Three human pancreatic cancer cell lines (Bxpc-3, Mia PaCa-2 and Panc-1) were purchased from American Type Culture Collection (ATCC, Rockville, MD, USA). The Mia PaCa-2 and Panc-1 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Rockville, MD, USA) supplemented with 10% fetal bovine serum, 100 U/ml penicillin G and 100 U/ml streptomycin. The Bxpc-3 cells were cultured in RPMI-1640 (Gibco) containing supplements as above. All three cell lines were incubated at 37°C under 5% CO2 in air. Bufalin, purchased from Sigma (St. Louis, MO, USA), was dissolved in dimethyl sulfoxide (DMSO) as a stock solution (10 mM) and stored at −20°C. The culture media containing different concentrations of bufalin were all freshly prepared at the time of each experiment. The final concentration of DMSO was <0.1%. Gemcitabine was purchased from Ely Lilly (Bad Homburg, Germany) and dissolved in normal saline to make a 50 mg/ml stock solution.

The MTT [3-(4, 5-dimethylthiazol-2yl)-2, 5-diphenyltetrazolium bromide (Sigma) assay was used to assess the cellular viability. Briefly, cells were planted on a 96-well plate at a density of 5x10³ cells per well and treated with drugs at different concentrations for 24, 48 and 72 h. A total of 20 μl MTT solution [5 mg/ml in phosphate-buffered saline (PBS)] was added to each well, and further incubated for 3–5 h. Then, the culture medium was removed and the MTT formazan was dissolved in 150 μl DMSO. The plates were agitated for 10 min, and absorbance was measured using an absorbance reader (BioTek ELx800, Winooski, VT, USA) at 490 nm.

Cells were distributed on a 6-well plate at a density of 5x10⁵ per well. After treatment with bufalin and/or gemicitabine for 48 h, cells were harvested and washed with PBS three times. Then the degree of apoptosis was detected by Annexin V/FITC binding assay according to the manufacturer’s instructions (BD Biosciences, Franklin Lakes, NJ, USA). The mixed solution was gently shaken and stored away from light at room temperature for 15 min. The stained cells were analyzed directly by flow cytometry using Cell Quest software (BD Biosciences).

ASK1 siRNA was designed by GenePharma (Shanghai, China) with human ASK1 cDNA. The sequences designed against three separate regions starting from nucleotide 1258, 2025 or 2960 were: si-ASK1 1258, sense 5′-GGCAGCGAGUAGAUAAUAUTT-3′ and antisense 5′-AUAUUAUCUACUCGCUGCCTT-3′; si-ASK1 2025, sense 5′-GUGGUUAGGUUUCCAGUAUTT-3′ and antisense 5′-AUACUGGAAACCUAACCACTT-3′; si-ASK1 2960, sense 5′-GGGCUGUACAAUCAUUGAATT-3′ and antisense 5′-UUCAAUGAUUGUACAGCCCTT-3′. A non-specific oligonucleotide served as the negative control. The cells plated in 6-well plates were transfected with 100 nM siRNA or negative control siRNA using Lipofectamine™ 2000 (Invitrogen, Carlsbad, CA, USA) following the manufacturer’s instructions. Briefly, when cells reached 60–70% confluence, a mixture of Lipofectamine 2000 and OPTI-MEM medium (Invitrogen) was incubated for 5 min, then incubated with siRNA for a further 30 min at room temperature to allow the complex formation. The transfection complex was added to each well ensuring distribution over the entire plate surface. The OPTI-MEM medium was replaced with DMEM at 4–6 h after transfection. The cells were incubated for 48 h prior to being harvested for further analysis.

Following treatment as described above, cells were washed with cold PBS and lysed in pre-chilled lysis buffer [1.0 mM ethylenediamineteraacetate (EDTA), 50 mM Tris-HCL (pH 7.4), 1% NP40, 0.1% SDS, 0.5% deoxycholate, 150 mM NaCl and 2% protease inhibitor cocktail]. Following centrifugation at 13,000 x g for 30 min, the supernatant was collected and quantitated using the BCA protein assay (Pierce Biotechnology, Inc., Rockford, IL, USA). Total protein (40 μg) was separated in 8–10% SDS-polyacrylamide denaturing gels and transferred to polyvinylidene difluoride membranes. After blocking in TBST (10 mM Tris-HCL pH 7.4, 150 mM NaCl and 0.1% Tween-20) with 5% non-fat milk for 1 h, the membranes were incubated with primary antibodies overnight at 4°C, followed by horseradish peroxidase-conjugated secondary antibodies. Immunoblotting for bcl-2, cleaved caspase-3, ASK1 (Cell Signaling Technology, Inc., Danvers, MA, USA), JNK and p-JNK (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) was performed. Immunoreactive bands were visualized by enhanced chemiluminescence (Amersham Biosciences Biotech, Piscataway, NJ, USA).

Four-week-old male nu/nu mice were purchased from the First Affiliated Hospital of Zhejiang University. All animal experiments were approved by the Laboratory Animal Regulations of the Ministry of Science and Technology of China. Each mouse was subcutaneously injected with 6x10⁶ Mia PaCa-2 cells in the back. Treatment was started when the subcutaneous tumors reached a minimum size of 100 mm³. The mice were randomly divided into four treatment groups: a) vehicle alone (control); b) bufalin (0.1 mg/kg, for 10 days); c) gemcitabine (125 mg/kg, three times/week for 2 weeks); d) bufalin and gemcitabine in combination. Each group consisted of six animals. The dose of 125 mg/kg twice a week for gemcitabine has been shown to be efficient in another pancreatic cancer xenograft model (17). To evaluate the tolerable therapeutic dose of bufalin in this animal model, we performed a preliminary dose-response experiment. Four times weekly i.p. injections of 1, 0.5, 0.2, 0.1 and 0.05 mg/kg bufalin were performed. Then, we determined the bufalin dose (0.1 mg/kg) for this study. The tumor size was measured every four days. The volume was calculated using the formula: volume = (length x width²) / 2. One month after the treatment, the xenografts were excised and stocked in 10% formalin.

Tissue sections (4 μm) were prepared using a microtome and placed on glass slides. For immunohistochemical examination, endogenous peroxidase was blocked in 3% H₂O₂ solution. Sections were incubated at 4°C with primary antibodies overnight. Following the removal of unbound antibodies, sections were incubated with biotinylated anti-mouse or anti-rabbit antibodies for 1 h, and then incubated with horseradish peroxidase complex for 10 min, followed by counterstaining with hematoxylin, dehydration and mounting. Sections without primary antibodies were used as negative controls for immunostaining. Random images obtained from each of the four groups were captured and analyzed at x400 magnification. The primary antibodies of anti-human Ki-67 and ASK1 were purchased from Cell Signaling Technology.

Results are presented as the mean ± standard error (SE), and all experiments were performed three times independently. The one-way analysis of variance (ANOVA) and the two-tailed Student’s t-test for unpaired samples were used to determine the statistical significance using SPSS 15.0. P<0.05 was considered to indicate a statistically significant result.

MTT assay was used to examine the cell growth inhibition efficacy of bufalin on three pancreatic cancer cell lines (Bxpc-3, Mia PaCa-2 and Panc-1). Cells were treated with bufalin at different concentrations (0–100 μM) for 24, 48 and 72 h. The results demonstrated that bufalin inhibited the growth of all three cell lines in a dose- and time-dependent manner (Fig. 1A). We subsequently investigated the effect of the combination with bufalin and gemcitabine on cell viability. Pancreatic cancer cell lines were treated with bufalin (0.01 μM) and/or gemcitabine (Bxpc-3, 0.5 μg/ml; Mia PaCa-2 and Panc-1, 5 μg/ml) for 48 h (18). Our results showed that the combination treatment with bufalin and gemcitabine inhibited the growth of all three cell lines more than either bufalin or gemcitabine used alone (Fig. 1B).

We investigated whether bufalin was capable of enhancing gemcitabine-induced apoptosis using flow cytometry analysis. The three pancreatic cancer cell lines were treated with different doses of drugs for 48 h in the same way as for the MTT assay. Exposure to bufalin (0.01 μM) induced apoptosis by up to 7.8% in Bxpc-3, 11.5% in Mia Paca-21 and 7% in Panc-1. Exposure to gemcitabine (0.5 or 5 μg/ml) in combination with bufalin (0.01 μM) for 48 h induced apoptosis by up to 16.8% in Bxpc-3, 21.8% in Mia Paca-2 and 17.4% in Panc-1 (Fig. 2A). To further examine the effect of inducing apoptosis by combination therapy, the expression of apoptosis-related proteins (bcl-2 and cleaved caspase-3) was evaluated. As shown in Fig. 2B, the combined treatment of pancreatic cancer cells with bufalin and gemcitabine significantly decreased the expression of bcl-2. Conversely, the expression of cleaved caspase-3 was significantly upregulated in the combination group compared with the control group and the bufalin and gemcitabine alone groups (Fig. 2B).

To further investigate the apoptosis induced by bufalin, we evaluated the expression of ASK1 in the three pancreatic cancer cell lines. As shown in Fig. 3A, relative to the control, treatment with bufalin induced a dose-dependent increasing expression of ASK1 from 0.001 to 0.1 μM in the three pancreatic cancer cell lines. However, bufalin (0.001 μM) treatment induced a higher ASK1 level in Bxpc-3 and Mia PaCa-2 cells than in Panc-1 cells (Fig. 3A). When cells were treated with bufalin at a dose of 0.01 μM, the expression of ASK1 protein did not differ significantly among the groups. In addition, ASK1 expression was analyzed at a different treatment time with 0.01 μM bufalin. A bufalin-induced time-dependent activation of ASK1 expression was observed in Bxpc-3, Mia PaCa-2 and Panc-1 cells (Fig. 3B). However, treatment with bufalin (0.01 μM) did not induce ASK1 expression until after 12 h in Panc-1 cells. The results also reveal that ASK1 expression was significantly upregulated after 48 h of treatment in the three pancreatic cancer cell lines (Fig. 3B). The Mia PaCa-2 cells treated with bufalin at 0.01 μM for 48 h were selected for our next experiment.

We analyzed whether gemcitabine could induce ASK1 expression and whether upregulation of ASK1 by bufalin could eliminate chemoresistance in Mia PaCa-2 cells, resulting in more marked gemcitabine-induced apoptosis. We found that the level of ASK1 protein was increased when Mia PaCa-2 cells were treated with gemcitabine for 48 h (Fig. 4A). We also tested whether the combined treatment with bufalin for 48 h could abrogate gemcitabine-induced ASK1 protein expression levels. Our results revealed that ASK1 expression induced by gemcitabine increased in the bufalin combination treatment (Fig. 4A).

ASK1 activation is a pivotal mechanism in a broad variety of cell apoptosis. To explore whether the ASK1/JNK pathway contributes to bufalin-induced apoptosis in pancreatic cancer cells, the expression of ASK1 and p-JNK proteins were investigated by western blot analysis. The expression of ASK1 was upregulated with the combined treatment in Mia PaCa-2 cells. Furthermore, such effect of p-JNK was also detected in Mia PaCa-2 cell treatment by bufalin with or without gemcitabine (Fig. 4A). Next, ASK1 siRNA was transfected into cancer cells. The expression of ASK1 in cells transfected with si-2096 decreased to 20% of that of the si-control. Immunoblotting was performed to examine the expression of p-JNK in the treatment group transfected with si-2096 or with si-control in Mia PaCa-2 cells (Fig. 4B). No difference in expression was observed with total JNK, demonstrating that bufalin and gemcitabine have no effect on the total JNK. The expression of p-JNK was significantly decreased following the combined treatment in si-2096 Mia PaCa-2 cells, suggesting that the increased level of p-JNK could be downregulated by si-ASK1 with the combination treatment in pancreatic cancer cells (Fig. 4B).

We used the Mia PaCa-2 subcutaneous xenograft as an in vivo model. Following the combination treatment with gemcitabine and bufalin for two weeks, the tumor volume was significantly reduced compared with that observed in the groups treated with control, gemcitabine and bufalin alone (Fig. 5A). Ki-67 nuclear antigen is used to determine cell proliferation activity, and is a key indicator of prognosis for certain malignant tumors. It was found to be significantly decreased in the combination group. In addition, ASK1 was found to be significantly upregulated in the combination group (Fig. 5B).