T8. GLUTAMATERGIC SIGNALLING AND AUTISM: A FAMILY BASED ASSOCIATION STUDY ON THE GLUTAMATERGIC NEUROTRANSMITTER SYSTEM

Abstract

Background Common single nucleotide polymorphisms (SNPs) explain around 50% of genetic risk underlying Autism-Spectrum-Disorder (ASD). In a previous study we showed that common functional SNPs of genes implicated in Fragile X Syndrome (FXS) were associated with ASD (Waltes et al., 2014). The FXS associated protein FMRP is regulated by the glutamatergic system, a patho-mechanism discussed for ASD. Objective We tested if functional common SNPs of glutamatergic genes were associated with ASD or ASD specific symptoms in two large ASD family cohorts (Autism Genome Project/AGP set, N=2734 families; German data set N=578 families). Since the genetic architecture between individuals with High IQ (HIQ= IQ>70) and Low IQ (LIQ= IQ≤70) differs (Vieland et al., 2010), we also split the two cohorts into high and low functioning (HIQ, LIQ) individuals. Methods 207 functional SNPs of 124 glutamatergic genes with a minor allele frequency over 5% were tested using Plink v 1.9 (DFAM) in the cohort with HIQ, LIQ and the combined cohorts, respectively. Phenotype association was studied by logistic ordinal regression correcting for gender, IQ, clinical site and population stratification. Phenotype measures were taken from the Autism Diagnostic Interview-Revised (ADI-R) scores for Social Interaction (Domain A), Verbal Communication (B1-B4; verbal individuals only), Non-Verbal Communication (Domain B2; B3; all individuals) and Repetitive Behavior (Domain C). Variants significant in both cohorts with effect sizes in the same direction were considered as replicated. Similarly, genes with any nominal significant variant in both cohorts were considered as replicated hits. Results We identified nominal significant associations of variants rs7206796 and rs3790112 (GNAO1) as well as rs3742926 (AKAP6) in both the German and the AGP LIQ cohort. In addition, significant but not overlapping SNPs of AKAP2 were identified (LIQ only). rs731826 (AKT1S1) was associated with Non Verbal Communication in the HIQ cohorts. In addition, we report 24 variants that were nominally associated with ADI-R scores in both cohorts. Genes that were involved in all five phenotypes tested were strongly related to the mTOR Pathway (e.g. genes MTOR and TSC2) or the second messenger system (e.g. G-proteins GNAS, phospholipases PLCB, protein phosphatases PPP1CA). A subset of genes was specific to each phenotype or significant in one of the subgroups (HIQ and LIQ) only. For example AKAP13 or RPS6K are associated with Verbal Communication in Low Functioning Individuals. Discussion Replication and extending previous findings we found that variants of genes that are associated with glutamate signaling, and specifically the mTOR pathway are modulators of ASD symptoms. Genes such as TSC1 or RPS6K are known to mediate glutamatergic signaling through mTOR, whereas AKAP proteins are important interactors of glutamate receptors (Sanderson et al., 2011). Both, the mTOR pathways and AKAP proteins have previously been associated with ASD (Chen et al ., 2014; Poelmans et al., 2013). Further functional analyses of glutamatergic variants are thus recommended to elucidate ASD etiology.