Abstract 2190: Transcriptomic signatures of obesity-driven colon cancer: A multi-species transdisciplinary approach

Abstract

Abstract Obesity is a highly prevalent metabolic disease and an established risk factor for colon cancer. However, critical biological pathways underpinning this relationship remain undetermined. Herein, we integrate human and murine colon tumor transcriptomic data to identify obesity-responsive pathways conserved between species. Human colon cancer samples (n=155) were collected as part of the clinical ColoCare Study, a prospective cohort study of patients newly diagnosed with colorectal cancer. Enrolled patients were diagnosed with stage I-III colon cancer and treatment naïve at time of tissue collection. Patients were classified as having a BMI <30 kg/m2 (i.e., non-obese) or BMI ≥30 kg/m2 (i.e., obese) at baseline. Age, sex, tumor stage, and collection site were adjusted for in all analyses. Patients were 63.3 ± 13.7 years, 51% male, and 34% had a BMI ≥30. C57BL/6J adult female mice were fed a low-fat control diet (Research Diets D12450J, n= 8) or a high-fat diet-induced obesity regimen (D12492, n=7) for ~23 weeks, at which time 700 Apc-null;KrasG12D/+;Trp53-null;Smad4-null;tdTomato (i.e., AKPS) organoids were orthotopically transplanted into the colonic submucosa via colonoscopy-guided injection. Obese mice weighed more than controls (controls 26.4 ± 3.7 g vs. DIO 45.9 ± 9.7 g) and had more mesenteric fat surrounding the colon (controls 201 ± 143 mg vs. 1176 ± 589 mg). Tumors were excised after 4 weeks of growth. RNA was extracted from bulk human tumor samples and from the epithelial-enriched EpCAM+ cell population in murine tumor samples for RNA sequencing. Genes differentially expressed between the reference and obese groups were identified using DESeq2. For gene set enrichment analysis, genes were preranked by log2 fold change for both human and mouse datasets. There were 76 and 132 differentially expressed genes (DEGs; FDRq<0.10) identified in the human and mouse datasets, respectively. None of these 208 DEGs were common to both datasets. Thirty-five Hallmark gene sets in the human dataset and 25 Hallmark gene sets in the mouse dataset were significantly enriched in the obese group; 18 of these gene sets were common to both mouse and human datasets. These overlapping pathways were largely related to inflammation (e.g., inflammatory response, IL6 JAK STAT3 signaling, TNFɑ signaling) and metabolism (e.g., cholesterol homeostasis, glycolysis, and fatty acid metabolism). Additional analysis revealed 204 genes common between the leading edges of both transcriptomic datasets. These transdisciplinary data suggest that obesity induces a set of metabolic and inflammatory processes and/or pathways in colon tumors which is conserved between mice and humans. Future studies will mechanistically evaluate the role of candidate genes and pathways in obesity-associated colorectal carcinogenesis in our AKPS orthotopic organoid transplant mouse model. Citation Format: Elaine M. Glenny, Tengda Lin, Victoria M. Bandera, Biljana Gigic, Aik Choon Tan, Babak Mirminachi, Saratchandra S. Khumukcham, Alessandro Carpanese, Christy A. Warby, Olena Aksonova, Caroline Himbert, Jennifer Ose, Chris Stubben, David Nix, Kenneth Boucher, Peter Schirmacher, Ildiko Strehli, Jolanta Jedrzkiewicz, Alexander Brobeil, Martin A. Schneider, Christoph Kahlert, Erin M. Siegel, Adetunji T. Toriola, David Shibata, Christopher I. Li, Jane C. Figueiredo, Jatin Roper, Cornelia M. Ulrich, Stephen D. Hursting. Transcriptomic signatures of obesity-driven colon cancer: A multi-species transdisciplinary approach [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2190.