Science —

New genetic mutations trigger many cases of autism

Many cases of autism in families without a history of the disease appear to be …

Viewed from some perspectives, autism is a genetic disease. Studies of identical twins show that, when one sibling is affected, the other is also, with a frequency approaching 90 percent. Several genes associated with autism have been identified due to their prevalence in individual families with high rates of autism. All of that, however, doesn't help explain why there are so many of what appear to be sporadic cases of autism, where a single individual in a family has the disorder; these account for the majority of all cases of autism.

Over the past several years, a case has been building that many of the sporadic cases are also genetic, the result of brand new mutations that appear only in the affected individual. A set of studies that appear in the journal Neuron make a compelling case that these new mutations can also account for a significant fraction of the total incidence of autism, but still aren't common enough to fully account for the incidence of the disease.

The new studies rely heavily on a project developed by a private foundation, the Simons Simplex Collection. The Simons Foundation has worked with academic researchers to obtain DNA and cell samples from thousands of unaffected parents and their autistic children; unaffected siblings are also included. This has allowed researchers to screen for changes in the DNA that are unique to the affected individual. In this case, the papers looked for what are called copy number variations, or CNVs, which occur when a chromosome has either ended up with an extra copy of part of the genome, or seen a part of the genome deleted. CNVs will often contain genes, so they create differences in the dose of these genes.

The studies found slightly different numbers, but the same general trend: new CNVs (deletions or duplications that are not present in either parent) are much more common in autistic children than they are in their unaffected siblings. One study found that they showed up 3.4 times more often; the other showed 3.9 times. In either case, there were many more changes in the dose of genes among autistic individuals. Moreover, most of these changes are rare, showing up in only one or a few individuals. This low frequency implies that, with a larger population, even rarer CNVs would be uncovered. The statistics are such that the authors estimate that there may be several hundred potential sites in the genome that could be linked to autism.

A number of areas that were previously associated with autism came out of the screens, but many of the sites were brand new. In one case, the authors found duplications of a region produced autism; previously, others had found that deletion of this region produced a disorder where affected individuals are unusually gregarious.

An accompanying paper looks at what the genes present in these CNVs do, In general, they seem to be involved in what you'd expect: synapse formation, the internal structure of nerve cells, and the generation of the axons and dendrites that help nerve cells form connections with one another. Thus, the findings help support the general picture that autism is associated with problems with wiring the brain up, rather than the production or function of nerve cells.

Where do these new results leave us in terms of the root causes of autism? They certainly support the general idea that even sporadic cases of autism are caused by new mutations. But on their own, CNVs don't seem to be occurring at a rate that is sufficient to account for all the sporadic cases of autism. There is the chance that many of the same genes we see within the CNVs, however, are damaged by smaller mutations, including single base changes, that can't be detected by the techniques used in these studies. Identifying these sorts of mutations, however, will probably require whole-genome sequencing of thousands of individuals, so we're unlikely to see it in the near future.

The studies also leave open the possibility of environmental influences. Many mutations associated with autism show what's called "variable penetrance," meaning that they may affect some individuals severely, but leave others without any obvious or serious symptoms. The differences could be the product of environment, and will make identifying even inherited cases much more challenging. The second area where environment could come into play is the mutation rate itself. Many environmental factors can cause or promote the accumulation of DNA damage, which can produce the sorts of new mutations seen in these studies.

Overall, the new studies strengthen the argument that autism is primarily a genetic disorder, and help explain why it has been so hard to identify the genes: most mutations are rare or completely unique to the affected individuals. The researchers involved are already screening more individuals from the Simons Simplex Collection, and preparing to start sequencing DNA from the same collection, so we should have an even clearer picture in the near future.

Neuron, 2011. DOI: 10.1016/j.neuron.2011.05.015, 10.1016/j.neuron.2011.05.002, 10.1016/j.neuron.2011.05.021  (About DOIs).

Channel Ars Technica