Stress in the microbiome-immune crosstalk

ABSTRACT

The gut microbiota exerts a mutualistic interaction with the host in a fragile ecosystem and the host intestinal, neural, and immune cells. Perturbations of the gastrointestinal track composition after stress have profound consequences on the central nervous system and the immune system. Reciprocally, brain signals after stress affect the gut microbiota highlighting the bidirectional communication between the brain and the gut. Here, we focus on the potential role of inflammation in mediating stress-induced gut-brain changes and discuss the impact of several immune cells and inflammatory molecules of the gut-brain dialogue after stress. Understanding the impact of microbial changes on the immune system after stress might provide new avenues for therapy.

KEYWORDS:

• Stress
• Th17 cells
• microbiota
• behaviors
• immune cells
• cytokines
• quorum sensing molecules

Stress

Stress is a part of modern life in developed societies. Although the definition of stress varies from individual to individual, it is considered to involve external encounters including mental or mechanical challenges, which can be actual or perceived. It is accepted that a low level of stress often is beneficial for the individual boosting the immune system, whereas continuous exposure to high level of stress can have detrimental consequences for the organism, leading to mental illness or other serious medical conditions.^1,2 The consequences of stress often differ from individual to individual. And it has been proposed that the timing, magnitude and duration of the encounter stress might be critical to prime individual to the potential consequences of a subsequent stress exposure.³ Thus, on one hand, childhood trauma is strongly associated with later in life development of depression, whereas on the other hand, repeated exposure to stress can be associated with resilience.

Information about the stressor is often perceived by sensory systems in periphery that relay the information to the brain. Many mediators of stress have been uncovered including the canonical hypothalamic pituitary adrenal axis (HPA)⁴ or the adrenaline/noradrenaline fight or flight response.⁵ These pathways aim at preparing the body to respond to the insult. In this review, we will focus on the effects of stress on the gut microbiome and reciprocally the effects of the gut microbiome on the stress response in the context of inflammation.

Stress affects the physiology of the entire body and significantly impacts the gut in an unpleasant way, often leading to dysbiosis, and leaky gut.

The gut microbiome is a fragile ecosystem composed of bacteria, protozoans, virus, fungus, archaea. Most research has concentrated on the bacteria, but new knowledge of virus and fungi in the context of the microbiome and health has also emerged. The definition of a healthy microbiome remains to be determined. Overall, the gut mycobiome is less taxonomically diverse and heterogeneous than the gut bacteriome and the virome between individuals. The geography accounts for 30.8%, the ethnicity for 15.6%, the diet for 9.8% and host factors and environmental factors account for the remaining variations of the microbiome variations across healthy individuals.^6,7

Bacteria

The human intestine harbors ~ 100 trillion bacteria that are crucial for health,⁸ they provide nutrients, process complex polysaccharides from the diet, and are required for a healthy immune system and neurodevelopment.^9,10 Bacteria in the gut form their own ecosystem but are also directly influenced by human physiological signals and vice versa. The composition of the gut microbiota varies among individuals but is estimated of ~ 1,000 species and > 7,000 strains. Bacteroidetes and Firmicutes are predominant, while Proteobacteria, Actinobacteria, Fusobacteria, Verrucomicrobiota and others are also present. Microbiota changes have been reported in depression (see section C).

Fungi

Fungi are an integral part of the human gut and represent a small (0.1%) but crucial component of the microbiome and although their function in the gut is not well-established, they regulate host homeostasis and pathogenesis.^11,12 Fungi are early colonizers of the gut and are influenced by diet intake and environment. The mycobiome is highly variable between individuals and over time, compared to the gut microbiota.^11,13 Some fungi are beneficial, such as Saccharomyces boulardii, which is used as probiotics, while others are associated with diseases such as autoimmune^14–16 or neurological disorders.^17,18 In diseases, it appears that the increased abundance of Candida albicans is associated with inflammatory bowel disease,^14,16,19 asthma,²⁰ liver disease,²¹ schizophrenia^22,23 and COVID-19.²⁴ Candida albicans is an interesting example of an interaction between a fungus and the host. Candida albicans drives anti-fungal T helper (Th) 17 cell-mediated immune responses. Mannans present in the wall of fungi are sufficient to increase Th17 differentiation and IL-23 responses, as well as toll-like receptor-, C-type lectin receptor-, and Nucleotide-binding oligomerization domain-containing protein (NOD)-like receptor-mediated anti-fungal actions.^25–27 Recent findings show that the mycobiome is altered in children and adolescents with depression.²⁸ Overall, it remains to be determined how fungi affect the host and whether their interactions with bacteria are critical for their role in diseases (e.g., fungi colonization in germ-free mice did not affect the host physiology,²⁹ and even less is known about the role of fungi in the stress response.

Viruses

Healthy human guts are also colonized by a great variety of viruses³⁰ including bacteriophages that infect bacteria, viruses that infect archaea, viruses that infect human cells or viruses present in the food. It seems that viruses resident of the gut do not have an envelope.

Like the bacteria and fungi, there is an inter-individual variation in the composition of the virome. Depressed patients have increased levels of Siphoviridae and Microviridae.³¹ Shifts in Caudovirales bacteriophage are also present in MDD,³² suggesting that stress, which is a major contributor to depression, might also affect phage biology. Although the relationship between viruses and stress is less studied than the relationship between bacteria and stress, it is accepted that stress by suppressing the immune system allows the resurging of latent viruses. Thus, for example, virus-specific antibodies and T cell responses to hepatitis B are suppressed by acute stress.³³ This might underline the known positive association between psychological stress and increased risk of viral infections.^34,35 But it is also possible that phages by modifying bacteria, shape bacterial communities which influence host physiology. Cocktails of phages have been used to treat bacterial infections,³⁶ and prophages seem responsible for the effects of food (e.g., sweeteners) in inhibiting bacterial growth for example.³⁷

Because stress is dependent on various factors and individual physiology is different, it is difficult to draw the line between beneficial and detrimental effects of stress for each individual, which might explain the complexity of the effects of stress on the gut-brain axis.

Stress and dysbiosis

Dysbiosis is defined as changes of the bacterial composition, bacterial metabolism, or bacterial distribution. This can translate into a loss of beneficial bacteria, an outgrowth of pathogenic bacteria, an overall reduction of bacterial diversity or any combination of those,³⁸ and a loss of bacterial diversity is often considered harmful to the host.

After stress, the blood including the immune system, and the autonomic systems brings stress signals to the gut that favor the bloom of pathogenic bacteria that promote dysbiosis and gut barrier permeability.³⁹ Thus, stressed people have unpredictable dysbiosis,⁴⁰ less health beneficial gut bacteria as exemplified in university students after chronic stress,⁴¹ and increased gut permeability after the speech stressor test.⁴² Furthermore, couples with hostile relationships have greater gut leakiness than couples with less hostile relationships,⁴³ suggesting that chronic stress can induce detrimental effects on the gut and the microbiome. In addition, stress affects multiple regions and habitats of the intestine within the lumen and the mucosal lining of the gut, showing a wide effect of stress in the gut.

The causes of dysbiosis are multiple and involve host factors such as genetics, health issues (e.g., inflammation or infections), lifestyle and environmental factors (e.g., diet, medications, hygiene). Diet is one of the major factor contributing to dysbiosis, and is affected by stress, offering a way whereby stress promotes dysbiosis. As an example, diet enriched in simple sugars is sufficient to disrupt gut permeability, induce inflammation and negatively change the host metabolism, and these effects are dependent on the microbiota as microbial depletion abolishes these effects.⁴⁴ It is in the common belief that people eat sugary treats when feeling stressed. Negative emotion is associated with increased taste-based eating⁴⁵ and unhealthy eating is promoted by mild stressors.⁴⁶ For example, cities whose NFL football team lost on Sunday saw an increased intake of saturated fat on Monday compared to a decrease in cities whose team won, and no change in cities whose NFL team did not play, or without a NFL team.⁴⁶ It has been proposed that the executive function in response to food cues is deactivated after stress favoring comfort foods.⁴⁷ Reciprocally, high consumption of sugars has been associated with cognitive impairments, amplifying the unhealthy eating induced by stress. In addition to the diet, stress also affects the host metabolism, which also impacts the gut microbiome.⁴⁸

The change of gut permeability after stress is amplified by inflammation, since the loss of gut permeability after stress allows the entry of commensal and pathogenic bacteria and food antigens to the lamina propria inducing local inflammation which often propagates to the blood circulation, inducing systemic inflammation.⁴⁹ However, the mechanisms, whereby stress promotes dysbiosis remain mainly unknown.

Stress hormones and the gut

Stress hormones (i.e., corticosterone metabolites, catecholamines) have been found in the stools after various stressors.⁵⁰ Stressed germ-free mice have increased level of corticosterone even though the basal level of corticosterone is unchanged compared to specific pathogen free mice in unstressed mice.^51–54 Stress hormones are released into the gut from the blood, but some are also produced in the gut by the enteric neurons, the gut epithelial cells or the enterochromaffin cells.⁵⁵ Stress hormones directly affect bacteria level. For example, the serum level of cortisol is a negative predictor of the level of Rumunococcus spp. in pigs.⁵⁶ Similarly, saliva level of cortisol is negatively correlated with the alpha and beta diversity of the gut microbiota in humans after stress⁵⁷. Catecholamines increase the bacterial growth of certain gram-negative strains⁵⁵ to the point that certain bacteria levels are increased by a 10,000 fold and infectiousness increases within 14 h in vitro.⁵⁷ Various studies have shown the effects of stress hormones on the gut microbes.^55,56,58,59 HPA activation such as maternal separation in rhesus monkeys or early-life stress in rats results in a reduction of the Lactobacilli Gram-negative bacteria,^60,61 and a change of the composition of the gut microbiota,^62,63 respectively.

Although glucocorticoid receptors have not been identified on bacteria, catecholamines receptors (QseBC) are present on certain bacteria, such as E. coli,⁶⁴ suggesting that the effects of stress hormones can directly influence bacterial metabolism (Figure 3). Indeed, in response to catecholamines, QseBC induces the transcription of flagellar regulon in E. coli.⁶⁵ Cortisol also impacts the community-wide transcriptome of the oral microbiome.⁶⁶ Other receptors such as BasRS and CpxAR^67,68 or KdpD and VicK (WalK)^69,70 have also been described as putative adrenergic receptors in bacteria. Catecholamines have antimicrobial activity on selective bacteria in vitro⁷¹ and in vivo⁷² reducing the growth of Staphylococcus aureus or Candida albicans, for example.⁷¹ However, growth of other bacteria is increased by catecholamines in minimum medium- or low-iron medium,^73,74 suggesting a role of catecholamines on iron utilization and amino acid biosynthesis. Catecholamines also affect bacterial chemotaxis, or their ability to migrate where conditions are beneficial for their growth.^75–82 The effect on motility seems dose dependent as low dose of noradrenaline decreases motility genes of P. aeruginosa PA14 whereas a high concentration of noradrenaline increases them,⁷⁸ showing the importance of the concentration of stress hormones for downstream effects on microbes. Catecholamines also influence the formation of bacterial biofilm, which is required for the bacteria to colonize the host and induce pathogenicity,^83–85 promoting the biofilm’s thickness of E. coli for example.⁸⁶ Other properties of bacteria are affected by catecholamines, such as sensitivity to antibiotics,⁸⁷ or production of metabolites.^88–90 In sum, the effects of stress hormones on bacteria are large, offering a possible explanation of the inter-individual variability of stress effects on microbiome.

In addition to modulate the composition and the properties of the gut microbiota, stress hormones also affect the host gastrointestinal properties. Thus, noradrenaline has been shown to affect colonic contractions.⁹¹ Increased gut permeability is only present in those with elevated levels of cortisol and besides cortisol, mast cells are also necessary to weaken the gut barriers to promote bacterial leakage. In addition, the adherence of bacteria to the gut mucosal lining and the uptake of bacteria into the Peyer’s patches are increased by adrenocorticotropic hormone (ACTH), cortisol and catecholamines.^92–95 Taken together, the stress hormones have a profound impact on the gut microbiome.

Microbiome and stress-related behaviors

Several reports point out to the role of the microbiota in stress-related behaviors such as anxiety and depression. For example, germ-free mice^52,96,97 or mice treated with antibiotics⁹⁸ exhibit reduced anxiety-like behaviors, whereas germ-free mice receiving stress-prone microbiota exhibit depressive-like behaviors.⁹⁹ Importantly, it has been demonstrated that stress exposure early in life or during adulthood affects the microbiota composition, suggesting that microbial populations shape the stress responsiveness of an organism.^62,100–104 Although there is not a consensus about which bacteria are affected by stress, exposure to certain bacteria changes behaviors. Thus, exposure to Citrobacter rodentium, Trichuris muris, and Campylobacter jejuni increases anxious behavior in rodents.^105,106 Segmented filamentous bacterium promotes depressive-like behaviors,¹⁰⁷ whereas exposure to Lactobacillus spp. and Bifidobacterium spp. reduces anxiety-like and depression-like behaviors.^108–111 Furthermore, diets that modify the microbiota such as prebiotics and probiotics have been shown to reduce stress-related behaviors.^112–117

Consistent with, changes of the microbiome composition have been associated with psychiatric diseases, in particular depression and anxiety^106,118,119 and administration of the probiotic Lactobacillus farciminis prevents the gut leakage and the activation of the HPA axis in depressed patients,^111,120 pointing to a role of the microbiota in depression. Furthermore, there is an increased risk of severe mental illnesses after antibiotic use, which reduces the gut microbiota diversity,^121–123 including a direct association between antibiotics use and subsequent development of depression.¹²⁴ Similarly, exposure to environmental factors or unhealthy diet known to impact the microbiome increases the incidence of depression.^125,126 Although modulation of the microbiota affects both cognition and mood,^{106,116,127–132} the mechanisms whereby the gut microbiota promotes depression remains to be fully understood. The findings about the changes of gut microbiota composition in depressed patients are inconsistent. Overall, depressed patients seem to have a shift toward pro-inflammatory bacteria,^133,134 with an enhanced Bacteroidetes/Firmicutes ratio.^135,136 One of the most consistent findings in depressed patients is an enrichment of the genus Bacteroides and Eggerthella and a depletion of the genera Blautia, Faecalibacterium, Coprococcus, and Sutterella.^{32,133,134,137,138} Transfer of human fecal microbiota from depressed patients to germ-free mice confers depressive-like behaviors to the recipient mice compared to mice receiving the microbiota from healthy patients, showing that the gut microbiome is sufficient for inducing depressive-like behaviors in mice.^136,139

Gut metabolites are also released by gut bacteria or by the host in response to microbial changes.¹⁴⁰ These include host molecules modified by bacteria such as bile acids or products of bacteria.¹⁴⁰ Thus, because the composition of the microbiome is changed in depressed patients, the microbial metabolome is also affected. For example, the level of the bile acids modulated by the microbiota is inversely correlated with the severity of depression symptoms,¹⁴¹ suggesting that bile acids could provide anti-depressant actions. Furthermore, in depressed patients, microbial metabolites such as short chain fatty acids (SCFA) including acetate, propionate, and butyrate, are reduced.^136,142 Butyrate, for example, has been shown to exert antidepressant actions when administered to rodents, reducing gut permeability and stress responsiveness.^143,144 Other microbial metabolites, such as increases of trimethylamine-N-oxide from choline, lipopolysaccharide (LPS), lactate and B vitamins have been associated with depression.^145–148 Neurotransmitters [e.g., serotonin or γ-aminobutyric acid (GABA)] are also produced by microbes and can modulate behaviors.¹⁴⁴ This suggests that in addition to the microbiome composition, the metabolism and the metabolites of the microbiota are critical to modulate the stress response. Although some fungi metabolites can be detrimental for health such as mycotoxins,¹⁴⁹ others have beneficial effects on health such as antibiotics,¹⁵⁰ immunomodulatory metabolites such as cyclosporin A¹⁵¹ or mood modulator such as psilocybin.¹⁵²

Quorum sensing molecules and stress

Chemical signaling molecules called quorum sensing molecules (QSMs) are used by bacteria to communicate, sense, and respond to environmental changes dependent on bacterial cell density.¹⁵³ They regulate synchronicity of bacterial group behaviors, known as quorum sensing.¹⁵⁴ In the gut, they regulate many of the interactions of the bacteria with the host environment to promote bacterial survival, including facilitation of bacterial attachment, nutritional supply, competition, and motility.¹⁵⁵ Quorum sensing signaling molecules (autoinducers) are differently produced between Gram-negative and Gram-positive bacteria. N-acyl homoserine lactones (AHL) are mainly produced by Gram-negative bacteria (Figure 1), while autoinducer peptides (AIP) are mainly used by Gram-positive bacteria.¹⁵⁶ AHL have been found in the upper intestine of both human and mice.¹⁵⁷ Autoinducer (AI)-2 is considered a universal QSM produced by both Gram-negative and Gram-positive bacteria and is produced by > 50% of sequenced bacterial species. AI-2 mediates interspecies communication throughout the bacterial kingdom,^158–160 is synthetized by LuxS¹⁶¹ (Figure 1) and engineered increase of AI-2 level in the gut favors the expansion of the Firmicutes phylum.¹⁵⁵ AI-2 regulates niche-specific behaviors (e.g., cell division, virulence, motility, and biofilm formation) of both commensal and pathogenic bacteria.^162,163 In the gut, Firmicutes use quorum sensing mediated by peptides [e.g., peptides produced as secondary signaling molecules after degradation of lipoproteins into the extracellular environment.¹⁶⁴ To survive in the gut, bacteria adapt their metabolism through metabolite-sensing mechanisms in a mutual relationship with the mammalian host.¹⁶⁵ Consistent with this, QSMs have been implicated in human diseases such as cancer and autism^{107,166–170} and more recently in the stress response.¹⁰⁷ Thus, AI-2, but not AHL is sufficient to promote susceptibility to stress-related behaviors in mice by acting on Segmented filamentous bacterium.¹⁰⁷ It is important to note that some bacteria can either not produce the autoinducer but can detect it, or can produce it but not detect it, the latter being especially true for AI-2. Altogether, stress has a profound impact on the gut microbiome, and reciprocally the gut microbiome is sensitive to stress signals.

Figure 1. AI-2 system in E. coli and AHL system in gram negative bacteria.

In E. coli, AI-2 is produced by LuxS, while AHL in Gram negative bacteria is produced by LuxL. In E coli, AI-2 is sensed by LsrB, which captures AI-2 into the bacteria. AI-2 is then phosphorylated by LsrK. Phosphorylation triggers AI-2 degradation. AHLs, in contrast, bind to LuxR to modulate the expression of genes required for virulence, biofilm formation, etc.

Gut-brain axis, immunity, and stress

Early colonization of the gut is pivotal for the maturation of the immune system.¹⁷¹ The gut is exposed to commensal bacteria and their by-products, food and environmental antigens and exhibits a 3-layers barrier to prevent these to reach the lamina propria or the circulation: 1) a thick layer of mucus composed of hyperglycosylated mucin-2, produced by goblet cells, which provide protection by forming a shielding barrier, 2) a monolayer of enterocytes linked with tight junctions to restrict trans-epithelial passage of microbes and metabolites and 3) a gut vascular layer of endothelial cells, pericytes and glial cells.¹⁷² Associated with those layers, the gut hosts more than 70% of the total immune cells of the body including innate and adaptive immune cells¹⁷³ (Figure 2). Dendritic cells in particular help maintaining the compartmentalization of the enteric microbiota.¹⁷⁴

Figure 2. Effects of the microbiota on intestinal immune response.

Signals (QSM, bacteria, microbial metabolites, neurotransmitters, neuropeptides) originating from the microbiota affect the intestinal epithelial cells, enterochromaffin cells, or diffuse through the intestinal barrier, to modulate immune cells (T cells, dendritic cells, etc.) present in the lamina propria. Microbial signals and immune cells often reach the bloodstream and migrate to the brain. Here are presented cells relevant for the stress behavioral effects.

The importance of the microbiota for maintaining immune responses in the gut is shown using microbiota-depleted mice or germ-free mice. Germ-free mice have immunological defects such as abnormal germinal center, smaller Peyer’s patches and lymphoid follicles¹⁷⁵ and reduced protective IgA.¹⁷⁶ These defects are corrected when a specific-pathogen-free single mouse is placed in the cage of germ-free mice. Because the gut immune cells monitor potential infections and are in contact with microbial antigens, gut bacteria promote the development of T helper (Th) 17 cells,¹⁷⁷ regulatory T cells (Tregs)¹⁷⁸ and memory T cells.^179–181 Thus, Segmented filamentous bacterium promotes Th17 cell differentiation¹⁷⁸ while Clostridium spp. induce Tregs.^182,183 In addition, commensal bacterial antigens shape immunoglobulin repertoires in the gut,¹⁸⁴ interacting for example, with the carbohydrate antigens of the ABO blood group to stimulate IgM antibody response¹⁸⁵ and protect from induction of mucosal IgE, which is associated with susceptibility to allergies.¹⁸⁶ Toll-like receptor (TLR)5 recognizes flagellin of bacteria and is responsible for containing colonization by flagellated bacteria during the neonatal period.¹⁸⁷ It is also thought that these interactions occur early in life to shape the immune response in adult.

However, the long-term impact of subtler dysbiosis early in life remains to be determined.

Neurons and astrocytes also participate in intestinal immunity. Thus, for example, astrocytes promotes group 3 innate lymphoid cell (ILC3) production of interleukin-22 and reciprocally immune cells confer survival signals to neurons to prevent inflammation-dependent bowel dysmotility.¹⁸⁸

There is a bidirectional communication between the brain and the gut during homeostasis, which involves multiple pathways including endocrine (HPA axis), neural (vagus nerve) and immune (mucosal, peripheral and central) pathways to detect noxious stimuli which trigger local and systemic inflammation.^189,190 The orchestration of the bidirectional communication in the stress response remains to be fully elucidated, including the descending pathways after stress affecting the gut microbial composition (Figure 3). Several ascending pathways controlling the communication from the gut to the brain have been identified. Thus, for example, microbiota changes often trigger local inflammatory responses, ranging from production of cytokines to differentiation of immune cells and production of antibodies or anti-microbial peptides (Figure 2). The immune cells are programmed to recognize microbes, and any inflammatory response will be induced when the immune system is in contact with microbes even commensal bacteria. One example is the lipopolysaccharide (LPS), present on Gram-negative bacteria, also known as endotoxin. Usually contained within the gut by the intestinal barriers,¹⁹¹ leakage of LPS into the lamina propria usually activates macrophages through TLR4,¹⁹² and TLR4 also mediates stress-related behaviors.^107,193 TLR4 might be one of the pathways whereby stress by modifying the composition of the gut microbiota, increases pro-inflammatory cytokines (e.g., IL-6, IL-1$\mathit{\beta }$, TNF and IFN$\mathit{\gamma }$) and chemokines,¹⁹⁴ often leading to the disruption of the gut barrier which increases gut permeability and the release of LPS into the circulation.^195,196 This is reversed by probiotic agents.^61,112,196 This finding is particularly relevant for the stress response, as circulating inflammatory signals (e.g., cytokines and LPS) are sufficient to promote stress-related behaviors in mice¹⁹³ and promote depressive symptoms in human.¹⁹⁷

Figure 3. Effects of stress on the microbiota and the immune response.

Stress hormones such as norepinephrine, epinephrine, and cortisol, impact the microbiota diversity and metabolites. Associated with stress, there is an increase of cytokine production but a decrease in various populations of immune cells. The increase of proinflammatory cytokines leads to increased differentiation of Th17 cells.

Other receptors such as NOD-like receptors shape the microbiota. Thus, NOD1 assists adaptive lymphoid cells to maintain intestinal homeostasis.¹⁹⁸ NOD2 is a bacterial sensor of intestinal epithelial cells and immune cells recognizing peptidoglycan-conserved motifs and induces anti-microbial and anti-inflammatory responses.¹⁹⁹ Once activated by commensal bacteria, NOD2 triggers survival and regeneration of the intestinal epithelium,²⁰⁰ controls the activation of antigen specific CD4 T cells including Th17 cells²⁰¹ and suppresses the expansion of proinflammatory bacteria.¹⁹⁹ NOD1 and NOD2 synergize with TLR4 to exacerbate sickness behaviors,²⁰² whereas NOD1 depletion promotes susceptibility to stress-related behaviors.²⁰³

The Aryl hydrocarbon receptor (AhR) is expressed on intestinal epithelial cells but also immune cells. AhR is a ligand-dependent transcription factor, essential to detect bacterial metabolites. It is important to note that AhR binds various ligands besides microbial metabolites, such as phytochemicals, AhR-active tryptophan metabolites, dietary or pharmaceutical components.^204–207 AhR has been recently shown to sense stress and to induce anxiety-like behaviors,²⁰⁸ suggesting that the change of microbiome after stress might be detected by AhR. Some AhR modulators act as antidepressant in female mice,²⁰⁹ although the mechanisms whereby they induce antidepressant actions remain to be determined.

Metabolite sensing G protein coupled receptors (GPCRs) are expressed on intestinal immune and nonimmune cells. GPCRs recognize microbiota-derived metabolites such as SCFA, amino acids, bile acids, lactate and promote the integrity of the intestinal barrier, inducing anti-inflammatory response and reducing the recruitment of pro-inflammatory immune cells by enhancing Tregs promotion. Overall, both the intestine and the immune system have a variety of receptors to sense the microbiome.

Consistent with this, transplantation of the microbiota of mice exposed to a stressor into germ-free mice induced exacerbated inflammatory responses to Citrobacter rodentium infection.²¹⁰ Some have proposed a role for regulatory T cells and the secretion of IL-10 in mediating the probiotic effects of microorganisms,^211–213 while other have focused on proinflammatory T cells to promote stress-related behaviors. Thus, stress-induced dysbiosis triggers the release of the proinflammatory cytokine, IL-17A, by Th17 cells to amplify the stress response.²¹⁴ The commensal bacteria, Segmented filamentous bacterium, for example, is sufficient to promote Th17 cell- induced depressive-like behaviors in mice and Segmented filamentous bacterium-specific Th17 cells have been shown to migrate to the hippocampus after stress.¹⁰⁷ Recently, Lactobacillus- dependent colonic IL-17A-producing $\mathit{\gamma \delta }$T cells were shown to promote chronic social defeat²¹⁵ and IL-17A-producing $\mathit{\gamma \delta }$T cells also promote anxiety-like behaviors,²¹⁶ pointing to the importance of IL-17A in promoting microbial-dependent stress-related behaviors. The importance of Th17 cells in mediating the effect of the microbiota after stress was further demonstrated using fecal transfer. Thus, fecal transfer of stools of depressed patients into germ-free-like mice induces depressive-like behaviors in the recipient mice,²¹⁷ whereas this is abolished in germ-free-like mice deficient of Th17 cells,³¹ confirming the importance of Th17 cells in mediating microbiota-dependent depressive-like behaviors. In addition to T cell change, stress also affects the number of myeloid cells in the blood²¹⁸ and this is dependent on the microbiota,²¹⁹ since the microbiota promotes the production of monocytes^220–223 and the trafficking of monocytes.¹⁴³

Phages from pathogenic bacteria also interact with the immune system. They are engulfed by dendritic cells, monocytes, and B cells, to induce type I interferon responses through TLR3 signaling.²²⁴ E. coli phages isolated from human feces induce Th1 and Tc1 responses in germ-free mice at mucosal sites while phages from other bacteria stimulate cytokine production via TLR9,²²⁵ suggesting that the changes of bacteriophages observed in depressed patients might also impact the immune responses of these patients.

Bacterial metabolism furthermore impacts the immune response and ultimately the brain. SCFAs effects on inflammation differ depending on the type and concentration of SCFAs. Butyrate, for example, has anti-inflammatory properties.^226,227 SCFAs levels strongly correlated with better cognitive test scores after stress in adolescents.²²⁸ SCFAs ameliorate depressive-like behaviors induced by chronic stress^143,229 or reward seeking behavior after psychosocial stress.¹⁴³ Thus for example, microbes through the activation of the SCFA-GPCR pathway^230,231 promote the production of glial-cell derived neurotrophic factor (GDNF),²³² which enhances the production of ILC3 and IL-22 to provide protection and restoration of the intestinal barriers.²³³ Bacterial metabolites induce the release by intestinal cells of local peptide neurotransmitters, including peptide tyrosine-tyrosine (PYY) and serotonin,^234,235 in addition to bacteria producing nitric oxide,²³⁶ acetylcholine,²³⁷ noradrenaline,²³⁸ GABA^239,240 and dopamine. Enteric neurons express specific receptors for some of these neurotransmitters,^234,235 while other neurotransmitters have been shown to affect directly the brain.^239,240 Several neurotransmitter released by gut microbes induce TLR signaling on epithelial, immune, and neuronal cells,^241,242 suggesting that once activated by microbes the immune system could influence neurotransmitter levels that affect the stress response. In addition, microbial metabolites alter the vagus nerve signaling to the brain, representing another route whereby bacteria by modulating cytokines and neurotransmitters send signals to the brain. Taken together, the microbiota-inflammation axis is critical to promote stress-related behaviors impacting well-being, and possibly influencing other health outcomes.

Impact of inflammation and the microbiota on the brain physiology

In addition to understand the impact of the microbiota on the immune system, germ-free mice have also been instrumental in understanding the impact of the microbiota on brain physiology ranging from early brain development to behavioral tasks.^243–245 Thus, structurally, the brains of microbiota-depleted mice exhibit hypermyelination of the prefrontal cortex,^246,247 increased hippocampal concentration of serotonin,⁵² altered expression of brain-derived neurotrophic factor (BDNF) in the hippocampus, cortex and amygdala,^51,96 increased neurogenesis in the hippocampus,¹⁰ reduced hippocampal dendritic spine density but increased basolateral amygdala dendritic spine density.²⁴⁸

Inflammation and changes of the microbiota have been shown to modulate the permeability of the blood-brain barrier. The depletion of the gut microbiota in germ-free mice is associated with decreased expression of tight junction proteins leading to structural alterations of the blood-brain barrier (BBB) and leaky BBB.²⁴⁹ Of note, colonizing germ-free mice with stool of specific-pathogen-free mice, decreases the permeability of the BBB,²⁴⁹ showing the importance of the gut microbes for the BBB integrity. Depletion of the gut microbiota with antibiotics in mice results in novel object recognition impairment²⁵⁰ and a decrease of hippocampal neurogenesis and memory retention.²²³ These impairments are reversed by administration of probiotic bacteria. Moreover, colonization with the butyrate producer Clostridium tyrobutyricum or treatment with sodium butyrate of germ-free mice increase tight junction protein expression improving BBB integrity.²⁴⁹ Sodium butyrate also prevents BBB breakdown after traumatic brain injury and promotes neurogenesis.^251–253

The neurotransmitters produced by the gut microbiota could also affect brain function. Thus, Lactobacillus and Bifidobacterium spp. metabolize the most abundant amino acid: glutamate, which is an excitatory neurotransmitter in the brain, to produce GABA, an inhibitory neurotransmitter.²⁴⁰ The conversion of tryptophan present in various food²⁵⁴ to the neurotransmitter tryptamine is achieved by decarboxylases secreted by Clostridium sporogenes. Yet, tryptamine is necessary for the release of serotonin by intestinal cells.^255,256 Consequently, germ-free mice have higher plasma tryptophan levels than specific-pathogen-free mice. The level of tryptophan in germ-free mice are normalized after colonization with the microbiota of a specific-pathogen-free mouse.⁵² Similarly, colonization with Bifidobacterium infantis increases plasma levels of tryptophan.²⁵⁷ The consequences of the conversion of tryptophan to indoles or other metabolites²⁵⁸ (e.g., kynurenic acid) by bacteria are anti-inflammatory²⁵⁹ and neuroprotective.²⁶⁰ For example, indoles are ligands of AhR that attenuates disease severity in the experimental autoimmune encephalomyelitis model.²⁶¹ In contrast, the production of quinolinic acid has neurotoxic effects disrupting the BBB and has been implicated in the etiology of psychiatric disorders and neurodegenerative diseases.^262,263 Therefore, understanding the role of tryptophan and glutamate-metabolizing bacteria may open new therapeutic avenues to 1) control BBB permeability, 2) enhance anti-inflammatory effects and 3) induce brain protection.

Microbiota is essential for the microglia development and function.^264,265 Microglia respond to factor produced by the microbiota as shown by the difference in structure and function of microglia of germ-free and specific-pathogen-free mice.^9,266,267 Microglia of germ-free mice are more ramified indicative of a resting state. Microglial morphology and density are also affected in germ-free mice and there is an attenuated inflammatory response,²⁶⁸ with male offspring microglia in early development being more affected than female offspring microglia.²⁶⁶ Yet, stress also affects microglia promoting their activation (e.g., increased branching of the microglia, production of proinflammatory cytokines) and activated microglia is associated with depressive-like behaviors.²⁶⁹ Whether microglia mediate microbiota-induced depressive-like behaviors remains to be determined.

Although the mechanisms whereby the microbiota affects the brain remain to be fully elucidated, there is convergent evidence for a role of the microbiota like the role of the immune system in shaping the brain function throughout the lifespan. Whether these changes by the microbiota require immune modulation remains to be determined.

Conclusions

Our understanding of the role of the microbiome in health and disease has advanced tremendously. Yet the characteristics, function and the composition of a healthy microbiome remains to be fully understood. Although the impacts of the microbiota in the stress response are starting to be uncovered, the molecular pathways responsible for these effects have not been yet fully understood. It is now clear that modulations of the microbiome have profound consequences on the stress response. Yet the bacteria responsible for stress effects are started to be identified, whereas other microbiome components such as phages, virus or fungus remain understudied. Many pathways converge on the gut microbiome, and it is unlikely that supplementation with only a cocktail of probiotics will be sufficient to provide relief from stress-related disorders. Diet appears as one of the modifier factor that might be able to control some of the negative effects of the stress response. Yet, with the climate change, it is projected that warmer climate will modify the soil-born fungi and viruses, which will likely impact the gut microbiome, and the stress response. The host response to microbial change involves a variety of metabolic and signaling pathways that could open new avenue of research to develop a new generation of microbiome-targeting drugs for both disease and prevention.

Acknowledgments

Work in Dr. Beurel’s lab was supported by the NIH (MH104656, MH110415). The author has nothing to disclose.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the National institue of Health [MH104656, MH110415].