Identifying biological pathways implicated in defined subgroups of phenotypic expression for Autism Spectrum Disorders and evaluating small molecule effects on expression of ASMT

Abstract

Autism Spectrum Disorder is a neurodevelopmental condition with evidence for genetic susceptibility. However, effect sizes for implicated loci are small, and current evidence does not explain the heritability. Phenotypic heterogeneity could be complicating genetic factor identification. We used multiple sources of behavioral and physiological data to identify ASD subgroups. Principal Components Analysis refined these data from an ASD dataset into 15 components best explaining phenotypic variance. Clustering identified two distinct groups, primarily based on phenotype severity. Using an independent dataset, we identified 15 data components and confirmed the severity-based dichotomy. There is significant familial clustering within groups, suggesting the clusters recapitulate genetic etiology. We performed separate family-based analyses of genetic data based on subgroup. Pathway analysis was performed. Five biological pathways uniquely associate with the ‘less severe’ subgroup. Ten genes potentially drive these associations. Five different pathways uniquely associate with the ‘more severe’ subgroup. 24 genes potentially drive associations with this subgroup. There is minimal overlap when comparing genes associated with different subgroups. We validated results in an independent dataset and see unique biological mechanisms associate with the ‘more severe’ and ‘less severe’ subgroups. Our results suggest meaningful subgroup definitions can help clarify the genetics of ASD. Uncovering pathways associated with subgroups further elucidated potential genes involved in trait expression. To progress toward understanding how these findings could be useful for treatment, functional characterization was necessary. Ample evidence indicates many drugs have altered efficacy and side effects in relation to genetic background. For many compounds used to treat ASD comorbid symptoms, individuals exhibit sleep problems. We evaluated the enzyme catalyzing the final reaction in melatonin synthesis, Acetylserotonin O-methyltransferase. We screened cell lines generated from patient DNA for differential expression effects against compounds presently used to treat symptoms. We replicated previous findings indicating homozygous presence of ASD risk alleles at promoter SNPs results in decreased gene expression. We also observe previously unreported expression effects attributable to heterozygosity at promoter SNPs, and a SNP in the 5'-UTR. Results show no significant changes in gene expression upon exposure to small molecule compounds for the non-risk haplotype.