Identification of circadian rhythm-related genes in colorectal cancer by integrating bioinformatics and multi-omics mendelian randomization

Circadian rhythm disorders are epidemiologically linked to colorectal cancer (CRC), but causal relationships and molecular mechanisms remain unclear. We applied a multi-omics Mendelian Randomization (SMR) framework to assess the causal effects of circadian rhythm-related genes on CRC. Instrumental variables were derived from blood-based QTLs for methylation (mQTL), expression (eQTL), and protein levels (pQTL), with validation using tissue-level eQTLs. Summary data for CRC came from the FinnGen R12 GWAS. Colocalization analyses confirmed shared genetic variants. Candidate genes were further evaluated for expression and prognostic value using TCGA and GEPIA2. Through blood-level SMR analysis, we identified 142 methylation loci, 11 genes, and 2 proteins associated with CRC in discovery cohort. Among them, 42 methylation loci, 3 genes, and 2 proteins were colocalized with CRC incidence. Our SMR analysis highlighted GRHPR as a key gene, which is negatively correlated with CRC risk at multiple molecular levels (m/e/pQTL), supported by robust colocalization evidence. Meanwhile, eQTL results showed that high level expression of UVSSA was positively associated with CRC risk. QTL results at the tissue level (Colon_Sigmoid and Colon_Transverse) support a causal relationship between the genes GRHPR and UVSSA and CRC. Crucially, survival analyses revealed that genetically predicted effects of NAF1 and ZNF365 on CRC risk were highly consistent with their prognostic values in clinical outcomes (OS and RFS). However, transcriptome expression profiling of the TCGA database CRC cohort revealed no significant differential expression of GRHPR between CRC and normal tissues, though its expression positively correlated with tumor purity. This integrated multi-omics MR analysis identified 12 circadian rhythm-related genes with potential causal relationships to CRC, highlighting GRHPR as a risk-reducing factor, and NAF1 and ZNF365 as potential dual-purpose biomarkers for both risk stratification and prognosis. The study provides new insights into regulatory mechanisms and signaling pathways, offering novel clues for future mechanistic research and targeted therapies.