Discovering the genetics of autism. (Medicine & Health).

NOT TOO LONG AGO, the parents of a child who was diagnosed with autism were overwhelmed with disheartening information. They were told that their offspring might have to be institutionalized, that he or she would never be able to be a productive member of society, that nothing could be done to change this prognosis, and, worst of all, that they themselves were likely to blame for their child's condition. Doctors and scientists knew little about autism, and the lack of a clear diagnosis or any real understanding of its cause led to finger-pointing and misinformation.

Today, while autism is far from completely understood, the future for families of newly diagnosed youngsters is hopeful. Scientists have determined that autism is most likely caused by a combination of genetic and environmental risk factors. Poor parenting has been ruled out as a contributing factor.

Autism is a generic term that designates several more-specific disorders, known as pervasive developmental disorders, or PDDs. According to a website developed at Duke University's Center for Human Genetics, "Exploring Autism: A Look at the Genetics of Autism" (www.brilliantcontent.org/autism/), a child with PDD has a pattern of delayed or unique language development and impaired social interaction when compared to one without the disorder. Specifically, individuals with PDD show symptoms in three areas: socialization (interaction with others), communication, and repetitive behaviors. One of the most-difficult aspects of autism for family members is the struggle for emotional connection with the affected child. Those with autism have particular difficulty interpreting the variety of social cues people use to infer how another person thinks or feels. This may be part of a larger problem with generalization. Children with autism can often learn discrete tasks or follow specific sets of direction fairly easily, but characteristically do not apply what they learn to other circumstances.

Research into the causes of autism has taken off over the last few years. More resources are being directed towards understanding and treating it than ever before. However, understanding its causes is an enormously complex undertaking, as researchers who are attempting to identify the genes that may increase an individual's risk of developing autism can attest. Some of this complexity comes from the fact that autism is often difficult to diagnose. Nearly 60 years have passed since Leo Kanner first described autism as a distinct clinical entity in 1943, and the definition of the disorder continues to be refined.

What is known is that autism describes a set of neurodevelopmental problems which result in significant disturbances in social, communicative, and behavioral functioning. Autistic disorder and the other PDDs (Rett disorder, Asperger disorder, childhood disintegrative disorder, and PDD-not otherwise specified) are clinical diagnoses, which are made based on developmental history and behavioral observations. The onset is in early childhood, usually before the age of three, and symptoms persist throughout adult life. Scientists estimate that the best-known of the PDDs, autistic disorder, occurs in between two and 10 out of 10,000 children, which makes it almost as common as Down syndrome. In addition, male children are three to four times more likely to have autistic disorder than females. Having a clear description of the symptoms is the first step in identifying its genetic basis, because it allows researchers to define clearly autistic patients in the population. Once they are sure they can determine who has autism, scientists can begin the search for the specific genes involved.

Evidence that some forms of autism have a genetic basis was first published in the 1970s. Since then, data from multiple studies have confirmed this hypothesis. These reports included observational studies, such as case reports and observation of families with more than one autism patient, usually siblings. As genetic research became more sophisticated, chromosomal and statistical evidence from twin studies also emerged, further supporting the genetic contribution to the development of autism. These studies show that autism occurs significantly more often in identical twins than in nonidentical twins. Since identical twins share all of their genetic material, this finding points strongly to a genetic component for autism.

Genetic research is one of the most-promising approaches to the diagnosis and treatment of disease today. This type of work has been jump-started by the completion of the Human Genome Project, giving scientists their first complete map of the 30,000 to 50,000 genes in the human body. The genetic basis for single-gene disorders is somewhat easier to determine than that of complex ones such as autism. Single-gene diseases, called Mendelian diseases, are passed on from generation to generation through a mutation (or change) on one specific gene. Genes themselves are small structures inside almost every cell of the human body that contain the blueprints of human life and tell bodies how to grow and develop. Genes determine many physical characteristics, such as hair and eye color and blood type. A well-known example of a trait with Mendelian inheritance is cystic fibrosis (CF), the gene for which was identified in 1989. Now, individuals who may have a family history of the disease can undergo genetic testing to see if they carry the mutation that causes CF. If both parents are found to carry it in their genetic material, geneticists can determine prior to birth whether a child conceived by those parents has CF.

Much more common in the population are complex genetic disorders. In them, the gene or genes involved do not directly cause the disease in question, but, rather, can make an individual more susceptible to that disease. Therefore, these genes are referred to as susceptibility genes. The susceptibility conferred by one gene can be increased or decreased by a number of factors, including other susceptibility genes, as well as a variety of environmental influences. Some examples of common complex genetic disorders are asthma, cardiovascular disease, Alzheimer's disease, and autism.

There are many reasons that scientists want to find the genetic risk factors involved in complex genetic disorders such as autism. Knowing the genes involved helps identify those individuals at an increased risk for the disease. For instance, if doctors knew that a particular individual had an elevated genetic risk for a certain disease, they could closely follow this individual, enabling earlier diagnosis and therefore a better chance at successful intervention. This is one of the most-compelling goals in the search for the genetic basis of autism. Because research has suggested that, the earlier in life an autistic child can be given specialized education and therapy, the better his or her functioning later in life is likely to be, early diagnosis of autistic disorder is of critical importance.

Another goal of genetic research is to investigate the potential of an approach called pharmacogenetics, the study of how an individual's genetic makeup can affect one's responses to drugs. The goals of pharmacogenetics are twofold: to develop medicines that can more specifically target and treat causes and symptoms of disease, and to make medications safer to use by allowing the understanding of how direct effects and side effects of a medication might be tied to specific factors in a person's genetic profile. For instance, if physicians know that a certain medicine causes unbearable nausea in 10% of the population and scientists can determine that those 10% all carry a specific version of a particular gene (called an allele), with genetic testing, doctors could avoid giving that medication to patients who are genetically predisposed to the negative reaction.

Pharmacogenetics may be able to help find more-effective drag therapies for use in autism. "We would like to be able to link subtypes of autism with more-specific intervention strategies," indicates Michael Cuccaro, associate professor of psychiatry and a member of the autism research team at Duke University Medical Center's Center for Human Genetics. "For example, some children may benefit more from medicines, some with behavioral intervention, although many will need both."

Almost since the first suggestions of a genetic component in autism, studies have been conducted to identify the genes involved. Research to date suggests that multiple genetic risk factors or genes contribute to cause the disorder. It may actually be not individual genes, but the interaction between specific genes, that causes autism. It is likely as well that different genes, or combinations of genes, may be responsible for different forms of autism, or for autism in different families. These forms may vary so slightly in how they affect an individual that the differences would be very difficult to categorize based on symptoms a child exhibits.

Looking for the cause of a complex genetic disorder is similar to trying to find someone's house without knowing the address. By narrowing down the area you are searching--from state to city to neighborhood to street--eventually you can find the actual address of a particular person. Just as gas stations or restaurants can be used as landmarks to help you find your way around when attempting to locate a specific house, genetic signposts called markers can help scientists find their way around a chromosome to locate a specific gene. (Markers are patterns of DNA that repeat in a predictable way at various locations along a given chromosome.)

In order to look for genetic risk factors for complex disorders, scientists test many markers on all the chromosomes, trying to find ones that appear consistently in family members who have a particular disorder, but not in those without it. Once a marker that is shared more often among autistic family members is found, certain statistical methods can tell a scientist how close the marker is to a gene. Testing more markers will narrow the search area of the gene (like closing in on the block a given house is on). Markers that are very close to a gene are said to be linked to that gene because the marker is rarely inherited without that gene also being inherited. Once scientists find a set of markers that are linked to a gene, they say they have found linkage.

Linkage indicates approximately where on a chromosome a gene is located. Scientists still need to determine the exact location of the gene. One common method to do this uses candidate genes, which are genes known to be in the target region. A gene is called a candidate if its known function in some way relates to the biology of the disorder. For instance, a gene that is known to affect the proteins that build muscles might be considered a candidate if scientists were looking for risk factors for a muscular disorder such as muscular dystrophy. This technique is like knocking on the door of every house on the block until you find the one where your friend lives. Scientists test candidate genes for changes that might cause the disorder. If there are not any changes in that gene in a person who has the disorder, the candidate gene is not a cause. If all the likely candidate genes are tested and none are found to be responsible for the disorder, researchers turn to genes whose functions are not yet known. Many genes may have to be tested before the right one is found.

The genetics of autism

Using these methods, researchers have been able to identify some promising chromosomal regions that may contain genes involved in autism. One of these regions is on chromosome 15. For over a dozen years, scientists have noticed that some individuals with autism have a chromosomal change involving a specific part of chromosome 15 as well. These individuals have extra copies, or duplications, of a small region of chromosome 15. This same region also contains genes associated with Prader-Willi and Angelman syndromes, two disorders whose patients display developmental delays and, in some cases, autistic-like behavior.

Research found markers on chromosome 15 that are seen more frequently in individuals with autism (those without the specific chromosomal duplication which led researchers to this region in the first place) than in those without autism. This finding strongly suggests that a gene or genes that contribute to autism may be in this region. Although the specific gene has not been found, the area in which to look has been dramatically narrowed. The next step is to look gene by gene through this very complicated chromosomal region to figure out which specific one may be involved with autism. In order to facilitate this work, researchers at the Duke University Center for Human Genetics have created a complete map, similar to a road map, of this region. They have published this map in order to help other researchers looking for genes related to autism on chromosome 15.

When researchers tested markers, using linkage analysis, on all chromosomes in the human genome, they found that some individuals with autism share certain markers on chromosome 7. This linkage had been previously suggested by work done by a research group primarily located in Europe, and has since been reconfirmed by other research teams. As a result of the linkage findings on chromosome 7, several candidate genes have emerged and are being examined as autism risk genes. The hope is that the chromosome 7 gene will be identified in the near future.

There is also evidence for linkage in an area on chromosome 2. The evidence that links this chromosomal region to autism is even stronger when researchers look only at the group of patients with the disorder who began to use phrase speech relatively late in their development (older than three years). One hypothesis that researchers have to explain this finding is that there may be a risk gene on chromosome 2 that is implicated primarily in a subset of autism families. This potential risk gene would not be involved in all cases of autism, but might contribute to risk in the subset.

The future looks bright for genetic researchers in autism and for autistic patients and their families. The discovery of genetic contributions to common diseases such as autism is already having an impact on primary health care practice. As a website devoted to disseminating information about genetic research into autism (www.exploringautism.org) reports, "This new emphasis on genetics is transforming the way we approach medicine. Defining the human genome will give us the ability to identify, treat, and ultimately prevent diseases in ways that we never imagined. During this century, we will be able to develop new therapies for diseases that have plagued humans for generations, from cancer to heart disease, from Alzheimer's disease to diabetes." Autism is a disorder that is ripe to reap the benefits of these genomic medicine milestones, which, it is hoped, will result in radical advancements in diagnosis, treatment, and care.

Margaret A. Pericak-Vance is the James B. Duke Professor of Medicine and director, Center for Human Genetics, Duke University Medical Center, Durham, N.C.