The human pancreatic cancer cell lines SW1990, BXPC-3 and PANC-1 were purchased from the Cell Bank of the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). SW1990/GEM is gemcitabine-resistant SW1990 cell line, which was retained in our laboratory. All the cell lines were authenticated by short-tandem repeat profiling and cultured in Gibco™ Dulbecco's modified Eagle's medium (DMEM) Thermo Fisher Scientific, Inc., Waltham, MA, USA) supplemented with 10% heat inactivated fetal bovine serum (FBS; Thermo Fisher Scientific, Inc.). Cells were incubated in a 5% CO₂ humidified incubator at 37°C. Construction and production of recombinant oncolytic vaccinia virus oVV-Smac and oVV were previously described (24). The Smac gene was inserted into the thymidine kinase (TK) region, disrupting the function of TK. Deletion of the TK gene inhibits viral replication in normal, non-dividing cells (25). However, cancer cells have a high concentration of functional nucleotides that enables oVV replication to occur in the absence of viral TK. Therefore, disruption of TK results in selective replication of the oVV in tumor cells. The T7 promoter was inserted before the exogenous genes to initiate their expression, and the gpt gene works as a screen gene engineered behind the exogenous genes. The whole expression cassette was constructed into the pCB vector, which is a shuttle plasmid for vaccinia virus packaging kindly provided by academician Xinyuan Liu (Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai, China). Each recombinant vaccinia virus was isolated through three rounds of plaque purification in 293 cells and purified by ultracentrifugation in a cesium chloride gradient. Moreover, virus titers were determined by TCID50 assay in 293 cells. Cells were infected with vaccinia virus at different doses at 37°C in a humidified atmosphere containing 5% CO₂.

PANC-1, SW1990 and BxPC-3 cells were dispensed in 96-well culture plates at a density of 5×10³ cells/well. After attachment, cells were infected with oVV, oVV-Smac with or without gemcitabine at given concentrations and times. The medium added together with PBS was used as a blank control. The cell survival rate was evaluated by a standard 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany), Medium was removed and fresh medium containing MTT (5 mg/ml) was added to each well. The cells were incubated at 37°C for 4 h, and after the supernatant was drawn off of each well carefully and an equal volume (150 µl) of DMSO was added to each well and mixed thoroughly on a concentrating table for 10 min. The absorbance of the plates was read at 595 nm with a GENios model DNA Expert Microplate Reader (Tecan Group, Ltd., Mannedorf, Switzerland). For combination index plots, CI is expressed as the log₁₀(CI) ± 1.96 SD, and the 95% confidence intervals (CIs) are shown where estimable, with the use of the algebraic approximation algorithm of the CalcuSyn program (Biosoft, Cambridge, UK). In the present study, CI values were calculated over a scope of levels of growth inhibition (GI) from 20 to 80% of the fraction affected.

Cells were harvested in lysis buffer (Beyotime Institute of Biotehnology, Jiangsu, China) containing 1% Complete Mini-Protease Inhibitor Cocktail (Roche Diagnostics, Basel, Switzerland) and 5 mM NaF. Protein extractions were quantified using the BCA kit (Thermo Fisher Scientific, Inc.) and heated for 10 min at 100°C. Protein (30 µg) was resolved on 12% SDS-PAGE and transferred to nitrocellulose membranes (Merk Millipore, Darmstadt, Germany). After blocking for 1 h at 37°C, the membranes were immunoblotted with different antibodies overnight at 4°C. Antibodies against Smac (dilution 1:1,000; cat. no. ab8114), vaccinia virus (dilution 1:500; cat. no. ab19970), GAPDH (dilution 1:2,000; cat. no. ab128915) were purchased from Abcam (Shanghai, China). Antibodies against caspase-8 (dilution 1:1,000; cat. no. MABC1606), caspase-9 (dilution 1:1,000; cat. no. MAB4709), caspase-3 (dilution 1:1,000; cat. no. AB3623), PARP (dilution 1:500; cat. no. AB16661), XIAP (dilution 1:1,000; cat. no. 07735), cIAP-1 (dilution 1:1,000; cat. no. ABC448), cIAP-2 (dilution 1:1,000; cat. no. AB3615), survivin (dilution 1:1,000; cat. no. MAB4617), livin (dilution 1:1,000; cat. no. ABC97), P-gp (dilution 1:500; cat. no. ABN455) and MDR1 (dilution 1:1,000; cat. no. MAB4162) were purchased from EMD Millipore Corp. (Billerica, MA, USA). Membranes were then washed with TBST and incubated with HRP-conjugated goat anti-rabbit (dilution 1:5,000; cat. no. HA1001) or anti-mouse antibody (dilution 1:5,000; cat. no. HA1006; both from for Huabio, Hangzhou, China) for 1 h at room temperature. Finally, blots were detected using ChemiDoc™ MP Imaging System (Bio-Rad Laboratories) with a Super Enhanced chemiluminescence detection kit (Applygen Technologies, Inc., Beijing, China).

Cells infected with oncolytic vaccinia viruses and/or gemcitabine were trypsinized and washed once with complete medium. Aliquots of cells (5×10⁵) were resuspended in 500 ml of binding buffer and stained with fluorescein isothiocyanate (FITC)-labeled Annexin V and propidium iodide (PI) (BioVision, Inc., Palo Alto, CA, USA) according to the manufacturer's instructions. Cell apoptosis and cell cycle were examined using FACS (FACStar cytofluorometer; BD Biosciences, San Jose, CA, USA).

RNA was extracted with Invitrogen™ Trizol reagent (Thermo Fisher Scientific, Inc.) following the manufacturer's instructions. cDNA was generated using the PrimeScript RT reagent kit (Takara Bio, Inc., Tokyo, Japan). The qPCR reactions were conducted in a total volume of 20 µl by using the following procedure: 1 cycle at 95°C (10 min), then 60°C (30 sec), followed by 39 cycles at 95°C (10 sec) and 60°C (30 sec). PCR amplicons were determined based on SYBR-Green I detection (Roche Diagnostic, Indianapolis, IN, USA), and the authenticity was certified by melting curve analysis. Quantitative PCR was operated using the CFX-96 qPCR system and iQ SYBR-Green SuperMix (Bio-Rad Laboratories). Relative gene expression was determined via the 2^ΔΔCt method. The primers used are as follows: Smac, 5′-GGAAGATCTCCTCCTGCATCC-3′ (forward) and 5′-CCGCTTAAGATACCGCTCGAG-3′ (reverse); GAPDH, 5′-CTTTGGTATCGTGGAAGGACTC-3′ (forward) and 5′-GTAGAGGCAGGGGATGATGTTCT-3′ (reverse).

All animal experiments were approved by the Institutional Animal Care and Use Committee, Zhejiang Provincial People's Hospital and all procedures were in accordance to the Guide for the Care and Use of Laboratory Animals (National Academies Press, Washington, DC). One hundred female BALB/c nude mice (4- to 5-weeks old, 20 g) were purchased from the Shanghai Experimental Animal Center (Shanghai, China). All 100 mice were housed in a specific pathogen-free environment, in which the temperature was maintained at 26–28°C, the humidity was 40–60%, and the daily light was maintained for 10 h (14 h without light). Ventilation was ensured 10 to 15 times per hour. When the mouse tumor reached a diameter of 2.0 cm or a volume of 2.5 cm³, or all mice in the PBS group died, we stopped the mouse experiment. The sensitive SW1990 cells or SW1990/GEM cells were injected subcutaneously into the lower right flank of female nude mice and the tumor xenograft model was established. Each group was composed of at least 8 animals and tumor growth was monitored and measured for every 4 days with a Vernier caliper. Tumor volume (V) was calculated according to the formula: V (mm³) = 1/2 × length (mm) × width (mm)². Once the subcutaneous tumors reached ~100 mm³, the nude mice were divided into 5 groups (8 mice in each group) randomly. Subsequently, mice were injected with gemcitabine, oVV, oVV-Smac, gemcitabine plus oVV-Smac, or PBS. oVV and oVV-Smac (2×10⁷ plaque forming unit/mouse) was injected once every day for a total of 4 times through intratumoral injection; gemcitabine was injected intraperitoneally at a total dose of 30 mg/kg body weight; and PBS 100 µl as control for a total of 4 times once every day.

Mice were sacrificed at 2 weeks post-injection according to ethical instructions by carbon dioxide. Tumors were separated, fixed using 4% paraformaldehyde, embedded in paraffin, finally cut into 4-µm sections for hematoxylin and eosin staining, immunohistochemical analysis and TUNEL assay according to the manufacturers' instructions. For immunohistochemical analysis, slides were incubated with primary antibody anti-Smac (dilution 1:100; cat. no. ab8114; Abcam, Shanghai, China) overnight at 4°C, and then incubated with biotinylated secondary antibody (dilution 1:1,000; cat. no. B2763; Thermo Fisher Scientific, Inc.) and further visualized using a diaminobenzidine (DAB) kit (Thermo Scientific, Inc.).

For comparison between 2 groups, significant differences were determined using the Student's t-test. Analysis of variance (ANOVA) followed by a Bonferroni multiple-group comparison test was applied for comparison of 3 or more groups. The analysis of the combined effects was performed with CalcuSyn software 2.0 (Biosoft, Cambridge, UK). Data are expressed as the mean ± SD. Statistical analysis was performed with IBM SPSS Statistics software version 20 (IBM Corp., Armonk, NY, USA). Statistical significance was defined at P<0.05.

The generation of oVV-Smac was performed by homologous recombination as described in our previous study (24). Fig. 1A depicts the construction scheme of oVV and oVV-Smac. Real-time quantitative polymerase chain reaction (qPCR) confirmed exogenous Smac expression. Three pancreatic cancer cell lines, namely PANC-1, SW1990 and BxPC-3, were infected with oVV and oVV-Smac at multiplicities of infection (MOI) of 10 for 24 h. As expected, a significant amount of Smac was observed in all oVV-Smac-transfected pancreatic cancer cell lines, but not in the oVV- or phosphase-buffered saline (PBS)-treated group (Fig. 1B). Similar results were obtained when the expression of Smac and vaccinia virus A27L was determined at the protein level by western blot analysis (Fig. 1C), suggesting that Smac was overexpressed in oVV-Smac-transfected cells both at the transcriptional and translational levels.

To control for any interference of the transgene and modified genome of vaccinia virus with the selective replicative ability of the recombinant oVV in different cell lines, a progeny assay was performed by infecting the three pancreatic cancer cells with various constructs including vaccinia virus (wild-type), oVV and oVV-Smac. The results indicated that both oncolytic viruses oVV and oVV-Smac replicated easily in the infected pancreatic cancer cells and yielded a high virus progeny (Fig. 1D). Thus, the selective replicative ability of oVV in cancer cells was not affected by the insertion of Smac and thymidine kinase (TK) deletion.

The 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay was performed 48 h post-infection to evaluate the cytotoxicity of oVV-Smac in the PANC-1, SW1990 and BxPC-3 cell lines. The results indicated a significantly higher inhibition of cell growth with oVV-Smac than with oVV (Fig. 2A and B); these effects were dose-dependent. Taken together, these findings demonstrated selective inhibitory effects of oVV-Smac on cancer cell growth in vitro.

To address the underlying mechanism of oVV-Smac-induced cytotoxicity, we evaluated oVV-Smac-associated apoptosis in vitro using flow cytometric analysis. The results indicated significant apoptosis in the PANC-1, SW1990 and BxPC-3 cell lines transfected with oVV-Smac compared with the percentage of apoptosis in the oVV- or PBS-treated cells (Fig. 3A).

We further evaluated apoptosis by assessing the expression of apoptosis-related proteins in SW1990 cells at 48 h post-infection using western blotting analysis. The results indicated a significant activation of caspase-3, −8 and −9, and increased poly(ADP-ribose) polymerase (PARP) cleavage in the oVV-Smac-treated cells (Fig. 3B; left blot). Additionally, the levels of X-linked IAP (XIAP), cellular IAP-1 (cIAP-1), cIAP-2, survivin and livin were also decreased in the oVV-Smac cells (Fig. 3B; right blot). Taken together, these findings indicated that oVV-Smac effectively induced apoptosis through the caspase and IAP pathways.

To determine whether oVV-Smac enhances the cytotoxic effect of gemcitabine, we analyzed the viability of pancreatic cancer cells using the MTT assay after co-treatment with oVV-Smac and gemcitabine. The PANC-1, SW1990 and BxPC-3 cells were treated with gemcitabine (1 or 5 µM) with or without oVV-Smac (0.1 or 0.5 MOI). The combination of oVV-Smac with gemcitabine significantly inhibited cell growth compared with treatment with gemcitabine or oVV-Smac alone (Fig. 4A upper panels). Next, the synergistic effect of gemcitabine combined with oVV-Smac on pancreatic cancer cells was quantified using the combination index (CI) analysis and expressed as CI vs. the fractional affect (Fig. 4A lower panels). In PANC-1 cells, at all the fractions considered, the Chou-Talalay CI was lower than one (log₁₀(CI) <0), indicating a potentiation effect of oVV-Smac when combined with gemcitabine, and vice versa. Additionally, investigation with SW1990 and BxPC-3 cells presented similar results (log₁₀(CI) <0). These results showed that the combination of gemcitabine and oVV-Smac had a synergistic tumor killing effect.

We further evaluated apoptosis using Annexin-V-FITC/PI double staining to evaluate the effect of oVV-Smac on gemcitabine-induced apoptosis (Fig. 4B). The apoptotic rate in the pancreatic cancer cells co-treated with gemcitabine and oVV-Smac was significantly higher than that in the cells treated with gemcitabine or oVV-Smac alone.

The question remains as to the mechanism by which oVV-Smac may influence gemcitabine-resistant pancreatic cancer cell lines. Therefore, we constructed a gemcitabine-resistant SW1990 cell line (SW1990/GEM). Subsequently, the expression of multidrug resistance-related proteins and IAPs in SW1990/GEM cells treated with oVV-Smac, Smac, or oVV was assessed by western blot analysis. Compared with the sensitive SW1990 cells, SW1990/GEM cells produced marked multidrug resistance-related proteins (P-gp and MDR1) and IAPs (XIAP, cIAP-1 and cIAP-2). Based on these results, Smac gene transfection has the capacity to reduce multidrug resistance-related proteins and IAPs, which could be further potentiated by Smac-armed oncolytic vaccinia virus (Fig. 4C).

Taken together, these findings indicated a synergistic repressive effect of the combined gemcitabine and oVV-Smac treatment on pancreatic cancer cell proliferation.

We developed two pancreatic tumor xenograft mouse models with sensitive SW1990 cells and SW1990/GEM cells using BALB/c athymic nude mice to evaluate the effect of co-treatment with gemcitabine and oVV-Smac in vivo (Fig. 5A). Antitumor efficacy was evaluated by plotting tumor growth curves over a 56-day observation period. The mean tumor volume was significantly decreased in mice injected with gemcitabine, oVV-Smac, and both compared with those injected with PBS (Fig. 5B and C). Furthermore, co-treatment of sensitive SW1990 and SW1990/GEM cells with gemcitabine and oVV-Smac was more effective than gemcitabine (P=0.001 and 0.002, respectively) and oVV-Smac alone (P=0.001 and 0.003, respectively). Co-treatment with gemcitabine and oVV-Smac was also associated with a higher survival rate when compared with treatment with PBS, gemcitabine, or oVV-Smac (Fig. 5D and E).

The tumor histopathological changes were further evaluated by hematoxylin and eosin (H&E) staining, immunohistochemistry (IHC) and TUNEL assay. The combined treatment with gemcitabine and oVV-Smac resulted in a higher cytotoxicity than single treatment as evidenced by H&E staining. Moreover, an intense expression of Smac in the tumor tissues was associated with the combined treatment as evidenced by IHC with anti-Smac (Fig. 6). Apoptosis was further examined using the terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) assay. Results from this experiment showed significantly higher apoptosis rates in the combination treatment when compared with either individual treatment (Fig. 6).