Intratumoral transcriptomic heterogeneity correlates with tumor location in colorectal cancer

Background/Objectives: Colorectal cancer (CRC) is the most prevalent cancer worldwide. Patients affected by CRC exhibit strong differences in prognosis and response to treatment, with 20% are metastatic at diagnosis and non-treatable. Here, we describe the molecular findings of transcriptome signatures and intratumoral heterogeneity. Methods: The cohort consisted of total RNA-seq of colorectal tumors and adjacent non-malignant colon tissue from 136 patients. In silico differential analysis and functional enrichment analysis were performed to identify altered pathways between conditions affecting biological functions. We evaluated CRC transcriptome deconvolution exploiting in silico single cell signatures on six major cell types (epithelial cells, fibroblasts/endothelial cells, myeloid cells, T/natural killer (NK)/NK T lymphocytes, B lymphocytes/plasma cells and mast cells) to detect the fractions and distribution of tumour-infiltrating cells in our CRC samples. Results: Differential analysis revealed 129 over and 281 down expressed genes between tumor compared to normal tissue implicated in 26 up and 17 down regulated pathways (FDR < 0.05). Localization analysis highlighted 11 up-regulated pathways in left-sided versus rectum and 68 down in right-sided versus rectum. Deconvolution permitted the estimation of cells composition, with 10 subpopulations being over-represented and 9 under-represented in tumor compared to normal tissue. Conclusion: Analysis showed that CRC exhibits heterogeneity of pathways depending on tumour location, such as inflammatory status, metabolism, oxidative stress and treatment response. Our results allowed to identify transcriptional signatures and their regulation, as well as to elucidate oncogenic pathways in the tumour microenvironment and intratumoural cellular heterogeneity involving functional diversification and environmental adaptation.