Abstract 12528: DNA Methylation Profiles of 3 Major CHIP Subtypes Provide Insight Into Underlying Epigenetic Mechanisms

Clonal hematopoiesis of indeterminate potential (CHIP) is an age-related phenomenon characterized by the presence of clonal populations of hematopoietic-derived cells in whole blood that harbor functional leukemogenic somatic mutations. Despite the well-established associations between aging and CHIP and aging and genome-wide DNA methylation (DNAm), the association between CHIP and changes in DNAm remains to be fully elucidated. To fill this knowledge gap, an epigenome-wide association study of CHIP was performed in Framingham Heart Study participants from the Offspring cohort (n total = 1594; 54% women; mean age 67 years) using whole blood derived DNA to elucidate the DNAm profiles of any CHIP and the three most frequently mutated CHIP driver genes DNMT3A, TET2 , and ASXL1/2 . Linear mixed models were used to test associations between CHIP status as the predictor variable and genome-wide DNAm of CpG sites as the outcome, with adjustment for age, age 2 , sex, smoking, 33 surrogate variables, and familial relatedness. From the analysis, 189 CpGs were found to be associated with any CHIP, and 139, 107, and 4 were associated with DNMT3A , TET2 , and ASXL1/2 CHIP, respectively (Bonferroni corrected P<1х10 -7 ). The CpGs associated with the DNMT3A , TET2 , and ASXL1/2 CHIP were distinct. Genome-wide DNAm directions of effect were similar for DNMT3A and ASXL1/2 CHIP but opposite to TET2 . Gene expression analysis identified CpG-gene transcript pairs for any CHIP and CHIP subtypes. The most significant gene sets were enriched in pathways related to immune response and hematopoietic cell lineage. Two-sample Mendelian randomization analysis was additionally conducted to investigate whether methylation at CHIP-associated CpGs causally influences risk of cardiometabolic traits and all-cause mortality. Evidence of a putative causal role of CHIP-associated CpGs in relation to several traits, including BMI and hypertension, and all-cause mortality was found. Taken together, we identified CpGs associated with CHIP and with the three major CHIP driver genes, elucidated their methylomic signatures, identified CpG-transcript pairs, and provided support for a potential causal role of CHIP-associated CpGs in relation to cardiometabolic traits and all-cause mortality.