Identification of DNA Methylation Alterations in Children with Autism Spectrum Disorders Conceived Using Assisted Reproduction

Background:  The use of fertility treatments has not been associated with significant risks to child health. Subfertility and fertility treatments (FT) have been associated with an increased risk of rare imprinting disorders such as Beckwith-Wiedemann and Angelman syndromes. The increased risk for these syndromes has been attributed to aberrant DNA methylation at imprinting centers. Increased rates of methylation alterations at imprinting centers have also been reported in animal models of FT. Recently, an increased risk of Autism Spectrum Disorders (ASD) in children conceived by subfertile couples and/or FT has been reported in some studies but not others. 

Objectives:  To assess whether aberrant methylation marks could be, at least in part, responsible for the ASD phenotype in patients conceived using fertility treatments (FT). 

Methods:  DNA from blood samples of controls (N=12), and ASD patients conceived with FT (N=12) or without FT (N=12) were run on the Illumina HumanMethylation27 microarray. To assess global DNA methylation the Mann-Whitney U test was used to compare the mean methylation of all probes between groups and also compared after dividing the probes into bins of 10% methylation intervals. For targeted analysis the Mann-Whitney U test with FDR was used to compare the average beta methylation values of ASD-FT samples against controls. A stringent novel individual analysis was used to generate a list of genes with significantly increased variance. At least one sample in each ASD group was required to have a 17% DNA methylation difference greater or less than the average of controls. 

Results:  The global analysis identified a statistically significant reduction in methylation was observed in the ASD-FT group compared to either controls (p=0.006) or the ASD group (p=0.015). A statistically significant increase in number of probes was observed in the 0-0.1 bin within the ASD-FT group compared to controls (p=0.001) and the ASD group (p=0.007). A statistically significant decrease in the number of probes was observed in the 0.1-0.2 bin in the ASD-FT group compared to controls (p=0.005) and to ASD group (p=0.008). Statistical analysis for targeted DNA methylation variants did not reveal any significant changes. In the targeted analysis, 17 and 84 genes were respectively identified in the ASD and ASD-FT groups, suggesting higher variability in DNA methylation in ASD-FT group. The gene list was compared to a list of known imprinted genes, and a loss of DNA methylation was observed at two CpG sites in a CpG island associated with the imprinted gene DIRAS3 in ASD-FT, but not in the ASD group. 

Conclusions:  We found a global loss of DNA methylation at CpG sites with reduced levels of methylation in the ASD-FT group. Our data also suggest higher levels of methylation variability in ASD-FT compared to the ASD group conceived without fertility treatments. These data provide an opportunity to study the role of DNA methylation dysregulation in ASD susceptibility.