A Multiomic Approach Integrating Genomic and Metabolomic Data Highlights Colorectal Cancer Pathways

Numerous genetic variants have been identified by genome-wide association studies as being associated with colorectal cancer (CRC) risk. Metabolome-wide association analysis was performed for 187 CRC-associated genetic variants using genomic data and untargeted ¹H nuclear magnetic resonance urine metabolomics from 1951 Airwave Health Monitoring Study participants. We identified statistically significant associations between seven CRC single-nucleotide polymorphisms (SNPs) and urinary metabolites. This included SNPs within or close to <i>RHPN2</i> with sucrose (<i>P</i> = 1.2 × 10^-7), <i>SLC6A18</i> with amino acids (<i>P</i> = 6.9 × 10^-5 with tyrosine, <i>P</i> = 9.9 × 10^-5 with leucine), and <i>MAP2K5</i> and <i>BMP2</i> with gut microbial metabolites (<i>P</i> = 1.6 × 10^-4 and <i>P</i> = 4.4 × 10^-4). The most significant correlation was followed by functional experiments in Caco-2 colon cancer cells. CRISPR-mediated knockout of a 48-nt <i>RHPN2</i> intronic region containing rs10411210 in colon cancer cells compromised cell growth. RNA sequencing was performed in the two sets of clones (3 edited and 3 unedited) followed by pathway enrichment, and gene ontology analysis depicted extensive deregulation of genes (448 up- and 195 downregulated) involved in cell division and several metabolic processes. Overall, these findings demonstrate that integrating genetic and metabolomic data highlights the importance of the <i>RHPN2</i> intronic locus in CRC potentially through metabolic processes affecting excretion of dietary and other metabolites.