It is well established that physical exercise, practised on a regular basis, decreases the risk of developing chronic diseases, improves cognitive function, and decreases the risk of dying prematurely. To be able to take full advantage of these many benefits, it is necessary to exercise regularly and preferably over long periods of time. Yet many people have a sedentary lifestyle, and motivation to exercise is low or non-existent. Motivation to exercise is regulated in the central nervous system and requires signals initiated by dopamine, a neurotransmitter involved in a host of functions including motor control, attention, memory, cognition, sleep, pleasure and motivation. Neurons that produce dopamine are found in regions of the brain called the ventral tegmental area and the substancia nigra. Dopaminergic neurons extend into other parts of the brain to regulate cognitive, emotional, and motivational aspects related to reward-associated behaviours.

Does the motivation to exercise depend solely on our brain and our state of mind regarding this activity? It seems not, according to a recent study carried out on mice which shows that motivation is partly attributable to bacteria present in the intestine. A surprising discovery that is the result of the combined efforts of several teams of researchers.

In order to identify new regulators of exercise performance, the researchers used a cohort of 199 mice with high genetic diversity. The cohort of mice was subjected to extensive genome, metabolome, microbiome analyses, and their exercise performance was evaluated (treadmill, exercise wheel). Genomic analyses suggest that genes contribute very little to the observed differences between the exercise performance of different mice.

Since previous work (see here, here, here, and here) suggested that the microbiome would have a potential role on performance during exercise, the researchers wanted to test whether the variability in the performance of different mice could be attributed to the microbiome, by performing “loss of function” (depletion of the microbiome) and “gain of function” (transplantation of the microbiome) experiments. Complete depletion of the microbiome with broad-spectrum antibiotics caused a decrease in the mice’s exercise performance by approximately 50%. On the contrary, transplantation of the microbiome from exercise-performing mice to “germ-free” mice (raised under sterile conditions and containing no microorganisms) increased the exercise performance of the recipient mice. In addition, the exercise performance of the recipient mice correlated with that of the donor mice. When the broad-spectrum antibiotic treatments were stopped, the exercise performance of the mice returned to normal, as did that of the germ-free mice when they were no longer kept under sterile conditions. Taken together, the results of these experiments suggest that the microbiome strongly contributes to the ability to exercise in mice.

In order to identify the class of microorganisms and more precisely which bacteria contribute to the increase in exercise performance, the mice were treated with narrower-spectrum antibiotics, and the intestines of germ-free mice were colonized with a single microorganism. Among the bacteria tested, those of the genera Eubacterium and Coprococcusimproved the exercise performance of mice, to levels comparable to those observed for mice that received a whole microbiome transplant.

At the mechanistic level, the researchers first tested whether the improvement in exercise performance by the microbiome was not caused by a favourable effect on muscle function. However, the results of several tests indicate that the microbiome has no significant effect on muscle physiology. The researchers’ attention then turned to motivation, one of the important factors contributing to exercise performance, along with musculoskeletal function.

One region of the brain that is particularly involved in motivation control is the striatum. As expected, levels of the main neurotransmitter involved in motivation/reward neural signals in the striatum, dopamine, increased after the mice exercised. However, this increase was much less significant in mice whose microbiome was depleted, indicating a role of the microbiome in the release of dopamine after exercise. Levels of two other important neurotransmitters in the striatum, namely glutamate and acetylcholine, did not change following exercise or microbiome depletion.

How can bacteria that colonize the gut boost dopamine levels in the brain? There are two possible pathways: 1) through circulating factors, i.e., metabolites produced by bacteria or 2) through afferent neural circuits. Proteomic analyses of blood samples did not identify any metabolites significantly associated with exercise performance that are related to the microbiome. The researchers therefore focused on the sensory neurons that innervate the intestine.

The researchers used a line of mice (Trpv1^DTA) in which a large part of the afferent vagus and spinal nerves that express the vanilloid receptor are absent. The exercise performance of Trpv1^DTA mice is low, comparable to that of normal mice whose microbiome has been depleted by antibiotics. Microbiome depletion in Trpv1^DTA mice did not alter exercise performance.

How can gut bacteria activate sensory nerves in the gut? The researchers showed that, in vitro, isolated spinal nerve neurons are activated by fecal extracts from normal mice, but much less by extracts from mice without microbiome. This result suggests that a metabolite from the microbiome is involved in the activation of sensory nerves. Metabolomics analyses identified candidates, several of the most potent of which were fatty acid amides (FAAs), such as N-oleoylethanolamide (OEA).

In order to prove that these compounds alone can boost exercise performance, the researchers introduced supplements of five FAAs to the diets of mice whose microbiome had been depleted by antibiotics. This supplementation restored signals generated by sensory nerves, increased levels of dopamine in the brain, and exercise performance. Then, the clever researchers transformed E. coli bacteria that normally do not produce FAA by introducing the genes responsible for the production of these metabolites. The intestines of germ-free mice were colonized with this bacterium modified to produce FAAs or with the parental line which does not produce FAAs. Exercise performance was improved by colonization with the FAA-producing bacteria, but not by colonization with the parent bacteria. Finally, the researchers showed that the effect of FAAs is mediated by cannabinoid type 1 (CB1) receptors, located in the sensory nerves in the intestine and which are connected to the brain via the spinal cord.

Studies done on mice don’t always translate to humans, but both have a similar endocannabinoid system connected to the ventral striatum. The results of this study suggest possible diet-based interventions to increase people’s motivation to exercise and optimize performance in elite athletes.