Two Mutations on Chromosome 16 May Cause 1% of Autism

A novel, recurrent microdeletion, as well as a reciprocal microduplication, on a specific region of chromosome 16 appears to confer substantial susceptibility to autism, researchers have reported in the New England Journal of Medicine.

The two mutations might account for approximately 1% of cases of autism, a frequency that is the same as that for the most common known cause of autism (duplication of the Prader-Willi/Angelman region), according to Lauren A. Weiss, Ph.D., of Massachusetts General Hospital's Center for Human Genetic Research, Boston, and her associates.

The investigators conducted a genomewide analysis of a sample of families in the Autism Genetic Resource Exchange to identify any recurrent deletions or duplications that conferred risk of autism in multiple families. One region on chromosome 16p11.2 carried deletions in four independent families with five autistic children. The same region showed duplication rather than deletion in another three independent families with seven autistic children.

The researchers then attempted to confirm their findings by searching for the same microdeletion in a genetic database of 512 children with autism, developmental delay, or mental retardation from Children's Hospital Boston. They found identical deletions with exactly the same boundaries in five autistic boys.

In contrast, they found no such deletions in a sample of 434 Children's Hospital patients who had undergone genetic testing because of dysmorphic features, multiple congenital anomalies, congenital heart disease, seizures, or other disorders unrelated to autism. Similarly, there were identical duplications at 16p11.2 in four children in the sample with autism, developmental delay, or mental retardation, but none in the sample of children with disorders unrelated to autism.

Dr. Weiss and her associates then tried to replicate their findings in a genetic database of more than 19,000 members of the general population in Iceland. They found three cases of the 16p11.2 deletion in the subgroup of 299 subjects who had autism or developmental disorders, “a finding that was consistent with the 1% frequency observed” in the Children's Hospital cohort with autism or developmental disorders.

In contrast, only two cases of the deletion were found in the 18,834 Icelandic control subjects, a rate that was 1/100th of that in the autistic Icelanders.

The microduplication mutation was not seen in any Icelandic subjects who had autism but was found in two people with bipolar disorder and in five unscreened control subjects.

Interestingly, a separate study of the Icelandic cohort showed that the deletion occurred “at a markedly increased rate” in people who had various language or psychiatric disorders. It was found in one subject each with schizophrenia; bipolar disorder; attention-deficit hyperactivity disorder; panic disorder, anxiety, depression, or addiction; and dyslexia.

“In total, we have observed the identical deletion of nearly 600 kb [kilobase] in 13 subjects with autism or developmental or language delay (10 confirmed de novo mutations, 2 confirmed inherited mutations from parents with ADHD or mental retardation, and 1 mutation of unknown inheritance), with the reciprocal duplication of the same region documented in 11 additional subjects,” said Dr. Weiss, also with the Autism Consortium, and her associates (N. Engl. J. Med. 2008 Jan. 9 [doi:10.1056/NEJMoa075974

“The fact that [the deletion mutation] is seen extremely rarely in the general population not only establishes a significant difference between rates in autism and control populations, but also unambiguously establishes that strong natural selection is acting against transmission of this deletion (as might be expected from an allele that increases the risk of autism by as much as a factor of 100), given how often it arises de novo in a single generation,” added the researchers, whose study was funded by several organizations, including the Autism Consortium. The researchers reported no conflicts of interest.

In an accompanying editorial, Evan E. Eichler, Ph.D., and Dr. Andrew W. Zimmerman said that after larger case-control groups have undergone genotyping, it is possible that some of the 50 other large de novo events observed by Dr. Weiss and her colleagues and other events described in recent studies might be specifically associated with autism. Dr. Eichler is affiliated with the University of Washington, Seattle; Dr. Zimmerman is with the Kennedy Krieger Institute and Johns Hopkins University, Baltimore.

“The discovery of significant associations for the rarer loci may require the screening of tens of thousands of DNA samples from patients rather than a few thousand,” they wrote [doi:10.1056/NEJMe0708756

“Deeper sample collection and new cost-effective genomic techniques may be needed to peel away the remaining layers of the onion,” they added.