Abstract
MiRNAs are small non-coding RNAs that post-transcriptionally regulate gene expression. They play important roles in cancer but little is known about the specific functions that each miRNA exerts in each type of cancer. More knowledge about their specific targets is needed to better understand the complexity of molecular networks taking part in cancer. In this study we report the miRNA-mRNA interactome occurring in pancreatic cancer by using a bioinformatic approach called miRComb, which combines tissue expression data with miRNA-target prediction databases (TargetScan, miRSVR and miRDB). MiRNome and transcriptome of 12 human pancreatic tissues (9 pancreatic ductal adenocarcinomas and 3 controls) were analyzed by next-generation sequencing and microarray, respectively. Analysis confirmed differential expression of both miRNAs and mRNAs in cancerous tissue versus control, and unveiled 17401 relevant miRNA-mRNA interactions likely to occur in pancreatic cancer. They were sorted according to the degree of negative correlation between miRNA and mRNA expression. Results highlighted the importance of miR-148a and miR-21 interactions among others. Two components of the Notch signaling pathway, ADAM17 and EP300, were confirmed as miR-148a targets in MiaPaca-2 pancreatic cancer cells overexpressing miR-148a. Moreover, a CRISPR-Cas9 cellular model was generated to knock-out the expression of miR-21 in PANC-1 cells. As expected, the expression of two miRComb miR-21 predicted targets, PDCD4 and BTG2, was significantly upregulated in these cells in comparison to control PANC-1.