Combining Anecdotal Observation with Scientific Discovery to Improve Translational Genetics Research in Autism Spectrum Disorders

Background: There are hundreds of genes implicated in risk for autism spectrum disorders (ASD); however, the clinical utility of this knowledge to help inform treatment remains limited. Understanding how currently implicated mechanisms contribute to variable expressivity of core symptoms and comorbidities in ASD is crucial to using genetics to inform more effective treatments. Generating hypotheses testing anecdotal evidence that highlights potential connections between comorbidities and ASD offers the opportunity to reveal important clinical and biological relationships. Previous studies have used this approach to successfully decipher pleiotropic genetic effects of ASD candidate genes (e.g., MET variation and expression of gastrointestinal symptoms in ASD). Sleep problems are common in individuals with ASD. Sleep has strong, neuronal-specific effects on the function of molecular, cellular and network mechanisms of synaptic plasticity. Synaptic pruning is also observed to be disrupted in individuals with ASD. Considering sleep is important to synaptic pruning it is possible that short sleep duration exacerbates atypical synaptic pruning, leading to more severe symptoms of ASD. Notably, genes encoding proteins essential for synaptic function are recurrently implicated in ASD.

Objectives: Our goal is to determine if dysfunction in synaptic genes underlies the expression of more severe symptoms observed in relation to short sleep duration in ASD.

Methods: We analyzed medical histories and whole-exome sequence data from 2,714 children in the Simons Simplex Collection. Linear regression was conducted to test if parent-reported sleep duration was associated with ASD-related symptom severity. Symptoms were compared between children in the lower 5^th (extremely short) and upper 95^th (extremely long) percentiles of the sleep duration distribution. Gene-set enrichment analysis for biological processes defined in Gene Ontology was conducted on ASD candidate genes included in the Autism Informatics Portal using the TopGO package in R. Sequence Miner was used to identify variants in genes assigned to the synaptic signaling biological process. Genetic risk scores were calculated, reflecting the number of variants (weighted by the likelihood of being detrimental as defined by eleven prediction algorithms) per individual. Linear regression was conducted to test if genetic risk was associated with ASD-related symptom severity. Influence of risk scores on the relationship of symptom severity and sleep duration was assessed with interaction tests.

Results: Shorter sleep duration was associated with increased social impairment, severity for numerous challenging behaviors and attention deficit disorder, depressive disorder, and obsessive compulsive disorder. Symptoms were more severe in children sleeping ≤420 minutes/night compared to ≥660 minutes/night. Increased burden of deleterious variants in synaptic signaling genes significantly influenced the relationship of short sleep duration with expression of depressive disorder.

Conclusions:

Results show a clear relationship between shorter sleep and more severe ASD symptoms, demonstrating the importance of identifying sleep problems in this patient population. We are currently exploring pleiotropic genetic effects influencing sleep duration and ASD symptom severity. Characterizing pleiotropic effects may help refine the mechanisms underlying expression of sleep problems in ASD. This, in turn, may inform more effective treatment options.