Mendelian Randomization-Based Discovery of Novel Protein Biomarkers and Drug Targets in Colorectal Cancer: Validation Through Prognostic Modeling, Single-Cell Analysis, and In Vitro Cell Experiments.

Abstract

RNA modification plays a crucial role in biological processes. This study uses Mendelian randomization to investigate the relationship between RNAm-SNPs and CRC to identify potential protein markers and therapeutic targets. We analyzed RNAm-SNPs and CRC using RMVar and GWAS datasets. eQTL and pQTL analyses were performed to assess associations with gene expression and protein levels. Two-sample MR and summary-data-based MR identified candidate proteins. Single-cell expression analysis, prognostic model construction, protein-protein interaction studies, and druggability evaluations were conducted to identify cell-type enrichment and prioritize targets. Cell experiments further validated key genes in CRC. Six proteins (STX10, TBCA, GRIA4, PPT1, Sperm-associated antigen 2, and IL-21) were linked to CRC risk. These genes are primarily active in fibroblasts and epithelial cells in colon tumor tissue. Three proteins (GRIA4, IL-21, PPT1) are established targets for psychiatric and tumor disorders and may serve as therapeutic targets for CRC. Predictive modeling showed potential for clinical decision-making, and cell experiments confirmed TBCA as protective and GRIA4 as a risk factor in CRC. This study identified protein biomarkers associated with CRC risk, uncovering potential screening biomarkers and therapeutic targets, providing insight into the disease's molecular mechanisms.