Autism linked to new kind of genetic defect


by Jim Schnabel

January 11, 2008

Two groups of independent researchers searching the genome in a new way have identified a gene defect that appears to underlie some cases of autism. The finding highlights the importance of a new category of genetic abnormality known as “copy number variation.”

The larger of the two studies appeared online on Jan. 9 on the website of the New England Journal of Medicine. The smaller one was published online on Dec. 21 on the website of Human Molecular Genetics. Both studies had sought an answer to an enduring mystery about the autism spectrum of disorders: On the one hand, research in identical twins has suggested that these disorders are caused far more by genetic problems than by environmental ones. On the other hand, in 90 percent of autism cases geneticists have been unable to pin the disorder on the usual genetic suspect—a mutated gene inherited from a parent.

The scientists looked for a new genetic suspect instead.

“Autism is a disorder with a high genetic component,” says Yiping Shen, a researcher at  Children’s Hospital in Boston who was second author on the New England Journal study. “But it’s not necessary that the genetic component be inherited from either parent.” As recent human-genome research has highlighted, some genetic abnormalities occur spontaneously during conception, when the separate chromosomes from sperm and egg cells are supposed to rearrange themselves into a single mixed genome but in places fail to do so properly. Many of these spontaneous or de novo errors result in a deleted genetic sequence or a duplicated one—a so-called copy number variation.

In the recent studies, both teams sifted through the DNA of hundreds of people with autism, searching for previously undiscovered copy number variations that could explain the disease. Both zeroed in on the short arm of chromosome 16, where a sequence of 25 genes varied in copy number in some autism cases, but was far less likely to do so in people without autism.

“Some of these genes are highly expressed in brain development,” says Shen. One of the next steps is to determine which of the 25 are responsible for causing autism. Once the precise culprits are found, their behavior will be studied in mice genetically engineered to have the same copy number variations.

“The genes for mice and humans,” Shen says, “are very conserved in this region,” meaning they are very slow to evolve, which suggests that the way the genes behave in mice should give solid clues to how they behave in people.

“It’s an exciting time,” says Jonathan Pevsner, an autism researcher at the Kennedy Krieger Institute in Baltimore. Although the just-described chromosome-16 abnormality applies to only about one percent of previously unexplained autism cases, it shows that copy number aberrations can affect autism and suggests that other aberrations might cause more cases of the disorder. Studies with larger groups of autism patients would have the statistical power to uncover more copy number aberrations.

“Is this kind of thing happening perhaps with less frequency, across the genome for other individuals with autism?” Pevsner says. Shen thinks it’s “quite possible” that different cases of autism could end up being linked to many other copy number abnormalities.

“It’s part of a new perspective the research community is developing, for cancer or arthritis or autism—whatever you’re interested in,” says Pevsner. “It’s a new model of how genetic variations can occur and how they can be associated with disease.”