Read our COVID-19 research and news.

Mice colonized with microbes from people with autism showed autismlike symptoms, including decreased social behavior.

Gut bacteria may contribute to autism symptoms, mouse study finds

Genes are a powerful driver of risk for autism, but some researchers suspect another factor is also at play: the set of bacteria that inhabits the gut. That idea has been controversial, but a new study offers support for this gut-brain link. It reveals that mice develop autismlike behaviors when they are colonized by microbes from the feces of people with autism. The result doesn’t prove that gut bacteria can cause autism. But it suggests that, at least in mice, the makeup of the gut can contribute to some hallmark features of the disorder.

“It’s quite an encouraging paper,” says John Cryan, a neuroscientist at University College Cork in Ireland who was not involved in the research. The idea that metabolites—the molecules produced by bacterial digestion—can influence brain activity “is plausible, it makes sense, and it will help push the field forward.”

Many studies have found differences between the composition of the gut microbiomes in people with and without autism. But those studies can’t determine whether a microbial imbalance is responsible for autism symptoms or is a result of having the condition.

To test the effect of the gut microbiome on behavior, Sarkis Mazmanian, a microbiologist at the California Institute of Technology (Caltech) in Pasadena, and collaborators put fecal samples from children with and without autism into the stomachs of germ-free mice, which had no microbiomes of their own. The researchers then mated pairs of mice colonized with the same microbiomes, so their offspring would be exposed to a set of human microbes early in development.

The researchers then ran these offspring through behavioral tests typically used to gauge autismlike symptoms in mice. They recorded how frequently a mouse vocalized and how often it approached and interacted with another mouse. They also tried to approximate the repetitive behavior seen in some people with autism by scattering marbles around a cage and counting how many a mouse buried. Compared with mice colonized with bacteria from children without autism, the mice that inherited a microbiome from a child with autism were less social and showed more repetitive behavior, the authors report today in Cell.

Mice with the autism-derived microbiome also had lower levels of several bacterial species that the researchers suspect could be beneficial. It’s known that microbes in the gut break down or modify the amino acids in food, and that byproducts can travel through the bloodstream and possibly into the brain. But researchers don’t know exactly which of the transplanted microbes interact with the brain to influence autismlike symptoms.

When the researchers dissected the mouse brains, RNA analysis of the two groups revealed differences in splicing—the way DNA’s message is processed before it’s translated into a protein—for 560 genes, including 52 that have been associated with autism. That’s an intriguing hint that the products of gut microbes might somehow change autism risk by influencing what forms of proteins are made in the brain, says Caltech biologist Gil Sharon, a postdoctoral researcher and first author on the new paper.

When the researchers looked at the contents of the mouse guts, they found differences between the two groups in the levels of 27 metabolites. In particular, mice harboring microbes from people with autism had lower levels of taurine and 5-aminovaleric acid (5AV), molecules that are known to bind to neurons and inhibit their activity. That finding fits with the theory that an imbalance between excitatory and inhibitory signals in the brain might underlie autism. The team also found with a different strain of mouse known to develop autismlike symptoms that feeding the animals either taurine or 5AV led to more social interaction and less repetitive behavior.

“There’s still a lot of missing links,” says Jun Huh, an immunologist at Harvard University who studies the relationship between bacteria and brain function. “But I think the real importance of this study is to show—for the first time—that there’s a causal relationship between the bacterial community and [autismlike] behavior.”

Shakuntla Gondalia, a gut microbiome researcher at Swinburne University of Technology in Hawthorne, Australia, says the next step should be to replicate the findings with fecal samples from people outside of the United States. Our resident microbes vary based on our environment and diet, she says, and the possible effect of this variation on autism risk remains a mystery.

These results are unlikely to yield new microbiome-based treatments right away, Cryan notes. The two metabolites highlighted in this study might turn out to be irrelevant to autism in people. Still, the research justifies a hunt for other metabolites deficient in the gut or brain of people with the disorder, he says. “This will give encouragement to the field that there is something there.”