An epigenome-wide association study in the case-control study to explore early development identifies differential...

Quantitative measures of autism spectrum disorder (ASD)-related traits can provide insight into trait presentation across the population. Previous studies have identified epigenomic variation associated with ASD diagnosis, but few have evaluated quantitative traits. We sought to identify DNA methylation patterns in child blood associated with Social Responsiveness Scale score, Second Edition (SRS).

We conducted an epigenome-wide association study of SRS in child blood at approximately age 5 in the Study to Explore Early Development, a case-control study of ASD in the United States. We measured DNA methylation using the Illumina 450K array with 857 samples in our analysis after quality control. We performed regression of the M-value to identify single sites or differentially methylated regions (DMRs) associated with SRS scores, adjusting for sources of biological and technical variation. We examined methylation quantitative trait loci and conducted gene-ontology-term pathway analyses for regions of interest.

We identified a region about 3.5 kb upstream of ZFP57 on chromosome 6 as differentially methylated (family-wise error rate [fwer] < 0.1) by continuous SRS T-score in the full sample (N = 857; fwer = 0.074) and among ASD cases only (N = 390; fwer = 0.021). ZFP57 encodes a transcription factor involved in imprinting regulation and maintenance, and this DMR has been previously associated with ASD in brain and buccal samples.

Blood DNA methylation near ZFP57 was associated (fwer < 0.1) with SRS in the full population sample and appears to be largely driven by trait heterogeneity within the autism case group. Our results indicate DNA methylation associations with ASD quantitative traits are observable in a population and provide insights into specific biologic changes related to autism trait heterogeneity.