Age of host impact on epigenetics in tumors derived from early or late onset Hispanic colorectal cancer patients

Introduction Over the past decade, the incidence of early-onset (EO) colorectal cancer (CRC) in adults under 50 has shown a significant increase compared to late-onset (LO) CRC rates, especially in the Hispanic/Latino (H&L) population. EO CRC exhibits unique molecular features compared to LO CRC, associating this distinct disease with a more aggressive presentation. Investigations on the effect of age within both the tumor and host environments could provide insight into key differences between EO and LO CRCs. This study, as part of Cancer Moonshot Program, explores how age of the host influences epigenetic changes in CRC tumors. Methods CRC patient derived tumor samples from an EO 36 year old (young patient, YP) and a LO 66 year old (old patient, OP) H&L male were implanted into both young (2 months, YM) and old male NSG mice (13-14 months, OM) to establish four PDX models: YP in YM, OP in YM, YP in OM, and OP in OM. Tumor samples were collected at endpoint (15mm length or width). Tumors were processed for epigenomic analysis using the Illumina EPIC Array Epigenetics Kit. Methylation data was used to identify differentially methylated regions (Δbeta>0.1) in YP in OM compared to YP in YM and OP in YM compared to OP in OM. Identified regions were mapped to the nearest gene using the human genome (GRCh38 Genome Reference Consortium Human Reference 38) from UCSC. The differentially methylated genes (Δbeta>0.2) were input into Ingenuity Pathway Analysis to identify relevant biological pathways. Results There were 8916 differentially methylated regions identified in YP in OM when compared to YP in YM. Of these regions, 5529 showed hypermethylation and 3387 showed hypomethylation. Regions were mapped to 5886 unique gene symbols, of which 1061 were hypermethylated and 999 were hypomethylated. YP in OM showed predicted activation of key genes, including TNF, WNT1, MYB, ERBB3, GLI3, EGR2, and EGR3, involved in the development and progression of gastrointestinal cancers. There were 28032 differentially methylated regions (Δbeta>0.1) identified in OP in YM when compared to OP in OM. If these regions, 17590 showed hypermethylation and 10442 showed hypomethylation. Regions were mapped to 11888 unique gene symbols, of which 4273 were hypermethylated and 2936 were hypomethylated. OP in YM showed predicted inhibition of key genes, including RELA, TNF, EGFR, YAP1, WNT3A, FGF2, IFNG, SP1, IGF1, ERBB2 and TEAD1, involved in cell transformation and tumor cell metastasis. Conclusions Our current findings show differences in epigenetic DNA methylation patterns in EO and LO CRC PDX models. Differentially methylated region analysis in both EO and LO CRC suggests that an older host age may correlate with activation of cancer related genes while younger host age may inhibit similar pathways, regardless of tumor age. Further studies are warranted to understand the epigenetics modifications that influence the observed differences.