Improving our understanding of gene-environment interaction (GxE) as an etiologic mechanism in autism has been identified as a research priority in the NIH Inter-Agency Autism Coordinating Committee's (IACC's) Strategic Plan for Autism Spectrum Disorders (ASD) Research as well as in the IOM's recent Report on Autism and the Environment. However, empirical investigations of GxE in autism remain scarce. Although several lines of evidence support its general plausibility, GxE research in autism faces several key obstacles: 1) lack of epidemiologic datasets uniting genetic and exposure data;2) absence of mechanistic data/hypotheses on GxE in vivo to guide epidemiologic analyses (such data increases the prior probability that tested GxE associations are real);and 3) low statistical power for testing GxE using conventional methods. Here we propose a case-control analysis of GxE aimed at surmounting these obstacles. The Study to Explore Early Development (SEED) is a multisite ASD case-control investigation recruiting 900 3-5 year old children with ASDs, 900 typically developing controls, and 900 children with non-autism developmental impairments. Biosamples for DNA and exposure information across a wide range of hypotheses are available. SEED therefore brings together genetic and environmental data in a large epidemiologic sample. To surmount the other obstacles highlighted above, we propose extending the GWAS approach to consider GxE in what has been referred to as a gene-environment-wide interaction study (GEWIS). To do this we will complete GEWIS genotyping on the first 500 cases and 500 typically developing controls with data available during this 2-year grant timetable, develop summary exposure variables related to prenatal exposures, and use recently proposed novel statistical techniques to search for genes in the context of heterogeneity by environment.
Our specific aims i nclude: (1) identify SNPs whose effects on ASD may vary across exposure categories related to maternal behaviors (smoking and alcohol use during pregnancy), infection, and maternal medication use using the Illumina 1M-Duo SNP panel and two complementary analysis methods: a 2-df test for genetic association in the context of heterogeneity, an empirical Bayes approach to efficiently estimate GxE interaction effects by exploiting GxE independence in controls. (2) Integrate SNP findings from the above approaches, further characterize exposure interactions via dose-response and specificity analyses, and pursue replication analyses in other ASD data sets. This work will accelerate future efforts centered around gene-environment interaction both in our full set of SEED participants and in other data sets for autism.
We plan to carry out a multi-center case-control study of 500 cases and 500 controls from the SEED study to identify genes whose effects on ASD may vary across exposure categories related to pre-natal infection, maternal medication use, or smoking and drinking during pregnancy.
| Andrews, Shan V; Sheppard, Brooke; Windham, Gayle C et al. (2018) Case-control meta-analysis of blood DNA methylation and autism spectrum disorder. Mol Autism 9:40 |
| Andrews, Shan V; Ellis, Shannon E; Bakulski, Kelly M et al. (2017) Cross-tissue integration of genetic and epigenetic data offers insight into autism spectrum disorder. Nat Commun 8:1011 |
