One problem, many paths

Though the diagnostic code scrawled on a doctor’s chart might suggest otherwise, each person who lives with an autism spectrum disorder has a very private disease. An avalanche of new genetic data shows clearly that there is no single culprit in autism. Each case stems from a unique jumble of genetic and environmental triggers, which makes figuring out one clear cause for every person’s disorder impossible.

This news may sound grim, but it contains a glimmer of hope. By uncovering huge numbers of genetic aberrations, scientists say, they have the opportunity to begin piecing together all of the disparate threads weaving through autism to find the commonalities.

A suite of new studies have identified numerous genetic changes that may have a role in the disorder, some of which could help scientists understand why boys are more vulnerable than girls, for instance. And some of the genes affected by these changes appear to be players in common networks of molecular activity in the brain. New work shows that many genetic changes impair nerve cell communication. Understanding this process and finding other common cellular activities that go awry may lead to powerful ways to combat autism, regardless of what caused it.

“Parents and families have been tremendously patient,” says child psychiatrist and geneticist Matthew State of the Yale University School of Medicine. “They’ve been promised a lot by geneticists for a long time, and it’s been tougher than any of us expected to deliver.” But the flood of studies in the last few months reflects tremendous progress, he says. “These are all, in their own way, making a chink in the armor.”

Pieces here and there

Two studies in the June 9 Neuron identified a host of genetic changes that together account for 5 to 8 percent of autism cases. Both studies examined DNA samples taken from carefully screened families, a roughly 1,000-family cohort called the Simons Simplex Collection. Each family included two unaffected parents and one high-functioning child diagnosed with autism spectrum disorder. For most families, an unaffected sibling was also included.

By including unaffected family members, the researchers could find abnormalities — specifically, duplications and deletions of DNA called copy number variations — that were not passed down from parents but arose spontaneously in the genomes of affected children.

One of the studies, coauthored by State and his colleagues, included 1,124 families and estimated the number of genome regions in which copy number changes were linked to autism at 130 to 234. The other study, done by Michael Wigler of Cold Spring Harbor Laboratory in New York and colleagues, looked at 887 families and put the number at 250 to 300.

“This really speaks to the immense heterogeneity of autism,” says geneticist Huda Zoghbi of Baylor College of Medicine in Houston, who was not involved in the studies. “We suspected it, but these data show it clearly.”

Results described by Wigler’s team may also help explain why autism spectrum disorders are much more common in boys. Autism strikes four boys for every girl, yet girls’ DNA actually harbors more of these rare autism-associated genome duplications and deletions, the researchers found. And these anomalies aren’t just more abundant in girls; each change interrupts more genes. For a girl with autism, each duplication or deletion disrupted a median of 15.5 genes, while for a boy, the number disrupted was just two.

Through some mysterious process, girls are just more resistant than boys to the genetic causes of autism, the results suggested. “Overall, it does look like a girl can have the same genetic insult as a boy, but not be diagnosed with autism,” Wigler says.

For the most part, each duplication or deletion was specific to each affected child. Although copy number variations of one particular region of chromosome 16 were observed in multiple children in both studies, changes to this region still accounted for only slightly more than 1 percent of autism cases.

“Some people will see that as glass-is-half-empty, but we see it as glass-is-half-full,” State says. “It shows that there are a lot more clues to be had.” State and his colleagues are now combing through 1,000 more samples from the Simons Simplex Collection, conducting more targeted studies to find out how some of these genes may contribute to the disorders.

While single, severe abnormalities in parts of the genome are clearly important for certain cases of autism, in some people the disorder could be caused by more numerous mild genetic changes. A new study from Zoghbi and her colleagues finds that children with autism are more likely than unaffected children to have double hits: mutations in two of 21 autism-implicated genes.

Just one of these mutations is not severe enough to completely scramble or eliminate a gene’s functions, nor is it strong enough to affect the parent, who oftentimes carried it as well. “Each parent is fine, but those two together now in a child can perhaps increase the possibility of having autism,” Zoghbi says. Once a certain mutational threshold is reached, the disorder appears, the researchers suggest in a paper to appear in an upcoming Human Molecular Genetics.

Linking up

Other researchers are taking a different tack: Instead of searching for DNA changes to identify related genes, they study how genes behave. For instance, a recent study by neurogeneticist Daniel Geschwind of UCLA focused on gene activity — measured by the amount of RNA molecules shuttling information from a particular gene to the protein-producing parts of cells.

Hundreds of genes behaved differently in the brains of people with autism compared with unaffected people, Geschwind and his colleagues reported online May 25 in Nature (SN: 6/18/11, p. 5). Many of these genes were involved in maintaining the complex and sensitive links called synapses, the junctions between nerve cells.

Wigler and his colleagues have also found that in children with autism, deletions and duplications tend to strike genes whose products are essential for nerve cell communication. The study, published in the June 9 Neuron, adds to a growing body of evidence that nerve cell signaling is profoundly altered in people with autism.

With better ways to uncover the biochemical networks in which genes, RNA and proteins are players, scientists can now study such nerve cell malfunctions with precision. Wigler and his team analyzed how genes in DNA regions that have duplications or deletions associated with autism interact with one another. Among other things, some of the affected genes tell nerve cells how to grow a signal-sending axon and how to find the right partner to send signals to. Others oversee the chemical messages that carry signals across synapses from cell to cell.

Another study from Zoghbi confirms that synapses are key. She and her team turned up an unexpected relationship between two proteins that both work at the junction where nerve cells connect. In a massive undertaking, the researchers tested whether each of 26 autism-related proteins latched on to nearly every other protein a human brain cell produces. This effort, reported online June 8 in Science Translational Medicine, identified more than 500 proteins that interacted with the autism-related proteins.

Two of the autism-related proteins, called SHANK3 and TSC1, had a surprisingly cozy relationship. Zoghbi and her colleagues didn’t think those two proteins would be related, but it turns out that they share at least 21 protein partners, implicating them as tight coconspirators at the synapse. These proteins probably act together in dendrites, the parts of a nerve cell that pick up messages from other cells, the researchers suggest.

“At the end of the day, we have to remember that features of autism suggest to us that neurons are not functioning well,” Zoghbi says.

Understanding this shared problem, and uncovering similar networks, may lead to ways to treat or prevent autism in the future. Figuring out ways to protect vulnerable brain processes may turn out to be more important than knowing exactly how things went wrong.