CNV Analysis and Exome Sequencing in Japanese Autism Spectrum Disorder Subjects

Background: Linkage and candidate gene association studies have implicated several genes and chromosomal regions in autism. However, consistent picture of a common susceptibility loci in autism is still lacking. Hence there is paradigm shift away from the previously held “common disease – common variant” hypothesis to a “common disease – rare variant” model, for the genetic architecture of autism. There is a growing consensus among geneticists that rare structural variants including genomic copy number variations (CNVs) may contribute to the autism etiology. Most recently, exome sequencing (ES) studies have also revealed the importance of rare single nucleotide variations (SNVs).

Objectives: In this study, we have examined the global CNV, with special emphasis on rare de novo and inherited CNVs in Japanese samples. We have also sequenced the exomes of individuals which include patients with sporadic autism and their parents, reasoning that these families with no previous history of ASD or related phenotypes would be enriched for de novo mutations.

Methods: The 203 families were recruited on collaboration with a non-governmental organization, Asperger Society Japan (http://www.as-japan.jp/) in central Japan. The Affymetrix Genome-Wide Human SNP Nsp/Sty 6.0 array was used to screen the samples.  PennCNV and Birdsuite package were used to identify autosomal CNVs from the genome wide SNP data. We have also sequenced the exomes of 60 individuals which include 20 patients with sporadic autism and their parents. For exome sequencing, Genomic DNA was captured using SureSelect Human All Exon v5 kit, and sequenced on HiSeq2000. Image analysis and base calling were performed by sequence control software real-time analysis and CASAVA software v1.8.  Single-nucleotide variants and small indels were identified using the GATK UnifiedGenotyper and filtered according to the Broad Institute's best-practice guidelines v3. 

Results: Among a total of 7305 CNVs detected in affected children after all QC corrections, 3728 were found to be affecting genes. Among the remaining 378 CNVs, 17 novel/ultra-rare de novo events were identified through a series of filtering and validating strategies. Among the de novo events, there were four novel, five partially novel and eight ultra-rare CNVs; many of these regions has genes involved in neurodevelopmental pathways. Among the inherited CNV, we found 53 ultra-rare events. All of the ultra-rare/novel CNVs were re-confirmed by relative qPCR. A total of 32 rare novel de novo events were observed though ES. This includes 15 missense mutations, a nonsense mutation, 2 frameshift mutations caused by indels, one splice site mutation, and 7 synonymous mutations. 25 de novo events had occurred within the coding sequences, while the remaining 7 de novo events were located within 20 bp of the exon/intron boundary. Majority of these mutations were predicted (PolyPhen or SIFT) to have a damaging effect on the respective protein structure/function. Several of these genes have been implicated in neural functions.  

Conclusions: We discovered several novel/ultra-rare de novo and inherited CNVs and SNVs, many of them are potentially deleterious; further highlighting the genetic heterogeneity of the disorder. Many of these variations affect genes that play major roles in crucial neurodevelopmental processes.