Gut-brain connection in autism

Many people with autism spectrum disorders also experience unusual gastrointestinal inflammation, but thus far scientists have not established whether and how those conditions might be linked.

Now, Harvard Medical School and MIT researchers, working with mouse models, may have found the missing link: Infections during pregnancy can lead to high levels of the inflammatory signaling molecule interleukin-17a (IL-17a), which can not only affect brain development in the fetus, but also alter the maternal microbiome in a way that primes the newborn’s immune system for future inflammatory attacks.

In four studies beginning in 2016, study co-senior authors Gloria Choi of MIT and Jun Huh of Harvard Medical School have traced how elevated IL-17a during pregnancy acts on neural receptors in a specific region of the fetal brain to alter circuit development, leading to autism-like behavioral symptoms in mouse models.

The new research findings, which were described in the journal Immunity last month, show how IL-17a can also act to alter the trajectory of offspring’s immune system development.

“We’ve shown that IL-17a acting on the fetal brain can induce autism-like behavioral phenotypes such as social deficits,” said Choi, the Mark Hyman Jr. Career Development Associate Professor in the Picower Institute for Learning and Memory and the Department of Brain and Cognitive Sciences at MIT. “Now we are showing that the same IL-17a in mothers, through changes in the microbiome community, produces comorbid symptoms in the offspring, specifically a primed immune system.”

The researchers caution that while the study findings are yet to be confirmed in humans, they do offer a hint that central nervous and immune system problems in individuals with autism-spectrum disorders share an environmental driver: maternal infection during pregnancy.

“There has been no mechanistic understanding of why patients with a neurodevelopmental disorder have dysregulated immune system,” said Huh, an associate professor of immunology at the Blavatnik Institute at Harvard Medical School. “With the new findings, we’ve tied these fragmented links together. It may be that the reason is that they were exposed to this increase in inflammation during pregnancy.”

Tracking timing

First, the research team set out to confirm that maternal immune activation (MIA) leads to enhanced susceptibility to intestinal inflammation in their offspring. To do so, they injected pregnant mice with poly(I:C), a substance that mimics viral infection. As expected, their offspring exhibited autism-like behavioral symptoms as well as gut inflammation when exposed to other inflammatory stimuli. These symptoms and inflammation were notably absent in the offspring of pregnant mice in an unaffected control group.

The neurodevelopmental aberrations the team tracked occurred while the fetus was still in the womb, yet it was not clear when the altered immune responses developed. To find out, the researchers switched mouse pups at birth so that those born to mice with immune activation were reared by mice from the inflammation-free control group, while pups born to mice from the control group were reared by mice with MIA. The team found that pups born to mice with MIA but reared by mice from the control group exhibited the autism-like symptoms but not the intestinal inflammation. Pups born to mice from the control group but reared by mice with MIA did not show autism-like symptoms but did experience intestinal inflammation. The results indicate that while neurodevelopmental pathways are altered before birth, the immune response is altered after it.

Microbiome-mediated molecular mechanism

Still one question loomed large: How did mice with MIA exercise this immune-altering effect on pups after birth? Other studies have found that the maternal microbiome can influence the immune system development of offspring. To test whether that was the case in the MIA model, the researchers examined stool from MIA and control mice and found that the diversity of the microbial communities was significantly different.

Then, to determine whether these differences played a causal role, the researchers raised a new set of female mice in a “germ-free” environment, meaning that the mice do not carry any microbes in or on their body. Then the scientists transplanted stool from MIA or control mice into these germ-free mice that subsequently got pregnant. Unlike with the controls, pups born to mice that received stool from mice with MIA exhibited intestinal inflammation. These results indicate that the altered microbiome of mice with MIA leads to the postnatal immune priming of offspring during rearing.

Among the notable differences the team measured in the intestinal inflammation response was an increase in IL-17a production by T cells, a major class of immune cells. IL-17a is the same cytokine that gets upregulated in mice with MIA. When the scientists compared the T cells of offspring from mice with MIA with those from control offspring, they found that in MIA-offspring, CD4 T cells were more likely to differentiate into Th17 cells, which in turn release IL-17a.

The finding prompted researchers to look at potential differences in how the CD4 T cells of the different groups may alter gene transcription. The CD4 T cells of mice exposed to the microbiomes of animals with MIA had higher expression of genes for T cell activation, suggesting they were more primed for T cell-dependent immune responses in response to infections.

“Thus, increase in IL-17a in moms during pregnancy leads to susceptibility to produce more IL-17a in offspring upon an immune challenge,” Choi said.

Having established that the immune system of the offspring can become dysregulated by exposure to the mother’s altered microbiome as a result of infection during pregnancy, the remaining question was how that microbiome becomes altered in the first place.

Suspecting IL-17a, the team tested the effects of antibodies that block the cytokine. When they blocked IL-17a in pregnant mice prior to immune activation, their offspring did not exhibit the intestinal inflammation later in life. This also held true when the researchers repeated the experiment of transplanting MIA stool to germ-free mice, this time including stool from pregnant MIA mice with IL-17a blockers. Again, blocking IL-17a amid maternal infection led to a microbiome that did not improperly prime the immune system of offspring.

Long-term questions

Huh said the results highlight how environmental exposures during pregnancy, such as infection, could have long-term health consequences for the offspring, a concern that has always been present but that may be exacerbated by the COVID-19 pandemic. Further study is needed, he said, to determine long-term effects on children born to mothers infected with SARS-Cov-2.

Choi added that emerging connections between inflammation and neurodegenerative diseases such as Alzheimer’s may also warrant further study given the team’s findings of how maternal infection can lead to enhanced inflammation in offspring.

Eunha Kim and Donggi Paik of Huh’s lab are the study’s co-lead authors. Other investigators included Ricardo Ramirez, Delaney Biggs, Youngjun Park, and Ho-Keun Kwon.

The work was supported by the National Research Foundation of Korea, the Jeongho Kim Neurodevelopmental Research Fund, The Simons Foundation Autism Research Initiative, the National Institutes of Health, the N of One Autism Research Foundation, and the Burroughs Wellcome Fund.

Adapted from an MIT news release.