Poster #76: Ivy Tran

Integrating Local African Ancestry and DNA Methylation to Investigate Epigenetic Age Deceleration in Colorectal Cancer

Ivy Tran BS¹, Yan Zhou PhD², Manju Ambelil MD², Rajeswari Nagarathinam MD², Olorunfemi Ayeotan BS¹, Juie Nahushkumar Rana PhD¹, Elaine Vaughan-Williams BS¹, Olivia Cole BS1, Zachary Reichenbach MD PhD3, Jayashri Ghosh PhD¹ 1Fels Cancer Institute for Personalized Medicine, Lewis Katz School of Medicine, Philadelphia, PA 19140, USA ²Fox Chase Cancer Center, Philadelphia, PA 19111, USA 3Department of Gastroenterology, Temple University Hospital, Philadelphia, PA 19140, USA

Colorectal cancer (CRC) is the fourth most common U.S. cancer and second leading cause of cancer death. Non-Hispanic African Americans (AA) experience approximately 20% higher CRC incidence and 40% higher mortality than non-Hispanic European Americans (EA), a disparity not fully explained by socioeconomic factors, suggesting tumor-intrinsic mechanisms. A potential mechanism is altered epigenetic aging. EAA is the difference between DNA methylation age and chronological age; positive values indicate age acceleration, whereas negative values indicate age deceleration. In other cancers, epigenetic age deceleration has been associated with poor prognosis, stemness, and increased proliferation. We analyzed 69 tumors using the Illumina EPIC v1 array and 58 matched genotypes using the Illumina Global Diversity Array. Ancestry was inferred using YRI, CEU, and CHB references. AA tumors exhibited significantly more negative EAA than EA tumors. This difference persisted after residualizing tumor age on matched normal age, and global African ancestry correlated with more negative Horvath age residuals, suggesting that ancestry-linked epigenetic variation contributes to a tumor-specific aging phenotype. Because single-nucleotide polymorphisms (SNPs) can influence DNA methylation by acting as methylation quantitative trait loci (meQTLs), we developed a novel integrative framework to localize genomic regions underlying this association. CpG methylation was modeled against residualized tumor EAA to define CpG-EAA peaks, while local African ancestry was modeled against residualized EAA across genomic windows to define local ancestry-EAA peaks. Intersecting these independently derived interval sets using a one-megabase proximity threshold identified convergent loci, including a chromosome 9 region containing NTRK2. NTRK2 encodes TrkB, the receptor for brain-derived neurotrophic factor. TrkB signaling is associated with CRC invasion, metastasis, treatment resistance, and poor prognosis and is pharmacologically targetable. Thus, this integrative method informs future studies aimed at functionally validating NTRK2 as a candidate mechanism through which ancestry-linked genomic and epigenetic variation may contribute to CRC disparities.