Constructing and Analyzing Competing Endogenous RNA Networks Reveal Potential Biomarkers in Human Colorectal Cancer.

Abstract

The role of the lncRNA-miRNA-mRNA competing endogenous RNA network in human colorectal cancer remains largely unknown, and accurate prognostics still elude us. This study aimed to identify differentially expressed mRNAs and lncRNAs between tumor and normal samples, delineate their interactions and find reliable biomarkers. We downloaded the RNA sequencing profiles and clinical information of 624 CRC patients from The Cancer Genome Atlas database. After expression difference analysis and interaction prediction, we identified 37 miRNAs, 5 lncRNAs, and 93 mRNAs to construct the ceRNA network (|log2 Fold Change| > 1, P-value < 0.05), and assessed relationships between them and clinical characteristics by t-test, Spearman correlation analysis, and Kaplan-Meier curve analysis. Besides, we validated PIGR and CD3D protein expression by immunohistochemistry staining. PIGR and CD3D mRNAs showed a negative correlation with tumor stage and their protein levels were lower in tumor tissues than in normal tissues. By survival analysis, MYC, F2RL2, and GINS2 positively correlated with the overall survival of CRC patients. Our study provides a novel comprehension of lncRNA-related ceRNA network in CRC and candidate molecules that serve as potential biomarkers of tumor stage and patient survival.