Autism Genome Project Consortium Implicates Rare CNVs in Autism

June 09, 2010

Members of the Autism Genome Project Consortium reported online today in Nature that they have identified a host of rare copy number variants (CNVs) that appear to contribute to autism susceptibility.

The team, which includes more than 100 researchers from centers around the world, used high-density microarrays to assess CNV profiles in about 1,000 families affected by autism and nearly 2,000 control individuals.

Their results suggest that while individuals with autism don't necessarily have more CNVs overall than unaffected individuals, they tend to carry more CNVs that affect genes — particularly those involved in processes such as brain cell communication and cellular proliferation.

The results substantiate the importance of genes in autism susceptibility, corresponding author Stephen Scherer, a molecular genetics researcher at the University of Toronto and director of the Hospital for Sick Children's Centre for Applied Genomics, said during a telephone briefing with reporters this week.

Those involved in the study also emphasized the potential of the new findings for helping to diagnose autism earlier and for finding pathways that might serve as targets for improved autism treatments.

Although a fraction of autism cases correspond to conditions — such as fragile X syndrome — resulting from changes to a single gene, most cases appear to be a consequence of much more complex and heterogeneous genetic patterns.

In an effort to uncover new autism risk genes and gain insights into commonly affected molecular pathways, consortium members used the Illumina Infinium 1M SNP microarray to evaluate CNV patterns in 1,275 individuals with autism spectrum disorders and their parents as well as 1,981 unaffected control individuals.

After their quality control steps, the team was left with data for 876 family trios, including 996 individuals with ASD, and 1,287 controls. The CNVs they identified were larger than 30,000 bases and present at less than one percent frequency overall.

Although the team did not find more CNVs overall in individuals with ASD, they found that more of these CNVs — especially deletions — affected gene coding regions in the ASD group than in unaffected individuals.

On average, individuals with ASD had about 19 percent more genic CNVs than control individuals. In addition, some 5.7 percent of children with ASD carried de novo CNVs not present in either of the child's parents.

Some, but not all, of the affected genes have been implicated in autism previously. Among the genes not found in past studies of autism: SHANK2, SYNGAP1, DLGAP2, and the DDX53-PTCHD1 locus.

"Most individuals with autism are probably quite unique," Scherer said, noting that even the most common CNVs detected turn up in less than one percent of the ASD cases tested.

Now that researchers have a catalog of genes affected by rare CNVs, Scherer added, it's possible to begin tying these genes together in pathways and looking at effects on brain function. For example, the researchers reported that the CNVs they detected often impacted genes involved in cellular motility, proliferation, and communication pathways.

By further characterizing these pathways, the team hopes to find clues to detecting autism earlier. Early detection and intervention, in turn, are expected to improve autistic children's social, intellectual, and language outcomes, noted co-author Geraldine Dawson, chief scientific officer for Autism Speaks and a member of the National Institutes of Health's inter-agency autism coordinating committee.

Such analyses are also expected to highlight biological pathways that might be targeted by new or existing drug treatments.

The current findings "give us an idea of what the picture [of autism genetics] may look like," Anthony Monaco, a researcher with the University of Oxford's Wellcome Trust Centre for Human Genetics and study co-author, told reporters.

Nevertheless, the researchers explained, since the rare CNVs found so far account for just 10 to 15 percent of autism risk, more research is needed to understand autism genetics and the interplay between autism susceptibility genes and environmental risk factors.

That will likely require studies using high-throughput sequencing of autism families to find additional genetic changes in genes and genomes, Monaco said, along with studies aimed at finding the best ways of translating genetic research into the clinic.

In an effort to get a better handle on the predictive statistics available from microarray data, Scherer and his colleagues in Ontario plan to use arrays to assess some 5,000 newly diagnosed autism cases in that province over the next few years.

Researchers in Oxford reportedly hope to do a similar pilot study involving at least 1,000 children with autism.

"By identifying the genetic causes of autism, we hope in the future to be able to improve the diagnosis and treatment of this condition which can affect children and their families so severely," Monaco said in a statement.

"[K]nowing about these genetic changes can help the families involved come to terms with why their child has autism, but it can also be important where there are siblings too in determining future risk."