Epigenome-wide DNA methylation association study of CHIP provides insight into perturbed gene regulation

Abstract

With age, hematopoietic stem cells can acquire somatic mutations in leukemogenic genes that confer a proliferative advantage in a phenomenon termed CHIP. How these mutations result in increased risk for numerous age-related diseases remains poorly understood. We conduct a multiracial meta-analysis of EWAS of CHIP in the Framingham Heart Study, Jackson Heart Study, Cardiovascular Health Study, and Atherosclerosis Risk in Communities cohorts (N = 8196) to elucidate the molecular mechanisms underlying CHIP and illuminate how these changes influence cardiovascular disease risk. We functionally validate the EWAS findings using human hematopoietic stem cell models of CHIP. We then use expression quantitative trait methylation analysis to identify transcriptomic changes associated with CHIP-associated CpGs. Causal inference analyses reveal 261 CHIP-associated CpGs associated with cardiovascular traits and all-cause mortality (FDR adjusted p-value < 0.05). Taken together, our study reports the epigenetic changes impacted by CHIP and their associations with age-related disease outcomes.

Fig. 1. Overview of Study Design.

This flowchart outlines the sequential steps of the study, from data collection to downstream analyses. CHIP Clonal Hematopoiesis of Indeterminate Potential, WGS Whole Genome Sequencing, WES Whole Exome Sequencing, TOPMed Trans-Omics for Precision Medicine program, eQTM Expression Quantitative Trait Methylation.

Fig. 2. Genome-wide Directions of Effect of Any CHIP and CHIP Subtypes.

Volcano plots showing the effect size (β) and -log₁₀(P-value) from the multiracial meta-analysis of epigenome-wide association studies (EWAS) for (a) any CHIP (Clonal Hematopoiesis of Indeterminate Potential), (b) DNMT3A CHIP, (c) TET2 CHIP, and the EWAS in the Framingham Heart Study (FHS) for (d) ASXL1 CHIP. Genes annotated to the CpG sites are shown. For panels (a–c), the effect size (β) and P-values were derived from fixed-effect meta-analysis of multiple cohorts. For panel (d), because ASXL1 CHIP was only available in the FHS cohort, the effect size (β) and P-values were derived from linear regression models. The color green indicates a significant negative association between CHIP and DNA methylation while purple indicates a positive association between the two variables. Yellow signifies non-significant associations between CHIP and DNA methylation. Two-sided tests were used for all analyses. P-values were adjusted for multiple comparisons using the Benjamini-Hochberg false discovery rate (FDR) method. Significant associations were defined as FDR < 0.05. Exact P-values, standard errors for the β, and 95% confidence intervals for significant results are provided in Supplementary Data 7-10. Source data are provided as a Source Data file.

Fig. 3. Functional Validation in CRISPR/Cas9-edited Hematopoietic Stem Cells Modeling CHIP.

Dot plots of methylation change from −1.0 (no methylation) to 1.0 (complete methylation) seen in engineered primary cell cultures compared to correlation of EWAS results ranging from −0.1 to 0.1. Following an initial CpG filtering using an uncorrected Student’s t-test (p < 0.05), significance was determined with a two-sided binomial test. a DNMT3A-associated CpG sites (n = 855 CpG sites) compared to DNMT3A engineered human stem cells (n = 4). b TET2-associated CpG sites (n = 312) compared to TET2-engineered human stem cells (n = 4). c ASXL1-associated CpG Sites (n = 139) compared to ASXL1-engineered human stem cells (n = 3). Source data are provided as a Source Data file. KO Knockout, mC methylcytosine.