Many physicians and parents report that their autistic children have unusually severe gastrointestinal problems, such as chronic constipation or diarrhea. These observations have led some researchers to speculate that an ailing gut contributes to the disorder in some cases, but scientific data has been lacking. Now, a provocative study claims that a probiotic treatment for gastrointestinal issues can reduce autismlike symptoms in mice and suggests that this treatment could work for humans, too.
The reported incidence of gut maladies in people with autism varies wildly between published studies—from zero to more than 80%—making it difficult to establish just how commonly the two conditions go together, says principal investigator Sarkis Mazmanian, a microbiologist at the California Institute of Technology (Caltech) in Pasadena. Overall, however, the evidence seems to point toward a connection. Last year, for example, a Centers for Disease Control and Prevention study of thousands of children with developmental disabilities found that kids with autism were twice as likely as children with other types of disorders to have frequent diarrhea or colitis, an inflammation of the large intestine.
For many years, Mazmanian and his and colleagues have been studying the effects of a nontoxic strain of the bacterium Bacteroides fragilis on diseases such as Crohn's disease, which causes intestinal inflammation and allows potentially harmful substances that should pass out of the body to leak through junctions between cells that are normally tight. Although the researchers don’t understand the mechanism, the bacterium appears to restore the damaged gut, possibly by helping close these gaps.
"The fact that we have an organism that repairs the gut makes it a very appealing" tool for testing whether gut abnormalities can contribute to autism, Mazmanian says. To explore that question, Mazmanian and colleagues at Caltech used a mouse model of autism that is thought to approximately recreate three of the disorder's hallmark deficits: lack of social interaction, decreased communication (mice normally emit ultrasonic, birdsonglike chirps), and repetitive behaviors such as compulsive grooming or burying marbles.
The first step of the experiment was to determine whether the mice showed signs of gastrointestinal inflammation or other gut abnormalities, says microbiologist Elaine Hsiao, a postdoctoral candidate at Caltech and lead author of the study. By the time the mice were 3 weeks old, the researchers found that their intestines were indeed as leaky as those of mice that had been treated with a chemical that induces colitis. Next, the researchers tested whether they could reverse the damage by feeding the mice applesauce laced with B. fragilis for a week. A second group of autism-mimicking mice as well as a group of healthy mice ate applesauce that did not contain the bacteria. Then the group waited to see what effect the bacteria would have on the rodents' intestines. "We didn't know what would happen—we were hoping the bacterium would survive in the gut," Hsiao says.
After 3 weeks, the team measured the levels of gut-derived molecules in the rodents' bloodstream and found that the treatment had stopped up their intestinal leakage. Bacterial counts from rodents' poop showed that although B. fragilis did not establish lasting colonies in the mice, they did "shake up the community," of microorganisms, bringing it closer to that of the normal mice, Hsiao says. After the treatment, the autism-mimicking mice also resembled their normal peers in two behavioral tests, the authors report today in Cell. The animals no longer compulsively buried marbles in their cages and increased their ultrasonic squeaking to typical levels. They did not increase their social interactions, however, Hsiao says.
"It’s really striking that any bacterial treatment—even a transient one—could have a lasting impact on behavior," Hsiao says. The most interesting thing about the results, she says, was not the correction of the autistic symptoms in the mice, but the clues the study provided about how the gut's microbial population may affect the brain and behavior. The researchers found that levels of a substance called 4-ethylphenylsulfate that is produced by gut bacteria increased 46-fold in the mice with autistic symptoms, but returned to normal after treatment with B. fragilis. When the team isolated that chemical and injected it into healthy mice, the rodents showed increased anxiety, another autismlike symptom, she says. Although the substance did not provoke the symptoms seen in the previous experiment, Hsiao says that the animals' altered response suggests that the substance could play a role in the disorder. Hundreds of other metabolic byproducts also changed in quantity after B. fragilis was administered and could have an effect, she adds.
By demonstrating that a widely used mouse model of autism does have gastrointestinal problems, and that these problems are associated with behavioral symptoms, the new research "shows us something fabulous," says Betty Diamond, an immunologist at the Feinstein Institute for Medical Research in Manhasset, New York. She cautions, however, that it would be premature to use B. fragilis or another probiotic as a treatment in humans. "We don't really understand" which bacterial species are important or how they colonize the gut, she says.
Although the findings are interesting, the study does not establish that the changing levels of microbes and the chemicals they produce caused any of the behavioral changes seen in the mice, says Emanuel DiCicco-Bloom, a neuroscientist at Rutgers University in New Brunswick, New Jersey. "I'd want to know more about the mechanism" by which the bacteria altered behavior in the mice before beginning to translate the findings to humans, he says. The group also didn't investigate how the bacteria affect a normal animal, because the microbes were administered only to autistic mice, he says. It's possible that B. fragilis could have a deleterious effect that the study didn't detect, he says. Combined with the inherent difficulty of extrapolating findings about human autism from a mouse, he says, "I think this is less well-established than it appears."
"We propose that after the repair of the leaky gut, neurotoxic molecules do not get into the system and cause behavioral abnormalities," Mazmanian says. But he agrees with DiCicco-Bloom that there are alternative explanations for why the mice changed their behavior—for example, "maybe bacteria are activating nerves in the gut that are communicating with the brain," he says. After resolving some of these questions, the group plans to initiate clinical trials in humans, Hsiao says. "We don’t want people to start applying this to humans" just yet, but "this opens the door to future research” in people.