Preimmune Recognition and Response to Microbial Metabolites

Abstract

It is now well understood that the eukaryotic host has evolved multiple mechanisms to monitor and respond to the diverse and biochemically active microbiota that thrives in a symbiotic fashion in the gut and other tissues. Generally, these mechanisms are based on traditional notions of innate and adaptive immune processes, which are mediated by recognition of, and response to, microbially derived macromolecules. Microbes themselves are metabolically active and contribute a vast array of small molecules, not present in germ-free model systems, with diverse putative and unknown biological function, and intensive work is ongoing to unravel their roles in physiological systems. Metazoans have evolved and maintain distinct gene regulatory networks to detect and respond to environmental, non-self-molecules (xenobiotics), and interestingly, recent investigation has shown that these pathways are operational in the detection and response to microbiota-derived small metabolites. These processes likely represent a general mechanism of host-microbe crosstalk, and they have clinical implications in drug and xenobiotic metabolism.

Eukaryotic-Prokaryotic Interactions

As a moment’s thought demonstrates, multicellular organisms have interfaced with their external environment continuously throughout evolutionary history. Part of the external environment must necessarily have included the myriad prokaryotic microbes and viral particles that are capable of occupying virtually all available niches, including sites external and internal to an organism. Necessarily, metazoans have had to evolve diverse and specific mechanisms to protect themselves from the threats posed by exogenous microbes, mechanisms now generally referred to as immunity. Indeed, our fundamental conception of immunology is based on the notion of defense against external invasion and subversion, and it is often illustrated and taught with corresponding martial metaphors. Traditionally, immunology is usually defined in terms of subfields of intrinsic, innate, and adaptive immunity, all of which are conceptualized as multifaceted and interactive cell-based defensive processes variably shared by most eukaryotic taxa and deployed in specific tissues.

Intrinsic defenses are represented by physical and chemical obstacles that can include secreted products, such as polymeric cuticle and mucus, epithelial barrier enhancing cellular junctions and cell walls, chemical defenses, such as antimicrobial cationic peptides, and reactive oxygen species (ROS), among others. Typically, intrinsic defenses are seen in epithelial tissues that interface directly with the external environment or line the lumen of gut or respiratory surfaces (37). Intrinsic defenses are often static, but can be induced in certain situations, such as the transcriptional activation of antimicrobial peptide (AMP) synthesis or posttranslational release of mucins (47). Intrinsic defenses, which are seen in all metazoans, remain a key component of human physiology, e.g., AMP secretion from Paneth cells in small intestinal epithelial crypt cells, mucins from goblet cells, ROS production via Nox/Duox enzymes, and extensive barrier structures in the epithelial monolayer (43, 47).

Innate immunity similarly has an ancient history and is present in lower invertebrates and plants. Innate immunity is based on receptor-mediated detection of macromolecules derived from, and characteristic of, microbes, a class of molecules now operationally referred to as “MAMPs”, for microbe-associated molecular patterns, which include bacterial lipopolysaccharides, peptidoglycans, flagellin, and viral nucleic acids (1, 3). MAMPs are bound by a functional class of transmembrane or intracytoplasmic receptors termed pattern recognition receptors (PRR) and includes the well-studied Toll-like receptors, Nod-like receptors, and others. PRR recognition of MAMPs results in activation of cytoplasmic signaling cascades and involve relays of phosphorylation and/or proteolytic cleavage and can result in transcriptional activation of inflammatory mediators (or intrinsic processes, such as AMP production) and/or variations on programmed cell death.

Finally, adaptive immunity, which is specific to higher chordates, occurs in animals where genetic recombinatorial machinery allows for the ability to detect and respond to a diverse array of foreign macromolecules (antigens), far beyond the dozen or so structures recognizable by the innate immune system. Adaptive immunity perceives and mediates responses via the activities of a specific class of effector cell (lymphocytes). B-cell dependent humoral and cell-mediated (T cell) components have been well. Prokaryotes profoundly affect adaptive immunity, as germ-free mice characteristically show hypoplastic lymphoid tissues and exhibit a spectrum of abnormalities in adaptive immune function (38).

A commonality of these various components of immunity is their ability to defend against both specific bacterial or viral whole organisms (e.g., intrinsic defenses to thwart active invasion and cell damage) and bacterial or viral products, including MAMPs/antigens (e.g., in simulation of inflammation, or B- and T-cell responses). Beyond these processes, multicellular organisms have evolved systems to perceive and respond to other forms of environmental stresses, including physiochemical threats from changes in abundance of O₂ (hypoxia), H⁺ (pH), and e⁻ (redox and electrophilic stress). Obviously, the ancestral and proverbial “primordial soup” was awash in extremes of chemistry, and understandably, cellular processes have evolved to respond and manage this physiochemical stress, including the hypoxia-inducible factor (HIF) pathway and redox signaling mechanisms. Similarly, small organic molecules were a feature of the prebiotic world, and simple metazoans would have been (and are) constantly exposed to a complex chemical milieu. Specific signaling pathways, such as nuclear factor E2-related factor (Nrf2), aryl hydrocarbon (AHR), and others, act as cellular or organismic transducers to coordinate defenses against such environmental challenges. In humans and higher vertebrates, these ancient and conserved pathways are fully operational, have been well studied in the context of exposure to environmental chemicals, and are, thus, of significant interest to pulmonary physiologists and toxicologists. It can be postulated that the vast array of extreme chemistry, small molecules, and metabolites produced by microbial fermentation of ingested foodstuffs can act as an internal “primordial soup”, from which we derive not just nutritional value, but a panoply of functional signaling molecules, which provides the host with a “preimmune” mechanism of mediating host-microbial interactions.

Symbiotic Microbiota

Recent years have seen increased interest in the microbiota—the microbial occupants invariably associated with metazoans (18, 22, 36, 59). Beginning with the endosymbiotic origin of mitochondria, wherein the enzymatic machinery of oxidative phosphorylation and, thus, energy production was outsourced to ancestral prokaryotes, numerous biochemical functions provided by microbes have been described, resulting in a diversity of symbiotic partnerships (4). The pervasiveness of these arrangements has suggested the concept of the “holobiont”. Holobiont was originally a term found in ecology literature, meaning “an assemblage of a host and the many other species living in or around it, which together form a discrete ecological unit” (48). The concept has gained traction in the host-microbial context, as the degree to which symbiosis occurs among living systems has become more apparent. Clearly, for example, the ecological role of termites in their woody environment would be moot if these insects were deprived of their symbiotic cellulolytic, gut-dwelling bacteria. Similarly, the existence of many herbivorous mammals and their ecological roles (and the environmental effects of modern cattle ranching) could not exist without a similar arrangement with the host and symbiotic gut microbes.

Interest in the role of the microbiota on mammalian physiology and, thus, human health has also burgeoned. The vast number of microbes (10 to 100 trillion) that reside in the mammalian gut and other anatomical locations serves numerous beneficial functions that includes stimulation of adaptive immune system development and competitive exclusion of pathogenic microorganisms (“colonization resistance”) (18, 22). Experiments in germ-free mice have shown a compelling role of the microbiota in influencing a wide range of innate and adaptive immune and metabolic processes (53). Quantitative and/or qualitative abnormalities of the microbiota—dysbiosis—have been associated with inflammatory bowel disease (IBD) (15), other allergic systemic immune conditions, metabolic/hepatic disorders, and even neuropsychiatric conditions (5, 19, 21). Bacteria administered as a therapeutic—probiotics—have been reported to dampen inflammation, improve barrier function, and promote reparative responses, and they have shown promise in both gut and systemic functional and inflammatory disorders (45, 61).

Mechanistically, the realization of the role of the microbiota in host physiology and pathology has been driven by advances in technical and informatics platforms that have allowed the accurate tabulation and classification of the microbial diversity in many biological systems. Current metagenomic studies have characterized the vast genetic diversity encoded in the microbiome, and mass spectroscopy-based methods have defined a prodigious “metabolome” of small molecules and fermentative products (54, 63). Thus, it is now apparent that microbial communities are capable of influencing the physiochemical environment they live in, whether the environment is a marine, aquatic, soil niche, or in the case of higher vertebrates, the gut lumen and other anatomic sites with stable microbial commensal populations. In mammals, specific microbially derived small-molecule metabolites are increasingly recognized as having regulatory functions in multiple host physiological processes (23, 34, 42, 50, 51). Metabolites influenced by the microbiota may be perceived by traditionally conceptualized immune receptors, for example, microbially modified metabolites taurine, histamine, and spermine can stimulate intracytoplasmic Nod-like NLRP6 inflammasome activation, and contribute to abnormal inflammatory signaling and colitis (34). In summary, there is increasing evidence of microbiota-encoded biochemical processes playing significant roles in human disease (17).

Thus, our interaction with the microbial world, traditionally defined by host perception of macromolecules/antigens, must now be expanded to include perception of xenobiotic-derived small molecules and metabolites. This review is intended to establish a context wherein the traditional view of immunity in the perception and response to microbes, must be extended to encompass biochemical/metabolic influences of the resident prokaryotes, and consideration of how the eukaryotic host has evolved to perceive them.

Xenobiotic Signaling

A xenobiotic, (as opposed to endobiotic) broadly defined, refers to substance that is exogenous to the body or to an ecological system (54). Xenobiotics can be considered to extend to dietary phytochemicals, environmental toxins, and pollutants, and the entire small-molecule pharmacopeia.

In the context of host-microbial interactions, chemical products of bacterial metabolism may be considered xenobiotics, and be confirmed as such if they are not detectable in germ-free whole animal systems.

Microbial metabolites can be detected by the host via a number of specific and dedicated membrane-bound and nuclear receptors. A primary example are the short-chain fatty acids, microbial fermentation products of less than six carbons, which include butyrate, propionate, and acetate. These molecules are well-known effectors of microbial functions, as butyrate is considered a major energy source for intestinal epithelial cells (12). Significantly, butyrate can also function as a potent regulatory molecule, affecting epithelial homeostasis via the G protein-coupled receptors (GPCRs) GPR43 and GPR41 (55). Furthermore, butyrate, a four-carbon carboxylic acid, is thought to represent a mimetic of histone modification and, thus, function as a histone deacetylase inhibitor at the chromatin level, bestowing the ability to stimulate gene expression via alteration of chromatin dynamics in immunocompetent cells (7, 27) (FIGURE 1).

FIGURE 1.

Physiochemical and xenobiotic response pathways Multiple eukaryotic signaling mechanisms enable the host to perceive microbial environmental changes and/or small molecules; see text for description. These include pathways that can detect physiochemical signals (Nrf2 and HIF-1α) and small molecules of microbial origin. bHLH, basic helix-loop-helix transcription factor; GPCR, G protein-coupled receptor; HIF-1α, hypoxia-inducible factor-1α; Nrf2, nuclear factor E2-related factor.

Additionally, microbial metabolites can act as ligands for intracytoplasmic nuclear receptors. For example, secondary bile acids—bile acids derivatized by members of the microbiota (deoxycholate and lithocholate)—are perceived by the farnesoid-X receptor (23). The AHR (6) and pregnane X nuclear receptors (60) detect indole tryptophan derivatives, another general class of microbial metabolites with potent host signaling effects, and can stimulate various gene regulatory programs that affect aspects of mucosal homeostasis. The AHR can detect and activate detoxification pathways in response to a range of xenobiotic molecules (46). Taken together, these examples define the existence of specific cellular receptors dedicated to the perception of microbia-derived small metabolites.

Hypoxic and Redox Signaling

Beyond microbia-derived metabolites, the host can detect microbiota-derived physiochemical stimuli. For example, bacterial respiration can stimulate epithelial responses via the oxidant sensor HIF (2, 58). Hif-1α is a cytoplasmic transcription factor that under basal normoxic conditions is constitutively modified by prolyl hydroxylases, allowing binding to the Von Hippel Lindau (VHL) ubiquitin ligase and consequent tonic proteasome degradation. During hypoxia, loss of prolyl hydroxylation releases Hif-1α from VHL, thus allowing Hif-1α accumulation, translocation to the nucleus, binding to specific hypoxia response element promoter motifs, and activation of a battery of genes to allow the cell to respond and adapt to hypoxic conditions, such as EPO, VEGF, and enzymes necessary for glucose metabolism. Early motile metazoans presumably would have used this pathway to spatially adapt to changing environmental conditions. As metabolically active prokaryotes are inseparable components of the environment, such organisms could have used the pathway to detect and respond to microbes. These events may exist in mammals; studies in germ-free mice have shown that the measurable Po₂ in the intestine is significantly higher than in conventional mice, and microbial growth can deplete luminal mucosal O₂ content (58). Interestingly, butyrate stimulates of O₂ utilization, greater anaerobiosis and compensatory Hif-1α activation that culminates in improved epithelial barrier function (28). Additionally, the commensal Bacteroides thetaiotaomicron can induce Hif-1α and result in colonization resistance to fungal infection (14).

Additionally, bacteria can stimulate enzymatic reactive oxygen species (ROS) production via NADPH oxidases and produce ROS as a byproduct of their own respiration, thus activating redox-signaling events. ROS are unstable intermediates of O₂ catabolism and are used as signaling messengers in virtually all multicellular life, including both plants and animals (32). In mammals, close contact of commensal Lactobacilli with the intestinal epithelial surface activates redox-signaling pathways, resulting in a wide range of downstream-signaling events (24–26). This redox signaling occurs by the transient reversible oxidation of low pK_a sensor cysteines present in the active site of a number of regulatory enzymes, including enzymes involved in MAPK activation (DUSPs) (62), NF-κB signaling (nedd8) (9), and cytoskeletal dynamics/motility (LMW-PTPases) (56). Overall, these conserved systems allow cellular reaction to microbial physiochemical stimuli and likely evolved contemporaneously with the earliest interaction between prokaryotes and metazoans, antedating the evolution of immunity (32).

Nrf2/ARE Pathway

The Nrf2/ARE signaling module is another evolutionarily conserved signal transduction pathway, one that responds to electrophilic and oxidative stresses (including ROS), whether derived from intrinsic (energy production) or extrinsic (xenobiotic) stimuli and, thus, represent a mechanism, whereby the host can react to a variety of environmental signals, including redox stress, electrophiles, and xenobiotics (33, 52) (FIGURE 2). Nrf2 is a member of the CNC (cap’n’collar) family of b-Zip (basic leucine zipper) transcription factors (41) and is posttranslationally regulated in a manner broadly analogous to Hif-1α activation. Under basal conditions present in the cytosol, Nrf2 is bound by the Keap1 subunit of a cullin-dependent E3 ubiquitin ligase that promotes Nrf2 cytoplasmic degradation. Prooxidant or electrophilic stress within the cytoplasm causes rapid oxidation of redox-sensitive cysteines on Keap1, leading to conformational change, suppression of proteasomal degradation, and consequent accumulation of Nrf2. Nrf2 then translocates to the nucleus, dimerizes with the small protein Maf, and binds to a well-defined antioxidant response element (ARE) promoter sequence, subsequently activating expression of an extensive gene network. These genes include antioxidants, efflux pumps, and detoxification enzymes, which when activated, serve to reestablish an optimal redox milieu and, thus, provide a cellular cytoprotective function (57).

FIGURE 2.

The Nrf2/ARE pathway: activation of the Nrf2/ARE signaling module At homeostasis, Nrf2 interacts with Keap1 (physically associated with a ubiquitin ligase), which results in constitutive ubiquitination and proteasomal degradation. Exposure to oxidant or electrophilic stress in the cytoplasm or microbially derived small molecules (e.g., 5-MIAA, indole acetic acid) results in transient and reversible modification of regulatory cysteine residues in Keap1, leading to Keap1 conformational change and Nrf2 release, stabilization, and cytoplasmic accumulation. Free Nrf2 translocates to the nucleus, where, as a dimer with the small DNA-binding protein Maf, it binds to antioxidant response element and transcriptionally activates the expression of a class of genes that includes cytoprotective factors and detoxification enzymes. Nrf2-induced gene products then act to reestablish cellular redox homeostasis and protect against subsequent insults. Nrf2, nuclear factor E2-related factor.

Model organisms with Nrf2 loss of function fail to upregulate these antioxidant or detoxification effector genes and are hypersensitive to a variety of exogenous insults, including UV radiation, sepsis, and inhaled xenobiotic oxidants and toxins (16, 29). Nrf2-null mice are hypersensitive to dextran sodium sulfate colitis, indicating protective function in barrier epithelia (30). Evidence is emerging that the Nrf2 pathway may play a role in interaction of the gut epithelia of multiple organisms with their resident microbes. In C. elegans, enzymatically generated ROS in response to bacterial infection stimulates the activation of the Nrf2 isolog SKN-1, resulting in cytoprotective effects (20). Drosophila Nrf2 (cncC) is stimulated by bacteria in the fly gut, and gut-directed overexpression of cncC, or suppression of Keap1 protects flies from oxidant-induced mortality (24). In mice, beneficial effects of probiotic Lactobacilli on epithelial injury are abrogated in Nrf2-null animals (24), and Peptostreptococcus produces the tryptophan metabolite indole acetic acid, which acts as an Nrf2 inducer and, consequently, stimulates beneficial barrier and restitution effects (64).

These observations have prompted increasing interest in the therapeutic exploitation Nrf2 as a mediator of cytoprotection (10). The rationale for exploiting Nrf2 is based on stimulation of hormesis. Hormesis is a concept originally derived from studies of unicellular eukaryotes (i.e., yeasts) wherein low, near-threshold levels of a stressor, such as UV exposure, is protective against more intense or prolonged stimuli (35). In this case, perception of the stressor, via the production and recognition of structural DNA alteration, results in upregulation of a battery of genes with DNA repair function. In a sense, a hormetic response is akin to a “training effect” well known to exercise physiologists (and athletes of all stripes!), wherein environmental stress results in adaptive responses, generally the result of compensatory gene expression. The microbiota, even when possessing a beneficial relationship with the host, nevertheless is an extrinsic influence, and xenobiotics produced by microbes are a potent signaling stressor. Thus, hormesis, as a response to xenobiotic and environmental stimuli, likely extends to the acquisition of, and adaptation to exogenous bacteria.

Systemic Nonimmune Microbial Signaling

Clearly, much microbial perception and signaling occur at epithelial surfaces. Epithelia, by definition, are derived from endodermal or ectodermal tissues, and thus function as interface with the outside world, whether the environmental matrix or luminal contents. Microbial preimmune signaling can affect systemic tissues. Recent data suggest the mammalian liver can function as a central coordinator of microbial xenobiotic perception (49). The liver is supplied by the portal vein, a vascular channel draining the vast majority of the intestinal watershed. This system is physiologically unique in that it is the only venous system that does not drain directly into the heart. Most digested macromolecules absorbed across the mucosa of the large and small intestine are delivered directly to the liver (a prominent exception is lipids, which are absorbed in mucosal lymphatic vessels and distribute systemically). In the liver, portal blood perfuses the hepatic sinusoids, and myriad biotransformations occur within the highly metabolically active hepatocytes. (In parallel, the bone marrow-derived Kupffer cells lining the sinusoids allow immune perception of MAMPs in the draining portal circulation) (FIGURE 3). Along with processing of dietary peptides and carbohydrates, the hepatocytes also metabolize ingested environmental small molecules, be they inadvertent toxicants, prescribed pharmacological agents, or, relevant to our discussion, small metabolites of microbial origin. This initial hepatic processing of ingested xenobiotics is referred to as “first pass metabolism”, a concept familiar to any introductory pharmacology student, but has not typically been considered in the context of host-microbial interactions.

FIGURE 3.

The gut-liver axis The microbiota produces a vast and diverse range of small-molecule metabolites that are absorbed across the intestinal barrier epithelia. The portal vein is a unique vascular channel that drains most of the intestinal tract and shunts the vast majority of digested dietary macronutrients and micronutrients, as well as ingested environmental toxins and administered pharmacology agents. The expansive metabolic capacity of the liver then mediates appropriate bioprocessing. What has been overlooked are macro and small molecules derived from the gut microbiota. It has long been understood that microbiota macromolecules (or MAMPs) are perceived by innate and adaptive signaling in the Kupffer and stellate cells. Recent data suggest microbial small molecules can be perceived by conserved xenobiotic response pathway operating in hepatocytes. Responses to the small molecules can influence host response to toxins or drugs, and perhaps mediate other physiological processes.

The Nrf2/ARE system operates in the liver and can transcriptionally activate drug-metabolizing enzymes (44, 65). Mice null in Nrf2, or biochemically unable to activate the pathway, fail to upregulate the typical effector genes and are hypersensitive to exposure to hepatotoxic drugs (13). Interestingly, Saeedi et al. (49) showed mice supplemented with oral Lactobacilli induced hepatic Nrf2 and typical Nrf2-responsive genes, resulting in animals resistant to subsequent exposure to several chemical oxidant challenges. Furthermore, specific bacteria were shown to activate hepatic Nrf2 via the metabolite 5-methoxy-indole acetic acid (or 5-MIAA). In parallel work with an invertebrate system, treatment of germ-free Drosophila with a fly-specific symbiont, L. plantarum, was effective in induction of fly Nrf2 (CnC) in the fat body of the fly, a systemic tissue that performs biochemical transformation homologous of the liver. Significantly, flies thus treated were resistant to the oxidant toxin paraquat (49). Thus, preimmune recognition of microbial metabolites can have effects on systemic detoxification pathways.

Implications

These data suggest changes in the microbiota community structure and abundance (dysbiosis), from many causes, including diet and antibiotic use, may result in consequent and idiosyncratic responses to potentially hepatotoxic agents. In the field of human medicine and pharmacology, all clinicians are aware of markedly different responses of individual patients to a standard dosage a given therapeutic. This includes unpredictable toxic effects, often manifesting as hepatocellular injury. Although it has been recognized that microbes can detoxify or otherwise biotransform ingested pharmacological agents by virtue of microbially encoded enzymes (8, 11, 31, 34, 39), it has been suggested that microbes could indirectly affect biotransformation of drugs via activation of host-encoded enzymes in the liver (and likely other tissues) via Nrf2 and other preimmune signaling. This implication is significant in that alteration of the microbiota could, thus, affect metabolism of parenterally administered agents, which is often the case in critically ill patients. Additionally, modulation of xenobiotic biotransformation in the liver due to variable exposure to microbial signals is a double-edged sword, as different enzymatic complements may act to detoxify one ingested agent, while also biotransform another into a toxic intermediate. For example, while acetaminophen is converted to an inert metabolite by one complement of inducible enzymes, other enzymes yield NAPQI, the highly hepatotoxic intermediate (40). Thus, it is difficult to conceptualize a specific probiotic approach to correct what could be described as a pharmacologically dysbiotic microbiota. However, preservation of a metabolic status quo could perhaps be preserved by autologous fecal microbiota transplant (FMT) in patients subjected to major drug-induced challenges (e.g., sepsis, transplant patients).

Overall, all metazoans, indeed, all living things, have evolved the capability of perceiving and responding/adapting to small molecules and physiochemical stimuli from the environment, which as we are rapidly realizing, includes the resident microbial communities. The realization that microbes have increasingly diverse mechanisms to affect our physiology will hopefully allow potential use of them as therapeutic agents. Additionally, such physiochemical, preimmune signaling by the microbiota may be considered a normal part of physiology in the gut and beyond.