Identification of Genetic Variation in Autism Using Multiplexed Massively Parallel Sequencing

Background:

The genetic basis of autism, in large part, is unknown and no one unequivocal susceptibility region has been identified in more than a few percent of cases. Previously, reduction of heterogeneity by examining specific sub-phenotypes of autism has resulted in stronger genetic signals. It is very likely that the autistic population with comorbid epilepsy represents a subset within the broader autism population that includes several distinct genetic insults. A common genetic link between autism and comorbid seizures has been suggested through the identification of a homozygous autosomal recessive frameshift mutation within contactin associated protein 2 (CNTNAP2), which has been shown to cause a form of focal epilepsy with autistic-like language regression within the Old Order Amish population. Another gene that has been linked to autism and epilepsy is reelin (RELN), which has been implicated through multiple lines of evidence.

Objectives:

To identify genetic variants in RELN and three members of the neurexin family, CNTNAP1, CNTNAP 2, and CNTNAP4, which are associated with autism and comorbid epilepsy, using next-generation sequencing technology.

Methods:

Multiplexed bar-coded resequencing, in which DNA barcodes are ligated to the fragments to be read, was used to simultaneously resequence a large number of samples within the same sequencing run on the Illumina Genome Analyzer. In this study, 94 individuals, 68 of whom have had at least one seizure, from the Autism Genetic Research Exchange (AGRE) cohort and 16 control samples from the National Institute of Mental Health (NIMH) collection, were resequenced across all exons from CNTNAP1, CNTNAP2, CNTNAP4, and RELN. Variants were subject to validation using capillary-based sequencing and TaqMan SNP genotyping assays. Genetic and bioinformatic techniques were used to assess the pathogenicity of the validated variants.

Results:

Exons from CNTNAP1, 2, 4, and RELN, were resequenced on the Illumina Genome Analyzer and median coverage was found to be at 40x per base across 94 individuals. Less than 10% of the originally targeted regions required additional sequencing by traditional capillary-based sequencing. Both previously identified and novel variants were found when compared to controls sequenced in this study and by aligning data from the 1,000 Genomes project. For genetic variants not found in the initial screen of controls, additional controls were screened at the variant position using ABI TaqMan assays. All novel variants identified by sequencing on the Illumina Genome Analyzer were validated when sequenced by a secondary capillary-based sequencing.

Conclusions:

The allelic spectrum of genetic variants across these genes supports that they harbor rare variants predisposing to autism with comorbid seizures.