Epigenetics of autism-related impairment: copy number variation and maternal infection

Abstract

Objective: Epidemiological data have suggested maternal infection and fever to be associated with increased risk of autism spectrum disorder (ASD). Animal studies show that gestational infections perturb fetal brain development and result in offspring with the core features of autism and have demonstrated that behavioral effects of maternal immune activation are dependent on genetic susceptibility. The goal of this study was to explore the impact of ASD-associated copy number variants (CNVs) and prenatal maternal infection on clinical severity of ASD within a dataset of prenatal history and complete genetic and phenotypic findings.

Methods: We analyzed data from the Simons Simplex Collection sample including 1971 children with a diagnosis of ASD aged 4 to 18 years who underwent array comparative genomic hybridization screening. Information on infection and febrile episodes during pregnancy was collected through parent interview. ASD severity was clinically measured through parent-reported interview and questionnaires.

Results: We found significant interactive effects between the presence of CNVs and maternal infection during pregnancy on autistic symptomatology, such that individuals with CNVs and history of maternal infection demonstrated increased rates of social communicative impairments and repetitive/restricted behaviors. In contrast, no significant interactions were found between presence of CNVs and prenatal infections on cognitive and adaptive functioning of individuals with ASD.

Conclusions: Our findings support a gene-environment interaction model of autism impairment, in that individuals with ASD-associated CNVs are more susceptible to the effects of maternal infection and febrile episodes in pregnancy on behavioral outcomes and suggest that these effects are specific to ASD rather than to global neurodevelopment.

Figure 1

Autism symptomatology and cognitive and adaptive functioning of children with ASD-associated copy number variants (CNVs) and history of maternal infection or fever during pregnancy. N=1971. Error bars = 95% Confidence Interval. As shown in graph C (p = .010), a main effect for presence of infections is demonstrated. As shown in graphs G (p =.019) and H (p = .049), main effects for presence of CNV are demonstrated. As shown in graphs A-E: (A) p = .006; (B) p = .006; (C) p = .017; (D) p = .012; (E) p = .014, significant interactions are observed. No significant interactions are observed in graphs F-H.