Abstract Colorectal cancer (CRC) is the third most common cancer and the third leading cause of cancer-related deaths in the U.S. Obesity - a worldwide public health concern - is a known risk factor for cancer including CRC. However, the mechanisms underlying the link between CRC and obesity have yet to be elucidated in part because of the molecular heterogeneity of CRC. In 2015, an international consortium classified CRC into four consensus molecular subtypes (CMSs). However, the molecular links between the CMSs and obesity have not been investigated. Here, we preformed a transcriptomic analysis to elucidate links between the distinct CMSs and obesity to better understand the pathophysiology of CRC. RNA-seq data and associated tumor and patient characteristics including body weight and height data for 231 patients were obtained from The Cancer Genomic Atlas - Colon Adenocarcinoma (TCGA-COAD) database. Tumor samples were classified into the four CMSs with the CMScaller R package; Body mass index (BMI) was calculated and categorized as normal, overweight, and obese. No significant differences were observed in sex, age, ethnicity, or anatomical location of patient tumors and stage of colon cancer across BMI categories. In contrast, a significant difference in CMS categorization between BMI categories was observed; more CMS3 tumors were observed in the obese than the normal BMI category. Gene Set Enrichment Analysis (GSEA) was conducted to elucidate CRC enriched pathways with a false discovery rate of P < 0.05 in the obese compared to normal BMI category for each CMS. Inflammation-related Hallmark get sets were enriched in CMS1 (8 gene sets), CMS2 (2 set sets), and CMS4 (8 gene sets), but not in CMS3 tumors. Obesity-linked CMS specific Hallmark gene set enrichment not related to inflammation was observed in CMS1 (KRAS signaling up) and CMS4 (E2F targets, Myc targets v1, DNA repair, MTORC signaling, G2M checkpoint) but not in CMS2 and CMS3. Differentially expressed genes (DEGs) between the obese and normal BMI patient tumor tissues for each CMS were identified with the DESeq2 R package and used to construct a protein-protein interaction network and determine 10 hub nodes (genes) with the STRING and CytoHubba apps, respectively, in Cytoscape. Nine of the 10 obesity-linked hub genes differed by CMS; the top ranked obesity-linked hub genes were Bassoon Presynaptic Cytomatrix Protein (BSN), DNA (cytosine-5)-methyltransferase 3A (DNMT3A), Melanoma-Associated Antigen 6 (MAGEA6) and UDP Glucuronosyltransferase Family 1 Member A6 (UGT1A6), for CMS1-4, respectively. Kaplan-Meier analysis for overall and relapse free survival was performed using the top 20 upregulated obesity-linked DEGs in each CMS. Higher expression of the DEGs in the CRC cohort GSE17536 was associated with significantly worse overall survival for CMS4 and relapse free survival for CMS2 and CMS3. Taken together, our data suggests that obesity is linked to CRC through CMS-dependent pathways. Citation Format: Michael W. Greene, Peter Abraham, Elizabeth A. Lipke, Peyton C. Kuhlers. Obesity and the consensus molecular subtypes of colorectal cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 2181.
Read full abstract