Abstract
BACKGROUND: Cancer is extensively studied, yet its origins and progression remain unclear. A key question is why tumors of the same type vary in aggressiveness and treatment response. In colorectal cancer (CRC), the third most common cancer, this variability led to the identification of molecular subtypes (CMS). However, the tumor microenvironment remains poorly understood and may be crucial for understanding carcinogenesis and drug resistance. A promising approach is analyzing cell interactions through ligand-receptor expression. This study used bioinformatics to examine CRC in different anatomical locations, identify diagnostic and prognostic biomarkers, and propose targeted drugs. METHODS: Expression data were obtained from the TCGA-COAD database. All samples were filtered based on the tumor's region of origin and purity. RNA-seq expression analysis was then conducted to assess molecular differences according to tumor location and purity, identifying region-specific ligands and receptors using the secretome list as a reference. Once these differences were identified, an interactome was constructed to depict cell interactions within the tumor microenvironment. The most relevant genes were then evaluated for their prognostic potential through survival analysis, and their susceptibility to pharmacological modulation was assessed to identify potential new drug candidates for CRC treatment. RESULTS: The integration of secretome data and the construction of the interactome proved to be a valuable approach for detecting novel biomarkers specific to right- and left-sided CRC. Through this approach, FGFR4, FLT1, and WNT5A were identified as key biomarkers involved in tumor carcinogenesis, modulating distinct processes in each region, such as fibroblast recruitment and cell division. Based on these biomarkers, Dovitinib and Nintedanib were predicted as potential therapeutic agents, as they target multiple identified markers. CONCLUSION: This study highlights FGFR4, FLT1, and WNT5A as key diagnostic and therapeutic biomarkers for CRC, with their relevance varying based on the tumor's site of origin. Leveraging these findings, we propose Dovitinib and Nintedanib as promising targeted therapies for CRC. These insights can enhance current treatment strategies and pave the way for future in vivo and in vitro studies, driving progress in CRC research and therapy.