Study seeks to pinpoint genes related to autism

A study just getting under way will help CHDD researchers answer these questions and provide a key piece for solving the biological puzzle underlying autism. The study is led by CHDD research affiliate Dr. Gerard Schellenberg, research professor of medicine, neurology and pharmacology, and is the second major component of the autism program project. Its long-term goal is to identify the genes that predispose an individual to autism.

To achieve this goal, Schellenberg and his colleagues, Dr. Ellen Wijsman, research associate professor of medical genetics, and Dr. Wendy Raskind, associate professor of medicine and a CHDD research affiliate, will analyze genetic information from families in which at least two siblings have autism. Schellenberg and Wijsman will focus on DNA sequences, and Raskind will look for chromosomal abnormalities. Although autism is not rare, families with two siblings who have the disorder are. To assemble the 200 families needed for the study, families from across the country will be invited to participate. Siblings of any age with autism are eligible. Genetic testing will take place at universities in Oregon, Alaska, Montana, Tennessee and Florida, as well as at the UW.

Sibling volunteers will undergo a full diagnostic workup conducted by the clinical core of the program project, led by Dr. Geraldine Dawson. This will include a battery of neuropsychological tests to evaluate possible subtypes of autism. Identifying subtypes could yield important clues to unraveling the question of genetic heterogeneity in autism­whether more than one gene abnormality leads to the same symptoms.

Initial evidence of autism's genetic nature came from family history research exploring statistical risk for the disorder. These studies have shown that when compared with the likelihood of an individual in the general population having autism, the chance of having autism is substantially greater for an individual who has a sibling with autism. The sibling risk of autism is 50 to 100 times greater than the risk observed in the general population­a difference that implies genetic factors, says Schellenberg.

Studies of twins further bolster the genetic hypothesis. The occurrence of autism is much greater in monozygotic twins who carry an identical set of genes than in dizygotic or fraternal twins whose genetic blueprints are usually quite different from each other. In the vast majority of identical twin pairs, if one twin has autism, the other twin also has autism. For fraternal twin pairs, the chances of both having autism are much like those of two siblings. Geneticists refer to the presence of a particular trait in both members of a twin pair as concordance.

"In the literature, the concordance rate of autism in monozygotic twin pairs ranges from nearly 70 to more than 95 percent," notes Schellenberg. "That's very high, although it doesn't appear to be absolute. It depends a little bit on how you define autism. The looser the criteria, the more concordance there seems to be."

The high concordance rate is good indication of the predominant role of genes in autism. However, Schellenberg points out, a concordance rate less than 100 percent suggests some additional, as yet unidentified, environmental factors may play some role.

The mode of inheritance is also an open question, although the disparity between the rate of occurrence in identical twins and in fraternal twins and siblings does suggest that multiple genes must interact to produce the disorder. "There might be six or more genes that have to come together in one individual in the right form to lead to autism," Schellenberg says. "Identical twins inherit all those same genes, but since you only get half of your genes from each parent, the chances of two non-twin siblings who are not so genetically similar getting six of the exact form of the genes necessary for autism becomes much smaller. That's why you don't see many families with multiple occurrences of autism."

Schellenberg's search for autism-related genes will encompass all chromosomes in the human genome, which contains an estimated 50,000 to 100,000 genes. Genes are discrete segments of the DNA molecule made up of specific sequences of the four chemical building blocks that constitute the molecule's backbone. Each gene provides the chemical code to produce a protein that is either part of the body structure or serves a function in the body's myriad chemical activities.

Many genes that code for single traits have been located based on knowledge of the protein they encode. However, in autism, as in most other complex genetic disorders, the identity of the proteins involved and their function is unknown. So researchers must use methods that rely heavily on statistics and pattern matching to locate genes. They compare the DNA of related individuals who show the traits determined by the gene and look for shared DNA sequences.

"We're really looking for anything that's inherited, and matching the pattern of inheritance in the family to the pattern of disease in the family," notes Schellenberg. He and Wijsman are very familiar with the careful study design and complex statistics required in this type of research. They have conducted a number of investigations using this method known as linkage analysis, including studies identifying two of the genes associated with Alzheimer's disease.

Researchers will search chromosomes of individuals with autism for shared DNA sequences.

To zero in on candidate sites for autism along the genome, Schellenberg and his colleagues will map polymorphic markers­well-known, distinct DNA sequences scattered on all chromosomes. Although these markers have no health manifestations, they are useful in genetic research because their form varies among individuals. The markers provide reference points for narrowing the search.

DNA for mapping will be isolated from blood samples collected from study participants­siblings with autism and their parents. "In linkage analysis, the first thing to determine is what form of these markers each person in the family has," explains Schellenberg. "If the affected pair share particular forms, more than you would expect by chance, and if that holds true in a large number of these sibling pairs, then the marker is in the vicinity of the gene. It's not the gene itself, but it is in the vicinity of the gene."

Comparing genetic material of a sufficient number of individuals will yield a pattern of shared markers leading to the discrete DNA sequences that are the genes themselves. Once the genes are located, Schellenberg and his colleagues intend to clone them so the proteins they encode can be determined, providing crucial insight into the fundamental cause of autism.

Top of Page | Next Story | CHDD Home