Autism and genetics
Ways to get autism that do not run in the family
April 28, 2007
Autism isn't just frustrating for parents. It is frustrating for scientists too.
There is clearly a genetic part to autism. It tends to run in families. And identical twins, who share the same genes, are much more likely to both have autism than are fraternal twins.
So genes are involved. But it isn't easy to nail down what is going on genetically with autism.
Autism (and its entire spectrum of disorders) is actually pretty common1 in 166 people have it. If you have one child with autism, the risk for the next child is only 2-6%. If autism were due to a single gene, we might expect numbers like 25% or 50%.
And then, if you have two children with autism, the chances that the third will be autistic are around 35%. This is usually interpreted as meaning that lots of different genes are involved (click here to learn more).
Another way to explain a lot of these results is if autism weren't in some of these folks' family trees to begin with. What if in some families, autism just appears out of nowhere?
This isn't as crazy as it sounds. DNA can and does change from generation to generation (click here to learn more).
Usually the differences aren't significant but sometimes they are. For example, 7 out of 8 cases of dwarfism are the result of completely new DNA changes.
To look for these changes, a team of researchers compared the DNA of autistic children to their parents. What they found at least 10% of the kids had missing or extra DNA compared with their parents. These DNA changes may explain the children's autism. And this number may be an underrepresentation.
Extra or missing genes can lead to autism
So how do you go about finding DNA changes that happen between generations? You can't just read all 6 billion letters of DNA of parents and children and figure out what is going on. This is too expensive and time consuming.
What the researchers decided to look at were big changes in DNA. Usually scientists focus on smaller changeschanges in a single letter or two of the code.
But lately they have begun to realize that bigger changes are significant too. Changes like big chunks of DNA that go missing. Or are duplicated.
The researchers looked at three groups. One group was 118 families with a single child with autism. The second group was 47 families with more than one autistic child. And the third group was 99 families who had no autistic children.
The team of researchers collected DNA from all the parents and children and compared them using something called CGH arrays. A CGH array is a way to look for big DNA changes in lots of places on a person's DNA all at once.
What the researchers were looking for was DNA that was different in autistic children as compared to their parents. And that is what they found.
Around 10% of the autistic kids in the families with one autistic child had big DNA changes compared to their parents. And most of these were deletionsmissing chunks of DNA.
Only 3% of the autistic kids had detectable DNA differences from their parents in families with multiple autistic children. Just 1% of kids in families with no autistic children had these sorts of changes.
These results tell us that one way to end up with autism is when big chunks of DNA go missing. Or get duplicated.
You might also think from these results that big DNA changes are more common in families with one autistic child. This may or may not be true.
The researchers were looking for DNA changes between parent and child. These changes aren't that common. Which means the odds of having two autistic children because of these sorts of changes is even rarer.
In a family with more than one autistic child, the odds are pretty good that the parents had the DNA changes already and passed it down. Now these might be the sorts of big changes described here. But the scientists would not see them because both parent and child already have these differences.
Finally, the results show that DNA changes in lots of different places can cause autism. This is consistent with most of the research done before.
Finding autism-linked genes
When scientists try to find autism-linked genes, it looks like they are all over the place. Scientists have found that 20 out of 23 chromosomes have regions that may be important for autism.
Having so many candidates makes it hard to find single autism-linked genes. This is because of the way scientists usually find disease genes.
What they tend to do is look at the DNA of lots of people with a disease. And compare it to close relatives who don't have the disease. By doing this on lots of families, the scientists can figure out where important changes are.
But it all becomes much harder if there is more than one gene. And even harder if there are 10 or 20.
The research here is useful in finding autism-linked genes because it can show scientists where to look in the DNA. Scientists can focus on those areas where they found the changes in other autistic families.
From this they may find smaller DNA changes that might contribute to autism in other families. And by identifying the genes, new treatments may become available too.
When a scientists knows a gene is involved in a disease, that may provide clues for how to treat the disease. For example, there is a form of leukemia called acute promyelocytic leukemia.
In this cancer, two genes get stuck together. One of the genes is called the retinoic acid receptor. It makes a protein called the retinoic acid receptor.
We know how to shut that protein off. So we can treat the cancer by shutting off the retinoic acid receptor. Without knowing about these genes being stuck together, there is no way this treatment would have been tried.
Perhaps by finding autism-linked genes we may come up with useful treatments. Or at the very least with useful tests so we can catch the disease early. And start treatment for these kids as soon as possible when it can help the most.