A new piece in the autism puzzle

Image by Broad Communications, courtesy of iStockPhoto

Autism is a frustrating condition for those who live it and for the researchers who seek to unravel it. Because the developmental disorder tends to run in families, genetic factors are likely to play a role in disease risk, but the quest to locate the responsible genes has been mostly fruitless. A new genome-wide scan led by researchers at the Broad Institute of MIT and Harvard and members of the Boston-based Autism Consortium has added a piece to the autism puzzle by identifying a region of the genome that is missing in some people with autism and duplicated in others. The findings, published in the January 10 issue of New England Journal of Medicine, help explain roughly 1% of autism cases, and contribute to the ongoing effort to lay bare the disorder’s genetic underpinnings.

Autism is a cognitive disorder with complex effects on behavior, such as social disability, impaired communication, repetitive behaviors, and restricted interests. Roughly 10% of cases stem from rare genetic conditions that involve either known DNA mutations or big chunks of the genome that are either present in excess or missing altogether. But the genetic factors behind the majority of autism cases remain hidden. On one hand, the presence of large genomic deletions or duplications in some cases of autism suggests that smaller mutations of this type might play a role in many of the unexplained cases. On the other hand, autism’s high heritability suggests that disease risk might instead be conferred by more common forms of genetic variation, such as single letter DNA changes, known as single nucleotide polymorphisms (SNPs). “Given what we know about autism, the truth has to lie somewhere between those two ends of the spectrum — many different forms of genetic variation are likely involved,” said Mark Daly, an assistant professor in the Center for Human Genetic Research at MGH, a senior associate member at the Broad, and senior author of the new study.

Daly and his colleagues decided to approach the problem from both angles using a new generation of Affymetrix gene chips. Developed in collaboration with several Broad scientists, these chips helped to characterize both SNPs and variations in the number of gene copies in the genomes of autistic children and their families. Compared to previous versions, the new chips have been enhanced to accurately identify both SNPs and small stretches of DNA that vary in copy number, Daly explained. In fact, the researchers waited to start the autism scan until these newer chips were available, placing the work among the first large-scale studies of copy number variation across the genome.

The approach is now starting to bear fruit. After scanning the genomes of over 750 families from the Autism Genetic Resource Exchange database, the researchers observed that a section of chromosome 16, roughly 600,000 nucleotides, or genetic “letters,” long, was missing in the genomes of five autistic children. By contrast, the deletion was found in the genomes of only five people among a total control sample of over 23,000 individuals, making the mutation a strong risk factor for autism.

The researchers gleaned other clues from the wealth of information in this large-scale genome scan. Because the missing DNA was intact in each child’s non-autistic parents, it appears to be a spontaneous mutation that is rarely passed down from generation to generation. As mutations go, the deletion arises relatively frequently, but strangely enough, is not found at a high level in the general population. This means that the deletion on chromosome 16 must be somehow strongly disfavored by evolution. A likely explanation is that the deletion’s behavioral effects make it unlikely for carriers to have children, so the mutation is not often passed on.

The ability to scan large numbers of people with the new chips was a critical factor in the study, because of the small number of carriers. If the study had been smaller, the one or two patients found to have the deletion would not have passed statistical muster. The newly identified deletion also sheds light on the complexity of autism. Though it is a strong risk factor for developing the disease, it also exists in non-autistic individuals, which means the relationship to disease is not one of simple cause-and-effect.

“The new high-throughput genome-wide studies have the power to expand our thinking about how genetics can contribute to disease,” said Daly. “It takes a simplistic model – that a mutation causes the syndrome – and adds layers to it because we can detect rare instances of the mutation not causing the syndrome.” Another layer is added to the story with the finding that some autistic children in the study had extra copies of the same section of DNA found missing in other autistic children. This result suggests that the causative proteins must be present in a narrow range to achieve normal cognitive function.

These discoveries are among the very first genetic pieces of the autism puzzle. Taken together, the deletions and duplications may account for more than 1% of autism cases, but the section of DNA in question contains 25 genes, of which one or more may have a true biological connection to autism, with others merely along for the ride. Working with researchers from throughout the Autism Consortium, a Boston-based group of scientists, physicians, and families committed to supporting families and promoting advances in autism research and treatment, Daly and his research team are searching for the relevant genes, hoping to understand how they rob the autistic individual of social and cognitive skills. Also, because the deletion is not usually inherited, it does not help explain why the disease runs in families. It does, however, help scientists zero in on genes that might be modified in more subtle ways that could account for the disease’s heritability in other families with autism.

Once the genes contributing to autism are identified, laboratory testing may help unravel the biological mechanisms behind the disorder and could one day provide a route for developing novel, effective treatments. For now, the copy number discovery is helping out in another realm — autism diagnosis. David Miller, an author on the paper and a clinical geneticist at Children’s Hospital Boston, tests for the deletion in children with developmental delays brought into his clinic. Presence of the deletion or duplication can help him inform parents on the cause of their child’s autism, and it helps parents understand the risk of their future children also having autism. Miller hopes that future discoveries on autism’s genetic causes may lead to a customized autism gene chip that can describe the spectrum of genetic changes likely to contribute to the disease.

In addition to identifying copy number variants, the genome scan characterized SNP variation in the samples. Clues about autism’s heritability may come from the ongoing SNP analysis, led by first author Lauren Weiss, a postdoctoral researcher with the Broad’s Psychiatric Disease Initiative who also led the copy number analysis. “This scan was designed to find both common SNP and copy number variation of modest impact and rare, high impact copy number variation in autism, and the chromosome 16 deletion is the first low-hanging fruit to come out of it,” said Weiss. “We hope there will be many more discoveries.”

Other contributors to the autism study include David Altshuler, director of the Medical and Population Genetics Program at the Broad Institute, associate professor of genetics and of medicine at Harvard Medical School and in the Department of Molecular Biology and Center for Human Genetic Research, in addition to several associate members at the Broad Institute: Christopher A. Walsh, the Bullard professor of neurology at Harvard Medical School and Beth Israel Deaconess Medical Center, chief of the division of genetics at Children’s Hospital Boston, and an investigator at the Howard Hughes Medical; James Gusella, director of the Center for Human Genetic Research at MGH, Bullard professor of neurogenetics in the Department of Genetics at Harvard Medical School; and Pamela Sklar, associate director of the Psychiatric and Neurodevelopmental Genetics Unit in the Center for Human Genetics Research at MGH, and director of neuropsychiatric genetics and associate professor of psychiatry at Harvard Medical School.

Broad Institute scientists contributing to the study also include Joshua Korn, Manuel Ferreira, Todd Green, and Douglas Ruderfer.