Identification of Rare Recurrent Genetic Variants in High-Risk ASD Families and Their Role in a Large ASD Case/Control Population

Background:  Genetic variation plays a significant etiological role in autism spectrum disorders (ASDs), and numerous studies documenting the relevance of copy number variants (CNVs) and single nucleotide variants (SNVs) in ASD have been published.

Objectives:  This study was designed with three goals in mind.  The first goal was to identify CNVs present in high-risk ASD families and to determine which of those CNVs contribute etiologically to ASD in the general population.  The second goal was to confirm the findings of several published ASD CNV studies using a larger case/control population, to determine the potential clinical utility of those CNVs in the genetic analysis of children with ASD.  The third goal was to determine if any SNVs identified as potential risk variants in high-risk ASD families supported their potential role as risk alleles in same case/control population.

Methods:  CNVs in high-risk ASD families were identified using the Affymetrix Genome-Wide Human SNP array 6.0.  SNVs were identified by sequence capture in regions of genetic linkage and in published ASD candidate genes.  CNVs and SNVs subsequently were evaluated in a set of 3000 ASD cases and 6000 controls using a custom Illumina iSelect array followed by molecular confirmation of significant variants.

Results:  We identified 153 putative ASD-specific CNVs in 55 affected individuals from 9 multiplex ASD families.  These CNVs were not observed in control samples from three generation Utah CEPH families.  Our case/control analysis revealed that 14 CNVs from high-risk ASD families were observed in unrelated ASD cases and had at least suggestive evidence for a role in ASD etiology.  We also identified CNVs not detected in our high-risk families using SNVs probes that we placed on the array, suggesting that some genetic regions can be impacted at both the structural and sequence levels.  Findings for published CNVs indicated that many appeared to increase the ASD risk only slightly, since these CNVs also were found in many control DNA samples.  One rare SNV was observed in two unrelated ASD cases and in none of the 6000 controls, suggesting that variants in this gene may be risk factors for ASD.

Conclusions:  Genetic variants identified in high-risk ASD families also appear to play a role in ASD etiology in unrelated ASD cases.  The absence of 10 of these variants from public ASD databases suggests that they represent previously unidentified risk variants.  These variants lay the groundwork for the development of a more sensitive test to use in the genetic evaluation of children with ASD.