Identification of glycolysis-associated long non-coding RNA regulatory subtypes and construction of prognostic signatures by transcriptomics for bladder cancer.
Chenyu MaoYuan GaoMingyu WanNong XuPublished in: Functional & integrative genomics (2022)
Glycolysis-targeted cancer therapy based on long non-coding RNAs (lncRNAs), owing to its high specificity and less toxicity, is at the preclinical stages. Our study aimed to examine the roles of the core glycolysis-associated lncRNAs in bladder cancer (BC). Glycolysis scores of BC were computed by single-sample gene set enrichment analysis (ssGSEA). Glycolysis-associated lncRNAs were screened by Pearson's correlation analysis. Unsupervised consensus clustering using ConsensusClusterPlus assessed the glycolysis-associated lncRNAs for the identification of molecular subtypes of BC. The Kaplan-Meier survival analysis, genomic mutations, and tumor microenvironment (TME) analysis were used to compare the characteristics of different subtypes. Key glycolysis-associated lncRNAs were screened by first-order partial correlation and univariate Cox proportional-hazards model analyses; finally, the lncRNA signature was constructed. Four glycolysis-associated lncRNA-regulated subtypes having differential overall survival (OS), clinical features, genomic mutation profiles, and TME profiles along with nuclear immunotherapeutic responses were identified. Nine lncRNAs localized in the nucleus were identified and transcription factors (TFs) significantly negatively associated with these were found to be enriched in multiple oncogenic signaling pathways. Among them, three lncRNAs (AC093673.5, AC034220.3, and RP11-250B2.3) exerted the most profound effects on glycolysis and constituted the lncRNA signature, which could substantially distinguish the risk levels among different BC patients. Four glycolysis-associated lncRNA-regulated subtypes were identified in this study, reflective of the biological characteristics and heterogeneity of BC. Three key glycolysis-associated lncRNA constituting a signature could predict the risk levels in BC, provide a reference for stratification, and be used as prognostic markers for BC diagnosis and treatment.
Keyphrases
- long non coding rna
- poor prognosis
- transcription factor
- genome wide identification
- cancer therapy
- network analysis
- machine learning
- genome wide analysis
- single cell
- signaling pathway
- newly diagnosed
- genome wide
- oxidative stress
- ejection fraction
- stem cells
- epithelial mesenchymal transition
- end stage renal disease
- prognostic factors
- wastewater treatment
- dna methylation
- bone marrow
- cell therapy
- diffusion weighted imaging