk with 0.1% Tris Buffered Saline-Tween-20 (TBS-T) for 60 min. The membrane was then incubated overnight at 4°C and with primary antibody, and then probed with secondary antibodies for 1 h at room temperature. Chemiluminescence was developed using the ECL Select Western Blotting Detection System and the band intensities were then detected using Image Quant LAS 500 (both from GE Healthcare Japan, Tokyo, Japan) and analyzed using NIH Image J software. The BxG30 cells were treated with several concentrations of MET and GEM (final concentrations of MET: 0, 0.825, 1.65 and 3.3 µg/ml; final concentrations of GEM: 0, 1, 2, 4 and 10 ng/ml).

The following antibodies were used for western blotting: pS6 rabbit antibody, S6 mouse antibody, HIF-1α rabbit antibody (all from CST; #2317S, #3716S) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) rabbit polyclonal IgG (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA; #sc-25778) as primary antibodies, and horseradish peroxidase (HRP)-conjugated anti-rabbit IgG antibody (CST; #7074P2) or HRP-conjugated anti-mouse IgG antibody (GE Healthcare Japan, Tokyo, Japan; NA931-100UL) as secondary antibodies. Each sample was analyzed 3 times independently.

On the day of the experiment, the culture medium was replaced with fresh RPMI-1640 medium containing 10% FBS. Culture dishes were placed in a modular incubator chamber (Billups-Rothenberg Inc., San Diego, CA, USA), a humidified airtight chamber with air flow valves. To simulate hypoxic conditions, the cells were incubated with 1% O₂, 5% CO₂ and 94% N₂ for 48 to 78 h.

ELISA was performed to quantify secreted VEGF from BxG30 cells using a Novex ELISA kit (Life Technologies Japan Ltd.) according to the manufacturer’s instructions. The BxG30 cells were incubated under hypoxic conditions as described above for 72 h with various concentrations of MET (final concentrations: 0.825, 1.65 and 3.3 µg/ml) and GEM (final concentrations: 1, 2 and 4 ng/ml). At the end of the ELISA procedure, stop solution was added and the absorbance of each well was measured at 450 nm using a Multiskan FC instrument (Thermo Fisher Scientific). Each sample was analyzed 3 times independently.

Dunnett’s test was used to compare the results from the treatment groups with the control group. One-way factorial ANOVA followed by the Dunnett’s test as a post hoc test was used to analyze multiple comparisons between the control group and each of the treatment groups. Differences were considered statistically significant at P<0.05.

The estimated tumor volumes and body weights of the mice were measured each week following implantation of the BxPC-3 xenografts. The final tumor weights were 0.59±0.05 g (mean ± SEM) in the control group (n=10), 0.32±0.07 g in the GEM-treated mice (n=11), 0.42±0.08 g in the MET-treated mice (n=11) and 0.23±0.06 g in the mice treated with GEM and MET (n=12) (Fig. 1A and B). Both the GEM-treated groups exhibited significantly reduced tumor weights compared with the control group. However, there were no differences in tumor weight between the mice treated with MET alone and the control animals. The minimal T/C% on day 42 was 55.6% in the GEM-treated mice, 88.7% in the MET-treated mice and 48.8% in the mice treated with GEM and MET (Fig. 1C).

The anti-tumor activity of MET against BxG30, a GEM-resistant PDAC cell line, was evaluated using a BxG30 xenograft model. The final tumor weights were 0.26±0.05 g in the control mice (n=14), 0.21±0.05 g in the GEM-treated mice (n=13), 0.15±0.06 g in the MET-treated mice (n=10) and 0.11±0.03 g in the mice treated with GEM and MET (n=14). Compared with the control animals, the final tumor volumes were significantly decreased only in mice treated with both GEM and MET (Fig. 2A and B). The T/C% was 80.2% in the GEM-treated mice, 54.0% in the MET-treated mice and 47.2% in the mice treated with both GEM and MET. The anti-tumor effect of GEM against BxG30 cells was clearly limited. The MET-treated animals exhibited a satisfactory inhibition of tumor growth, although the minimal T/C% was <50% on day 42 only in the group of mice treated with both GEM and MET (Fig. 2C). These data indicated that combination therapy with GEM and MET exerted marked anti-tumor activity against GEM-resistant PDAC.

It was decided that the specimens were inadequate for use in further experiments, as some specimens harvested from both the BxPC-3 and BxG30 xenograft model were necrotic in some places with ulceration of the skin and/or bleeding from tumors.

We initially evaluated the cytotoxic effects of MET and GEM in BxG30 cells using an in vitro WST cytotoxicity assay. Both agents inhibited cell growth in a concentration- and time-dependent manner (data not shown). The IC₅₀ values (means ± SEM) of MET and GEM were 3.36±0.13 µg/ml and 3.75±0.22 ng/ml, respectively. The IC₅₀ values of MET and GEM were connected by a dotted line in a classical isobologram to evaluate potential synergistic effects (Fig. 3), but none were observed.

The phosphorylation of ribosomal protein S6 (pS6), one of the most important downstream effectors of the mTOR signaling pathway, was assessed by western blot analysis. The relative expression of pS6 and total S6 (pS6/S6) was markedly decreased in a dose-dependent manner in the cells treated with MET. S6 phosphorylation was inhibited by incubation with MET, but not with GEM (Fig. 4A and B). The pS6/S6 ratios of each MET concentration were 84.08±29.66% (0.825 µg/ml), 78.48±6.74% (1.65 µg/ml) and 42.39±2.74% (3.3 µg/ml) relative to the control, respectively. Treatment with 3.3 µg/ml of MET significantly inhibited the activation of S6 compared with the control cells (P<0.05). HIF-1α is a well-known downstream target of mTOR and its expression level was also evaluated by western blot analysis. Hypoxic conditions clearly induced the overexpression of HIF-1α. When the BxG30 cells were treated with various concentrations of MET, the expression of HIF-1α was suppressed in a dose-dependent manner. Treatment with >1.65 µg/ml of MET significantly suppressed HIF-1α expression, even under hypoxic conditions (Fig. 5).

The production of VEGF by the BxG30 cells was evaluated by ELISA under hypoxic conditions. The results revealed that VEGF production was suppressed by MET treatment in comparison with the untreated cells (P<0.01). However, no significant difference was observed in the secretion of VEGF between the GEM-treated and untreated cells (Fig. 6).