Human tumor tissue samples and adjacent normal controls were obtained by surgical resection from nine patients with pancreatic cancer, at the Department of General Surgery, Changhai Hospital of Shanghai, Second Military Medical University. All samples were snap-frozen and stored in liquid nitrogen (−176°C). Signed informed consent was obtained from all patients, and the study was approved by the Ethics Committee of the hospital.

The human pancreatic cancer cell lines (PANC1, PaTu8988 and SW1990) were kindly provided by the Second Military Medical University of Shanghai. The cell lines were cultured with DMEM (Hyclone Laboratories; GE Healthcare Life Sciences, Shanghai, China) supplemented with 10% fetal bovine serum (FBS; Gibco, Thermo Fisher Scientific, Inc., Waltham, MA, USA) in a humidified 5% CO₂ incubator at 37°C.

Total RNA was extracted from tissues and cultured cells with Trizol Reagent (Invitrogen; Thermo Fisher Scientific, Inc.) following the manufacturer's protocol. RNA concentration and integrity were determined by spectrophotometry. Reverse transcription was performed using the PrimeScrip™ RT reagent kit (Takara Bio, Inc., Otsu, Japan). Quantitative RT-PCR was carried out using the SYBR^® Premix Dimmer Eraser kit (Takara Bio, Inc.). GAPDH was used as an internal control. The primer pair used for the amplification was as follows: PVT1 forward, 5′-TGAGAACTGTCCTTACGTGACC-3′ and reverse, 5′-AGAGCACCAAGACTGGCTCT-3′; GAPDH forward, 5′-TTGGTATCGTGGAAGGACTCA-3′ and reverse, 5′-TGTCATCATATTTGGCAGGTT-3′. PCR reactions were performed on the ABI7300 system (Applied Biosystems; Thermo Fisher Scientific, Inc., Sunnyvale, CA, USA). The relative expression fold change of mRNAs was calculated using the 2^−ΔΔCt method.

The specific sequences of small interfering RNAs for PVT1 were: siPVT1-1 sense, GCUUGGAGGCUGAGGAGUUTT and antisense, AACUCCUCAGCCUCCAAGCTT; siPVT1-2 sense, CCCAACAGGAGGACAGCUUTT and antisense, AAGCUGUCCUCCUGUUGGGTT. They were synthesized by GE Healthcare Dharmacon (Lafayette, CO, USA). A non-specific siRNA sequence was used as a negative control. For ectopic expression, we amplified the entire PVT1 sequence with RT-PCR using the following primers: Forward, TGCTCTAGACTCAAGATGGCTGTGCCTGTCAGCT and reverse, CCGGAATTCAGTAGAAAAAGAATTTAATAGACACG, and then cloned it into pCD513B-1 (SBI Pharmaceuticals, Co., Ltd., Tikyo, Japan) at XbaI and EcoRI sites. All PCR products were confirmed by DNA sequencing.

Cells in logarithmic growth phase were diluted and seeded at 2.5×10⁵ cells/well into 6-well plates. The siRNAs or plasmid were transfected into pancreatic cells using Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. The cells were incubated in a humidified chamber for 48 h before use in the following assays.

A cell viability assay was performed using a CCK-8 kit (Dojindo Molecular Technologies, Inc., Kumamoto, Japan) following the manufacturer's protocol. Cells were seeded in 96-well plates at a density of 5,000 cells/well and cultured with medium containing 10% FBS for 10, 24, 48 and 72 h. After the cells were washed twice with cold PBS, 10 µl CCK-8 and 100 µl serum-free medium were added into each well and the plates were incubated for a further 45 min at 37°C. The absorbance was assessed at a wavelength of 450 nm using Wellscan MK-3 ELISA (Labsystems Dragon, Helsinki, Finland).

Protein was extracted from transfected cells with RIPA lysis buffer at 4°C for 10 min. Equal amounts of protein were loaded and separated by 10% SDS-PAGE and transferred to PVDF membranes (EMD Millipore, Billerica, MA, USA). The membranes were blocked with 5% non-fat dried milk in TBST for 1 h at room temperature and then incubated with primary antibodies at 4°C overnight. After washing with TBST, the membranes were further incubated for 1 h at room temperature with corresponding horseradish peroxidase-conjugated secondary antibody in an appropriate dilution (goat anti-mouse; 1:5,000; cat. no. A0216; and goat anti-rabbit; 1:10,000; cat. no. A0208) and then washed three times with the same buffer. The immunoreactive protein bands were visualized using an ECL kit (Pierce; Thermo Fisher Scientific, Inc.). The antibodies used were: rabbit anti-TGF-β (cat. no. 3712; Cell Signaling Technology, Inc., Danvers, MA, USA), rabbit anti-MMP2 (cat. no. YT2798) and rabbit anti-MMP9 (cat. no. YT1892; both from ImmunoWay Biotechnology Co., Plano, TX, USA), an EMT antibody sampler kit (cat. no. 9782), a Smad2/3 antibody sampler kit (cat. no. 12747), mouse anti-β-Tubulin (cat. no. 4466), and rabbit anti-β-actin (cat. no. 4970; all from Cell Signaling Technology, Inc.) and all above antibodies diluted with 1:1,000. β-actin and β-tubulin were used as internal controls. Signals were visualized using an enhanced chemiluminescence system.

Briefly, migration assays were conducted using 8.0-µm culture insert chambers (EMD Millipore). Invasion assays were performed using the Corning^® Matrigel^® Invasion Chamber (Corning Inc., Corning, NY, USA). FBS-containing medium (10%) was placed in the bottom chamber to act as a chemoattractant. Then, 5×10⁴ cells in a 100-µl volume of serum-free medium were placed in the upper chamber. Invasive cells on the lower surface of the membrane were those that had invaded the Matrigel or had migrated through the polycarbonate membrane. After incubation at 37°C for 24 h, the cells were fixed in paraformaldehyde solution and stained with crystal violet. The top surface was gently scrubbed with a cotton bud and the remaining cells at the bottom surface were counted under a microscope (Olympus GX41; Olympus Corp. Tokyo, Japan) in five randomly selected fields at a magnification of ×20.

Transfected cells were grown to 95% confluence. A wound was scratched in the monolayer with a 200-µl pipette tip. Suspended cells and debris were washed with PBS three times, then the cells were incubated in medium containing 2% FBS at 37°C. At 0 and 24 h the wound image was captured using an OLYMPUS inverted microscope at a magnification of ×100, respectively, to evaluate the cell migratory potential.

Matrigel (BD Biosciences, Franklin Lakes, NJ, USA) was diluted in serum-free media to a concentration of 0.04 µg/µl, and then 50 µl of this Matrigel was transferred into 96-well plates. Transfected cells (4×10⁴) in a 100-µl volume of serum-free medium were plated onto the Matrigel-precoated 96-well plates in triplicate. After incubation at 37°C for 4 h, the medium was then carefully removed, and the plate was washed with PBS to remove the non-adherent cells. The Matrigel-precoated plate without cells served as the negative control. After washing, adherent cells were assessed with a CCK-8 kit (Dojindo Molecular Technologies, Inc.) following the manufacturer's protocol. The relative optical density (OD) value was determined at 490 nm using Wellscan MK-3 ELISA. The OD values reflected the proportion of cells in the Matrigel-coated 96-well plate.

All the presented data and results were confirmed in at least three independent experiments and were expressed as the mean ± SD. Statistical comparisons were performed using Student's t-test and one-way ANOVA. The data was analyzed using GraphPad Prism 7 (GraphPad Software, Inc., La Jolla, CA, USA). P<0.05 was considered to indicate a statistically significant difference.

To confirm the role of PVT1 in pancreatic cancer, we first examined its expression profile in 9 pairs of matched human pancreatic tumors and adjacent normal tissues by qRT-PCR. The results indicated that the expression level of PVT1 in tumor tissues was significantly higher than that in adjacent normal tissues (Fig. 1A). In addition, it was revealed that PaTu8988 cells expressed the highest level of PVT1, whereas SW1990 cells expressed the lowest level of PVT1 among the 3 cell lines (Fig. 1B). Since PaTu8988 and BxPc-3 cells expressed a high level of PVT1, these two cell lines were selected to investigate the effect of specific depletion of PVT1 (Fig. 1C and D), and SW1990 was selected to study the effect of overexpression of PVT1 assays (Fig. 1E).

To determine the role of PVT1 in pancreatic cancer cells, we utilized a CCK-8 assay to estimate cell viability. We found that PVT1 knockdown induced lower viability of PaTu8988 and BxPc-3 cells at 10, 24, 48 and 72 h after transfection (Fig. 2A and B). In contrast, PVT1 overexpression enhanced cell viability at 10, 24, 48 and 72 h in SW1990 cells (Fig. 2C). In addition, we found that PVT1 knockdown suppressed the ability of adhesion in PaTu8988 and BxPc-3 cells (Fig. 2D and E) while PVT1 overexpression enhanced the ability in SW1990 cells (Fig. 2F). Collectively, these results revealed that PVT1 may promote viability of pancreatic cancer cells and tumor cell adhererence to the extracellular matrix (ECM).

Next, we examined the ability of migration by wound healing assays in pancreatic cancer cells. The migration rate was 0.292±0.053, 0.125±0.012 and 0.143±0.005 (P<0.01) in siPVT1-NC, siPVT1-1 and siPVT1-2 PaTu8988 cells (Fig. 3A and D), and 0.291±0.002, 0.224±0.025 and 0.245±0.034 (P<0.05) in siPVT1-NC, siPVT1-1 and siPVT1-2 BxPc-3 cells, respectively (Fig. 3B and E), while 0.185±0.015 and 0.664±0.011 in pCD513B-1 and pCD513B-1-PVT1 (P<0.01) SW1990 cells (Fig. 3C and F). These results indicated that knockdown of PVT1 significantly decreased the migration ability of PaTu8988 and BxPc-3 cells. To confirm the aforementioned results, we conducted Transwell assays to examine migration ability. The number of migrated cells was 58±12, 16±10 and 18±9 (P<0.01) in siPVT1-NC, siPVT1-1 and siPVT1-2 PaTu8988 cells, and 57±6, 20±8 and 18±9 (P<0.01) in siPVT1-NC, siPVT1-1 and siPVT1-2 BxPc-3 cells, respectively (Fig. 3G, I and J). Furthermore, the number of migrated cells was 20±10 and 127±12 (P<0.01) in pCD513B-1 and pCD513B-1-PVT1 SW1990 cells (Fig. 3H and K). All these data revealed that PVT1 promotes the ability of migration in pancreatic cancer cells.

Subsequently, we examined the ability of invasion using BD Matrigel invasion assays in pancreatic cancer cells. PaTu8988 and BxPc-3 cells were transfected with siPVT1-NC, siPVT1-1 and siPVT1-2 plasmids, and SW1990 cells with pCD513B-1 and pCD513B-1-PVT1 for 72 h. The number of invasive cells was 82±5, 24±9 and 20±5 in siPVT1-NC, siPVT1-1 and siPVT1-2 PaTu8988 cells, and 68±3, 28±7 and 24±11 in siPVT1-NC, siPVT1-1 and siPVT1-2 BxPc-3 cells, respectively (Fig. 4A-C), demonstrating that downregulation of PVT1 significantly inhibited the invasion ability of PaTu8988 and BxPc-3 cells. In addition, the number of invasive cells was 20±5 and 80±8 in pCD513B-1 and pCD513B-1-PVT1 SW1990 cells, respectively (Fig. 4D and E) thus revealing that upregulation of PVT1 significantly promoted the invasion ability of SW1990 cells. Matrix-metalloproteinases (MMPs) have been confirmed to play pivotal roles in tumor cell invasion through adherence and degradation of the basement membrane and ECM (20). To determine whether PVT1 can regulate the activity of MMPs, we investigated the expression level of MMP2 and MMP9 through western blotting. In the PaTu8988 and BxPc-3 cells, knockdown of PVT1 suppressed MMP2 and MMP9 (active and total) protein levels (Fig. 4F). Conversely, in SW1990 cells overexpression of PVT1 increased MMP2 and MMP9 (active and total) protein levels (Fig. 4G). Collectively, these data revealed that PVT1 promotes the invasion ability of pancreatic cancer cells.

During tumor development, tumor cells constantly communicate with the surrounding microenvironment, which can guide tumor cells to undergo a phenomenon termed EMT. EMT is regarded as a pivotal step for the promotion of tumor invasion and metastasis and plays a potential role in the progression of pancreatic cancer (21). Thus, we assessed whether PVT1 was involved in EMT to influence cancer metastasis. We examined the expression of EMT marker genes and EMT-related transcription factors using western blotting. PVT1-siRNAs increased E-cadherin expression and decreased vimentin, N-cadherin, β-catenin, Snail and Slug expression in PaTu8988 and BxPc-3 cells (Fig. 5A and B). Conversely, ectopic expression of PVT1 decreased E-cadherin expression and increased vimentin, N-cadherin, β-catenin, Snail and Slug expression in SW1990 cells (Fig. 5C). These observations strongly demonstrated that PVT1 promoted a transition from epithelial to mesenchymal phenotype.

Multiple studies have suggested that the TGF-β/Smad signaling pathway is a central regulator in cancer cell viability, metastasis and the EMT process. Thus, we speculated that PVT1 could regulate TGF-β/Smad signal transduction in human pancreatic cancer cells, focusing on p-Smad2/3, Smad4 and TGF-β1 since they are pivotal signaling molecules in the TGF-β/Smad signaling pathway. PVT1 knockdown increased Smad4 expression and decreased p-Smad2/3 and TGF-β1 expression in PaTu8988 and BxPc-3 cells (Fig. 6A and B). Conversely, SW1990 cells transfected with PVT1 expression plasmid decreased Smad4 expression and increased p-Smad2/3 and TGF-β1 expression (Fig. 6C). However, TGF-β1 knockdown downregulated p-Smad2/3 after we transfected BxPc-3 and SW1990 cells with pCD513B-1-PVT1 (Fig. 6D and E), indicating that PVT1 upregulates p-Smad2/3 mainly through TGF-β1 upregulation. Knockdown or overexpression of PVT1 had no influence on total Smad2/3 protein. These data revealed that PVT1 promoted EMT in pancreatic cancer cells possibly via TGF-β/Smad signaling.