Advances in Human Microbiome Science: Gut–Brain Interaction

Overview

On March 15, 2016, the Academy's Microbiome Science Discussion Group convened researchers for Advances in Human Microbiome Science: Gut–Brain Interaction, the second of three symposia on the causal relationships between microbiota and disease—this one focused on the microbiome–gut–brain axis. Commensal human colon microbiota are integral to numerous functions that maintain health. While research on these organisms has traditionally focused on disorders of the gut, there is growing interest in their connection to the central nervous system (CNS). The interconnectedness of the gut and the brain—the association between dysregulation of the gut microbiome and psychiatric disorders, neurodegeneration, and impaired brain development, for example—raises the possibility of targeting the microbiome to treat neurological diseases. The meeting featured presentations by scientists studying the gastrointestinal system and the brain and a panel discussion on translating discoveries into therapeutics.

"Friends with brain benefits" was the description John Cryan of University College Cork gave gut microbiota in his talk on microbial regulation of neural function. He discussed the intriguing links between stress, microbiota, brain health, and aging. Maternal separation is a well-defined mouse model of anxiety and depression; it also exhibits gastrointestinal symptoms of gut inflammation, elevated proinflammatory cytokines, gut-barrier permeability, and increased colonic transit. Indeed, the model is used to study both depression and irritable bowel syndrome (IBS). In another animal model of anxiety, prenatal stress produces heightened stress responses and a low-diversity microbiome in adult mice. In humans, maternal stress alters the infant microbiome, which may also be affected by Cesarean section, formula feeding, maternal infection, and antibiotic use. In older adults, stress is associated with proinflammatory immune responses, a hallmark of aging, and changes in gut-barrier function, along with cognitive impairment, anxiety, depression, and social isolation. Health in older adults correlates with a diverse gut microbiome, which is strongly associated with diet.

Gut–brain signaling involves multiple mechanisms. (Image courtesy of John Cryan and Ted Dinan)

These links prompted Cryan's team to study stress resilience and mental health through the lens of the early-life microbiome and to study the changing microbiome as a driver of aging. Differences in immune function in stressed mice correlate with microbiome changes that persist into aging. There may be particular vulnerability in adolescence, when the brain's connectivity changes, its dopaminergic system is maturing, and psychiatric disorders often develop. In animal studies, mice that consume antibiotics in adolescence have increased anxiety-like behavior and altered cognition, but mice that receive antibiotics at younger ages show no differences in anxiety, cognition, stress, or immune responses in adulthood (although they are hypersensitive to visceral pain). Cryan noted that it is difficult to study the lifetime effects of a changing microbiome, because the consequences of past disruption are difficult to detect as such after diversity recovers.

"The field is grappling with the concept of how could bacteria in the lumen of the gut signal to the brain," Cryan said. Several mechanisms may be involved, including gut-barrier function, intestinal immunity, vagus nerve signaling, and neuroactive metabolite production. "Maybe the microbiome is the one pulling the strings of the brain," Cryan said. "Your state of gut will markedly affect your state of mind."

Robert Yolken of the Johns Hopkins School of Medicine introduced the gut's involvement in psychiatric disorders including schizophrenia, depression, and bipolar disorder. In genome-wide association studies most genetic markers for these disorders identify immune system genes. Yolken noted the shift in thinking required for neurologists and psychiatrists, used to regarding these as disorders of the brain only, to consider a role for systemic inflammation (found in 30%–40% of patients) and the gut–brain axis, and to treat gastrointestinal symptoms as non-secondary. The microbiome affects blood–brain barrier permeability in mice, and if the same effect is found in humans the microbiome could potentially be manipulated to change the brain's exposure to systemic inflammation and drugs. If inflammation is causal in psychiatric disorders, anti-inflammatory medications could avoid the cognitive and behavioral side effects of drugs that target brain signaling pathways.

Several psychiatric medications have unknown mechanisms of action that may involve the microbiome. Yolken's team found differences in the throat microbiome—more genes for a Lactobacillus phage found in people with autoimmune disorders and type 2 diabetes—in patients taking valproate, a drug that changes the microbiome and behavior in mouse models of autism. If the drug's efficacy in psychiatric disorders is the result of microbiome alterations, less-toxic drugs with the same effects could improve treatment options. Microbiome changes may also explain metabolic symptoms of psychiatric disorders such as weight gain and low energy; patients with schizophrenia had more genes for pathogenic organisms, decreased metabolism, and increased lipid and sugar transport in the gut. The researchers have begun testing interventions targeting the microbiome, conducting clinical trials of probiotics to prevent a second manic attack in people hospitalized for mania and of a derivative of a compound in broccoli for schizophrenia. The broccoli metabolite has been shown to be beneficial in autism, and patients in the schizophrenia intervention group take a combination of the precursor to the metabolite and the enzyme (a microbial product) needed to produce it.

The gut–immune–brain interactome model in psychiatric disorders. (Image courtesy of Robert Yolken)

IBS symptoms triggered by gut infection and inflammation (bacterial gastroenteritis) can last years, observed Premysl Bercik of McMaster University. Studies suggest a role for dysbiosis in the resulting functional gut disorders, finding for example improved IBS symptoms after a course of antibiotics and, in a meta-analysis, benefits of probiotic treatment. Some patients with IBS have substantial differences in microbiota compared to control subjects.

Bercik described his lab's work to unravel the microbiome–gut connection in mouse models. His team found changes in neurological phenotypes of pain sensitivity and locomotor activity and in amygdala and hippocampus brain chemistry in mouse models after antibiotic and probiotic treatments. Germ-free mice have heightened anxiety phenotypes whether exposed to maternal separation or not, but something interesting happens when the mice are colonized with the microbiota of a normal mouse. Those with a history of maternal separation acquire a distinct microbiome—predicted to produce different levels of key neuroactive metabolites—and exhibit increased anxiety-like and depression-like behaviors. These phenotypes do not change in germ-free control mice after microbiota colonization. Bercik suggested that early-life stress affects the hypothalamic–pituitary–adrenal (HPA) axis, changing gut function and the metabolome, with metabolites then triggering brain changes and anxiety.

Mice colonized with human IBS microbiota show characteristic phenotypic changes in the gut, anxiety-like behavior, immune activation, and altered metabolomic profiles, demonstrating a causal role for microbiota in behavior in mice. The mice also show changes in the expression of genes associated with immunity and anxiety, some of which are part of both inflammatory and neural pathways. For example, there is heightened glucocorticoid receptor signaling, which is linked to gut hypersensitivity and to anxiety phenotypes. In early clinical trials in humans, probiotic treatment improved IBS-associated depression scores and gut symptoms and reduced the activation of brain areas that control emotion.

Gut–brain interaction after early-life stress. (Image courtesy of Premysl Bercik)

The enteric nervous system (ENS) is a neural network extending from the esophagus to the anus, its circuits intertwined in the epithelium. This "second brain" controlling the gut completes most functions autonomously but is also innervated by and communicates bidirectionally with the CNS. "Gastrointestinal function is under neural control," said Vassilis Pachnis of the Francis Crick Institute. His team is investigating how the gut environment influences the development and activity of the ENS.

The ENS develops from undifferentiated neural progenitor cells with neuronal and glial lineages; progenitors invading the gut establish columns of cells that innervate through the mucosa and are important for gut motor activity. This activity, measured via electrical signals called migrating motor complexes, reveals altered architecture of ENS circuitry in germ-free mice—indicating a role for microbiota in ENS development, which continues postnatally after the microbiome is established. Pachnis's team found that mucosal glial cells are mostly absent in germ-free mice but appear upon microbial colonization of the gut. Indeed, glial cells from the ENS's myenteric plexus migrate to the gut mucosa throughout life, perhaps prompted by changes in the gut microbiome. Enteric glial cells can revert to neural stem cells capable of generating neurons and glial cells, and thus repairing damage to ENS networks. The team's research is focused on elucidating the mechanisms by which the conditions in the gut microbiota and immune system regulate the activation of glial cells, their acquisition of stem cell properties, and the structure of ENS circuitry.

Sarkis Mazmanian of the California Institute of Technology turned to the CNS and the gut–brain connection in autism spectrum disorders (ASD), which feature deficits in social interaction, language, and communication, as well as behavioral abnormalities such as repetitive and stereotyped behaviors. Among the comorbidities of autism—diagnosed in 1 in 68 children in the U.S.—are intellectual disabilities, immune and metabolic dysfunction, gastrointestinal symptoms, and traits such as aggression, hyperactivity, and anxiety. Intrigued by a correlation between the severity of children's gastrointestinal symptoms and the severity of ASD-associated behaviors, Mazmanian's team began testing mouse models to find out whether probiotics could rescue autism-like phenotypes.

The maternal immune activation (MIA) mouse model produces offspring with core behaviors of autism and associated neuropathologies such as defective migration of Purkinje cells. In humans, infection during pregnancy is linked to ASD. The altered microbiome of MIA offspring can be rescued by the probiotic Bacteroides fragilis, a microbe with anti-inflammatory effects in the GI tract that treats IBS, arthritis, and colon cancer in mouse models. When MIA offspring receive the probiotic at weaning, they do not develop ASD-associated phenotypes of anxiety, vocalization defects, or repetitive stereotype and compulsive behaviors but retain deficits in social interaction. Surprisingly, the researchers did not find differences in immune activation in the mice after probiotic treatment; Mazmanian hypothesizes that changes in the metabolome and gut barrier may instead explain the phenotypic rescue. Data suggest that B. fragilis repairs tight junctions in the gut lining; children with ASD commonly have problems with gut-barrier function known as leaky gut, which may allow harmful metabolites from an altered microbiome to reach the brain. In a proof-of-concept study to show that a single metabolite could alter behavior, Mazmanian's team demonstrated that 4-ethylphenylsulphate induces anxiety-like phenotypes in healthy mice. This compound is similar to a metabolite used as a urinary biomarker in ASD and is less abundant after probiotic treatment in MIA offspring. These data point to "an active process between the microbiome and the nervous system [that] ... may be metabolic," Mazmanian concluded.

Ted Dinan of University College Cork returned to the topic of stress and the microbiome. He reviewed the mechanisms by which gut microbiota may regulate neural function, but cautioned that although studies support a role for microbiota in stress responses, "it's still an assumption that gut microbes influence the brain and behavior" in humans. Nonetheless, differences in microbial diversity and richness in patients with major depression cannot be explained by diet, and rats with humanized microbiota from patients with depression acquire anhedonia-like and anxiety-like behaviors not found in animals with microbiota from healthy controls. In a blinded study, people who reported feeling less anxious when taking a probiotic had lower cortisol levels in saliva upon waking—a measure of stress and depression—and improved cognition compared to when they were taking a placebo. Dinan and Cryan introduced the term psychobiotics in 2013 to describe bacteria with mental health benefits, and both are investigating the effect.

Psychobiotics in the literature and the media. (Image courtesy of Ted Dinan)

Gut microbiota produce neurotransmitters, cytokines, and short-chain fatty acids that could act in the brain, but it is unclear, Dinan noted, how they would do so directly. "I'm not suggesting for one moment that neurotransmitters produced by microbes in some magical way move from the gut to the brain. I think that's highly unlikely. But that doesn't mean they are not capable of influencing brain function," he said. The levels of these molecules in the gut are transferred along with depressive symptoms by microbiota transplant. Rats with the microbiota of patients with depression, like the patients themselves, have altered tryptophan metabolism (a neurotransmitter precursor the brain needs to produce serotonin and melatonin) and elevated C-reactive protein. Dinan's team linked intake of probiotics to reduced stress, improvements in depressive phenotypes, and altered GABA receptor expression in key brain areas such as the amygdala and hippocampus in rodents. These effects did not appear in animals that had the vagus nerve removed. The team is currently studying these probiotics in early trials in healthy volunteers and patients with major depression.

Dinan pointed to the need for more research into possible interventions for the gut microbiome and noted that antibiotics, not just probiotics, are beneficial in treating some disorders. Gut microbiota are implicated in weight gain and other side effects of antipsychotics and other medications. Given that medications can also change the microbiome dramatically, Dinan predicted that defining these effects will soon be a condition of drug approval.

In the final presentation, Emeran Mayer of the University of California, Los Angeles, began with arguments against simple translation from animal models to humans. Much focus is on the gut's influence on the brain, less is on bidirectional pathways and brain signaling to the gut via the autonomic nervous system and the HPA axis. Stress-associated hormones can also act on microbes and alter their gene expression.

The brain biochemistry and behavioral responses that can be measured in rodents are only approximations of human brain imaging signatures and the subjective emotional, cognitive, and pain responses that humans report. Mice exhibit "evoked reflexive behaviors" while humans have "spontaneous complex emotional feelings, thoughts, and behaviors"; many of the brain regions important in psychiatric disease are small or not present in mice. The brains are so different that Mayer likened the comparison to that of a linear-processing computer with a supercomputer capable of early artificial intelligence. Furthermore, there is no evidence of consistent emotional or cognitive effects in humans after fecal microbial transplant, dietary change, courses of antibiotics or probiotics, or other microbiome-altering factors. The increasing prevalence of autism is an anomaly among disorders linked to the gut microbiome: autoimmune and metabolic diseases are not becoming more prevalent, and neither are psychiatric disorders. "The heterogeneity of human populations resulting from genetic variability, differences in stress reactivity, trait anxiety, and depression ... greatly differ from homogenous inbred mouse populations," Mayer said.

Nonetheless, Mayer's lab has found evidence in humans of changes in specific brain regions after probiotic treatment, although the researchers did not detect changes in mood or gut function. Early data also suggest an association of IBS gut microbiota with grey matter signatures and the structure of some brain regions, but causation has not been established. Mayer proposed a systems approach to the study of gut–brain interactions, accounting for microbiota, transcription, protein networks, metabolites, immune function, noncoding RNA, and neural networks—the outputs of the "gut connectome" communicating with the neural connectome. The team is mapping the connections among networks to identify the most important "nodes" in the system that affect clinical outcomes.

A systems approach to IBS and gut–brain interaction. (Image courtesy of Emeran Mayer)

Use the tabs above to find multimedia from this event.

Presentations available from:
Premysl Bercik, MD (McMaster University Medical Centre, Canada)
John F. Cryan, PhD (University College Cork, Ireland)
Ted Dinan, PhD (University College Cork, Ireland)
Emeran Mayer, MD (University of California, Los Angeles)
Vassilis Pachnis, MD, PhD (The Francis Crick Institute, UK)
Robert Yolken, MD (John Hopkins School of Medicine)
Panel discussion

This eBriefing is part of a series on the causal relationships between microbiota and disease, also including:

Platinum Sponsor

The Microbiome Science Discussion Group is proudly supported by

How to cite this eBriefing

The New York Academy of Sciences. Advances in Human Microbiome Science: Gut–Brain Interaction. Academy eBriefings. 2016. Available at: www.nyas.org/GutBrain-eB