Focus banner

New Genetic Pieces Found in Autism Puzzle

Analytic Methods Combined to Trace Rare and Common Factors of Disease Susceptibility

“Why does my child have autism?” It’s a question parents have long been asking doctors. While the clinical features of this brain development disorder are well known, its genetic underpinnings are not.

Mark Daly
Photo courtesy MGH

Mark Daly and colleagues used genomewide arrays to scan DNA samples from autism patients versus controls in search of common, single letter differences in the genetic code. This approach identified a specific gene active in nerve cells in which variations may increase autism susceptibility.

In one of the most comprehensive studies of autism to date, an international team led by investigators at Harvard, MIT and Johns Hopkins identified three new sites of DNA variation in autism. One was a common single letter variation in the genetic code.

Armed with data from more than 1,000 patient families and improved technologies for combing through the genome, the researchers brought together two different analytic approaches to search for both common and rare genetic variations in the disease. Two rarer variants, on chromosomes 6 and 20, were identified in addition to the single letter difference, which appears on chromosome 5 and reflects variation in a gene called SEMA5A, active in nerve cells. Importantly, the team discovered that their findings correlated with a putative functional difference in the autistic brain: the protein product of the SEMA5A gene is diminished in brains of autism patients.

The three newly identified sites add to earlier findings by the team, revealing structural variants on chromosome 16 that were associated with autism in the same set of patient data. Upon further unraveling, this constellation of genetic clues will likely advance the understanding of autism biology. The multipronged analytic approach of the study may also be useful in investigating the genetics of other complex diseases that can be inherited.

Combination Approach
Autism is believed to be a disorder of many genes, and there are likely to be mutations of different frequencies that play a part, said senior author Mark Daly, HMS associate professor of medicine at Massachusetts General Hospital and associate member of the Broad Institute of Harvard and MIT. Rare mutations may contribute strongly to autism cases within a family, while common mutations occurring in a larger population may have weaker or more incremental effects on susceptibility.

“The best opportunity we have is to study families with multiple affected offspring,” Daly explained. Such data allows for analysis of both rare and common genetic factors.

“This autism study was an exemplar of what we can find both by linkage and association in families.”

—Aravinda Chakravarti

The regions on chromosomes 6 and 20 were identified by linkage analysis, a method comparing afflicted and nonafflicted members of individual families. The SEMA5A gene was identified by genomewide association, a method making comparisons across families.

The combined linkage and association approach “hasn’t been done often,” Daly said, but “it’s not a novel concept.” The purpose is simply to increase the chance of identifying DNA variants of different frequencies that contribute to the disorder.

“We need to understand the full frequency spectrum,” explained Aravinda Chakravarti, also a senior author and director of the Center for Complex Disease Genomics at Johns Hopkins University School of Medicine. “This autism study was an exemplar of what we can find both by linkage and association in families.”

In addition to this combined analytic approach, the record number of autism families with genetic data, recent strides in genomics technology, and the availability of higher resolution DNA-scanning methods all contributed to the scope of the study.

Biological Understanding
In order to use the findings of this study to advance the biological understanding of autism, the newly identified DNA variations will need to be further characterized. For instance, the sequence details of the autism-associated variant of SEMA5A must be described, and the regions of variation on chromosomes 6 and 20 must be narrowed down to specific genes that confer autism susceptibility. Then the function of these genes and their variants can be investigated. “That’s our clearest path to identifying the biological processes that are causal in autism,” Daly said.

Right now there are not enough genes identified in autism to formulate a clear hypothesis about the biology of the disorder. “We have so many more genes to discover than the few that we’ve actually put our hands on,” Daly said, “We’re very much at the beginning.”

How will researchers know when they have finally identified enough genes?

“When the genetic findings are true enough and numerous enough that they start identifying significant, statistically robust connections in … other datasets,” Daly said, referring to sets of biological data gathered independently of genetic studies. “Then we’ll recognize that we’re deriving biological insight from the genetic studies.”

The research appears in the Oct. 8 issue of the journal Nature, with co–first authors Lauren Weiss of Harvard and MIT and Dan Arking of Johns Hopkins.

Students may contact Beth Luise, administrative assistant for Mark Daly, at for more information.

Conflict Disclosure: Mark Daly declares no conflicts of interest. Aravinda Chakravarti is a paid member of the scientific advisory board of Affymetrix, Inc., a relationship managed under conflict-of-interest rules at Johns Hopkins University.

Funding Sources: Work in Mark Daly’s lab was supported by the Autism Consortium and the Simons Foundation. In Aravinda Chakravarti’s lab, work was supported by the National Institute of Mental Health and the Simons Foundation. The authors are solely responsible for the content of this work.


Copyright 2009 by the President and Fellows of Harvard College