Pancreatic tumor and normal adjacent tissues were obtained from three pancreatic cancer patients who underwent surgery in the Department of General Surgery, Shanghai First People’s Hospital, Shanghai Jiaotong University from October 2012 to May 2013. According to the tumor-node-metastasis (TNM) staging system (7th edition) of the American Joint Commission on Cancer (AJCC), all patients had stage I or II (T1-3, N0-1, M0) pancreatic cancer. None of the patients had received chemotherapy or cytotoxic agents within the last 12 months prior to the inclusion into this study. None of the patients showed clinical signs for active infectious diseases. Patients with stage III and IV pancreatic cancer with locally advanced unresectable foci (T4) and/or systemic metastases (M1) were excluded. Informed consent was obtained from all study participants. The present study was approved by the Ethics Committee of Shanghai First People’s Hospital, Shanghai Jiaotong University, Shanghai, China.

Primary endothelial cell cultures derived from human pancreatic tumor masses or adjacent normal tissue were prepared according to the method reported by Naschberger et al (19). Briefly, after tissues were cut into 1-mm³ blocks, they were digested with collagenase II (17,100 U/g) in 5 ml of endothelial basal medium (EBM-2) supplemented with 0.5% fetal bovine serum (FBS) (EBM-low; both from Lonza, Cologne, Germany) at 37°C with 5% CO₂ for 1 h under constant agitation. Following digestion, cells were filtered (cell strainer 100 µm; BD Biosciences, Franklin Lakes, NJ, USA) and isolated by centrifugation at 500 × g for 5 min at 20°C. The cell pellet was resuspended in 5 ml of EGM-2-MV (EGM-2-MV BulletKit; Lonza), and the cell suspension was added to flasks pre-coated with 1.5% gelatin. The growth medium was refreshed every 2 days for 5–7 days until the cell confluence reached 70–80%. After the cells were detached with 1–2 ml of Accutase (Life Technologies, Carlsbad, CA, USA), they were washed with EGM-2-MV. The cells were resuspended in 60 µl of MACS buffer (1X PBS pH 7.2, 0.5% bovine serum albumin, 2 mM EDTA), and then were incubated with 20 µl of CD31 beads (CD31 MicroBead kit; Miltenyi Biotec, Bergisch Gladbach, Germany) at 4°C for 15 min. After washing with MACS buffer, the cells were resuspended in 1 ml MACS buffer, which was applied to a MACS separation column (Miltenyi Biotec), and the endothelial cells were isolated following the manufacturer’s instructions. The cell concentration was adjusted to 1−2×10⁴/cm² and cultured on 25 T flasks pre-coated with 0.5% gelatin. Cells were cultured until complete confluency was reached, and the medium was refreshed every 2 days.

Total RNA was isolated using the mirVana™ microRNA isolation kit according to the manufacturer’s instructions (Ambion, Austin, TX, USA). In brief, the sample was first lysed in a denaturing lysis solution, and the lysate was then subjected to phenol-chloroform extraction. After purification over a glass-fiber filter, the total RNA integrity was confirmed with an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA).

The miRNA expression profiles in CECs were compared with those of normal endothelial cells (NECs) derived from the same patient. Affymetrix FlashTag^® Biotin HSR was used for miRNA labeling and hybridization onto the Affymetrix GeneChip^® MicroRNA 3.0 Array (both from Affymetrix, Santa Clara, CA, USA) following the manufacturer’s instructions. The miRNA expression profile was scanned through a GeneChip^® Scanner 3000 (cat. #00-00212, Affymetrix), and primary data were analyzed using GeneChip-compatible™, Command Console Software 3.1 and Expression Console Software. The CEC and NEC miRNA expression profiles of the same patient were compared to identify differentially expressed mRNAs. In addition, the array results were also compared against the miRNA databases, microrna.org and TargetScan, to identify candidate miRNAs that are relevant to angiogenesis.

The differential expression of miR-200c, miR-182, miR-139-5p and miR-200b was validated by qPCR analysis using the stem-loop TaqMan MicroRNA Assay (Applied Biosystems, Foster City, CA, USA), according to the manufacturer’s instructions. Briefly, mature candidate miRNAs were reverse transcribed into cDNA from 10 ng of total RNA with mature miRNA-specific looped RT primers from the TaqMan MicroRNA assays kit and reagents from the TaqMan MicroRNA Reverse Transcription kit (Applied Biosystems) following the manufacturer’s instructions. Real-time PCR was performed on the 5′-extended cDNA with Applied Biosystems TaqMan 2X Universal PCR Master Mix and the appropriate 5X TaqMan MicroRNA Assay Mix (both from Applied Biosystems) for each miRNA of interest. For each sample, the threshold cycle (Ct) was calculated by the ABI 7500 Sequence Detection System software. Standard curves were used to determine miRNA concentration in the samples, which were then normalized to small nuclear RNA U (RNU) RNA.

The specific inhibitors for miR-139 or miR-200c (MiScript microRNA Inhibitor; Qiagen, Hilden, Germany) were added into CEC cultures to evaluate the effect of these miRNAs on CEC proliferation. In addition, since miR-1 regulates angiogenesis (i.e., enhanced tube formation and migration of human-derived cardiomyocyte progenitor cells) (20), a miR-1 inhibitor (Qiagen) was used as a positive control. An inhibitor with scrambled sequence that has no homology to any known mammalian gene (Qiagen) was used as a negative control. Briefly, cells were seeded in a 24-well plate at a density of 0.4−1.6×10⁵ cells/well in 500 µl of culture medium containing serum and antibiotics. After 4 h, cells were incubated with 50 nM miRNA inhibitor diluted in 50 µl of culture medium without serum and with 1.5 µl HiPerfect transfection reagent (Qiagen). The cells were incubated with the transfection complexes under normal growth conditions, and gene expression was monitored after 12 h by qPCR. qRT-PCR data were calculated as 2^−ΔΔCt after normalizing to the control.

Cell proliferation was assessed using the MTT assay (Sigma, St. Louis, MO, USA) 48 h after siRNA inhibition following the manufacturer’s instructions.

Transwell migration assay were conducted 48 h after siRNA inhibition using a fluorometric cell migration assay kit with polycarbonate membrane inserts (5-µm pore size; Cell Biolabs, San Diego, CA, USA) using a modified protocol described by Chim et al (21). Briefly, cells were serum-starved overnight in Dulbecco’s modified Eagle’s medium (DMEM) prior to initiation of the experiment. The lower chambers were filled with 1 ml of medium containing 10% serum. Cells (4×10⁴) were resuspended in 200 µl of Opti-MEM and added to the upper chamber. After 24 h at 37°C, migrating cells were counted after staining with crystal violet.

After cells were serum-starved overnight in DMEM, they were seeded in 24-well plates that were coated with Geltrex™ reduced growth factor basement membrane matrix (Invitrogen, Carlsbad, CA, USA) and incubated at 37°C for 30 min with Opti-MEM. Medium was then removed and replaced with the medium containing 10% serum and incubated at 37°C for 24 h. Four randomly selected fields of view were analyzed, and tube formation was quantified by measuring the length of tube-like structures and the number of branching points. Tube length was assessed by drawing lines along the tube-like structure and measuring the length of the line in pixels using a modified protocol as previously described (21).

For cell proliferation and migration as well as tube length and total loops, data are presented as mean ± standard deviation (SD). One-sample t-tests were performed to evaluate the cell proliferation with the mean of the control group set to 1. Differences among the negative control, miR-1, mir-139 and miR-200c inhibitor groups were also assessed by one-way ANOVA with a post-hoc LSD test as for pair-wise comparisons. One-sample t-tests were also performed to evaluate the mean tube length or total loops set as 1. Moreover, differences between the negative control and miR-1 inhibitor groups were compared using the two-sample t-test. Statistical assessments were two-tailed, and P-values <0.05 were considered to indicate statistically significant results. SPSS 18.0 statistics software (SPSS, Inc., Chicago, IL, USA) was used for all statistical analyses.

The miRNA expression profiles of CECs were compared with those of the NECs in three pancreatic patients. As shown in Table I, 14 miRNAs were upregulated by >20-fold in the CECs as compared to NECs, including hsa-miR-139-5p, hsa-miR-182, hsa-miR-183, hsa-miR-192, hsa-miR-194, hsa-miR-200a, hsa-miR-200b*, hsa-miR-200b, hsa-miR-200c, hsa-miR-203, hsa-miR-25*, hsa-miR-27a*, hsa-miR-375 and hsa-miR-92a-1*.

We compared the primary miRNA data with available miRNA databases (http://www.microrna.org; www.targetscan.org) to identify candidate angiomiRs.

Of the miRNAs with >20-fold increased expression, miR-182, miR-183, miR-192, miR-194, the miR-200 family, miR-203, miR-27a*, miR-375 and miR-92a-1* have been implicated in regulating tumor angiogenesis (16,22–32). We chose to validate the upregulated expression of miR-200b, miR-182, miR-139 and miR-200c using qPCR analysis as these candidate angiomiRs had the greatest increase in fold-expression in the CECs (Table I). As shown in Fig. 1, the upregulated expression of miR-139 and miR-200c was confirmed in CECs relative to NECs.

To determine whether either miR-139 or miR-200c influences CEC proliferation, specific inhibitors against them were utilized. As shown in Fig. 2, the efficiency of the miRNA inhibition was confirmed in three different pairs of NEC and CEC cultures. Each miRNA inhibitor suppressed the expression of their target miRNA by at least 80%. We next determined whether inhibition of either miR-139 or miR-200c could influence CEC proliferation using MTT assays. As shown in Fig. 3, cell proliferation was not altered upon inhibition of miR-139 or miR-200c as compared to the negative control group. Similar results were obtained after inhibition of miR-1, which is a known regulator of angiogenesis (20). However, cell proliferation was significantly decreased in all groups as compared to the untransfected control cells (all P<0.05). These data suggest that neither miR-139 nor miR-200c influences pancreatic CEC proliferation.

The effects of miR-139 and miR200c on CEC migration were next evaluated using their specific inhibitors. As compared to the untransfected cells, cell migration in all the samples was significantly decreased (all P<0.05, Fig. 4). Notably, significantly reduced CEC migration was observed after transfection with the miR-1, miR-139 and miR-200c inhibitors compared to the negative control group (all P<0.05). These data suggest that miR-139 and miR-200c regulate CEC migration.

The effects of miR-139 and miR200c on CEC tube formation were next evaluated as an in vitro measure of their effects on angiogenesis. Representative images of the cells from all three CEC cultures are shown in Fig. 5A–C. As shown in Fig. 5D and E, the average tube length and total loop number were significantly decreased with miR-139 and miR-200c inhibition in all three CEC cultures compared to those of the negative control group (all P<0.05). No such changes were observed with miR-1 inhibition. Thus, miR-139 and miR-200c may regulate vasculature formation during angiogenesis.