Identification of a Somatic Mutation-Derived Long Non-Coding RNA Signatures of Genomic Instability in Renal Cell Carcinoma

Abstract

Background: Renal cell carcinoma (RCC) is a malignant tumor with high morbidity and mortality. It is characterized by a large number of somatic mutations and genomic instability. Long non-coding RNAs (lncRNAs) are widely involved in the expression of genomic instability in renal cell carcinoma. But no studies have identified the genome instability-related lncRNAs (GInLncRNAs) and their clinical significances in RCC. Methods: Clinical data, gene expression data and mutation data of 943 RCC patients were downloaded from The Cancer Genome Atlas (TCGA) database. Based on the mutation data and lncRNA expression data, GInLncRNAs were screened out. Co-expression analysis, Gene Ontology (GO) enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis were conducted to explore their potential functions and related signaling pathways. A prognosis model was further constructed based on genome instability-related lncRNAs signature (GInLncSig). And the efficiency of the model was verified by receiver operating characteristic (ROC) curve. The relationships between the model and clinical information, prognosis, mutation number and gene expression were analyzed using correlation prognostic analysis. Finally, the prognostic model was verified in clinical stratification according to TCGA dataset. Results: A total of 45 GInLncRNAs were screened out. Functional analysis showed that the functional genes of these GInLncRNAs were mainly enriched in chromosome and nucleoplasmic components, DNA binding in molecular function, transcription and complex anabolism in biological processes. Univariate and Multivariate Cox analyses further screened out 11 GInLncSig to construct a prognostic model (AL031123.1, AC114803.1, AC103563.7, AL031710.1, LINC00460, AC156455.1, AC015977.2, ‘PRDM16-dt’, AL139351.1, AL035661.1 and LINC01606), and the coefficient of each GInLncSig in the model was calculated. The area under the curve (AUC) value of the ROC curve was 0.770. Independent analysis of the model showed that the GInLncSig model was significantly correlated with the RCC patients’ overall survival. Furthermore, the GInLncSig model still had prognostic value in different subgroups of RCC patients. Conclusion: Our study preliminarily explored the relationship between genomic instability, lncRNA and clinical characteristics of RCC patients, and constructed a GInLncSig model consisted of 11 GInLncSig to predict the prognosis of patients with RCC. At the same time, our study provided theoretical support for the exploration of the formation and development of RCC.