Autism spectrum disorders (ASD) are a common phenotype with a complex etiology. While a rare single gene or genomic interval may be sufficient to lead to ASD, most patients likely owe their disease to a combination of both genetic and environmental variation. One area that bridges genetic and environmental influences is epigenetic variation. Seeking evidence for epigenetic variation specific to ASD, we screened 49 ASD males and their unaffected fathers for methylation status of 1,505 CpG dinucleotides corresponding to 807 genes. Using 33 pairs as a training set and 16 pairs as a validation set, we statistically identified 116 differentially methylated loci (DML) that gave the highest predictive power to classify ASD boys from their fathers. Using these DML, we screened, in a blinded fashion, DNA isolated from whole blood in an additional 29 ASD son/unaffected father pairs and correctly classified ~80% of the ASD affected individuals. We validated a subset of these loci using bisulfite sequencing and ruled out age-dependent effects. To follow-up on this extraordinary observation, we propose here to conduct a comprehensive genome-wide methylation analysis that will interrogate 27,578 CpG dinucleotides associated with more than 14,000 genes in DNA isolated from whole blood in 300 ASD Simons Simplex 4-person families (father, mother and discordant male sib pair). Using this data, we will then construct a smaller custom set of DML and screen an additional 900 ASD probands and their families. This analysis will directly test our provocative preliminary data and potentially identify DML that could compromise a peripheral biomarker assay for ASD.
Autism is a common disorder whose causes are poorly understood. Both genetic and environmental influences are thought to act together causing autism. Epigenetics is an area that bridges genetic and environmental influencing and commonly is studied by examining dynamic methylation changes to the DNA. We have screened a total of 78 autistic boys and their fathers by this method in 807 genes and find 116 methylation changes that together can correctly identify nearly 80% of the affected boys from their fathers. We propose to confirm these exciting data in a much larger set of genes in 1,200 autistic boys and their families, including unaffected brothers. These data not only could provide new insight into the causes of autism but could also result in a screening test for autism. )
