About 90 percent of autism spectrum disorders have suspected genetic causes but few genes have been identified so far. Now a large international team, including six present and former University of Washington researchers, has identified several genetic links to autism, chief among them a variant of semaphorin 5A, a gene whose protein product controls nerve connections in the brain.

Semaphorin 5A had already been known to help guide growing neurons to the right connecting points in the brain during fetal development. To verify that semaphorin 5A plays a role in autism, the researchers looked at brain tissue samples from the Autism Tissue Program and the Harvard Brain Bank, and found the amount of the semaphorin 5A protein to be significantly reduced in autism brains compared to non-autism brains. The finding suggests that autism could result from differences in how nervous system connections are made in the brain.

Publishing this week in the journal Nature, the team also reports additional evidence that many rare and common genetic variants contribute to autism.

"The biggest challenge to finding the genes that contribute to autism is having a large and well studied group of patients and their family members both for primary discovery of genes and to test and verify the discovery candidates," said Aravinda Chakravarti, one of the lead authors of the study. He is a professor of medicine, pediatrics and molecular biology and genetics at the McKusick-Nathans Institute of Genetic Medicine at Johns Hopkins. "This latest finding would not have been possible without many research groups and consortia pooling together their patient resources. Of course, they would not have been possible without the genomic scanning technologies either."

Using genome wide scanning technologies, the team "read" the genomes of family members among whom more than one individual was diagnosed with an autism spectrum disorder. In all, they studied 1,031 nuclear families and 1,553 affected children, looking at 500,000 spots in the genome, so-called single nucleotide polymorphisms, for differences that stood out in the DNA. The information was collected from two sources: the Autism Genetic Resource Exchange and the US National Institute of Mental Health.

Specifically, they performed two types of genetic analysis to identify genes with both rare and common genetic variations that might contribute to autism. By studying siblings with autism, the authors teased out four regions of the human genome -- on chromosomes 6, 15, 17 and 20 -- where rare variants in yet unidentified genes appear to contribute to autism susceptibility. Additionally, after examining the patterns of genetic similarities and differences in unrelated people with autism, the researchers discovered a common variation near only one gene, semaphorin 5A.

That only one common variant in came out of this large sample study indicates there probably are many more that contribute to autism, but none have large effects, the researchers believe. But there also may be numerous rare gene variant containing genes. Identifying these will require even larger collaborative studies that are currently underway.

The identification of semaphorin 5A, and seven additional likely -- but not yet proven -- genes involved in nervous system development, cell structure and other cell functions, only was possible through an extensive collaborative effort that included in patient samples from the Autism Consortium in Montreal, the Autism Genome Project and additional samples from Finland and Iran.

The researchers said these discoveries are an important step forward, but just one of many that are needed to fully dissect the complex genetics of autism.

UW scientists involved in the research are Annette Estes and Jeff Munson of the university's Autism Center; Ellen Wijsman of the department of biostatistics and medicine; and Olena Korvatska of the department of medicine. Former UW researchers Geraldine Dawson of Autism Speaks and Gerard Schellenberg of the University of Pennsylvania also participated.

The research was funded by the National Institute of Mental Health and the Simons Foundation.

###