Construction of a prognostic glycolysis‐related lncRNA signature for patients with colorectal cancer

Abstract

Aerobic glycolysis is a common metabolic phenotype in tumors that helps cancer cells adjust to severe living conditions and can aid metastasis in several types of carcinomas, including colorectal cancer (CRC). Long non‐coding RNAs (lncRNAs) can influence tumor biology and have been previously used to assess patients' outcomes and to identify potential therapeutic targets. However, despite the importance of glycolysis‐related lncRNAs (GRLs) in the development of CRC, studies on their use as prognostic markers are still limited. Herein, we applied a series of bioinformatic analyses to screen potential prognostic lncRNAs for colorectal cancer. Out of all lncRNAs screened, nine GRLs were selected to constitute a prognostic signature. Based on the signature, two molecular subtypes were classified with distinct prognostic outcomes and excellent diagnostic accuracy (The 1‐, 3‐ and 5‐year AUC are 0.756, 0.716, and 0.721, respectively). The prognostic value of this signature was further validated using another cohort. The enriched molecular pathways, immune infiltration, and mutation landscape were also significantly different between the two groups. The different drug sensitivity results between the two groups suggest a potential strategy for precise treatment. Furthermore, we confirmed that AFAP1‐AS1 could regulate aerobic glycolysis and metastasis of CRC cells. Overall, we developed a glycolysis‐related lncRNA (GRL) signature and suggested that this signature could offer a predictive value and identify potential therapeutic targets for cancer therapy.

Machine learning-based glycolysis-associated molecular classification reveals differences in prognosis, TME, and immunotherapy for colorectal cancer patients

“…Therefore, it is crucial to investigate the subtypes of glycolysis in CRC to uncover potential therapeutics. Currently, despite some studies reporting CRC’s glycolytic signature employing several genes and long non-coding RNAs ( 28 , 29 ), there is no research focusing on the gene-based molecular subtypes of glycolysis in CRC. In this study, we used the unsupervised clustering method to categorize each CRC patient using genes associated with glycolysis.…”