Identification and Validation of Mitophagy-Related Biomarkers in Colorectal Cancer: An Integrated Analysis of Single-Cell Transcriptome and Mendelian Randomization.

Mitophagy is essential for cancer formation and invasion, but its role in colorectal cancer (CRC) remains unclear. We obtained sequencing data and mitophagy-related genes (MP-RGs) from public databases. Differential expression analysis and weighted gene coexpression network analysis (WGCNA) identified mitophagy-related differentially expressed genes (DE-MPGs). Mendelian randomization (MR) analysis identified candidate genes with genetically supported causal relevance to CRC. Biomarkers were identified using machine learning, receiver operating characteristic (ROC) analysis and expression studies. Single-cell RNA sequencing (scRNA-seq) analyzed biomarker expression profiles in various CRC cell types. Quantitative PCR (qPCR) validated biomarker expression in clinical CRC samples. 147 DE-MPGs were identified. MR analysis revealed seven genes with potential causal contributions to CRC susceptibility. Three genes, SGCE (IVW: OR = 1.00041, p = 0.011), ATP8B2 (IVW: OR = 0.99920, p = 0.042), and RANGAP1 (IVW: OR = 0.99861, p = 0.002), were selected as biomarkers. Immune microenvironment and checkpoint differences were observed between CRC and controls. Biomarker expression varied among cell types. qPCR showed decreased SGCE and ATP8B2 and increased RANGAP1 in CRC. SGCE, ATP8B2, and RANGAP1 can serve as mitophagy-related biomarkers with genetically supported causal relevance to CRC, providing new insights for CRC diagnosis and therapy.