Autism is likely to be linked to several genes

Researchers are close to identifying several genes that influence different aspects of autism.

By Hugh McIntosh

Ever since a study revealed that if one identical twin had autism, the other was likely to have it too, researchers have been searching for genes that cause autism. Now, after 20 years of looking, scientists believe they?re closing in on a handful of genes and chromosomal 'hot spots' that may be responsible for different aspects of the disorder.

Identification of specific autism-related genes would reveal the proteins the genes produce?knowledge that will boost researchers? ability to diagnose autism and to discover more effective treatments for the disorder, which is characterized by communication problems, social impairment, and unusual or repetitive behaviors.

The discovery of such genes has been hampered by the complex nature of autism. Because the symptoms of people with autism vary dramatically in degree and form, researchers believe the condition might involve two or more of a large number of genes. In fact, a person with autism may have mutations in several of perhaps 20 possible genes. Thus, two people with the disorder might have mutations in two completely different sets of genes.

In the past two years, scientists have identified several candidate genes. Some might alter the effects on the brain of neurotransmitters, others might compromise the immune system enough to allow viral infections that may cause autism, and another may influence embryonic development of the nervous system.

Chromosome 15

Among the most promising findings, say experts, are reports that an autism gene may be on the long arm of chromosome 15, near the centromere?an indented point that holds the two sides of a chromosome together. This region is a well-known spot for genetic abnormalities, including duplications of parts of the chromosome?s DNA. Short duplications cause no apparent harm. But longer duplications are associated with about a 50 percent risk for autism.

Last year, the research team of child psychiatrist Edwin Cook, MD, of the University of Chicago, reported that the autism risk associated with longer duplications appears to come through the mother.

Cook suspects that genes in this region of chromosome 15, which encode receptors for the neurotransmitter gamma-amino butyric acid (GABA), might be involved with autism. In fact, three genes for three GABA receptor subunits all are good candidates because they are associated with seizures and anxiety, which are common among autistic children, Cook says.

His team found moderately strong evidence for an association between one GABA subunit and autism?about one of every 70 children studied had a chromosome 15 duplication including this gene. Their study is published in the American Journal of Human Genetics (Vol. 62, No. 5, p. 1077?1083).

Duke University researchers say an autism gene might be a little farther away from chromosome 15?s centromere, just beyond the GABA receptor genes. A genetic screening of about 50 families with autistic children turned up three positive markers in this area, says molecular geneticist John Gilbert, PhD. An autism susceptibility gene may lie between these markers and the GABA receptor genes. Exactly what that gene might do is still unknown.

To examine the link between chromosome 15 duplications and autism more closely, researchers at the University of California, Los Angeles, hope next year to launch a nationwide study of 100 children with these duplications. Researchers will look for molecular differences that might explain why half the people with this duplication have autism and half do not, says geneticist Carolyn Schanen, MD, PhD. They will also try to determine whether autistic children with this duplication are more likely to have severe language problems.

'In the kids that I know?it essentially wipes out language to have this extra piece,' Schanen says. In the study, a psychologist will travel around the country to assess the children?s phenotype, as well as to collect blood samples for genetic analysis.

A serotonin transporter gene?

Researchers have long found that many autistic persons have elevated blood levels of the neurotransmitter serotonin. This finding suggests that people with autism have a defect in the gene that produces serotonin transporter?a substance that sweeps serotonin from the space between two nerve cells, thus ending its effect on the cells.

In the general population, the transporter gene occurs in either a long or a short form. Last year Cook and his colleagues reported finding that children with autism inherited the short form more frequently than expected, based on typical inheritance patterns. This pattern of inheritance, called 'preferential transmission,' suggests that the gene is a susceptibility gene?one that plays a role in whether a person gets the disorder.

Research groups in France and Germany, however, found preferential transmission of the long form. Though conflicting, these results represent some of the stronger evidence to date for a genetic role in autism.

However, Yale University neurochemist George Anderson, PhD, says that, to him, the findings of the three studies suggest that the serotonin transporter gene may not be a susceptibility gene. Rather, it may be a 'quantitative trait locus,' which affects the degree of a genetic trait in someone who already has the disorder.

Cook agrees with that possibility but offers another explanation for the findings: The three samples studied may have contained different mixes of autism subtypes. The two other studies, he says, may have enrolled higher proportions of children referred for treatment of severe behavior problems such as aggression. In contrast, his group?s sample contained a high proportion of people referred for communication or socialization problems rather than severe behavior problems. To test this possibility, his group has begun collecting data about aggression in people in their sample.

Nervous-system genes

Another autism candidate gene may lie among the genes involved with early development of the nervous system. This idea emerged from a Swedish study of 100 people whose mothers had taken thalidomide during pregnancy.

The study found that five of the 15 people exposed to thalidomide during days 20 to 24 of gestation had autism, says embryologist Patricia Rodier, PhD, of the University of Rochester. This suggests that the damage leading to autism occurred during the development of the hindbrain, long before the cortex and other parts of the forebrain developed.

These and other findings have led Rodier and her colleagues to look for mutations in several well-known developmental genes.

'If you have mutations in some of these genes that are critical in the early stages of the development of the nervous system, that in itself may be sufficient to cause the kinds of neuroanatomical changes that we think underlie autism,' Rodier says. 'But it could also be that that just makes you more sensitive to environmental agents at that time.'

For example, the investigators hypothesize that embryonic environmental factors?such as the presence of thalidomide?may act during early embryonic development on the genes being studied.

Immune deficiency?

Gene-environment interaction is the idea behind another candidate autism gene, called C4B, on the short arm of chromosome 6. This gene produces complement C4 protein, which works with the antibody immunoglobulin A to fight viruses. Deficiencies in either of these substances reduce the immune system?s ability to respond to viral infection.

'Autistic children are chronically ill, which is an indication that they have a deficiency in their immune system,' says immunologist Roger Burger, PhD, at Utah State University in Logan.

Immune deficiency might contribute to some cases of autism by allowing a virus to damage the brain or trigger an auto-immune response that causes brain injury, Burger says. Damage might also occur in utero if the mother has an immune deficiency. People with autism have an unusually high frequency of a form of the C4B gene that produces no protein.

Working together

While the University of Utah research team and others follow interesting leads, the National Institutes of Health (NIH) is funding five major collaborative groups to conduct genome screens of families with autism. Earlier this year, one of these consortiums reported in the journal Human Molecular Genetics (Vol. 7, No. 3, p. 571?578) that they?d found a hot spot on chromosome 7.

'We think it?s a susceptibility gene,' says geneticist Anthony Monaco, MD, PhD, of the University of Oxford. 'But what gene?or what type of gene, we really have no idea.' The suspect region includes genes expressed in the brain during development.

Supporting their finding is evidence from another collaborative group headquartered at Duke University. The researchers screened about 50 families and turned up 'a number of interesting regions' on chromosome 7, says Duke geneticist Margaret Pericak-Vance, PhD.

A group at the University of Iowa has evidence for a link between autism and chromosome 13. The group is also exploring the idea that there is a broad autism phenotype that includes people with milder autism-like symptoms, as well as those with classic autistic disorder.

'We?re seeing some pretty large pedigrees where there are maybe two autistic kids and an autistic first cousin, and then in between a lot of people who we think have the broad phenotype,' says child psychiatrist Joseph Piven, MD. 'These pedigrees?look more like single-gene disorders' than multigene disorders.

Although a single gene might explain autism in one family, the number of tantalizing prospects turning up from the genome screens suggests several genes are involved in autism, says Marie Bristol-Power, PhD, coordinator of the NIH autism network. She notes that the five collaborative groups together are expected to enroll more than 1,000 families with autistic children, generating plenty of statistical power to root out the genetics of this complex disorder.

Hugh McIntosh is a writer in Chicago.