Temporal Gene Expression Profiles and Behavioral Regression in Children with ASD with Postsynaptic Density Gene Disruptions

Background: Approximately one-third of children with an autism spectrum disorder (ASD) experience developmental regression within the first three years of life. However, the mechanisms underlying this behavioral phenotype remain unknown. Recently, disruptions in postsynaptic density (PSD) genes have been found to result in higher rates of developmental regression in language and social engagement skills (Goin-Kochel et al., under review). PSD genes play a role in regulating synaptic function in human neocortex (Bayés et al., 2011) and show differential timing in expression, with some genes preferentially expressed during synaptic formation and others during synaptic remodeling and differentiation (Swulius et al., 2010). Thus, disruptions in PSD genes preferentially expressed later in synaptic development may lead to normal initial behavioral development followed by onset of abnormal development and/or behavioral regression. Understanding the relationship between developmental gene expression timing and phenotypic ASD profiles may elucidate mechanisms of behavioral regression in ASD.

Objectives: To explore the following correlations in individuals with ASD who have PSD gene disruptions: 1) gene expression timing and presence of behavioral regression and 2) gene expression timing and age of ASD symptom onset.

Methods: Participants were 33 children from the Simons Simplex Collection with PSD gene disruptions (as defined by Iossifov et al., 2014) who meet strict criteria for ASD. Age of typical maximum expression for each PSD gene was extracted from the BrainSpan transcriptome exon microarray data (http://www.brainspan.org). Eight children with parentally reported history of developmental regression in language and social engagement skills were compared to 25 children without regression. First, groups were compared using independent samples t-test on the age of typical maximum expression of disrupted genes observed between the first prenatal trimester and third postnatal year. Second, age of typical maximum expression was used to predict regression groups using logistic regression. Finally, for all participants, using regression analysis, age of typical maximum gene expression was used to predict parentally reported age of ASD symptom onset.

Results: Age of maximum expression did not predict regression group (p=.46), and regression groups did not differ significantly in age of maximum expression (t(37)=.72, p=.48). However, age of maximum expression significantly predicted age of ASD symptom onset (F(1, 37)=5.02, p=.03, R²=.119) such that the later the age of maximum expression, the later the age of reported onset of ASD symptoms. That is, participants with disruptions in PSD genes typically expressed later in development tended to have later reported ages of ASD symptom onset.

Conclusions: The observed relationship between age of ASD symptom onset and age of maximum gene expression in individuals with likely gene-disrupting mutations to PSD genes suggests the importance of gene expression timing in symptom expression timing. Previous mouse model studies demonstrate preferential expression of PSD genes in different phases of synaptic development (Swulius et al., 2010); these findings suggest that the differential phases of expression can impact symptom onset. Better understanding of the normal developmental timing of gene expression in genes disrupted in individuals with ASD may aid in explaining the phenotypic variability regarding symptom onset among children with ASD.