A genetic autism study of unprecedented scope and power has uncovered more than two-dozen high-confidence risk genes for the disorder. It offers compelling evidence that spontaneous, or de novo, mutations contribute to autism in at least 27 percent of families in which the parents and siblings are unaffected.

The new research has also established conclusively that 'higher-IQ' autism, which mostly affects boys, has a different genetic basis from 'lower-IQ' autism, which commonly affects both boys and girls.

The researchers, whose findings was published by Nature, sequenced the whole exomes - the protein-coding regions of the genome - of 2,515 families from the Simons Simplex Collection, a large repository of genetic, biological and phenotypic data from 'simplex' families, which consist of one child with autism, unaffected parents and usually at least one unaffected sibling.

The study, carried out at three different universities across the country, pinpointed seven genes that have mutations in three or more children with autism, implicating these genes in the disorder with near certainty. It also identified another 20 genes with mutations in two children. Each of these genes has more than a 90 percent chance of being a true autism gene, the researchers reported in their paper, 'The contribution of de novo coding mutations to autism spectrum disorder.'

"We have a set of genes for which now, if people see a likely gene-disrupting mutation when sequencing a young child, there's a high risk of the child developing autism, and that, to my mind, is pretty powerful stuff," says Evan Eichler of the University of Washington, a Simons Foundation Autism Research Initiative (SFARI) Investigator who leads one of the laboratories that contributed to the study. "Recognizing this early on may allow for earlier interventions, such as behavioral therapies, improving outcomes in children."

The exome analysis was also carried out in the laboratories of SFARI Investigators Michael Wigler of Cold Spring Harbor Laboratory in New York and Matthew State of the University of California, San Francisco.

Although the researchers at the three different universities might be assumed to be competitors, in an unusual move they have published the results of their exome-sequencing studies in a single joint paper. "That seemed to us to make the most sense scientifically, though not from the standpoint of getting credit," Eichler says. "We felt that it would create the biggest benefit for the community of autism researchers."

Because the study was performed on such a large number of families, it has a statistical power that has advanced several hypotheses about autism genetics beyond the realm of serious doubt, Wigler says. For instance, the study identified a collection of genes that likely play a role in autism in individuals with lower nonverbal intelligence quotients (IQ), but not those with higher nonverbal IQ. "High-IQ autism is almost entirely a male disease, which has been a bit of a puzzle," Wi¬¬gler says. "The paper clearly indicates that the genetic causality differs from that of girls and lower-IQ boys. That had been speculated, but now it's clear, and it's very important."

The exome sequencing also showed that girls with autism are particularly likely to have mutations in genes expressed during early embryonic development. This finding, combined with the fact that autism is less prevalent in girls than boys, suggests that girls may somehow be protected against mutations in autism genes that are expressed in later stages of development, but not those expressed in early embryonic development.

The study identified about 300 new autism candidate genes, about half of which are likely to be true autism genes. The findings are a vindication of the approach to understanding autism genetics via large-scale sequencing of simplex families, according to State. "It makes the definitive point that there's a way forward in gene discovery in autism," he says. "For more than a decade, the field has been searching for a systematic way to identify autism risk genes."

The researchers have estimated that about 500 genes are "highly vulnerable," Wigler says, meaning that "if such a gene is damaged by mutation in a child, there is a high probability that the child will develop on the autism spectrum." Just one of the two copies of a gene needs to be damaged, almost always leaving a good copy, Wigler notes. "In the future, that good copy might be manipulated pharmacologically to restore cognitive function."

The new study bolsters the idea that the hundreds of autism risk genes converge on comparatively few biological pathways. For instance, many of the genes identified by the exome sequencing interact with the protein at the heart of fragile X syndrome, the most widespread form of inherited intellectual disability in boys. Another set of the genes regulate chromatin, a protein complex that helps package DNA in the cell nucleus. "The mutations don't seem to be randomly distributed," State says. "Understanding the particular roles these genes are playing is going to be tremendously exciting for the field."

The findings concerning the new candidate genes and the higher-confidence genes will be aggregated at SFARI Gene, a free online genetics database developed by SFARI for the autism research community and those studying related disorders. The site offers a curated database of all autism candidate genes, with assessments of the strength of the evidence for each gene, together with information on protein interactions, copy number variants and animal model resources.

This study, while far-reaching, examined only de novo mutations and the exome, which constitutes just 1.5 percent of the human genome. Further examination of the Simons Simplex Collection will likely uncover many more genetic causes of autism, according to Eichler. "If we can explain 27 percent of simplex autism from this small sliver of a sliver of DNA, there's a lot more to be mined from complete genome sequencing of the SSC," Eichler says. "Stay tuned."