Dysregulation of miR-15a and miR-214 in human pancreatic cancer

To identify dysregulated miRNAs, we used qRT-PCR to measure the expression of seven mature miRNAs (miR-15a, miR-27a, miR-100, miR-125b, miR-181a, miR-200a and miR-214) in 10 pancreatic cancer tissues and their adjacent benign tissues. These seven mature miRNAs were chosen based on recent reports that identified them as having important functions in cancers. After normalization to U6 RNA expression as a control, the differential expression patterns of the miRNAs in cancer and benign pancreatic tissues were determined.

Among the miRNAs studied, we found that four miRNAs were frequently overexpressed in the cancer tissues studied: miR-100, miR-125b, miR-200a and miR-214 (Table 1 and Figure 1A). In particular, miR-214 expression was elevated in 8 of 10 (80%) cancer specimens. Only one miRNA, miR-15a, showed decreased expression in cancer tissues compared with matched benign pancreatic tissues; this effect was evident in 7 of 10 (70%) samples (Figure 1B).

Among the four upregulated miRNAs, miR-214 was previously reported to be associated with mouse pancreas development [14]. However, there are no reports on the function of miR-214 in human pancreas development or in the chemoresistance of pancreatic cancer. This is the first report implicating the dysregulation of miR-214 in pancreatic cancer. As for miR-15a, a tumor suppressor that has been reported in various cancers, its functions in pancreatic cancer are unknown; however, it was the only one downregulated in our examination. Therefore, miR-214 and miR-15a were chosen for further study.

To investigate the potential functions of miR-15a and miR-214 in pancreatic cancer, we first measured the viability of cells transfected with miR-15a/miR-214 mimics or their controls (mimics-NC) using the CCK-8 assay. BxCP-3 pancreatic cells were used in our examination. The transfection efficiency of both miR-15a and miR-214 and their corresponding controls in BxCP-3 cells was measured by qRT-PCR assay. The results were analyzed using the paired Student's t-test. MiR-214 was upregulated more than 14-fold in BxCP-3 cells after transfection, whereas miR-15a was upregulated about 6-fold (Figure 2A); this result indicated better transfection efficiency of miR-214. We then assessed cell viability. The CCK-8 assay showed that overexpression of miR-15a significantly decreased the viability of BxCP-3 cells compared with the control (p < 0.05) (Figure 2B). These results indicate that the expression level of miR-15a is important for pancreatic cancer cell growth. Because miR-15a was downregulated in pancreatic cancer, we hypothesized that miR-15a might function as a tumor suppressor in the disease, a role it has been shown to play in other cancers [15‐18].

Documented evidence indicates that miR-214 functions as either an oncogene or a tumor suppressor in different cancers. It was also reported that miR-214 negatively regulates HeLa cell proliferation and increases the ability of T cells' viability [19, 20]. However, we observed no obvious effect of miR-214 overexpression on cell viability (p > 0.05) (Figure 2B), which implies that miR-214 might have other roles in pancreatic cancer. A previous study showed that miR-214 can promote cell survival and cisplatin resistance in human ovarian cancer [21]. Because overexpression of miR-214 was observed in pancreatic cancer tissues, we questioned whether this phenomenon might be related to tumor cell survival and drug resistance in pancreatic cancer.

To address this issue, we investigated the expression patterns of miR-214 in BxCP-3 cells treated with GEM. GEM is currently the first-line treatment for advanced pancreatic cancer, and it acts by inhibiting tumor cell proliferation and inducing apoptosis [22‐25]. Prior to determining the effects of GEM on miR-214, we examined the effect of GEM on cell viability at 24, 48 and 72 hrs using the CCK-8 assay. Cell viability decreased in a time-dependent manner in response to GEM treatment (Figure 2C). After 72 hrs of 10 μM GEM treatment, cell viability decreased to approximately 20% compared with untreated cells. Next, we detected the expression pattern of miR-214 in cells treated with GEM. We found that miR-214 was dramatically downregulated after treatment with GEM. MiR-214 levels decreased by 60% at 24 hrs and remained low for 72 hrs (Figure 2D), indicating that miR-214 was responding to the drug treatment. We then investigated whether overexpression of miR-214 could modulate the sensitivity of BxCP-3 cells to GEM-induced apoptosis. After 72 hrs of GEM treatment, we found that the viability of BxCP-3 cells transfected with miR-214 mimics was significantly higher (about 22%) than that of the NC and MOCK negative control groups (Figure 2E). These results suggest that miR-214 might be involved in the chemoresistance of pancreatic cancer cells.

To further study the mechanisms of both miR-15a and miR-214 in pancreatic cancer cells, we predicted and validated potential targets for both miRNAs. Putative target genes that were identified by one or more of five different target prediction algorithms (PicTar, TargetBoost, TargetScanS, MiRanda and miRbase) were screened for the location and number of putative binding sites as well as their biologic relevance. Among the candidate targets of miR-15a chosen for experimental validation were PIM1, CDC25A, BCL2L2, WNT3A, SMAD7, LRP6 and FGF7, each of which has been reported to play a role in cell proliferation (Table 2). Using the same methods, seven candidates: RASSF5, PIM1, BAX, BIK, NEO1, ACVR1B and ING4, were predicted as the putative targets of miR-214 and chosen for experimental validation (Table 3). The wild-type 3'-UTR of each gene was cloned into the 3'-UTR of a Renilla luciferase reporter gene of a modified psiCHECK2 expression vector, and the resultant constructs were transfected into 293T cells using Lipofectamine 2000. Luciferase expression in cells expressing the WNT3A and FGF7 reporters was significantly suppressed (18% and 20%, respectively) when co-transfected with miR-15a mimics (Figure 3A and 3C). These data indicate that WNT3A and FGF7 might be targets of miR-15a. In addition, miR-214 repressed the luciferase activity of the ING4 reporter construct by 13% (Figure 3B and 3C). Expression levels of the remaining reporter constructs were unaffected by miRNA co-transfection.

WNT3A is a member of the Wnt/β-catenin signaling pathway. Dysregulated Wnt/β-catenin signaling has been linked to various human diseases, including cancer. WNT3A promotes the activation of survival and proliferation pathways through the phosphorylation of the kinases ERK and Akt. Here, we demonstrated that WNT3A may also be a direct target of miR-15a. Moreover, we identified FGF7, a fibroblast growth factor, as another potential target of miR-15a. FGF7 was reported to play an important role in pancreatic organogenesis, and FGF10/FGFR2 signaling recently emerged as a promising new molecular target for pancreatic cancer [26]. MiR-15a directly targets multiple genes relevant in pancreatic cancer and therefore may serve as a novel therapeutic target in pancreatic cancer.

The tumor suppressor ING4 belongs to the ING family of genes, which comprises type II tumor suppressor genes [27, 28] involved in cell cycle arrest, transcriptional regulation, DNA repair and apoptosis. Downregulation of ING4 has been reported in various tumors, including gliomas, breast tumors and stomach adenocarcinoma. Hepatocellular carcinoma (HCC) patients with low ING4 expression had poorer overall survival and disease-free survival than those with high expression [29]. Xie et al. found that upregulation of ING4 could suppress lung carcinoma cell invasiveness and reduce tumor microvessel formation [30]. It was also reported that miR-650 targets ING4 to promote gastric cancer tumorigenicity [31]. In the present study, we found that ING4 is a potential target of miR-214, which was overexpressed in pancreatic cancer and could modulate the sensitivity to GEM-induced apoptosis in BxCP-3 cells. Expression levels of miR-214 could potentially serve as prognostic markers; however, the utility of miR-214 as a therapeutic target in human pancreatic cancer remains to be determined.

BxPC-3 human pancreatic cancer cells were maintained in RPMI 1640 medium containing 10% fetal bovine serum (FBS; Gibco BRL). 293T cells were maintained in DMEM containing 10% FBS.

Ten samples of pancreatic cancer tissues and their adjacent benign tissues were obtained from patients at the Second Affiliated Hospital of Sun Yat-sen University. All specimens were immediately snap-frozen in liquid nitrogen and stored at -80°C. Patient characteristics are available for all patients. Written informed consent for the biological studies was obtained from the patients involved in the study or from their parents/guardians. The study was approved by the Ethics Committee of the affiliated hospitals of Sun Yat-sen University.

Total RNA was isolated with Trizol (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. qRT-PCR was performed as previously described [32] using the Hairpin-it™miRNAs Real-Time PCR Quantization Kit (GenePharma, Shanghai, China) containing a stem-loop-like RT primer and PCR primers specific to the various miRNAs or the U6 RNA internal control (Table 4). The expression of miRNAs in tumor tissues relative to that in adjacent benign tissues was determined using the 2^-ΔΔCT method [33]. Briefly, the △C_T of each miRNA was determined relative to that of the U6 endogenous control RNA, which was robustly and invariantly expressed across all samples. MiRNA expression levels in each of the 10 microdissected pancreatic cancer tissues were compared against matched benign pancreatic tissues, and each sample was assessed in triplicate for each miRNA.

Target gene prediction was performed to meet the following two criteria. First, miRNA targets were analyzed using following algorithms, TARGETSCAN http://​www.​targetscan.​org/​, PICTAR http://​pictar.​mdc-berlin.​de/​, TargetBoost, and Miranda (Miranda IM - Home of the Miranda IM client. Smaller, Faster, Easier) and miRBase http://​microrna.​sanger.​ac.​uk/​sequences/​index.​shtml. Second, to reduce the likelihood of false positives, only putative target genes predicted by at least two of the programs were accepted.

BxPC-3 cells (1 × 10⁴ per well) were plated in 96-well plates in RPMI medium 1640 and 10% FBS that was supplemented with sodium pyruvate at 37°C in a humidified atmosphere of 5% CO₂. Cells were transfected with 100 nM miRNA duplex (Ambion) or scrambled duplex (negative control, Ambion) using Lipofectamine 2000 (Invitrogen). For the cell viability study, cytotoxicity was determined in BxCP-3 cells treated with GEM using the CCK-8 assay. Cells were plated at 1 × 10⁴ per well in a 96-well plate and allowed to adhere for 8 hrs. The cells were then cultured in the absence or presence of 10 μM GEM for 24, 48 or 72 hrs. After GEM treatment, cell viability was measured using the CCK-8 assay.

The 3'-untranslated terminal region (3'-UTR) segments (Table 2, Table 3) of 59 bp of the 3'-UTR of the target genes were synthesized by Sangon (Shanghai) and inserted into the psi-CHECK-control vector (Promega) for miRNA functional analysis.

Transient transfection was performed in 293T cells with 100 nM miR-15a or miR-214 mimics and 0.1 μg of psi-CHECK-control or psi-CHECK-3'UTR fluorescence reporter constructs. Fluorescent activities were measured consecutively using Dual-Luciferase assays (Promega) 24 hrs after transfection, according to the instructions of the manufacturer.