The unusual suspects

With no obvious culprit in sight, geneticists do broader sweeps to identify autism’s causes

It’s a high-stakes version of the board game Clue. Scientist-detectives probing the origins of autism must contend with an enormous cast of characters. Within the past year, researchers have found dozens, possibly hundreds, of rare genetic mutations that may contribute to the disorder, and a handful of common mutations may also be involved.

A boy diagnosed with autism makes eye contact with a therapist while working on his social and communication skills. Drew Perine/AP Photo/The News Tribune
GETTING THE MESSAGE | Projections called axons transmit messages from one nerve cell to another (left) at the synapse (right), a junction between nerve cells held in place with the aid of various cell-adhesion molecules. Studies have linked autism with mutations in a number of genes involved in cell adhesion at the synapse. A disruption in any one of the proteins (dark purple) that these genes code for may throw off the system so that nerve cells don’t connect normally. C. Betancur et al/Trends in Neurosciences, adapted by E. Feliciano
GETTING NETWORKED Some people with autism may be missing all or parts of genes involved in brain development and function. Newly discovered genes affected by these deletions (purple) work in networks with other genes (orange) that have already been linked to autism. These networks help guide important cell functions, including cell proliferation and central nervous system development.
SHUFFLING DNA Autism often arises in low-risk families via mutations that show up in a child but are not present in the parents’ DNA. A chromosome inversion, where a single chromosome undergoes breakage and rearranges itself (left), is one example. The autism-related gene CNTNAP2 (highlighted at right by a fluorescent probe) is split in an inverted chromosome. Nicole Wright/Yale School of Medicine, American Journal of Human Genetics

Faced with this staggering lineup of genetic suspects, scientists have turned to new DNA sequencing technologies and other methods to track clues within the brain and pin down the who, where and how underlying autism.

Nobody expects to find Colonel Mustard in the kitchen with a knife. The latest clues have made it clear that with autism, there will turn out to be multiple culprits.

“There’s not going to be a simple explanation for autism,” says neurogeneticist Daniel Geschwind of the University of California, Los Angeles. “The genetics are very complex, and there are likely to be many different genetic and biological mechanisms involved.”

Researchers have long known that genes play a role in autism, a disorder marked by impaired social interaction and communication. Studies of twins suggest that as many as 90 percent of autism cases may have a genetic link. The problem, in many cases, is that scientists don’t know what to make of those findings.

“What you hope for is that you find a mutation and then every time you see the mutation, a person’s got some evidence of autism,” says Yale University neurogeneticist Matthew State.

But that’s not what scientists see. Studies have linked a handful of common gene variants to autism, but most of the genetic mutations implicated in the disorder are rare. In many cases, the mutations found in kids with autism are also seen in kids without autism. And some of the gene variants that raise the risk of autism have also been linked to other psychiatric disorders, such as manic-depressive illness or schizophrenia.

To try to untangle these complications, researchers are using the genetic findings as a starting point, followed up by studies to see how and where the products of these potentially rogue genes work in the brain. Lately a more complete picture has begun to emerge.

“What genetics is doing for us is telling us where to look and what are the fundamental processes that are involved,” State says.

Identifying common MOs

While it’s still early in the game, recent findings tie many of the rare mutations to genes in common biological pathways and networks — most of which help control the way the brain develops and functions. Some of the genes belong to pathways used by brain cells to communicate, while others are involved in the growth of nerve cells, anchoring cells together or signaling within a cell.

Such findings provide targets for new treatments and approaches, the scientists say. Already, some labs are working to find chemical markers of the disorder that could be used for early diagnosis.

Other laboratories are using new technologies to sequence all protein-coding regions and even all of a person’s DNA to identify parts of the genome that diverge in people with autism and healthy subjects. As researchers home in on genomic differences that can predispose an individual to autism, ways to tailor treatments and interventions may emerge.

“It’s someplace where we’ve been 30 or 50 years behind other areas of medicine,” says State, a specialist in child psychiatric disorders. “If you go to a cardiologist, they understand the pathology that they look at in the clinic at a molecular and cellular level. In autism and other child psychiatric disorders, we have not had that.”

Multiple rare syndromes

Autism is not one disorder, but a spectrum of disorders. People with autism often have great difficulty with communication and social interaction. Some cannot speak or maintain eye contact. Many have repetitive routines. A common trait is obsessive attention to certain details. Symptoms can be mild to severe, and about half of kids with autism have some kind of intellectual disability. In the United States, an estimated one in 110 children has an autism spectrum disorder.

For years, most studies focused on finding common variations in a handful of genes. While researchers found a few genetic variants linked to autism, the cumulative effects were disappointingly small, accounting for just a fraction of cases.

“The hypothesis had been that in order for a common disease to arise, you would need to have common genetic variations in perhaps just a few genes,” says Stephen Scherer, director of the Center for Applied Genomics at the Hospital for Sick Children in Toronto. “But that idea is shifting.”

One reason for the shift: Following the completion of the Human Genome Project in 2003, it became apparent that the genome is much more variable than previously thought. Studies showed that the number of copies of a particular gene differs from one individual to the next. Many of these variations involve either segments of DNA that are entirely missing from the genome or the same segment repeated several times.

Through such gains or losses, called copy number variations, whole stretches of DNA can be erased or repeated (SN: 7/3/10, p. 12). Most copy number variants appear to be harmless. But some can remove parts or all of a gene. Studies have implicated copy number mutations in a number of diseases, including schizophrenia. And in 2007, researchers at Cold Spring Harbor Laboratory in New York found evidence that such anomalies may also be associated with the symptoms of autism spectrum disorders.

Researchers then began looking for rare variants in autism — and found them. In a widely acclaimed study in the July 15 Nature, an international group of researchers compared genomes of nearly 1,000 autistic people and about 1,300 healthy controls. The scientists found dozens of genes were involved. Most of the variants were sections of DNA that were either duplicated or missing. Some of the genetic changes were inherited from the kids’ parents, while other variants were new, arising from alterations of DNA in the egg or sperm of one of the parents, or in the offspring themselves.

Those findings support an emerging consensus within the scientific community, says Scherer, who co­authored the study. Namely, that autism, instead of having just one or two genetic risk factors, probably has hundreds.

“I think it says something fundamental about autism, that you can think of it as a collection of rare syndromes,” he says.

Scientists now suspect that, while the number of different genes involved is large, the protein products of these genes participate in a much smaller number of common pathways that regulate brain development and function. Two genes may encode two proteins that seem totally unrelated but, in fact, interact closely in a particular pathway. A deficiency in either’s function could result in the same outward defect.

That seems to be what’s happening, Scherer says. While each of the variants may account for only a small fraction of autism cases — no single variant can be said to account for more than 1 percent — collectively the rare variants may account for a large fraction of the cases.

Scientists are now working to identify all of the genes involved and create a catalog of autism risk genes. Studies to date have already identified 100 or so strong candidate genes. And Scherer and his colleagues recently completed a second study with an additional 1,500 families, with analysis of the data now underway.

Even closer scrutiny

Researchers also are working to find even more subtle mutations. Scientists at the Broad Institute in Cambridge, Mass., are using next-generation DNA technologies to compare the genomes of autistic individuals with those of healthy controls. The method will allow the researchers to read each gene in every region of the genome, making it possible to pick up on variations that are much more rare and to detect submicroscopic insertions and deletions that contribute to autism.

“It’s clear that there is a lot more to discover,” says Mark Daly, who oversees the DNA sequencing studies at the Broad Institute, a genetic research center operated by Harvard University and the Massachusetts Institute of Technology.

For example, scientists have yet to find a single autism-related mutation capable of disabling any one copy of a gene in a systematic way. Nor have they figured out ways to interpret the effects of copy number variants, which can take out dozens of genes in a single deletion.

“It’s been challenging to extract really specific biological insights from those observations because we don’t know which, or which sets, of a 25-gene deletion is relevant to autism,” Daly says.

Such discoveries may soon emerge. By searching for single changes in the DNA, called individual point mutations, scientists hope to “fill out” the rest of the genetic components that contribute to autism risk, Daly says. “It’s not that we didn’t know that this is what we wanted to look for. We never had the technical ability to search for these types of mutations in a systematic and unbiased fashion.”

Other scientists are looking at differences in gene activity to find biochemical markers for early diagnosis. Geneticist Stephen T. Warren of Emory University in Atlanta says genes active in white blood cells may reflect what’s happening in the brain. Using blood samples from twin brothers — one with autism and one without — his group is looking at more than a dozen sites in white blood cells where DNA is chemically altered, or methylated, affecting whether a gene is turned “on.” Such differences may serve as a genetic signature for some cases of autism.

Warren’s group is now investigating whether any of the markers are present before the onset of autism. If so, the irregular gene activity patterns might be used to help identify high-risk children as infants, instead of at age 2 or 3, and allow early intervention therapies to be initiated. Studies show that the younger the child, the more flexible the brain. With intense behavioral intervention, new mental pathways might be created to overcome some of autism’s effects.

Finding conspirators

As for the genes themselves, researchers have found dozens over the past few years that appear to raise the risk of autism. Many of the genes can be tied to common biological pathways and networks, providing insight into what the genes do and how they operate to control — or interfere with — specific brain processes. Scientists are treating such findings as possible signposts that can help navigate the many potential routes to autism.

For example, many of the studies point to molecules that help form and maintain brain connections. Other findings point to the processes used by nerve cells to pass signals to one another. Mutations in genes with names such as SHANK3, SHANK2 and NRXN1 have been linked to autism. All of these implicated genes code for proteins found at synapses, the junctions where one nerve cell releases chemical signals to communicate with another, and are crucial for normal signaling between nerve cells.

As the list of genetic suspects grows, the challenge lies in unraveling how problems in any one of the genes could send the brain’s circuits awry. To find answers, scientists are looking at a handful of rare syndromes where autism patients share a common mutation leading to the same outcome.

Four years ago, scientists found one such example. Researchers studying a rare mutation that causes an epileptic disorder in Old Order Amish children observed that all the children with two copies of a recessive gene — called contactin associated protein-like 2, or CNTNAP2 — developed frequent seizures in early childhood, followed by features of autism.

Further studies followed, linking CNTNAP2 and autism. In January 2008, three groups of researchers independently reported that they had identified defects in the gene in larger groups of subjects. The three groups used different strategies and different populations to look for possible autism links, and found different mutations — some common and some rare. Still, the labs arrived at the same conclusion: Variations in CNTNAP2 predisposed carriers to autism.

State’s group is now pursuing studies to figure out exactly how. CNTNAP2 normally makes a type of protein called a neurexin, which is located in neurons. The protein helps brain cells link up during the development of the nervous system. It also appears to help growing cells adapt and alter axons, projections through which brain cells send electrical impulses essential for normal brain function.

In test-tube experiments and in zebra fish, State’s team is engineering cells that lack the gene completely. “If we can find out what goes wrong in the protein that we know leads to this rare syndrome of bad seizures and autism, then we can use that as a benchmark and can begin to ask questions of the mutations that we are seeing in other people,” he says.

Tracing lines of evidence

Meanwhile, Geschwind and his group are working to catch the gene suspects in action. Studies show that during early development, CNTNAP2 is highly active in parts of the human brain important for language processing. Other studies have implicated CNTNAP2 in certain language disorders. Together, the studies suggest that a disrupted version of the gene could throw a monkey wrench into the works during the earliest stages of development.

Using functional MRI, the scientists are looking to see whether the brains of people with and without autism activate differently in specific regions during language-related tasks. The next step will be to see whether activation patterns can be linked to genetic variations. Other labs are using similar approaches to show how autism-associated genes might work on distinct brain regions.

Still, solving the puzzle of autism will be daunting. The fact that identical twins do not always share the disorder suggests that environmental factors as well as genes are at play, Geschwind says. The task is made even more difficult because the genetic aberrations that have been implicated in autism do not present a clear, one-to-one relationship with patients’ symptoms or abilities. Some of the submicroscopic changes found in autistic children, for example, are also seen in one of their parents, though the parent may have only mild symptoms or no symptoms at all. Other mutations found in autistic children are seen in kids without autism as well.

And new questions will arise. As research pieces together a more complete picture of the various genes involved — showing which they are, where they work and how they control brain circuits — it may reveal ways to tie particular features of autism to specific derangements in DNA. Like the movie version of Clue — which had multiple endings with different killers committing the same crime — the findings will probably finger multiple molecular missteps leading to the same behaviors and outcomes.

Understanding how the brain functioning of one child with autism differs from that of another could help in developing treatments tailored to specific behaviors or problems. Whether kids with similar symptoms but different genetic variations will respond to the same treatments is unknown, Geschwind says. “But that’s a question that we have to ask and begin to look at.”


SIDEBAR: Autism spectrum disorders
Autism is often used as a catchall phrase to describe a spectrum, or range, of disorders that affect a person’s ability to communicate and interact socially. Autism spectrum disorders include conditions with a wide variety of symptoms that range in severity and come in many different combinations. The autism spectrum disorders include:

Autistic disorder
Also known as “classic autism,” this disorder affects a person’s ability to communicate, form relationships and respond appropriately to the environment. Some people with autistic disorder are high-functioning and can speak and interact, while others are more severely affected and nonverbal.

Asperger syndrome
Individuals with Asperger syndrome do not have a delay in spoken language development, but they can have serious deficits in social and communication skills. People with Asperger syndrome often have obsessive, repetitive routines and preoccupations with a particular subject, such as trains.

Childhood disintegrative disorder
Children with this disorder typically develop normally for two to four years before the condition, which resembles autistic disorder, arises. Previously mastered language, social and toileting skills may be lost.

Rett syndrome
This disorder almost always affects girls. Babies with Rett syndrome develop normally until 6 to 18 months of age, when development slows and their heads no longer grow normally. Affected children don’t develop normal speech and may exhibit unusual hand movements and walking patterns.

Pervasive developmental disorder/not otherwise specified
Also known as atypical autism, this diagnosis is often used when some, but not all, of the symptoms of classic autism or another disorder are seen. Like other autism spectrum disorders, it is characterized by social and speech problems.

Sources: National Institutes of Health, National Rett Syndrome Foundation