The gut’s resident bacteria, collectively called the gut microbiota, can have marked effects on brain function and on behaviour — but the mechanisms underlying this interplay remain largely unknown. Writing in Nature, Chu et al.1 define these mechanisms in unprecedented scope and detail. The authors report that mice lacking a complex microbiota exhibit altered fear-associated behaviour, changes in gene expression in cells in the brain, and alterations in the firing patterns and rewiring ability of neurons. The work represents a leap forward in our understanding of the interplay between the gut and brain.
Animals update their responses to environmental cues throughout their lives. This process of behavioural adaptation is driven by underlying cellular and molecular changes in the brain. Chu and colleagues analysed how changes in the gut microbiota affect one such adaptation: fear conditioning.
First, the authors trained mice to associate a tone with an electric shock, and measured how strongly that association was formed. The association developed normally both in control animals and in animals that had been treated with antibiotics to deplete their gut microbiota. The researchers then performed an extinction task, in which they repeatedly played the tone without an electric shock before measuring the rate at which the animals updated their behaviour (such an update indicates that the fear response has been extinguished). The microbiota-deficient mice were unable to update their response, and showed persistent fearful behaviour long after control animals had adapted. Chu et al. found the same phenomenon in mice that had been raised germ-free in sterile isolators and so had never developed a gut microbiota.
The current study is not the first to examine the effects of the microbiota on fear conditioning — previous work has shown a decrease in the acquisition of this response in germ-free mice compared with controls2,3. But Chu and colleagues are the first to report a specific deficit in fear extinction (Fig. 1). What truly sets their work apart, however, is the breadth and depth of the mechanistic findings that they subsequently went on to gather.
Extinction of the fear response is heavily dependent on the function of the brain’s prefrontal cortex4. Chu et al. performed in vivo imaging of this brain region in their animals to analyse both neuronal activity patterns and the formation and elimination of structures called dendritic spines, which are involved in the formation of synaptic connections between neurons. During the fear-extinction test, control animals showed less dendritic-spine elimination and more spine formation than did microbiota-deficient animals. The ability to create synapses and to maintain appropriate existing synapses is a key part of synaptic plasticity — a process crucial to learning and memory, in which the strength of synaptic connections changes in response to changes in neuronal activity. A higher ratio of spine formation to elimination might therefore partially explain why control animals were better able to appropriately extinguish the fearful stimulus.
Tight control of gene expression is also crucial for proper regulation of synaptic and behavioural plasticity. Previous work has indicated that changes in the microbiota alter the gene-expression profile of the prefrontal cortex as a whole5, but Chu and colleagues performed RNA sequencing on single cells throughout the region, enabling them to identify gene-expression changes in individual cell types. These data show that microbiota depletion has a more pronounced effect on excitatory than on inhibitory neurons, setting the stage for future research in which the microbiota could be targeted to alter the characteristics of specific neuronal populations.
The authors’ single-cell sequencing also reveals gene-expression changes in microglia, the brain’s resident immune cells. Previous studies6,7 have shown that altering the microbiota causes changes in microglial gene expression and function. Chu and colleagues found high expression of genes associated with an immature state in the microglia of their microbiota-deficient animals — a change that might affect the cells’ ability to function normally.
In the past decade, it has become clear that microglia have a crucial role in synaptic connectivity. By engulfing and degrading unwanted synapses, the cells ensure that neuronal connections are pruned or maintained as needed8. Changes in this process can alter neurodevelopment9 and are implicated in psychiatric disease10. The researchers’ RNA sequencing revealed changes in genes related to the role of the microglia in synapse organization and assembly. Although Chu et al. did not directly assess changes in the engulfment of synapses, their results lay the groundwork for future research into how interactions between the microbiota and microglia affect synapse density in the brain.
Finally, Chu and colleagues profiled gut metabolites (the molecules produced from metabolic processes) to identify molecules that might drive the gut–brain interactions they had observed. The authors found four metabolites that were significantly less abundant in microbiota-deficient mice than in controls. They therefore posit that the microbiota affects neurons and microglia in the brain through metabolites that are released into the circulation.
The gut microbiota is highly metabolically active, and the theory that the gut and brain communicate through circulating microbiota-derived metabolites is a popular one11. Manipulations of microbial metabolites have been shown to affect a range of behaviours, from autism-like actions12 to those involving reward-seeking for drugs13. Experiments that manipulate levels of the metabolites identified by Chu et al. could improve our understanding of gut–brain communication.
Such research could also reveal a route to translating the current findings into clinical advances. The potential applications are wide-ranging, because alterations in cognition and synaptic plasticity are seen in nearly all neuropsychiatric disorders. Perhaps most germane to the current study would be the treatment of post-traumatic stress disorder, in which people cannot extinguish memories of frightening or traumatic experiences. Chu and colleagues’ work raises the possibility of targeting the gut microbiota and its metabolites as a strategy for helping such individuals. Much remains to be done, but this study is an important step in our mechanistic understanding of the gut–brain axis.