Abstract
INTRODUCTION: Colorectal cancer (CRC), characterized by its complex genetic heterogeneity and varied responses to treatment, is a leading cause of cancer-related mortality worldwide. The role of N1-methyladenosine (m1A)-related genes in tumor biology remains underexplored. This study aimed to investigate the prognostic value of m1A-related genes in CRC, characterize their role in tumor molecular subtyping, and explore their influence on the tumor microenvironment (TME) and immune infiltration. METHODS: To identify prognostic markers, univariate Cox analysis was performed using multiple datasets, including TCGA and GEO, identifying 43 m1A-related genes. Four distinct molecular subtypes of CRC were defined based on the expression of these genes using non-negative matrix factorization (NMF). Immune infiltration analysis was conducted, and the TIDE algorithm was used to predict response to immune checkpoint inhibitors (ICIs). Furthermore, a prognostic model based on m1A-related genes was constructed and validated across multiple datasets. RESULTS: The results demonstrated that the four CRC molecular subtypes exhibited significant differences in survival outcomes and clinical characteristics. Stromal cells showed higher m1A scores, suggesting a regulatory role in the TME. There was a positive correlation between m1A-related gene expression and immune checkpoint genes. Moreover, the constructed prognostic model showed robust predictive performance and outperformed other recently published models. DISCUSSION: The findings suggest that m1A-related genes are not only valuable biomarkers for CRC prognosis but also have significant implications for the immune landscape and could serve as potential targets for therapeutic intervention, particularly in the context of immunotherapy. For instance, SLC12A2 was found to enhance invasion, proliferation, and migration of colorectal cancer cells while inhibiting apoptosis. Further studies are needed to understand the functional roles of m1A modifications across different cell types within the TME.