In silico prediction and experimental validation of MIR17HG long non-coding RNA, MIR17HG-derived miRNAs and GPC5 expression profile in Breast Cancer

Abstract

Abstract BACKGROUND Breast cancer as one of the most causes of cancer-related mortality in women has attracted the attention of researchers. Recently, biological biomarkers play important roles in the early diagnosis of breast cancers in clinics. They considered non-invasive biomarkers for cancer diagnosis and play an important role in the prevention of tumor development.OBJECTIVE We aimed to investigate lnc RNA MIR17HG which is the host gene for generating miR17-92 cluster. We employed bioinformatics and experimental approaches to evaluate the expression level of variants of lnc RNA MIR17HG and its derived miRNAs (miR18a-5p and miR20a-5p) and also its neighbor gene (GPC5). In addition, we evaluated the correlation of candidate genes to predict the similarity function of targeted genes in breast tumors and finally, we surveyed the efficacy of selected genes as new potential diagnostic biomarkers in discriminating against breast cancer patients and non-cancerous.METHODS We used bioinformatic tools to analyze TCGA data in order to predict results at the first step. The expression levels of candidate genes were assessed within tumors and adjacent normal tissues by qRT-PCR. Their impacts as diagnosis breast cancer biomarkers were evaluated by ROC curve analysis. The relation of candidate genes is also evaluated by Pearson's correlation coefficients.RESULTS According to our findings, MIR17HG and its derived miRNAs showed up-regulation and GPC5 showed down-regulation in BC. They also have a positive linear correlation in breast tumors and could discriminate between cancer and non-cancerous breast tissues.CONCLUSIONS Our data analysis showed differentially expressed of MIR17HG and its derived miRNAs and GPC5 in breast tissues compared to adj-normal tissue. Also, we demonstrate a linear correlation between candidate genes. In addition, selected genes can potentially act in discriminating tumor tissues and adj-normal tissue as breast cancer diagnosis biomarkers.