Divergent transcriptomic signatures and microenvironmental markers of <i>KRAS</i> and <i>BRAF</i> mutations in colorectal cancer: A multi-omics integrative analysis.

Abstract

3668 Background: KRAS and BRAF mutations are critical drivers in colorectal cancer (CRC), acting as key nodes in the MAPK signaling pathway. While their genomic prevalence is well-documented, the specific transcriptomic programs they trigger and how these programs differ between KRAS and BRAF variants remain a subject of intense translational interest. This study utilizes a multi-omics approach to define mutation-specific signatures that could inform personalized therapeutic strategies. Methods: We performed an integrated analysis of SNV/Indel data and gene expression profiles from 192 CRC patients. Patients were stratified into BRAF -mutant (n=20), KRAS -mutant (n=102), and MAPK-wild-type (WT, n=64) groups. Frequency was calculated as the percentage of unique samples mutated per gene. Differential expression analysis (DEA) was utilized to identify significant (p &lt; 0.05) gene expression changes specific to each driver mutation. Results: The most frequently mutated genes were TP53 (79.2%), APC (56.8%), and KRAS (53.1%). BRAF mutations were identified in 10.4% of the cohort. KRAS -mutant tumors exhibited a specific upregulation of SPP1 (osteopontin) and TGFBI . These markers are associated with extracellular matrix remodeling and TGF-β signaling, suggesting that KRAS mutations may actively contribute to an immune-excluded tumor phenotype. BRAF -mutant CRC was characterized by a distinct signature involving ABI3BP upregulation and a profound downregulation of CTNNBL1 (p &lt; 0.00001). Additionally, a cluster of small nucleolar RNAs ( SNORA2B , SNORA9 ) was significantly elevated, pointing toward altered ribosome biogenesis or non-coding RNA regulation unique to the BRAF -mutant subset. While both mutations activate the MAPK pathway, the downstream transcriptomic output is significantly divergent, with KRAS favoring myeloid-recruiting signals ( SPP1 ) and BRAF favoring structural and RNA-processing alterations. Conclusions: Our multi-omics integration reveals that KRAS and BRAF mutations in CRC are not transcriptomically redundant. The discovery of SPP1 as a KRAS -associated marker and CTNNBL1 as a BRAF -linked marker provides novel avenues for targeted therapy and patient stratification. Specifically, KRAS -mutant patients might benefit from therapies targeting the SPP1 -TGF-β axis in combination with standard care.