A total of 20 cases of primary PC patients and their paired adjacent non-tumor tissues were obtained from the Affiliated Hangzhou Hospital of Nanjing Medical University. All procedures performed in studies involving human participants were in accordance with ethical standards of the institutional and/or national research committee and with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. Informed consent for publication was obtained from all individual participants included in the present study. Human PC cell lines hTERT-HPNE, MIA-PaCa-2, PANC-1, and Aspc-1were purchased from the Cell Biology Institute of Shanghai (Shanghai, China). All cell lines were cultured in DMEM medium supplemented with 10% fetal bovine serum (FBS) and maintained at 37°C in a 5% CO₂ atmosphere. To construct overexpression plasmid, we synthesized sequences of human pre-miR-23a (miR23a-P1 (HpaI): TATCACATTGCCAGGGATTTCCTTCAGAGAGGAAATCCCTGGCAATGTGATTTTTTTC, miR23a-P2 (XhoI): TCGAGAAAAAAATCACATTGCCAGGGATTTCCTCTCTTGAAGGAAATCCCTGGCAATGTGATA), then inserted it into the linear plasmid PLL3.7. miR-23a inhibitors and inhibitor NC were purchased from Guangzhou RiboBio Co., Ltd., (Guangzhou, China). Before transfection, cells were seeded at 24-well plates and transfected with blank, PLL-3-blank, PLL-3-miR-23a, miR-23a-inhibitor, inhibitor NC using Lipofectamine™ 2000 Transfection Reagent (cat. no. 11668-500; Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) applying with 100 pmol oligonucleotides in each well according to the manufacturer's instructions.

Total RNA was isolated from the tissues and cells utilizing TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.). The primers for qRT-PCR were indicated in Table I. GAPDH or U6 snRNA served as an internal control for mRNA and miRNA expression, respectively, and quantified relative to control using the delta-delta CT method.

Protein extraction from tissues and cells were performed according to the manufacturer's instructions. Immunoblotting was conducted using the following primary antibodies: anti-PLK-1, Bax, Bcl2, cyclinB1, E-cadherin, Vimentin (Abcam, Cambridge, USA), anti-GAPDH (Abcam), and the second antibody horseradish peroxidase conjugated against mouse or rabbit IgG (1:2,000 dilution; ZhongShan, Guangdong, China). Signals were exposed to a film (FujiFilm, Tokyo, Japan) with ECL Plus (Beyotime Institute of Biotechnology, Haimen, China) according to the manufacturer's instructions.

CCK8 Kit (YEASEN, Shanghai, China) was used to analyze cell proliferation rate of cells with different treatments. In brief, cells were suspended and seeded into 96-well plates, after culturing for 24, 48 and 72 h. CCK8 was added into the medium following the manufacturer's instruction. The absorbance was detected at 450 nm. Cell proliferation rate was presented with the value relative to control group.

We added suspended and stained (Cell-Tracker-Red; Invitrogen; Thermo Fisher Scientific, Inc.) MIA-PaCa-2 cells on a confluent HMC monolayer, which was stained with the fluorescence dye Cell-Tracker-Green (Invitrogen; Thermo Fisher Scientific, Inc.). 24 h for migration analysis, and 48 h for invasion analysis.

Cells were grown in 6-well plates to about 50–60% confluence and transiently transfected with blank, PLL3-blank, PLL-3-miR-23a, miR-23a-inhibitor and inhibitor NC respectively. Cells were trypsinized and collected at 48 h post-transfection before washing with PBS twice, treated by Annexin V-EGFP Apoptosis Detection kit according to the manufacturer's instructions (Nanjing KeyGen Biotech., Co., Ltd., Jiangsu, China), and analyzed with a flow cytometer (FACS calibur; BD Biosciences, Franklin Lakes, NJ, USA).

PLK1-3′UTR-WT and PLK1-3′UTR-mut plasmids were obtained from GeneChem, Inc., (Daejeon, Korea). Cells were cotransfected with non-relative control RNA duplex and PLK1-3′UTR-WT, or cotransfected with non-relative control RNA duplex and PLK1-3′UTR-mut plasmids were served as control groups. Cells cotransfected with PLL3-miR23a and PLK1-3′UTR-WT or PLK1-3′UTR-mut were defined as case groups. The pRL-TK (Promega Corporation, Madison, WI, USA) was used as a normalization control. Cells were collected at 24 h after transfection, and luciferase activity was measured using a dual-luciferase reporter assay kit (Promega Corporation) and recorded by chemiluminescence meter (Promega Corporation).

All experiments were carried out under the ethical approval of Ethics Committee for Animal Experimentation of The Affiliated Hangzhou Hospital of Nanjing Medical University. Four-week-old female BALB/c nude mice were purchased from ShangHai and maintained in specific pathogen-free conditions. To build the subcutaneous human MIA-PaCa-2 tumor xenograft model, MIAPaca-2 cells were subcutaneously inoculated (5×10⁶ cells suspended in a 1:1 mixture of serum free-Dulbecco's modified Eagle's medium/Matrigel (BD Biosciences per site); total volume 0.1 ml) bilaterally into the flanks of athymic mice. When tumors were grown at approximately 100 mm³, mice were injected with miR-23a agomir or agomir NC (Guangzhou RiboBio Co., Ltd.) following the manufacturer's recommendation. The animals were monitored for activity, physical condition, determination of body weight, and measurement of tumor volume (1/2 × length × width²) were made twice a week. After 4 weeks of treatment, tumors were harvested and weighed. To measure the tumors accurately, mice were rendered unconscious with and digital calipers were used to measure the tumor size over time.

All data were obtained from at least three independent experiments, and presented as the mean ± standard deviation. Datasets with only two groups were analyzed using a Student's t-test. Differences between multiple groups were analyzed using one-way analysis of variance with the Tukey-Kramer post-hoc test. P<0.05 was considered to indicate a statistically significant difference. All of the statistical calculations were performed using the SPSS software v15.0 (SPSS, Inc., Chicago, IL, USA).

Sequence analyses of the mature sequence of miR-23a showed that miR-23a could be the potential functional regulator of 3′UTR of PLK mRNA. To assess whether PLK1 was a potential target of miR-23a, we performed qPCR to detect expression levels of miR-23a and PLK-1 in three human PC cell lines (PANC-1, MIA-PaCa-2 and Aspc-1) and one human pancreatic ductal epithelial cells hTERT-HPNE. As shown in Fig. 1A and B, the expression level of miR-23a and PLK-1 mRNA were significantly increased in three PC lines compared to hTERT-HPNE cell. However, the expression pattern of miR-23a and PLK-1 showed a reverse trend in three PC cell lines. For example, compared with MIA-PaCa-2 and Aspc-1 cell lines, the highest expression of miR-23a was shown in PANC-1 cells, while PLK-1 level was lowest in PANC-1 cells. To further verify the direct relationship between miR-23a and PLK1, we constructed PLK1-3′UTR-WT and PLK1-3′UTR-mut plasmids to perform double luciferase reporter assay in MIA-PaCa-2 cell line. Results indicated that up-regulation of miR-23a dramatically repressed the luciferase activity by 50% when cotransfected with PLK-1 3′UTR-WT in PC cells, but there was no change in the PLK-1 3′UTR-mut group (Fig. 1C and D). Therefore, these findings suggest that PLK-1 is a potential target of miR-23a in PC cells.

We further explored whether miR-23a could affect biological functions of PC cells. Cell proliferation assay suggested that miR-23a overexpression significantly suppressed the proliferation of PC cell at 24~48 h post transfection, while transfection with miR-23a inhibitor moderately enhanced the multiplication capacity of PC cells compared with control groups (Fig. 2). According to the statistical analysis indicated that MIA-PaCa-2 cell growth rate was the lowest in miR-23a transfection group compared with the control groups, and the highest in group transfected with miR-23a inhibitor after transfection for 24, 48, and 72 h. However, there was no significant change among the three control groups at any point in time (Fig. 3A-C).

We further investigated the effects of miR-23a on the capability of migration and invasion in MIA-PaCa-2 cells. As shown in Fig. 2C, miR-23a overexpression prominently reduced cell migration compared with blank and PLL3.7 group, while miR-23a inhibitor increased the migration ability of MIA-PaCa-2 cells. Similar results were acquired in cell invasion assay (Fig. 2C, lower panel). Furthermore, transfection efficiency was also verified through detecting the transcriptional level of has-miR-23a to support our findings (Fig. 3D). Taken together, we demonstrate that miR-23a acts as a tumor suppressor in PC cells.

We sought to examine whether miR-23a could inhibit cell growth by modulating cellular apoptosis. Flow cytometry assays showed that miR-23a overexpression promoted cell apoptosis significantly (52.22% vs. 4.92%), while transfection with miR-23a inhibitor decreased cell apoptotic rate compared with NC group (4.32% vs. 10.68%) (Fig. 4A). To investigate the underlying mechanism of proliferation inhibition and apoptosis promotion induced by miR-23a, the expression level of several proliferation and apoptosis related proteins, including oncogene PLK-1, apoptotic activator Bcl-2-associated X protein (Bax), Bcl-2, tumor suppressor E-cadherin, cell cycle-associated protein Cyclin B1 and Vimentin were determined, respectively. As expected, the protein expression of PLK-1 was decreased or increased in cells transfected with miR-23a mimic or inhibitor, respectively (Fig. 4B). Moreover, miR-23a overexpression inhibited Bcl-2, Cyclin B1, and Vimentin expression, while promoted Bax and E-cadherin expression, while opposite effects were observed in cells transfected with miR-23a inhibitor. Therefore, our findings demonstrate that miR-23a mediates cell apoptosis probably by targeting PLK-1 and regulating downstream effectors.

Furthermore, we tried to observe the critical anti-carcinoma role of miR-23a in vivo. As indicated in Fig. 5A, the tumor masses distinctly diminished in group of MIA-PaCa-2 cells injected with miR-23a agomir compared to control and negative control groups at 24 days after initial injection. We also stripped out the neoplasm at 31 days post-injection, the result showed that tumor size was the lowest in miR-23a agomir-treated group than control groups (Fig. 5B). Moreover, according to the statistic analyses of tumor growth curve, we could see from Fig. 5C, control and negative control group presented a similar increasing tendency, but up-regulation of miR-23a obviously slowed tumor growth. We further explored the transcriptional levels of miR-23a and PLK-1 of tumor mass in each group, and found PLK-1 expression was significantly suppressed in miR-23a agomir-treated group (Fig. 5D), and these data further verified the essential inhibition of miR-23a on cell proliferation in vivo.

To further determine the correlation between miR23a and PLK-1in human PC samples, we examined transcriptional levels of miR-23a and PLK-1 in 20 cases of primary PC patients and their paired adjacent non-tumor tissues by using qPCR. Among 20 pairs of PC patient tissue samples, 19 (95%) cases showed remarkable up-regulation of miR-23a in cancer tissues (Fig. 6A and B). However, miR-23a expression was negatively correlated with tumor grade, tumor size and lymph node metastasis (Table II). As shown in Fig. 6C, 10 (50%) cases showed up-regulation of PLK-1 in cancer tissues compared with non-tumor tissues, indicating its high expression was potential correlation with the progression of PC. Conversely, 10 (50%) cases presented down-regulation of PLK-1 in tumor samples compared to non-tumor tissues. Importantly, pearson's correlation coefficient analysis indicated that miR-23a and PLK-1 expression was negatively correlated in PC tissues (Fig. 6D). Thus, these findings in human PC tissues present some contradictions with our in vitro data.