By Julia Karow
A significant portion of autism may be caused by de novo mutations that were not inherited from either parent, according to a new trio-based exome sequencing study by researchers at the University of Washington School of Medicine.
The study, published online in Nature Genetics this week by the groups of Evan Eichler and Jay Shendure, found mutations in some of the same genes that have been implicated in other neuropsychiatric disorders.
The researchers sequenced the exomes of 20 patients with sporadic autism spectrum disorder and their parents. In four of these patients, they found potentially causative de novo mutations, including mutations in genes that have previously been associated with autism, intellectual disability, and epilepsy.
"It's an interesting paper. It's where the field is going, and it's a first taste of what we are going to see a lot of this year," said Joseph Buxbaum, director of the Seaver Autism Center at Mount Sinai School of Medicine. Buxbaum is involved in another study, also involving Baylor College of Medicine and the Broad Institute, that sequenced the exomes of 1,000 autism cases and the same number of controls and about 150 family trios and expects to publish initial results on the trios soon.
Another study, headed by researchers at the Hospital for Sick Children in Toronto, is sequencing the exomes of 1,000 Canadian autism patients.
According to Buxbaum, there is mounting evidence that many mutations in genes involved in autism and other brain diseases with early onset interfere with reproductive fitness and are unlikely to be passed to the next generation. It is thus likely that they occur de novo, which the results of this study seem to confirm.
A trio exome sequencing study on intellectual disability by Dutch researchers last year came to the same conclusion — it found that a large fraction of previously unexplained cases can be attributed to de novo mutations in protein-coding genes (IS 11/16/2011).
Because in family trio studies, the parents serve as controls to filter out benign variants, the University of Washington researchers were able to get meaningful results from a relatively small number of exomes. "This is not going to explain all of autism" or other neuropsychiatric diseases, said Shendure, an assistant professor of genome sciences at UW, "but it provides a potentially efficient mechanism for explaining some fraction of those diseases and identifying candidate genes."
For their study, the researchers analyzed samples from 20 families with a single child that was diagnosed with autism spectrum disorder, most of them from the Simons Simplex Collection.
Using Roche NimbleGen's SeqCap EZ Exome in-solution capture method, the scientists enriched exonic DNA and sequenced it on an Illumina Genome Analyzer IIx with single or paired-end 76-base pair reads, which yielded sufficient coverage to call variants in about 90 percent of the target sequence.
Besides filtering out parental variants, they also filtered those variants previously observed in dbSNP, the 1000 Genomes pilot project, and about 1,500 exomes previously sequenced at the University of Washington. All remaining de novo candidate mutations were validated by Sanger sequencing.
[ pagebreak ]
Overall, the scientists identified 21 de novo mutations, about half of them protein-changing. Four patients carried potentially causative de novo mutations — in the genes FOXP1, GRIN2B, SCN1A and LAMC3 — based on the gene being previously implicated in autism or a related disease, or in a biological pathway involved in autism.
According to Shendure, it is impossible based on this study alone to estimate what fraction of autism is sporadic, or caused by de novo mutations. For one thing, the sample size was too small. Furthermore, some of the other de novo mutations they found may also be causative but cannot be identified as such yet, while some causative de novo mutations might have been missed because they reside outside the exome or in portions of the exome that were missed.
The study's results may provide insights into the genetic basis of inherited autism as well, Shendure said, because some of the same genes may be involved both in sporadic as well as inherited forms of the disease.
He and his colleagues are now scaling the study up to a larger number of trios, also provided by the Simons Simplex Collection, though he declined to mention how many.
Buxbaum said that the study he is involved in only found one known disease gene in 100 autism family trios — an indication that there might be many more autism genes that have yet to be discovered.
While both studies are research studies, they could pave the way to an improved molecular diagnosis of the disease. Array-CGH can already provide a molecular cause in 15 to 20 percent of autism patients, Buxbaum said, and the approach is recommended by both the American College of Medical Genetics and the American Academy of Pediatrics.
Autism patients can benefit from a molecular diagnosis in several ways. For a start, families "often want to know, they want to understand, and it helps them," he said. Also, families with patients who have mutations in the same gene sometimes form advocacy groups in order to help push the science forward and learn about "clinical issues that they might have to face," he said.
Studying patients with the same or related mutations as a group might also be more meaningful than classifying them purely by phenotypes, which tend to be broad, Shendure said.
In addition, a molecular diagnosis can help counsel families, for example about the risk of additional children also developing autism.
And long term, genes related to autism can provide new drug targets and ultimately lead to new therapies, Buxbaum said. His own center, for example, has been studying the SHANK3 gene, mutations in which can lead to autism, in mouse models for two years, and is talking to pharmaceutical companies about clinical trials.
Have topics you'd like to see covered by Clinical Sequencing News? Contact the editor at jkarow [at] genomeweb [.] com.