Host-Microbe Interactions: Communication in the Microbiota-Gut-Brain Axis

Abstract

Animals harbor a diverse array of symbiotic microorganisms that coexist in communities across different body sites¹. These microbes maintain host homeostasis and respond to environmental insults to impact host physiological processes. Trillions of indigenous microbes reside in the gastrointestinal tract and engage with the host central nervous system (microbiota-gut-brain axis) by modulating immune responses, interacting with gut intrinsic and extrinsic nervous system, and regulating neuromodulators and biochemicals². These gut microbiota to brain signaling pathways are constantly informed by each other and are hypothesized to mediate brain health across the lifespan. In this review, we will examine the crosstalk of gut microbiota to brain communications in neurological pathologies, with an emphasis on microbial metabolites and neuromodulators, and provide a discussion of recent advances that help elucidate the microbiota as a therapeutic target for treating brain and behavioral disorders.

INTRODUCTION

Over recent decades, foundational microbiota studies from animals and humans reveal microbes play critical roles in promoting health and modulating the development and function of the central nervous system^1,2. While most of this review will focus on evidence of the bacteriome in the microbiota-gut-brain signaling, emerging studies reveal the gut virome (viruses) and mycobiome (fungi) can modify the bacteriome and influence host physiology. For example, change in bacteriophage composition has been associated with Parkinson’s disease³. Fecal virome transplants, prepared from fecal filtrate composed of viral particles, have been used to improve Clostridioides difficile infection, necrotizing enterocolitis, and small intestinal bacterial overgrowth^4,5. In a murine model of chronic variable stress, researchers observed reduced social and increased depression-like behavior, alterations in immune responses, and bacteriome changes. Transfer of the fecal virome prior to the induction of chronic stress protected against stress associated disruptions in behaviors, immunity, and bacteriome⁶. Notably, this protection was associated with the restoration of the bacteriome’s capacity for S-adenosylmethionine synthesis⁶. In a separate study, Caudovirales was associated with better performance in memory tasks in humans, mice, and flies⁷. Similarly, changes in gut mycobiome composition, such as abundance of Saccharomyces cerevisiae and Candida albicans, has been associated intestinal disorders, cancer, metabolic disorders, and neurological disorders including multiple sclerosis, Rett syndrome, autism spectrum disorder (ASD), schizophrenia, and Alzheimer’s disease⁸. Mucosa-associated fungi support intestinal epithelial function via IL-22 signaling and promote social behavior via IL-17 pathway in mice⁹. As such, the virome and mycobiome play an important role in the modulation of the microbiota-gut-brain axis, suggesting that they should be considered when designing microbiome-directed therapies. For the remainder of the review, we will describe the effects of the microbiota in reference to bacteria mediated impacts on host physiology.

In humans, a departure from the homeostatic microbiota composition has been associated with gastrointestinal dysfunction, neuropsychiatric disorders, diabetes, obesity, and autoimmune diseases^10–13. In addition, patients with ASD, mood disorders (anxiety and depression), amyotrophic lateral sclerosis Parkinson’s, and Alzheimer’s have alterations in the gut microbiome composition and metabolite levels, suggesting microbiota changes contribute to nervous system abnormalities^14–17 (Figure 1B). For example, using the Rotterdam Study (1,054 participants) and the Amsterdam HELIUS program (1,539) to study the fecal microbiota composition in relation to depressive symptoms, researchers identified 13 different genera that were associated with increased presentation of depressive symptoms¹⁸. Correspondingly, in invertebrate animals, such as bees, planarians, and drosophila, environmental stressors can change their microbiome and alter stress response, regeneration rates, developmental patterning, memory, food preference and aggressive behaviors^19–24. In other animals, such as birds, diet-induced gut microbiome changes are associated with exploratory behavior and problem-solving performance^25,26. Germ-free (GF) mice, which are raised devoid of microbes, exhibit brain and behavioral changes including increased blood-brain barrier (BBB) permeability, disrupted microglia maturation patterns, reduced thalamocortical axonogenesis, altered neurochemical expression, modified synaptic morphology, abnormal cognitive behaviors and heightened stress responses compared to conventionally raised mice with a complete microbiome^27–34. Further, in preclinical studies using murine models of neurodevelopmental, mood and neurodegenerative disorders, disruption of the gut microbiome has been linked to specific neurological dysfunctions and behavioral abnormalities across different life stages^{14–16,35,36}. Notably, selective colonization with a specific bacterium or a consortium of bacteria in stress, ASD, and ALS mouse models yield beneficial brain and behavioral effects^16,35,36. For example, Lacticaseibacillus rhamnosus regulates γ-aminobutyric acid (GABA) receptor expression in the brain, promotes gut health, and reduced anxiety- and depression-related behaviors in mice and humans^33,37. Together, the microbiota is recognized to be a modifier of human and animal biology.

Figure 1. Gut microbiota modulation of metabolites and nervous system.

(A) Gut bacteria regulate levels of neurotransmitters that act on the vagus nerve, levels of microbial metabolites that cross the blood brain barrier, and levels of precursor molecules that generate neuromodulators to influence brain and behaviors. (B) Gut regulated metabolites and neuromodulators are associated with mouse models of neurological disorders.

As we consider the role of the microbiota in the gut-brain axis it is important to highlight biological determinants of the microbiota such as body site, age, and sex can differentially influence host physiology^29,38,39. For example, in mice, the paternal and maternal microbiome can impact placental development, offspring growth and development trajectories, and fitness^28,40–42. Nonobese diabetic mice show a sex-specific microbiome profile that mediates sex-biased autoimmunity, sex hormone expression levels and metabolic changes, whereby transplanting male specific pathogen-free (SPF) microbiota into female recipients correlated with increased testosterone levels in female mice and protection against invasive insulitis⁴³. Similarly, diet-induced changes to the microbiota are sex-dependent leading to differential effects on locomotion and anxiety-related behaviors⁴⁴. In addition to the gut microbiota, metabolites are also regulated by sex. In a study that profiled the metabolites along the gastrointestinal tract and systemic sites (peritoneal fluid, serum, liver, spleen, and urine) of GF, gnotobiotic-colonized with OMM12 consortium of bacteria, and SPF mice, the metabolome was strongly determined by body site, and secondarily controlled by the microbiota, age, or sex⁴⁵. Together, these findings suggest that sex, along with other biological factors can influence the microbiome and metabolome, and further studies are needed to elucidate how sex-induced changes in the microbiome and metabolome confer health and disease responses in the host.

While we are only at the beginning of understanding the complex communication between the microbiota and brain, in this review, we will examine evidence of how this bidirectional signaling impacts brain functionality and behaviors in health and disease.

Immune and neuronal communication between the gut microbiota and the brain

One mechanism of gut-brain signaling is through microbial interactions with the immune system. The dense population of immune cells in the gastrointestinal tract are in constant communication with microbes and their metabolites to inform immune homeostasis and direct pro-inflammatory or anti-inflammatory responses. Many bacterial metabolites including neuromodulators, bile acids, choline, and short-chain fatty acids (SCFAs) are immunomodulatory^31,46. Microbiota-host interactions at the level of the gut can lead to the release of cytokines and chemokines that enter blood and lymphatic circulation, and influence immune signaling across the body, including the brain^27,31,47. Indeed, in addition to significant brain and behavioral abnormalities, germ-free mice have an underdeveloped immune system and impaired neuroimmune function^48,49. An important component of the neuroimmune system are microglia, which are the brain’s resident immune cells. In germ-free mice, microglia exhibit an immature phenotype and impaired function, which were prevented by conventionalization of the microbiota or supplementation of microbial metabolites such as SCFAs³¹. Further, as reported in germ-free mice, the microbiota regulates development and integrity of the BBB, which serves to regulate the passage and exchange of molecules and nutrients between the circulatory system and brain parenchyma, as well as protect the brain from toxins, pathogens, and inflammation during injury and disease^27,50. Yet, a compromised BBB can result in the crossing of microbial products, and bacterial or viral infections entering the brain via the bloodstream or cerebrospinal fluid, and subsequently induce an immune response against the brain. In profiling the human microbiota, patients with multiple sclerosis and Parkinson’s disease display an altered gut microbiota composition compared to healthy controls^51,52. These shifts in the gut microbiome are associated with immune and neurobehavioral abnormalities. For example, in mice with overexpression of a-synuclein, a model of Parkinson’s disease, absence of the microbiota ameliorates motor impairments and altered microglia activation. Whereas administration of SCFAs promotes motor dysfunction and α-synuclein-stimulated microglia activation¹⁵ (Figure 1B). In a separate study using the experimental autoimmune encephalomyelitis (EAE) model for multiple sclerosis, mice with a deficient or depleted microbiota displayed resistance to brain inflammation and had reduced EAE severity⁵³, thus suggesting the gut microbiome can inform neuroimmune responses that then modify brain pathology and behaviors.

A second mechanism of gut-brain communications is mediated by gut intrinsic neurons (enteric nervous system - ENS) and extrinsic neurons (spinal and vagal afferents) (Figure 1A). The ENS is positioned at the interface of the microbiota and intestinal cells, to respond to the microbiota and its metabolites. To convey sensory information to the brain, ENS communication is relayed from intestinofugal neurons to sympathetic ganglia to primary afferent neurons. Studies of the ENS using germ-free animals show abnormal ultrastructural and neurochemical development, altered neuronal excitability, and deficits in intrinsic sensory signaling^54,55. In addition, colonizing germ-free mice with a conventional microbiota induces maturation of the ENS, and this was dependent on serotonin and 5HT4 receptor⁵⁶. The vagus nerve is the 10th cranial nerve and facilitates bidirectional communication between the gut and brain (Figure 1A). The vagus nerve relays information about gastrointestinal motility, secretion and absorption, epithelial permeability, and is influenced by the digestive environment and composition of the gut microbiome⁵⁷. Recent discoveries reveal the vagus nerve expresses a variety of G-protein coupled receptors (GPCRs)^58–60. Nutrient and sensory signaling through these receptors that are localized to the vagal nerve-brain pathway informs complex behaviors such as food preference, motivation, and reward^60–63. In addition, studies using subdiaphragmatic vagotomy reveal the vagus nerve mediates the effects of the gut microbiome on serotonin and dopamine neurotransmission in the brain and, emotional and social behaviors that are implicated in neurodevelopmental and mood disorders^33,64,65. These immune and neuronal microbiota-gut-brain communications are also highly influenced by endocrine and bioactive molecules. Thus, raising the questions, how do microbial metabolites mediate microbiome signaling to the brain, and how do microbial metabolites influence neurological states. In subsequent sections, we will examine how microbially derived molecules interact with the host and contribute to the brain and behaviors.

Communication pathways of microbially-modulated molecules

Short-chain fatty-acids and amino acids

Another major pathway of microbial influence on the nervous system is through molecules modulated by the microbiome (Figure 1A). Among the different classes of small molecules, SCFAs, which include propionate, butyrate, and acetate have been most examined and reviewed by Dalile and colleagues⁶⁶. SCFAs, products of bacterial fermentation of non-digestible polysaccharides, play a key role in gut-brain signaling pathways, targeting immune, endocrine, and neural cells, which is supported by evidence that germ-free animals and antibiotic-treated animals have considerably lower SCFA levels^67,68. One mechanism that SCFAs can mediate microbiota-gut-brain communication to influence brain physiology and behavior is through epigenetic modifications, whereby all SCFAs have histone deacetylase inhibitory effects that alter learning and memory, depression-like behaviors, and social behaviors⁶⁶. Additionally, the SCFA receptor FFAR3 is expressed on the vagus nerve and SCFA administration can regulate neuronal activity in the dorsal vagal complex, parabrachial nucleus and hypothalamus⁶⁹, suggesting multiple levels of regulation by SCFA on gut-brain signaling. Further, decreased fecal SCFA levels and correlating brain and behavioral changes have been reported in Parkinson’s disease and anorexia nervosa¹⁵. In animal models, SCFAs and SCFA-producing bacteria have neuroprotective and anti-inflammatory effects in neurodevelopmental, autoimmune and neurocognitive disorders, including autism, multiple sclerosis, Parkinson’s, and Alzheimer’s^15,70–72 (Figure 1B). Overall, SCFAs have broad influences on nervous system development and function, but further research is needed to understand how individual SCFAs influence the brain either directly or indirectly, and how these interactions may result in interventions for neurological disorders.

Dietary protein and amino acids are metabolized by intestinal epithelial cells and gut bacteria to mediate microbe-host interactions. Amino acids (AA) have diverse roles in the intestine including, synthesis of molecules for nutrition, host physiology, and intestinal structure and function (eg. proline, neurotransmitters, mucins, immunoglobulins), fermentation of proteins and catabolism of AA for energy (e.g., SCFA, branched-chain fatty acids, phenolic and indolic compounds), and production of AA metabolites that are neuromodulators (eg. GABA, dopamine, serotonin, and nitric oxide). Absence of the microbiota induces lower levels of intestinal AA, as seen in GF animals compared to conventional controls⁷³ (Figure 1A). Indoles derived from dietary tryptophan metabolism by the gut microbiota, such as indole propionic acid, indoxyl-3-sulfate, indole-3-acetate, indole-3-aldehyde, tryptamine, and 3-methyl-indole are aryl hydrocarbon receptor (AhR) or pregnane X receptor (PXR) agonists^74,75. In the EAE mouse model of multiple sclerosis, indoxyl-3-sulfate, in combination with type 1 interferon signaling, activate AhR in astrocytes to reduce inflammation in the central nervous system and EAE clinical disease score⁷⁴ (Figure 1B). Using a sciatic nerve crush mouse model of nerve injury, researchers demonstrated indole-3-propionic acid, produced by Clostridium sporogenes, can promote axon regeneration and recover sensory responses via neutrophil chemotaxis and PXR signaling⁷⁵ (Figure 1B). While many microbial metabolites have beneficial properties, 4-ethylphenyl sulfate, a metabolite of dietary tyrosine, which was increased in a maternal immune activation mouse model³⁶, impaired oligodendrocyte development, decreased oligodendrocyte to neuron interactions, reduced myelination of axons, and induced anxiety-like behaviors in mice³⁵. Together, amino acids and amino acid metabolites maintain intestinal homeostasis and have important functions in regulating neuroimmune signals and neurophysiology. However, further research is needed to understand the bioavailability of amino acids in different neurological conditions, elucidate the exact molecular and cellular mechanisms of how various amino acids interact with the brain, and examine how manipulating dietary amino acid levels may appropriately be used for interventions in neurological disorders.

Neuropeptides and Neuromodulators and receptors

The gut microbiota can influence the levels of peripheral neuroactive molecules and hormones that modulate host physiology. GF mice have significantly reduced intestinal levels of the neurochemicals serotonin (5-HT) and GABA, detectable intestinal levels of dopamine, norepinephrine, and epinephrine⁷⁶ (Figure 1A). Over 90% of the body’s 5-HT is synthesized in the gut, where specific bacteria regulate 5-HT biosynthesis⁷⁷. Intestinal 5-HT is used for a variety of biological functions including immune modulation, bacterial colonization, and host lipid metabolism⁷⁸. However, given that 5-HT does not cross the BBB, it is unlikely that gut derived serotonin directly impacts central neural circuits. Therefore, it will be important to understand how peripheral 5-HT are involved in gut-brain communication and whether it affects neurological function. GABA, the major inhibitory neurotransmitter in the nervous system, is synthesized by Escherichia spp., Lactobacillus spp., Bacteroides spp.^37,79, and likely others as the fecal microbiome encode genes for glutamate decarboxylase, the enzyme that converts glutamic acid to GABA. Given that microbes and sensory neurons that innervate the gut epithelium express GABA receptors, intestinal GABA can influence the microbiota and host.

However, how intestinal GABA may impact microbial function and mediate microbe-host crosstalk remains to be uncovered. A variety of bacteria produce catecholamines, for example, the genus Escherichia produces norepinephrine, and Bacillus synthesizes norepinephrine and dopamine. While enterochromaffin cells respond to catecholamines⁸⁰ (Figure 1A), how bacterial derived catecholamines influence host physiology is unclear.

Molecules modulated by gut bacteria communicate with neuronal and non-neuronal cells through an array of transporters and GPCRs^81–83. GPCR interactions with ligands are vital for signaling sensory information and maintaining homeostasis. Free fatty acid receptors FFAR2 and FFAR3 are GPCRs that respond to SCFAs to regulate activation of spinal cord neurons, gut-associated sympathetic neurons, and microglia^31,84,85. Notably, many microbiota-modulated metabolites are ligands for canonical neurotransmitter receptors such as HTR (serotonin), DRD (dopamine), and HRH (histamine) families^81–83. A number of GPCRs are present on vagal neurons, suggesting microbiota metabolites can signal to the brain through peripheral neurons⁵⁸. GPCRs are implicated in a large number of diseases including cancer, depression, and Alzheimer’s, making them the targets of approximately 40% of all prescribed medications⁸⁶. Thus, understanding the interactions between microbial molecules and host GPCRs along the gut-brain axis will be important for developing microbial metabolite-based therapeutics.

Microbiome and microbial products as therapeutic targets

Pioneering microbiome-centered therapeutic strategies requires an understanding of how microbes interact with the host across various contexts. Current microbiome-based therapeutic modalities, which span prebiotics, probiotics, antibiotics, dietary changes, and fecal transplants, necessitate further investigations to determine their effectiveness as interventions for neurological abnormalities. A primary challenge is defining a healthy versus a disease microbiome that can be reproducibly targeted for microbiome-based therapeutics. Notably, individual demographics - genetics, diet, and lifestyle confound results and affect reproducibility. Further, an incomplete understanding of how different systems within the host responds to microbial changes inhibit effective strategies. That said, the future of microbiome-based therapeutics is promising, as studies continue to refine our understanding of host-microbe interactions along the gut-brain axis.

Fecal microbiota transplantation (FMT) has been most efficaciously employed in treating C. difficile in patients with an 87% resolution after the first FMT⁸⁷. In rodent studies, microbiota transplantation from a healthy control ameliorated gut dysbiosis, cognitive decline, age-associated behavioral impairments, social deficits and depressive behaviors, and in some cases by augmenting the abundance of SCFA-producing microbes^88–91. Other FMT studies have demonstrated that these procedures may apply to other pathophysiologies such as atherosclerosis⁹². Together, FMT can have broad therapeutic implications, including but not limited to neurological disorders. However, it should be noted that these are preclinical findings and further investigations are needed to understand how FMT should be administered, including how to select the donor microbiota, and what are the consequences across systems in the host.

Probiotic supplementation is a form of microbial-based intervention that has historical roots, as humans have been consuming probiotics for thousands of years, mainly in the form of fermented products. As described throughout this review, probiotic use in preclinical and clinical studies confer brain and behavioral benefits. However, the unregulated nature of probiotics, suggest further research on standardizing how to administer the bacteria are needed. For example to ensure that the indicated amount of living bacteria is surviving the journey from manufacture to the intestine, researchers designed a mineral coating to encapsulate Bacteroides fragilis that can withstand O2 exposure, UV radiation, ethanol, and the acidic environment⁹³. Supplementing the coated B. fragilis sufficiently ameliorated body weight loss, intestinal length decreases, TNF-α levels, and structural damage of the colon in DSS-induced mouse model of colitis. Though exciting, this study raises the questions of how does one versus a consortium of bacteria fair in a single capsule, what are the effects of the probiotics on local versus the systemic environment, and are there host specific responses that are shaped by genetics, lifestyle, and diet.

Another strategy of microbiota-centered intervention involves modulating metabolite levels. The microbiota can regulate metabolites, such as carnitine mimics, that can cross the blood brain barrier and localize to specific brain regions⁹⁴. In preclinical studies, administration of metabolites in disease conditions show beneficial effects. In a mouse model of amyotrophic lateral sclerosis, Blacher et al. found decreased nicotinamide metabolism, and supplementing nicotinamide improved motor function¹⁶. In mouse model of peripheral nerve injury, microbially produced indole-3 propionic acid promoted axonal regeneration and thermal sensory responses⁷⁵ (Figure 1B). Finally, in a mouse model of maternal malnutrition, supplementing pregnant dams with SCFAs promoted placental growth and vascularization⁴⁰. Together, rescuing deficiencies in microbial metabolites offers an avenue for microbes to modulate host physiology in numerous disease conditions.

Others have found some microbial metabolites have negative effects on the host. The microbial metabolite 4EPS was previously reported to be elevated in a mouse model of atypical neurodevelopment. In a recent study, 4EPS, produced by bioengineered bacteria, entered the brain, impaired oligodendrocyte maturation, and induced anxiety-like behavior in mice³⁵ (Figure 1B). For example, in mice with impaired gut barriers, the microbial enzyme dipeptidyl peptidase 4 (DPP4) decreased active glucagon-like peptide 1 (GLP-1) and impaired host glucose homeostasis⁹⁵. Inhibiting microbial DPP4, increased active GLP-1 levels and improved glucose metabolism in diabetic mice, suggesting the selective titer of metabolites function can regulate host responses.

Overall, studies with a complete library and characterization of metabolites and their corresponding receptors in varying biological contexts will inform strategies for precision medicine. Our dynamic relationship with microorganisms continues to reshape our understanding of biology in health and disease conditions. Unraveling the mechanisms governing the interplay between host and microbes will reveal the fundamental relationships between the microbiota, gut and brain and offer valuable insights into novel therapeutic approaches for neurological diseases.