Construction of an initial microRNA regulation network in breast invasive carcinoma by bioinformatics analysis

•We identified 68 aberrantly expressed miRNAs in breast invasive carcinoma (BRCA).•We found 125 strongly correlated miRNA/target gene pairs in BRCA.•We validated 6 miRNA/target gene pairs.•We identified 9 miRNAs and 12 miRNA/target pairs to distinguish BRCA subtypes.

IntroductionmicroRNAs (miRNA) are involved in many biological processes. They repress target gene expression and play a vital role in breast invasive carcinoma (BRCA). Although many miRNAs are identified to be aberrantly expressed in BRCA and deemed as tumor markers, only sporadic individual studies report their target genes and the pathways involved.MethodsmiRNA and mRNA expression data were collected from the Cancer Genome Atlas (TCGA) pilot project. Aberrantly expressed miRNAs and mRNAs in BRCA were identified by comparing tumor samples with normal adjacent tissues. Differentially expressed miRNAs and mRNAs in different breast cancer subtypes were also analyzed. miRNA/target correlations were predicted by calculating the spearman correlation coefficients between miRNA and mRNA, and validated by luciferase assay.Results31 up-regulated miRNAs, 37 down-regulated miRNAs, 1105 up-regulated mRNAs and 1222 down-regulated mRNAs were identified in BRCA; 125 miRNA/target correlations were predicted, 6 of them were validated. In addition, we also found 9 miRNAs and 143 mRNAs differently expressed between estrogen receptor positive and negative breast cancers, and 4 miRNAs and 46 mRNAs differently expressed between progesterone receptor positive and negative breast cancers. Twelve miRNA/target correlations determined the breast cancer subtypes.ConclusionWe developed a new systematic analytic method for analyzing TCGA database, which took into account both miRNA and mRNA data to dissect the miRNA regulation network in BRCA.