A Computationally Constructed lncRNA-Associated Competing Triplet Network in Clear Cell Renal Cell Carcinoma

Abstract

Long noncoding RNAs (lncRNAs) are revealed to be involved in the tumorigenesis and progression of human malignancies mediated by microRNA (miRNA) via the competing endogenous RNA (ceRNA) mechanism, a newly proposed “RNA language.” However, the lncRNA-associated competing triplet (lncACT) network among ceRNA transcripts in clear cell renal cell carcinoma (ccRCC) is currently lacking. We carried out differential expression analysis to identify aberrantly expressed lncRNAs, miRNAs, and mRNAs by analyzing the RNA-seq data of 420 ccRCC tissues and 71 noncancerous kidney tissues obtained from The Cancer Genome Atlas (TCGA). Then, a ccRCC-specific ceRNA network was built using computational algorithms, including miRcode, TargetScan, miRanda, and miRTarBase. In total, 1491 dysregulated lncRNAs were found between normal renal tissues and ccRCC (fold $\text{change}>4$ and false discovery $\text{rate}<0.01$ ). A ceRNA network that comprised of 46 DElncRNAs, 11 DEmiRNAs, and 55 DEmRNAs was established by integrating the lncRNA/miRNA and miRNA/mRNA interactions into lncACTs. Several lncRNAs were identified to be significantly associated with clinical features of ccRCC patients. Notably, four key lncRNAs (TCL6, HOTTIP, HULC, and PCGEM1) were tightly correlated with both patients’ clinical characteristics and overall survival (log-rank $P<0.05$ ), indicating their potential important roles in ccRCC. HOTTIP may be a potential prognostic and therapeutic molecular marker for ccRCC patients. Collectively, our results provide a comprehensive view of the lncRNA-associated ceRNA regulatory network for a better understanding of the mechanisms and prognosis biomarkers for ccRCC.