Genomic and Electrophysiologic Parameters Contribute to Clinical Endophenotypes in Autism and Epilepsy Populations

Background:  Autism spectrum disorders (ASD) are known to have complex inheritance patterns such as copy number variants (CNVs), single gene disorders, and rare mutations of common synaptic genes. Alteration in interneuron enriched genes may explain why perturbations of GABAergic circuitry have been implicated in common neurodevelopmental disorders such as autism (ASD), intellectual disability, and epilepsy. Defects in GABAergic migration, cell numbers, and circuit formation are found in animal models of autism and epilepsy.

Objectives:  Our central hypothesis is that temporal and spatial specific expression of GABAergic signaling pathways contributes to the influence of sleep disruption on epileptiform discharges and seizure expression in autism populations.

Methods:  A unique database of autism epilepsy subjects (130), autism alone subjects (97), and epilepsy alone (150) was compiled by recording the genes in CNVs from clinical microarray data (CMA). These data from CNV genes were compared to all exon mutations from autism genetic studies published to date. The Gene ID numbers and UniProt IDs were recorded and then searched through the Gene Ontology (GO), KEGG, Pathways Commons, and WikiPathways databases to record interactions. Comparisons were made between ASD alone, ASD + partial epilepsy, and ASD + generalized epilepsy groups.

Results:  

ASD + generalized epilepsy had a higher rate of detectable CNVs on clinical microarray testing (51%, p=0.05) than ASD alone (33%) or ASD + partial epilepsy (31%). The time at autism diagnosis was similar among the three groups. ASD + generalized epilepsy subjects had an older age of seizure onset (75 months, p=0.05) than ASD + partial epilepsy onset (52 months). ASD + generalized epilepsy subjects were significantly older when regression occurred (35 months, p=0.02) than ASD + partial epilepsy (25 months) or autism alone subjects (17 months). Individuals with epilepsy alone had an even younger age of seizure onset (39 months, p=0.007) and differed from autism populations in terms of polysomnogram parameters (total sleep time, waking after sleep onset, and sleep efficiency, p=0.04 to 0.007).

Meta-analyses of all four NGS papers yielded similar pathway results (including chromatin biology, cell to cell junction, cell adhesion categories, etc) to those already published [All adjusted p values were = or <0.009]. In contrast, only 6 genes among the 400+ genes from CNVs found in autism-epilepsy subject were found in common with all exon mutations published to date. Upon pathways analyses, some results were similar among all autism groups. However, pathways analyses suggest that there is a distinct set of enriched genes in CNVs from autism-epilepsy subjects including MAP kinase signaling, PAR1 mediated thrombin signaling, regulation of ARF GTPase signaling, diurnally regulated genes with circadian orthologs, prostaglandin synthesis and regulation, and transmission across chemical synapses. Notable genes deleted in CNVs from autism-epilepsy subjects including GABRB3, GABRG3, CHAT, and SLC18A3 (acetylcholine vesicle transporter) [all involved in interneuron function] were only found in subjects with ASD + generalized epilepsy.

Conclusions:  

Taken together, these data suggest that distinct subgroups within the autism and epilepsy populations have associated genomic variants which may impinge strongly on function of unique neural circuits and expression of clinical endophenotypes.