In this pair of studies, the researchers identify and characterize a number of mutations in the gene that regulate brain levels of serotonin, a neurotransmitter involved in many biological processes including breathing, digestion, sleep, appetite, blood vessel constriction, mood and impulsivity.
About 25 percent of people with autism have elevated levels of serotonin in their blood. Selective serotonin reuptake inhibitors (SSRIs), drugs used to treat depression, anxiety and obsessive-compulsive disorders, also improve some of the symptoms of the disorder. These findings have led scientists to propose that serotonin plays an important role in autism.
In the August issue of the American Journal of Human Genetics, Sutcliffe, Blakely and colleagues report that several mutations within the serotonin transporter (SERT) gene, which regulates serotonin levels in the brain, may be risk factors for autism.
One variation in the SERT gene has been extensively studied and previously led to an inconclusive association to autism. No other variation stood out as a strong risk factor for the disease. Sutcliffe's own work had detected a strong linkage between autism and a 'spot' on chromosome 17 – the neighborhood where the SERT gene resides. A few common variants or versions of this gene were known, but did not seem to impart increased risk of autism.
"We failed to see evidence for a common version of the SERT gene that is the same in a majority of people," said Sutcliffe, associate professor of Molecular Physiology and Biophysics and the lead investigator on the study. "So, either this was not the gene, or there had to be different genetic variants that were acting differently in different people."
Sutcliffe, Blakely and colleagues decided to dig deeper into the DNA sequence of the SERT gene to identify these rare mutations and to assess their role in autism risk. Using DNA samples from 120 families likely to possess a genetic risk factor on chromosome 17, the team found 19 different SERT mutations (or variants) in families with multiple affected males, consistent with the well-known sex-bias seen in autism incidence.
Four of these variants were in 'coding' regions, or parts of the gene that get translated into protein. The other 15 variants were in 'noncoding' regions, which are edited out of the final protein product but may have important regulatory roles in the expression of the gene. "These coding mutations tracked with an increased severity of rigid-compulsive behaviors," Sutcliffe explained. These types of behaviors are a common characteristic of autism and related disorders like obsessive-compulsive disorders.
The findings underscore the relationship between autism and disorders like OCD and may explain why SSRIs are effective in treating these conditions, he said.
Strengthening the case for autism-linked SERT variants, in the August online issue of the Proceedings of the National Academy of Sciences, Blakely, Sutcliffe and colleagues describe regulatory problems in SERT gene variants, suggesting a possible mechanism for how SERT mutations may disrupt serotonin signaling in autism.
"We show that there are specific signaling pathways that cannot talk to SERTs with these mutations," said Blakely, Allan D. Bass Professor of Pharmacology and senior author on the PNAS study.
Initially, Harish Prasad, Ph.D., a senior scientist in the Blakely lab, examined 10 different SERT variants to see how well they functioned. With the exception of one variant common to both studies, most of these variants had not been previously linked to any clinical conditions.
While the variant SERTs could perform their basic function of 'vacuuming' up excess serotonin, intracellular signaling pathways that normally fine-tune SERT activity were unable to control five of the 10 mutant SERT proteins examined.
"We were stunned because the cell just can't 'talk' to these SERT proteins in a normal way," Blakely said. "Although it's impossible to extrapolate from a molecule to a person," he said, "it is striking that these mutations, which do not allow proper communication with SERT, show up in a disorder fraught with communication problems."
Interestingly, drugs that target these intracellular pathways, the p38 MAPK and the PKG pathways, have been investigated in a number of disorders unrelated to autism, such as inflammation and cancer. Targeting these pathways might offer a new alternative for treating autism with medications.
"This is a potential therapeutic area that we hadn't envisioned before," Blakely said.
Based on these findings, Blakely and Sutcliffe predict that there will one day be a way to test autistic children for these gene variants, similar to the testing done for cystic fibrosis, a disease linked to a single gene but triggered by many different mutations.
"Autism has such a high genetic risk, but these new findings suggest that there may be many variants of individual genes at work," Blakely said.
With such genetic testing, said Sutcliffe, "you might be able to predict which kids would respond positively to particular SSRI medications."
"We now have concrete evidence in our families that the SERT gene is a risk factor in autism," Blakely said. "Perhaps more importantly, we also have new pathways that could have some therapeutic end points, and that, to us, is really good news."
The studies were supported by the National Institutes of Health, the National Alliance for Autism Research, and Vanderbilt's Kennedy Center, Center for Human Genetics Research, Center for Molecular Neuroscience and the General Clinical Research Center.
Proceedings of the National Academy of Sciences