Abstract
Epitranscriptome studies have shown that critical RNA modifications drive tumorigenicity; however, the role of 5-methylcytosine (m(5)C) RNA methylation remains poorly understood. We extracted 17 m(5)C regulators and clustered distinct m(5)C modification patterns by consensus clustering analysis. Gene set variation and single-sample gene set enrichment analysis were applied to quantify functional analysis and immune infiltration. The least absolute shrinkage and selection operator was employed to develop a prognostic risk score. Kaplan-Meier with log-rank test was used for survival analysis. Differential expression analysis was performed with the "limma" R package. Wilcoxon signed ranked test or Kruskal-Wallis test was used to compare groups. We observed that m(5)C RNA methylation was commonly upregulated in gastrointestinal cancer and related to prognosis. Clusters were identified for m(5)C patterns, with distinct immune infiltrations and functional pathways. The risk scores of m(5)C regulators were independent risk factors. Differentially expressed mRNAs (DEmRNAs) in m(5)C clusters were involved in cancer-related pathways. The methylation-based m(5)Cscore showed a significant effect on the prognosis. Patients with a lower m(5)Cscore exhibited more therapeutic efficiency on anti-CTLA4 therapy in liver cancer, while the combination of anti-CTLA4 therapy and pd1 was more efficient for patients with a lower m(5)Cscore in pancreatic cancer. We uncovered dysregulations of m(5)C-related regulators in gastrointestinal cancer and their associations with overall survival. Some immune cells were differently infiltrated in distinct m(5)C modification patterns, indicating their potential impacts on gastrointestinal cancer cell-immune. Moreover, an m(5)Cscore, derived from DEmRNAs in specific clusters, can serve as a classifier for immunotherapy.