Skip to main content

Spectrum: Autism Research News

Grouping similar traits in autism uncovers new genetic culprits

by  /  11 August 2008
THIS ARTICLE IS MORE THAN FIVE YEARS OLD

This article is more than five years old. Autism research — and science in general — is constantly evolving, so older articles may contain information or theories that have been reevaluated since their original publication date.

Group theory: Some genes may have bigger effects in people with specific physical traits, researchers say.

Sorting diverse autism cases into subgroups that share similar traits ― such as language delay or intelligence quotient (IQ) ― reveals two genomic regions associated with the disorder, researchers have found.

The study, published online in July, is the second from the Autism Genome Project, a multisite effort to identify autism-related genes1. The groupʼs first study implicated several chromosomal regions by looking across the spectrum of autism disorders in 976 ‘multiplexʼ families that have two or more children diagnosed with autism.

The new study takes a finer-grained approach by grouping autism cases according to particular traits or ‘sub-phenotypesʼ that may result from the same genetic abnormalities.

“The idea is that not everybody with autism has the same genetic risk factors,” says Andrew Paterson, a geneticist at the Hospital for Sick Children in Toronto, who led the study. “Some of those different genes may have larger effects in subsets of individuals with particular phenotypic characteristics.”

The researchers used IQ scores and delayed onset of language to create several subgroups from more than 2,000 individuals with autism. They also used scores from the Autism Diagnostic Interview-Revised (ADI-R), a standard diagnostic tool, to quantitatively measure two separate aspects of autism: difficulty with social interactions and repetitive behaviors.

Using a technique called linkage analysis – which helps localize the genes for a trait to a particular chromosomal region – the researchers identified a region on chromosome 15q in a subgroup of those with an IQ score of at least 70, the standard cutoff for mental disability.

They found another region on chromosome 11p in the subgroup of those who didnʼt speak their first phrase until after 36 months of age.

In part because of the large sample size, the signals pointing to these chromosomal regions are unusually strong, lending the results more credibility than those from other studies of this kind.

“This is very exciting news,” says Maricela Alarcon, assistant professor at University of California, Los Angeles, whose own study of 133 families divided by language traits implicated the same regions, but with less certainty2.

Fine print:

Each of these regions could contain hundreds or thousands of genes, says Paterson. The next step is to pinpoint any genes relevant to autism.

For example, the region associated with the IQ subgroup may contain genes related to IQ, to autism or both. “It may be that IQ is just working as a marker for some other aspect of the phenotype that may not be as easy to measure using questionnaires,” says Paterson.
Because many aspects of autism occur together, it can be difficult to separate those traits, he says.

brain scan chromosome 22

Grouping-Traits-Img2.jpg


Signs of intelligence: Linkage analysis reveals an association with chromosome 15 for an IQ of 70 or above, and with chromosome 11 for language delay.

Choosing the traits for the subgroups is also complicated because the information is collected at more than ten different sites, which may differ in how they measure the traits. To minimize the differences, the researchers picked broad categories, which some researchers say are too coarsely defined.

“The phenotypes that they selected for investigation, or that they were basically limited to, are suboptimal,” notes Susan Santangelo, a genetic epidemiologist at Harvard Medical School.

Instead, Santangelo says, a more powerful approach would have been to quantify a trait on a numerical scale.

Although the study used two quantitative traits ― social impairment and repetitive behavior ― as measured by ADI-R, the researchers did not find any chromosomal regions for those traits. “The quantitative traits basically didn’t help them at all,” says Santangelo.

Quantitative traits are most useful when they can be assessed in people both affected and unaffected by a disorder, she notes. This gives a wider range of scores, which increases the potential for linkage analysis to detect the genes involved.

The ADI-R, however, cannot be meaningfully applied to control participants. “[They] would just zero out on most of those measures,” she says.

Another measure, the Social Responsiveness Scale, can be used to test both those with autism and controls. That scale is relatively new, however, and researchers are just beginning to use it to assess people whose samples are stored in gene banks such as the Autism Genetic Resource Exchange. One study that used the scale last year localized autism-related genes to chromosomes 11 and 173.

References:


  1. Liu X.Q. et al. Biol. Psychiatry Online publication ahead of print (2008) PubMed 

  2. Spence S.J. et al. Am. J. Med. Genet. B. Neuropsychiatr. Genet. 141, 591-598 (2006) PubMed 

  3. Duvall J.A. et al. Am. J. Psychiatry 164, 656-662 (2007) PubMed 


TAGS:   autism