Novel mtDNA methylation-associated prognostic signatures in colorectal cancer.

Background: Given the high morbidity and mortality of colorectal cancer (CRC), as well as its challenging treatment, the present study was conducted to identify reliable prognostic signatures for predicting clinical outcomes. Methods: This study identified key genes through differential expression analysis and cluster analysis on CRC transcriptomic data from public databases and mitochondrial DNA methylation-related genes (MDM-RGs) from previous literature. Cox analysis and machine learning were used to define prognostic genes and build a prognostic model, followed by the establishment of a nomogram to assess the risk score as a prognostic indicator. Furthermore, the impact of prognostic genes was explored by employing immune infiltration, antitumor immunoassay, medication sensitization assay, and prognostic gene dependency analysis. In addition, the expression of MDM-RGs in CRC was assessed by qPCR and Western blotting. Results: Consequently, this study identified three prognostic MDM-RGs (TINAG, EPHB2, and FCN3), and a risk model was constructed based on these three prognostic genes. A nomogram was created to predict CRC prognosis clinically. Significant differences in risk scores were observed across subgroups, particularly in stage and T stage groups. This study observed disparities in immune cells involving lymphocytes and memory cells, with the identification of 92 pharmaceuticals showing intergroup significant differences. Sorafenib, Salubrinal, and Roscovitine were positively correlated with the risk score, whereas WO2009093972 was negatively correlated. Additionally, this study identified several target genes such as FBXO25 with TINAG, CCDC28A with EPHB2, and SH2D6 with FCN3, with subsequent validation achieved through qPCR and western blotting. Conclusions: In conclusion, this study identifies three prognostic genes, providing new insights into CRC pathogenesis and potential therapeutic strategies.