Open Access Highly Accessed Open Badges Research

Evidence of novel fine-scale structural variation at autism spectrum disorder candidate loci

Dale J Hedges1*, Kara L Hamilton-Nelson1, Stephanie J Sacharow1, Laura Nations1, Gary W Beecham1, Zhanna M Kozhekbaeva1, Brittany L Butler1, Holly N Cukier1, Patrice L Whitehead1, Deqiong Ma1, James M Jaworski1, Lubov Nathanson1, Joycelyn M Lee1, Stephen L Hauser2, Jorge R Oksenberg2, Michael L Cuccaro1, Jonathan L Haines3, John R Gilbert1 and Margaret A Pericak-Vance1

Author Affiliations

1 Hussman Institute for Human Genomics, University of Miami Miller School of Medicine, 1501 NW 10 Ave, M-860, Miami, FL 33136, USA

2 Department of Neurology, School of Medicine, University of California, 220 Montgomery St., 5th Fl., San Francisco, CA 94143-0248, USA

3 Center for Human Genetics Research, Vanderbilt University, 519 Light Hall, 2215 Garland Ave. South, Nashville, TN 37232-0700, USA

For all author emails, please log on.

Molecular Autism 2012, 3:2  doi:10.1186/2040-2392-3-2

Published: 2 April 2012



Autism spectrum disorders (ASD) represent a group of neurodevelopmental disorders characterized by a core set of social-communicative and behavioral impairments. Gamma-aminobutyric acid (GABA) is the major inhibitory neurotransmitter in the brain, acting primarily via the GABA receptors (GABR). Multiple lines of evidence, including altered GABA and GABA receptor expression in autistic patients, indicate that the GABAergic system may be involved in the etiology of autism.


As copy number variations (CNVs), particularly rare and de novo CNVs, have now been implicated in ASD risk, we examined the GABA receptors and genes in related pathways for structural variation that may be associated with autism. We further extended our candidate gene set to include 19 genes and regions that had either been directly implicated in the autism literature or were directly related (via function or ancestry) to these primary candidates. For the high resolution CNV screen we employed custom-designed 244 k comparative genomic hybridization (CGH) arrays. Collectively, our probes spanned a total of 11 Mb of GABA-related and additional candidate regions with a density of approximately one probe every 200 nucleotides, allowing a theoretical resolution for detection of CNVs of approximately 1 kb or greater on average. One hundred and sixty-eight autism cases and 149 control individuals were screened for structural variants. Prioritized CNV events were confirmed using quantitative PCR, and confirmed loci were evaluated on an additional set of 170 cases and 170 control individuals that were not included in the original discovery set. Loci that remained interesting were subsequently screened via quantitative PCR on an additional set of 755 cases and 1,809 unaffected family members.


Results include rare deletions in autistic individuals at JAKMIP1, NRXN1, Neuroligin4Y, OXTR, and ABAT. Common insertion/deletion polymorphisms were detected at several loci, including GABBR2 and NRXN3. Overall, statistically significant enrichment in affected vs. unaffected individuals was observed for NRXN1 deletions.


These results provide additional support for the role of rare structural variation in ASD.