A new study of nearly 1000 people with autism has confirmed that the genetics of the disease are much more idiosyncratic than some had thought. Rather than a few genes that raise the risk of autism throughout the population, scientists are finding dozens of genes that spur disease, many of them in just one or two people. It appears that these variants do share certain characteristics, however: Many play a role in cell proliferation and cell signaling in the brain.
Although scientists are heartened by this new map of autism genetics, they also say that they have a long road ahead in discerning how these genetic changes cause this particular disease.
Over the past few years, researchers studying autism and schizophrenia have found that the genomes of patients with one or the other are riddled with so-called copy-number variants: deletions or duplications of stretches of DNA that can encompass many genes. Indeed, the most dramatic of these copy-number variants are even visible with a microscope, as abnormalities in the chromosomes of children afflicted with undiagnosed intellectual disabilities. Some of these disease-causing changes happen spontaneously during embryonic development, whereas others are inherited.
The latest study, published online today in Nature, is the second phase of the Autism Genome Project consortium, which comprises more than 120 scientists in 11 countries in North America and Europe. (The first, published in 2007 in Nature Genetics, was a broad analysis of gene changes and copy-number variation, with fewer families and less detailed analysis of rare copy-number variants.) Here the scientists scanned the genomes of 996 children with autism-spectrum disorders, a group of conditions that affect social and communication skills, at high resolution and compared them with the genomes of the children's parents and to 1287 people without the disease.
Just over 5% of those affected had at least one "de novo" copy-number variant that had appeared in the sperm or egg cell forming the embryo, or spontaneously in the fertilized egg. And copy-number variants inherited from parents were also much more common in the autistic cohort: Overall, rare copy-number variants were 19% more likely to disrupt genes in autistic children than in controls.
Deletions of genes had a big effect. Nearly all of these were extremely rare, showing up in, at most, a handful of families. "Most individuals that have autism will have their own rare form," genetically speaking, concludes senior author Stephen Scherer, a geneticist at the Hospital for Sick Children in Toronto, Canada.
That said, the team found that genes deleted in autistic patients tended to perform similar tasks. Many were involved in aspects of cell proliferation, such as organ formation. A number participated in development of the central nervous system and others in maintaining the cytoskeleton, which protects the cell and helps it move.
"These are not random hits in the genome" and clearly have some connection to autism, says Jonathan Sebat, a geneticist at Cold Spring Harbor Laboratory in New York state. On the one hand, "you look at this and you immediately become enthralled," he says, especially because when it comes to parsing genetics, autism has lagged behind other psychiatric disorders because studies have been smaller. But at the same time, these are likely to be the same types of genes involved in schizophrenia and other related diseases, he says, so understanding how they cause autism will likely be tough.