Autism: New Findings From Big Gene Study

Preliminary results from the largest-ever autism genome study, published today in the online edition of Nature Genetics, offer promising new avenues of research into autism spectrum disorder (ASD), which affects 1 in 150 children in the United States. The findings, based on an analysis of genetic data from 1,200 families, are significant because they come from the Autism Genome Project (AGP), a consortium of over 120 scientists from 19 countries who joined forces in 2002. The unprecedented collaboration dramatically increased the pool of genetic samples from families with autism, giving researchers a better shot at teasing out possible underlying causes of the disorder. NEWSWEEK's Julie Scelfo spoke with UCLA's Dr. Dan Geschwind, a professor of neurology, psychiatry and human genetics at the David Geffen School of Medicine, who heads the school's chapter of the Autism Genetic Research Exchange (AGRE). Excerpts:

NEWSWEEK: What is the Autism Genetic Research Exchange?
It's the U.S.'s largest and most well annotated pool of genetic data for the study of autism. Cure Autism Now started funding it in 1997, and three or four years ago NIH began funding it as well. Currently, I'm the chief scientific adviser.

Why is AGRE important?
In the past it was difficult for researchers to collect genetic samples from children with autism because it involves extensive psychological testing and drawing blood, which is hard for ordinary folks but especially difficult for kids with autism. The folks at Cure Autism Now, which started AGRE, made it easier by sending a phlebotomist and researcher directly to the home instead of asking families to bring their kids to the hospital. The result was a repository of data scientists like me could use to study the genetic basis of autism.

Why hadn't individual researchers pooled their data before?
The usual model in this field is investigators collect their own data from families and don't share it with other people. There's a reason for that: the way things are funded, investigators have to show their independence, their genius. I don't think that model holds much any more. The future of science is really in large collaborative efforts, like the human genome project. What it does is make samples of data available to people all over the world who couldn't otherwise afford to collect it. It's a very efficient way of doing research.

And the Autism Genome Project expanded the pool of data and scientists worldwide?
The AGP is made up of, basically every group in the world with a significant interest in autism genetics. It's coordinated by Autism Speaks. We meet, combine samples, and do joint analyses. It's large studies like this that are going to help us understand what the genetic architecture and landscape of autism is.

So what are the preliminary findings?
What this shows is that there are rare chromosomal changes, what we call copy number variations, [associated with autism] rather than a single base pair or DNA sequence.

What does that mean?
Humans often have common genetic variations that are linked to disease. The thinking has been that this model should apply to autism as well, and it does—common genetic variations do contribute to autism. But we've also discovered that a significant proportion of ASD results from rare mutations that are specific to people with ASD, and don't occur in normal people. Finding rare mutations is a huge scientific challenge because it requires an intense, expensive genome investigation.

Haven't scientists known for a long time that autistic spectrum disorders are genetic?
Yes. We know, based on family and twin studies, that autism is the most heritable of any neuropsychiatric disorder.

But do scientists know which specific genes are involved?
Yes, we've begun to identify them. It's a very exciting time. Let's say you know there are some needles in some haystacks and you have to go find them. That's what the genome is like. The question is what do we find in people with autism that is not in other people. Or what do we find more of? So the next step, if we're looking for a proverbial needle in a haystack, is to identify regions of the genome, which are parts of chromosomes, where the genetic risk factors lie.

What does the new study mean for parents of kids with autism?
It's kind of a first step in beginning to find what the landscape of autism genetics is going to look like. What, from a genetic standpoint, are the blocks that actually build autism? Our goal is to identify them, which will hopefully, not only help in diagnosis, but in predicting how the disease shape people's lives. We think there are many different forms of autism and by understanding the genetic basis, it could give us a way to test what form of ASD somebody has.

Would this have any implications for treatment?
It's not currently possible, but the goal would be to identify, as early as possible, children who are at risk of autism. And if they're at risk, you can intervene early. The brain is very plastic early on, and we have a lot of evidence that the earlier one intervenes, the better chance we have of making an impact on the outcome.

That's the one certainty in autism right now: interventions stand a better chance the earlier they're implemented, right?
Yes, but some children don't respond to early interventions. We need to figure out why. And genetics will help with that. That's the hope.