Developmental methylomics of autism spectrum disorder

Autism spectrum disorder (ASD) is a neurodevelopmental disorder that affects ~1% of the population. Progress has been made in elucidating the genetics of ASD through large-scale genome-wide association studies (GWAS) and whole-exome sequencing (WES) studies that have identified several loci associated with ASD. However, a substantial fraction of ASD status cannot be explained by genetic sequence variation. There are multiple reasons to expect that DNA methylation (DNAm) may account for part of this unexplained variation. First, part of the ASD-related genes identified via DNA sequence variation include genes involved in chromatin modification and DNAm. Second, ASD likely originates during prenatal development, a period of dynamically regulated changes in DNAm in the brain. As this remodeling may result in epimutations that can dysregulate brain function, disruption of the DNAm regulation in utero represents a plausible ASD risk mechanism. Third, ASD is associated with several neonatal and other environmental risk factors. Because DNAm can be modified by environmental factors, it may mediate the effect of these risk factors on ASD. The overall aims of this project is to enhance our understanding of DNAm in ASD etiology, and use DNAm marks at for early detection of individuals at risk for ASD. For this purpose we will generate methylome-wide data using samples from ASD cases age 18-25 years and matched controls from an existing Swedish case- control study called Population-based Autism Genetics and Environment Study. In addition, we will use stored neonatal blood samples to generate a second methylation profile for these same individuals at birth. Thus, we will have methylome-wide data from blood for two time-points from all participants, accompanied by longitudinal phenotype information spanning birth to current date obtained from the Swedish registers. We will use a sequencing-based approach to assay the DNAm status of nearly all 28 million common CpG sites in the human genome and will perform a battery of novel statistical analyses including methylome-wide association studies (MWAS) of whole blood and individual cell-types in blood; analyses integrating DNAm information with neonatal risk factors and already existing GWAS and WES data; and studies exploring the role of DNAm in the ASD sex-bias. Significant findings will be replicated in four existing and independent blood sample collections, and studied in the newly generated methylation/expression data from ASD brain samples. Finally, we propose to use neonatal DNAm markers to create multi-marker methylation risk scores (MRSs) for predicting ASD risk.