Enrichment of Methylation Quantitative Trait Loci Among Genes Associated with Autism Spectrum Disorder

Background:  Many of the genes and SNPs previously associated with autism are intragenic and/or do not have a clear functional consequence. A potential function may be expression regulation via epigenetics. The availability of genome-wide DNA methylation data (DNAm), a type of epigenetic mark, together with genome-wide genotype data, has allowed for the discovery of methylation quantitative trait loci (meQTLs): single nucleotide polymorphisms (SNPs) that appear to control DNAm levels at particular loci. Examining meQTL data with respect to previously reported ASD-related variants or genes may provide a functional context for their ASD associations.

Objectives:  The purpose of this work was to examine the extent of meQTLs in previously reported ASD-related genes and regions compared to non ASD-related genomic locations.

Methods: In our sample of children with both genome-wide SNP data and genome-wide DNAm data, we performed chromosome-wide SNP-DNAm association testing for each of 401,126 CpG sites to identify SNPs (meQTLs) significantly associated (p < 1E-15) with DNAm levels. This reflects 1.43 x 10¹¹ comparisons, and thus an extreme significance threshold is required. We then tested for enrichment of ASD-related genes, as defined in the Simons Foundation Autism Research Initiative Gene Database (SFARI Gene), among identified meQTLs using a permutation-based test. Samples for meQTL identification were obtained from the Study to Explore Early Development (SEED), a national multi-site autism case-control study of children aged 2-5 years. Genome-wide DNAm was measured using the Illumina HumanMethylation450 BeadChip in peripheral blood-derived DNA among 610 individuals. After applying rigorous quality control metrics, 607 samples remained for downstream methylation analyses. High quality genome-wide genotyping data, measured by the Illumina Omni1 Quad and Affymetrix Axiom arrays, was available for 591 of the same individuals.   

Results:  Across the genome, we identified 1,044,683 meQTL SNPs corresponding to 35,779 CpG sites of DNAm.  Typically SNP clusters were associated with CpG DNAm clusters. We considered 629 ASD-related genes/regions. We are now performing the ASD-meQTL enrichment tests and plan to present our findings at the conference.

Conclusions:  We have identified SNPs that control DNA methylation levels across the genome and will show enrichment of ASD gene associations and meQTLs. This may guide functional insights for previously identified ASD-related variants.