Epigenetic signatures of asthma: a comprehensive study of DNA methylation and clinical markers

Abstract

Background: Asthma, a complex respiratory disease, presents with inflammatory symptoms in the lungs, blood, and other tissues. We investigated the relationship between DNA methylation and 35 clinical markers of asthma.

Methods: The Illumina Infinium EPIC v1 methylation array was used to evaluate 742,442 CpGs in whole blood from 319 participants from 94 families. They were part of the Netherlands Twin Register from families with at least one member suffering from severe asthma. Repeat blood samples were taken after 10 years from 182 individuals. Principal component analysis on the clinical asthma markers yielded ten principal components (PCs) that explained 92.8% of the total variance. We performed epigenome-wide association studies (EWAS) for each of the ten PCs correcting for familial structure and other covariates.

Results: 221 unique CpGs reached genome-wide significance at timepoint 1 after Bonferroni correction. PC7, which correlated with loadings of eosinophil counts and immunoglobulin levels, accounted for the majority of associations (204). Enrichment analysis via the EWAS Atlas identified 190 of these CpGs to be previously identified in EWASs of asthma and asthma-related traits. Proximity assessment to previously identified SNPs associated with asthma identified 17 unique SNPs within 1 MB of two of the 221 CpGs. EWAS in 182 individuals with epigenetic data at a second timepoint identified 49 significant CpGs. EWAS Atlas enrichment analysis indicated that 4 of the 49 were previously associated with asthma or asthma-related traits. Comparing the estimates of all the significant associations identified across the two time points yielded a correlation of 0.81.

Conclusion: We identified 270 unique CpGs that were associated with PC scores generated from 35 clinical markers of asthma, either cross-sectionally or 10 years later. A strong correlation was present between effect sizes at the 2 timepoints. Most associations were identified for PC7, which captured blood eosinophil counts and immunoglobulin levels and many of these CpGs have previous associations in earlier studies of asthma and asthma-related traits. The results point to a robust DNA methylation profile as a new, stable biomarker for asthma.

Keywords: Asthma; Epigenetics; Epigenome-wide association; Methylation; Microarrays.

Fig. 1

Heatmap showing correlations between the clinical asthma markers and principal components generated from the asthma marker data. An explanation of the variables is provided in Table 1

Fig. 2

EWAS atlas enrichment analysis of the 221 unique CpGs identified at TP1 across the 10 EWASs. The plot shows the number of previously identified Differentially Methylated CpGs (DMC) that are present within the CpGs identified in our analysis. The EWAS Atlas job can be found via this job ID: ea5abe0e33f28b27ab3948e2ccd4044c

Fig. 3

EWAS atlas enrichment analysis of the 49 unique CpGs identified at TP2 across the 10 EWASs. The plot shows the number of previously identified Differentially Methylated CpGs (DMC) that are present within the CpGs identified in our analysis. The EWAS Atlas job can be found via this job ID: a3e03507f39d7af4c7a82d0b0725308e