M5C-Related lncRNA Predicts Lung Adenocarcinoma and Tumor Microenvironment Remodeling: Computational Biology and Basic Science.

Abstract

Purpose: Epigenetic RNA modification regulates gene expression post-transcriptionally. The aim of this study was to construct a prognostic risk model for lung adenocarcinoma (LUAD) using long non-coding RNAs (lncRNAs) related to m5C RNA methylation. Method: The lncRNAs regulated by m5C methyltransferase were identified in TCGA-LUAD dataset using Pearson correlation analysis (coefficient > 0.4), and clustered using non-negative matrix decomposition. The co-expressing gene modules were identified by WGCNA and functionally annotated. The prognostically relevant lncRNAs were screened by LASSO regression and a risk model was constructed. LINC00628 was silenced in the NCI-H460 and NCI-H1299 cell lines using siRNA constructs, and migration and invasion were assessed by the Transwell and wound healing assays respectively. Results: We identified 185 m5C methyltransferase-related lncRNAs in LUAD, of which 16 were significantly associated with overall survival. The lncRNAs were grouped into two clusters on the basis of m5C pattern, and were associated with significant differences in overall and disease-free survival. GSVA revealed a close relationship among m5C score, ribosomes, endolysosomes and lymphocyte migration. Using LASSO regression, we constructed a prognostic signature consisting of LINC00628, LINC02147, and MIR34AHG. The m5C-lncRNA signature score was closely related to overall survival, and the accuracy of the predictive model was verified by the receiver operating characteristic curve and decision curve analysis. Knocking down LINC00628 in NCI-H460 and NCI-H1299 cells significantly reduced their migration and invasion compared to that of control cells. Conclusion: We constructed a prognostic risk model of LUAD using three lncRNAs regulated by m5C methyltransferase, which has potential clinical value.