With a three-year $3.2 million grant from the National
Institute of Mental Health, Johns Hopkins scientists will
lead the largest hunt for genetic contributors to autism, a
neuropsychiatric condition whose causes are almost as
mysterious today as when the condition was first described
in 1943.
The researchers will apply new genome-searching
technologies to available samples and information from 465
families, including 979 individuals with autism, to
identify genetic factors that contribute to the
condition.
"Autism is quite likely to result from the combined
effects of multiple, very subtle genetic changes that
differ considerably from family to family, since no single
reliable genetic cause has been found yet," says Aravinda
Chakravarti, principal investigator of the project and
director of the
McKusick-Nathans Institute of Genetic
Medicine at Hopkins. "We'll be looking for combinations
of genetic mutations and extra or missing gene copies that
are much less common, even in the affected group, than most
scientists are used to considering. This is a huge
undertaking."
Recent research suggests that as many as one of every
500 births is affected by autism, which is characterized by
social and communication deficits and restricted and
repetitive interests. Understanding the condition's genetic
roots may reveal important clues to its biology and hence
targets for treating some of its effects or trying to
prevent it.
"The molecular genetic study of autism provides one of
the best scientific opportunities in medicine: the chance
to identify the missing or abnormal signals that prevent
full development of a small set of nerve cells in the
brain," says project member Edwin Cook, director of the
University of Chicago's Laboratory of Developmental
Neuroscience and a child and adolescent psychiatrist who
sees patients with autism. "Once we know the disrupted
signals, we can begin the process of rational development
of medical treatment of autism."
One aspect of the new project is to use "DNA chip" or
microarray technology from project collaborator Affymetrix
that will broaden the hunt for genetic changes behind
autism. With these new microarrays, designed to detect the
specific building blocks present at about 500,000 locations
scattered throughout the genome, the researchers hope to
find specific single changes in the genetic sequence
— changes known as single nucleotide polymorphisms or
SNPs (pronounced "snips") — either associated with
autism or perhaps with the condition's risk, severity or
constellation of symptoms.
"Conventional family studies have failed to detect
autism genes," Chakravarti says. "Our planned studies are
far more powerful at gene detection and will give a more
complete assessment across all human genes. Because autism
is so genetically complex, we're talking about identifying
genes that might contribute to only a few percent of
cases."
A second part of the upcoming studies is to look for
genes that are present in extra copy numbers. Normally,
each cell carries two copies of each gene, one from the
mother and one from the father, each on its own chromosome.
But some preliminary evidence suggests that autism might
stem in part from "dose defects" — extra or fewer
copies of genes that arise because sections of chromosomes
are improperly copied and inserted into or deleted from the
genome. Such events could alter normal production of the
genes' products, which might damage cells, Chakravarti
says.
But unlike those with disorders such as Down syndrome,
people with autism don't have extra chromosomes, just
— and only perhaps, at this point — extra
segments of chromosomes. So existing techniques to count
and look at the structure of intact chromosomes aren't
enough.
Instead, the team will use digital karyotyping, a
technique recently developed by project member Victor
Velculescu, assistant professor of oncology in the
Johns
Hopkins Kimmel Cancer Center. The researchers will use
about 400,000 short stretches of DNA to detect extra copies
and identify any differences between family members with
autism and those without. The techniques will also be used
to see whether any extra segment copies were inherited
— passed from the parents to the child — or
appear spontaneously in the child and aren't found in the
parents' genetic material.
To find answers in all these data, David Cutler will
lead the project's computational biology aspect. His team
will develop some new computer programs and apply some of
their existing ones to sift through the data to find and
"score" associations between traits and the results of
genetic studies. Most studies searching for human disease
genes have tended to look at 1,000 or fewer markers because
no one was sure they'd be able to make sense of the results
of bigger studies.
"Now we know we can do it," says Cutler, an assistant
professor in the McKusick-Nathans Institute of Genetic
Medicine.
Leaders of the project's various components are
Chakravarti, Velculescu, Cutler and Dan Arking, all of
Johns Hopkins; and Cook, of the University of Chicago. The
family samples and information are publicly available from
the NIMH-funded Autism Genetics Initiative Data Archive and
the Autism Genetic Resource Exchange.