The human pancreatic cancer cells, including MIA-PACA-2, ASPC-1, BXPC-3, CAPAN-1, CAPAN-2, CFPAC-1, HPAF-II, HS766T, PANC1 and SW1990 were obtained from the Americacn Type culture Collection (ATCC, Manassas, VA, USA). PANC1, SW1990, CAPAN-2, HS766T and MIA-PACA-2 cell lines were maintained in DMEM with 10% FBS (Gibco, Carlsbad, CA, USA). ASPC-1, BXPC-3, CFPAC-1, HPAF-II cell lines were maintained in 1640 with 10% FBS. CAPAN-1 cell line was maintained in IMDM with 10% FBS. The cells were incubated in a humidified atmosphere of 5% CO₂ and 95% air.

Total RNA was extracted from pancreatic tissues, adjacent normal tissues and cell lines using TRIzol reagent (Invitrogen, USA) according to the manufacturer's instructions. cDNA was synthesized using the Prime-Script RT reagent kit (Tiangen, Beijing, China). qRT-PCR was performed using SYBR Green PCR Master Mix (Takara, Dalian, China) on an ABI 7500 fast real-time PCR system (Applied Biosystems, Foster City, CA, USA). Expression data were uniformly normalized to the internal control U6 and the relative expression levels were evaluated using the ∆∆Ct method. The primers for NDRG1 were 5′-CTCTGTTCACGTCACGCTGT-3′ (forward) and 5′-CTCCACCATCTCAGGGTTGT-3′ (reverse) and for GAPDH were 5′-GGACCTGACCTGCCGTCTAG-3′ (forward) and 5′-GTAGCCCAGGATGCCCTTGA-3′ (reverse) according to the human NDRG1 and GAPDH cDNA sequences in GenBank. The GAPDH mRNA level was used for normalization.

Cell lysates were prepared by SDS lysis solution. Protein concentration was measured using a BCA protein assay kit. Equal amount of protein was separated by electrophoresis on a 10% SDS-polyacrylamide gel. The proteins were electrotransferred from the gel to nitrocellulose membrane. The membrane was blocked with 5% non-fat milk solution for 1 h, and then was incubated with primary monoclonal antibody against NDRG1 (Abcam), β-actin (Calbiochem) at 4°C overnight. β-actin was used as an internal control. After washing with TBS-T, the membrane was incubated with secondary antibodies against rabbit immunoglobulin G. The membrane was washed and detected by the enhanced chemiluminescence (ECL) detection system (Thermo Fisher Scientific) according to the manufacturer's instructions.

The primer sequences used for NDRG1 expression were: NDRG1 forward, 5′-TTAGATCATGTCTCGGGAGATGCAGGAT-3′; and reverse 5′-TTGAATTCCTAGCAGGAGACCTCCATGG-3. The PCR product was then cloned into pcDNA3.1(+) (Invitrogen) using standard techniques. Either the NDRG1 expression vector pcDNA3.1(+)/NDRG1 or empty vector pcDNA3.1(+) was transiently transfected into MIA-PACA2 and PANC1 cells using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. For the cell proliferation assay, cells were seeded into a 96-well plate (2×10⁴ cells/well), and positive colonies were selected with G418 (Invitrogen) supplemented with growth medium. NDRG1 short hairpin RNA (shRNA) target sequence was: 5′-GCATTATTGGCATGGGAAC-3′. Double-stranded DNA coding NDRG1 shRNA was cloned into pRNATH1.1 Adeno shuttle vector containing a cGFP marker (Genscript). Adenovirus was packaged in Ad-293 cells (Agilent) and purified by CsCl₂ gradient ultracentrifugation. Viral particle titer was determined by plague assay. For adenoviral transduction, MIA-PACA-2 and PANC1 cell was transduced with 100 multiplicity of infection of adenoviral control or shRNA for 24–48 h. Stable clones were confirmed by real-time RT-PCR and western blotting.

Cells from each group (NDRG1 overexpression and control, NDRG1 shRNA and control) were seeded into 6-well plates (5×10⁵ cells/well). The confluent monolayer was starved overnight, and then a single, linear scratch was created using a 20-µl pipette tip. After wounding, the cells were washed gently with PBS to remove cell debris and placed in fresh DMEM supplemented with 0.1% FBS to block cell proliferation. Images were captured using a phase contrast microscope at 0 and 48 h. Wound closure was expressed as a percentage of the wound area at 0 h.

Invasion assays were performed using the BD BioCoat™ Matrigel chamber in 24-well plates (BD, USA). Resuspension solution (100 µl) containing 1×10⁵ cells in DMEM with 1% FBS was added to the upper chamber, and 600 µl DMEM supplemented with 10% FBS was added to the bottom chamber. After a 48-h incubation at 37°C in a 5% CO₂ incubator, cells in the upper well were wiped off using a cotton swab. Cells in the lower chamber were fixed, stained with H&E, and counted under a light microscope. The migration assay was performed in a similar manner except that the chambers were covered without Matrigel.

Cell proliferation assay was performed with Cell Counting Kit-8 (Dojindo, Kumamoto, Japan) according to the manufacturer's instructions. Briefly, indicated pancreatic cancer cells were seeded in 96-well plates (1×10⁴ cells/well) 24 h post-transfection and cultured in the growth medium. Cells were examined at 0, 24, 48, 72, 96 and 120 h. CCK-8 (10 µl) was added to each well at different time-points. After an incubation of 2 h at 37°C, absorbance was measured at 450 nm.

Statistical analysis was performed using GraphPad Prism version 4.02 software (GraphPad Software Inc., La Jolla, CA, USA) or SPSS 13.0 software (SPSS, Inc., Chicago, IL, USA). Values are expressed as the mean ± standard deviation. Comparisons between multiple groups were made using a one-way analysis of variance, followed by t-test. P<0.05 was considered to indicate a statistically significant difference between values. All experiments were conducted at least in triplicate.

In order to select suitable cell lines for subsequent upregulation or downregulation of NDRG1 and to observe these effects on proliferation, invasion, migration of cancer cells, we assessed NDRG1 expression in pancreatic cancer cell lines. As shown in Fig. 1, the expression of NDRG1 mRNA and protein were slightly mismatched. The top three highest mRNA expressions were CAPAN-2, HPAF-II, SW1990 but in NDRG1 protein expression, CAPAN-2, HPAF-II, PANC1 is the highest. Whereas, the lowest expression of mRNA and protein were MIA-PaCa2, BxPC-3, CAPAN-1 and MIA-PaCa2, CAPAN-1, HS766T, respectively (Fig. 1).

We employed PANC-1 and MIA-PaCa2 transfected with NDRG1 expressing plasmid and shRNA as cell models for the following in vitro study. After transfection with 48 h, cells were lysed and subjected to qRT-PCR and western blotting for detecting NDRG1 expression. As indicated in Fig. 2, NDRG1 mRNA and protein were significantly elevated after transfection with NDRG1 expressing plasmid in PANC-1 and MIA-PaCa2 cell lines in comparison with the vector control cells, while the levels of NDRG1 were reduced remarkably with NDRG1 shRNA transfection as compared with controls. These data suggest that overexpression and knockdown of DRG1 in pancreatic cancer cell lines were established and ready for in vitro experiments.

Wound scratch assays and Transwell assays were performed to evaluate the influence of NDRG1 on pancreatic cancer migration and invasion. As shown in Fig. 3, the distance changes were calculated by defining the change of control cells as 100%. Significant wound closure inhibition was observed in the NDRG1 overexpression PANC-1 (1.00 vs 0.60±0.05, P=0.0035) and MIA-PaCa2 (1.00 vs 0.58±0.055, P=0.0001) cells. However, those with NDRG1 siRNA transfection migrated faster than control cells (PANC-1: 1.00 vs 1.83±0.1, P=0.0003; MIA-PaCa2: 1.00 vs 1.48±0.18, P=0.01). Furthermore, Transwell assays with or without Matrigel were used to examine PANC-1 and MIA-PaCa2 cell properties of invasion and migration. In Transwell assays, the numbers of migrated and invaded PANC-1 (migration: 302.67±23.03 vs 688.33±33.5, P=0.00008; invasion: 225.67±26.10 vs 400±43.86, P=0.004) and MIA-PaCa2 (migration: 63±5.71 vs 111±7.12, P=0.0017; invasion: 72.67±12.50 vs 173.33±7.77, P=0.0002) cells significantly increased after NDRG1 knockdown compared to controls. While the numbers of migrated and invaded PANC-1 (migration: 328.33±47.06 vs 152.33±21.55, P=0.004; invasion: 691.3±66.03 vs 356±21.63, P=0.001) and MIA-PaCa2 (migration: 86±2.94 vs 44±9.09, P=0.003; invasion: 229.7±23.71 vs 143.7±10.21, P=0.004) cells were less than control after NDRG1 overexpression. Taken together, these findings demonstrated NDRG1 prevented pancreatic cancer cell migration and invasion. In CCK-8 proliferation assay, OD450 values were significantly higher than control cells at day 5 after NDRG1 knockdown in PANC-1 and MIA-PaCa2 cell lines. In contrast, the readouts of OD450 declined significantly after NDRG1 overexpression in both cell lines (Fig. 3). These results indicated that NDRG1 had the potential of inhibiting pancreatic cancer cell proliferation, migration and invasion.

Next, we determined the p-STAT3, MMP2, MMP9, PTEN, PI3K and p-AKT after NDRG1 regulation. By western blot assay, p-STAT3, MMP2, MMP9, PTEN, PI3K and p-AKT level were tested after NDRG1 plasmid or shRNA transfection. In both PANC-1 and MIA-PaCa2 cell lines, we observed that p-STAT3, MMP2, MMP9, PI3K and p-AKT expression were negatively correlated to NDRG1 level while PTEN was positively correlated to NDRG1 overexpression. Specifically, Fig. 4 shows stronger p-STAT3, MMP2, MMP9, PI3K and p-AKT band and much fainter PTEN band in NDRG1 downregulation cell lysates, whereas the opposite were revealed in NDRG1 upregulated pancreatic cancer cells.