The Intersection between Bacterial Metabolism and Innate Immunity

Abstract

Background

The innate immune system is the first line of defense against microbial pathogens and is essential for maintaining good health. If pathogens breach innate barriers, the likelihood of infection is significantly increased. Many bacterial pathogens pose a threat to human health on account of their ability to evade innate immunity and survive in growth-restricted environments. These pathogens have evolved sophisticated strategies to obtain nutrients as well as manipulate innate immune responses, resulting in disease or chronic infection.

Summary

The relationship between bacterial metabolism and innate immunity is complex. Although aspects of bacterial metabolism can be beneficial to the host, particularly those related to the microbiota and barrier integrity, others can be harmful. Several bacterial pathogens harness metabolism to evade immune responses and persist during infection. The study of these adaptive traits provides insight into the roles of microbial metabolism in pathogenesis that extend beyond energy balance. This review considers recent studies on bacterial metabolic pathways that promote infection by circumventing several facets of the innate immune system. We also discuss relationships between innate immunity and antibiotics and highlight future directions for research in this field.

Key Messages

Pathogenic bacteria have a remarkable capacity to harness metabolism to manipulate immune responses and promote pathogenesis. While we are beginning to understand the multifaceted and complex metabolic adaptations that occur during infection, there is still much to uncover with future research.

Introduction

The innate immune system is the body’s first line of defense against pathogens. It consists of anatomical barriers like mucous membranes and skin, effector cells such as macrophages and neutrophils, and soluble mediators like antimicrobial peptides (AMPs) and complement factors [1, 2]. Upon coming into contact with harmful microbes, epithelial surfaces such as the skin and mucous membranes serve as barriers that protect the internal host environment from external threats [3]. For instance, the keratinocytes of the skin establish a permeability barrier that prevents pathogens from entering while simultaneously activating the immune response via recognition of microbial products termed pathogen-associated molecular patterns (PAMPs) [4]. Recognition of PAMPs signals the release of soluble factors, including AMPs and inflammatory cytokines, which recruit effector cells, such as monocytes, macrophages, and neutrophils to the infection site [2, 4]. Neutrophils enter from the vasculature, where they eliminate the pathogen by phagocytosis and production of a wide range of antimicrobial molecules [5]. Monocytes and macrophages phagocytose pathogens but also release an extensive repertoire of pro-inflammatory cytokines that enhance immune defenses and activate the adaptive immune system [2, 6, 7]. The coordinated activities of neutrophils, monocytes/macrophages, and additional innate cells not considered in this review (NK cells, dendritic cells, etc.) contain infection while the adaptive immune response develops [2, 5].

A crucial aspect of innate immunity is the ability of immune cells to react to environmental stimuli and quickly launch a response that effectively manages infection without causing excessive harm to neighboring tissues [2]. Immune cells achieve this task through metabolic reprogramming, which involves adjusting cellular metabolism to meet nutritional demands imposed by the environment [8]. A well-established example of metabolic reprogramming in immune cells occurs in macrophages upon encountering a pathogen [8]. Activated macrophages undergo a metabolic switch from the tricarboxylic acid cycle (TCA) and oxidative phosphorylation (OXPHOS) to aerobic glycolysis. This shift in metabolism increases the synthesis of purines and pyrimidines, which are used as substrates to generate nicotinamide adenine dinucleotide phosphate hydrogen (NADPH), an essential cofactor for the NADPH oxidase [9]. In addition, by prioritizing glycolysis and limiting the TCA cycle and OXPHOS, macrophages can generate ATP more rapidly and prevent the accumulation of intracellular reactive oxygen species (ROS), which have the potential to harm the cell [8].

Like immune cells, bacterial pathogens can shift metabolism upon encountering the innate immune system, suggesting metabolic changes may reflect a physiological adaptation of the pathogen that can also modulate the immune response and promote bacterial survival [10, 11]. Bacterial metabolic adaptations are often driven by horizontal gene transfer events that improve microbial sensing of environmental cues and promote enhanced regulation of metabolic pathways, especially in fluctuating conditions [10]. This evolutionary process has allowed several bacterial pathogens to develop a broader repertoire of mechanisms to obtain vital nutrients like transition metals, glucose, free peptides, and lipoic acid compared to nonpathogenic bacteria [10, 12, 13]. Furthermore, some bacterial pathogens possess remarkable mechanisms to detect and respond to oxygen levels within the host, switching between respiration and fermentation as needed [10, 14, 15]. Thus, many bacterial pathogens have a flexible metabolism, which helps overcome nutrient limitations within the human body and evade immune defenses [9].

The relationship between bacterial metabolism and innate immunity is also substantiated by the ability of immune cells to detect and respond to bacterial metabolic byproducts, several of which exhibit immunomodulatory properties [14]. For example, bacterial fermentation products such as short-chain fatty acids (SCFAs) act as immunomodulators in many immune cell types, including tissue-resident macrophages in the gut [16]. These macrophages are essential in safeguarding the gut against invading pathogens and mitigating the onset of inflammatory diseases [16]. Exposure of macrophages to SCFAs inhibits mTOR and ultimately enhances antibacterial activity [16]. Emerging evidence indicates there exists a critical interplay between bacterial metabolism and immune system functions that promotes the survival of the pathogen during infection [9, 14, 17]. For instance, pathogenic bacteria such as Staphylococcus aureus, Salmonella enterica, and Mycobacterium tuberculosis adapt their metabolic pathways, including glycolysis, TCA cycle, and lipid metabolism, to meet the energy requirements within the host environment [17, 18]. Additionally, these bacteria use or generate metabolic byproducts with immunomodulatory properties, which enable them to manipulate the innate immune response and establish a successful infection [18]. However, the current state of knowledge on the effects of bacterial metabolism on the innate immune responses in the context of infection is comparatively limited relative to immunometabolism, which primarily considers the metabolic changes of immune cells that drive antimicrobial and pro-inflammatory responses [17, 19].

This review will discuss recent insights on bacteria-immune system cross-talk through the lens of bacterial metabolism. We will consider how several pathogenic bacteria shape metabolic pathways, including glycolysis, lipid metabolism, the TCA cycle, and the synthesis of lipid derivatives to evade components of the innate immune system, including epithelial barriers of the gut and skin, phagocytic leukocytes, and complement. We will conclude by reviewing the relationship between antibiotics, metabolism, and innate responses, highlighting several unanswered questions and possible future directions for the field.

Bacterial Metabolism and Pathogenesis at Epithelial Barriers

The epithelial cell layers of the skin and mucous membranes are some of the first lines of defense against invading pathogens [3, 20]. Epithelial cells sense pathogenic microbes and trigger a pro-inflammatory response to aid in preventing dissemination [21]. The epithelial layer typically contains many commensal microbes (microbiota), and because of these interactions, epithelial cells have acquired the ability to discriminate between pathogenic and nonpathogenic bacteria [21–23]. The microbiota also produces metabolites that offer benefits at host epithelial surfaces [21, 24]. For instance, commensal bacteria in the gut and skin break down complex substrates that host cells are unable to assimilate, producing sub-metabolic products, including secreted proteins, organic acids, fatty acids, indole, bacteriocins, and nitric oxide (NO), all of which protect the gut and skin by increasing mucus secretion, AMP synthesis, and expression of tight junctions [21, 25]. Additionally, several bacterial metabolites control the function of inflammatory cells, preventing excessive inflammation, a key factor in promoting tissue health and homeostasis [21, 25, 26]. However, disruption to the skin or gut microbiota can adversely affect tissue homeostasis and organ function, leading to detrimental inflammatory responses that establish an environment suitable for bacterial pathogens to colonize and cause disease [24, 27]. In this section, we will use Escherichia coli and Listeria monocytogenes as examples of pathogens that respond to the metabolic status of the gut to regulate the expression of virulence factors and promote infection. We will also discuss how the skin pathogens, Cutibacterium acnes and S. aureus, exploit the nutrients present in the skin to produce metabolites that shift immunological responses to favor pathogen survival. Lastly, we will consider how the skin environment selects for bacterial mutants capable of promoting a metabolic profile that is conducive to infection.

Gut Metabolites Promote Virulence Gene Expression and Pathogenesis

E. coli is considered part of the human microbiota, but some strains, such as enterohemorrhagic E. coli (EHEC), are linked to the emergence of severe gastrointestinal diseases [27, 28]. The ability of EHEC to cause infection relies on the acquisition of the locus of enterocyte effacement (LEE), a pathogenicity island that encodes several virulence factors, including a type III secretion system (T3SS) and the Shiga toxin [27, 29]. Recent research has uncovered that arginine sourced from the host is a crucial nutrient needed to activate LEE expression in EHEC, thereby stimulating Shiga toxin secretion and improving bacterial adhesion to epithelial cells [28]. The resulting increase in LEE activity has been linked to microbiota disturbances, as it provides EHEC with an enhanced ability to colonize the gut, provoke inflammation, and cause disease [27]. EHEC obtains arginine through the arginine transporter ATP-binding protein (ArtP). A ΔartP mutant has significantly reduced LEE expression, toxin secretion, and adhesion to epithelial cells [28]. Notably, the ΔartP mutant grows identically to the WT strain; therefore, the reduction in pathogenicity is likely due to decreased expression of the LEE locus [28].

Transcriptomic analyses demonstrated that arginine activates several genes in the LEE locus. The activation of arginine metabolism in EHEC is mediated by the transcriptional regulator ArgR, which triggers the expression of genes involved in the binding and transport of arginine, such as gene clusters, artPIQM-artJ and argT-hisJQMP [30]. Since an artP mutant reduces LEE expression, it was possible that ArgR was involved in the regulation of LEE genes [28]. Indeed, further characterization of ArgR demonstrated that it binds to the LEE promoter region, regulating its transcription. Interestingly, this phenotype is not exclusive to EHEC, and the deletion of argR in other Enterobacteria, such as Citrobacter rodentium, also reduced infective capacity in vivo [28]. Together, these studies suggest that arginine, a metabolite with immunomodulatory properties, may activate the expression of virulence determinants in various gut pathogens [28]. However, some key questions remain unanswered, such as whether the physiological levels of arginine found in humans are sufficient to elicit a similar response by other gut pathogens and how microbial communities might compete for this important amino acid in vivo [28].

L. monocytogenes is another significant gastrointestinal pathogen that causes the foodborne disease listeriosis [31]. The rapid modulation of L. monocytogenes metabolism allows it to resist the low pH in the stomach and evade immune responses [32]. Once it has crossed the intestinal barrier, L. monocytogenes disseminates through the lymph nodes to the spleen and liver, where it replicates intracellularly [31]. L. monocytogenes partially achieves its metabolic plasticity by sensing the redox state of the cell through the Rex regulator [32, 33]. Rex senses the availability of NAD+ and NADH to respond to changing energy demands [33]. Recent transcriptomic analysis of an L. monocytogenes Δrex mutant during aerobic growth uncovered that genes involved in fermentative metabolism, such as alcohol dehydrogenase (lap), pyruvate formate lyase (pflA and pflBC), and lactate dehydrogenase (ldhA), were the most elevated, suggesting that Rex represses fermentative metabolism [31]. Consistent with this observation, metabolite quantification showed that a Δrex mutant secreted more lactate and formate [31].

Furthermore, a Δrex mutant is more resistant to the antimicrobial effects of acidified bile. Additionally, a Δrex mutant overexpresses the virulence factors, InlA, InlB, and bile salt hydrolase, which are responsible for the invasion of epithelial cells and bile detoxification, respectively [31]. Coincident with these findings, a Δrex mutant led to increased invasion of epithelial cells and hepatocytes as well as improved resistance to bile relative to the WT and complemented strains. These studies indicate that metabolic regulation via Rex is crucial for the invasion of host cells and protecting against gastrointestinal tract stressors [31]. Surprisingly, the Δrex mutant was not attenuated in the gastrointestinal tract, whereas bacterial burdens in the spleen, liver, and gallbladder were markedly reduced, suggesting that Rex is only necessary for replication in these organs [31]. Together, these studies support the idea that activation of fermentation is required for L. monocytogenes to thrive within the gastrointestinal tract yet must be reduced to promote infection of other organs [31].

The studies mentioned above highlight the relevance of sensing the metabolic status of the host to pathogenesis in the gut. The ability of E. coli and L. monocytogenes to adapt to metabolic cues from the external environment provides a critical advantage for survival and invasion of the epithelial barrier. Notably, the intestine is an incredibly complex environment, and we are only beginning to understand the role of bacterial metabolism in shaping pathogenesis within this environment. Future research aimed at identifying how gut metabolites promote the pathogenesis of other common pathogens of the GI tract has the potential to uncover emerging themes on the impact of metabolites on virulence factor gene regulation, barrier disruption, and inflammation.

The Role of Skin-Derived Metabolites in Modulating Inflammation and Bacterial Pathogenesis

The influence of bacterial metabolism on innate immunity extends beyond the gastrointestinal tract, as bacterial metabolites can also interfere with the antimicrobial response at epithelial barriers such as the skin [21]. Lipids are one of the most abundant macromolecules in the skin, and they play a crucial role in maintaining organ homeostasis and regulating inflammatory responses [26]. Sebocytes in the skin specialize in producing sebum in hair follicles [34]. Blocked hair follicles can result in acne vulgaris, a dermatological condition in which C. acnes, a commensal skin bacterium, triggers inflammation in the sebaceous glands [34]. C. acnes is known to secrete triacylglycerol lipases, which hydrolyze triglycerides, releasing free fatty acids and glycerol. The subsequent accumulation and catabolism of free fatty acids reduce oxygen levels in the sebaceous glands, which ultimately induces fermentation in C. acnes [35]. Genome-scale metabolic pathway analyses suggest that C. acnes produces propionate via a unique pathway known as the Wood-Werkman cycle, which facilitates NAD+ regeneration [35]. The Wood-Werkman cycle is highly efficient at fermenting glycerol, resulting in the secretion of high concentrations of propionate (Fig. 1) [35].

Fig. 1.

S. aureus and C. acnes metabolism manipulate the immune response of skin cells to promote bacterial survival. Left panel: S. aureus depletes fumarate from the extracellular milieu, inducing glycolysis in keratinocytes (1). Activation of glycolysis in keratinocytes and the subsequent generation of mitochondrial ROS lead to the induction of necroptosis (2), causing tissue damage, which in turn facilitates bacterial invasion (3). Right panel: C. acnes triggers the synthesis of essential lipids in the skin cells, which are used to produce free fatty acids and glycerol. The fermentation of glycerol by C. acnes leads to the production of propionate, which further stimulates the production of pro-inflammatory cytokines in sebocytes.

Sebocytes sense propionate in the extracellular milieu and trigger the production of pro-inflammatory cytokines [36]. Recent experimental evidence suggests that sebocytes treated with spent media from C. acnes grown in the presence of glycerol produced higher levels of IL-6, IL-8, tumor necrosis factor (TNF)-α, and CXCL10 than sebocytes treated with a spent media control [36]. The increased pro-inflammatory cytokine production in sebocytes is recapitulated upon adding toll-like receptor (TLR) ligands with propionate [36]. Propionate was found to enhance cytokine production in sebocytes through epigenetic regulation of histones 8 and 9 [36, 37]. This regulatory process activates the free fatty acid receptors, which sense SCFAs, thereby regulating the immune response based on the metabolic status of the cell [38]. Free fatty acid receptors 2 and TLR2 are increased upon propionate stimulation, suggesting the elevated cytokine response is mediated, at least in part, through TLR2 activation [36, 39].

C. acnes also stimulates lipid production in host cells to prolong inflammatory responses via propionate secretion (Fig. 1) [35, 40]. This idea was recently put forward in a study demonstrating that keratinocytes increase lipid synthesis in the presence of spent media from cultures of C. acnes [40]. Using lipidomic analyses, the authors measured higher amounts of triacylglycerol, ceramides, cholesterol esters, and free fatty acids than the control group. Similarly, they observed increased lipid levels in the tissue of mice infected in the skin with C. acnes, suggesting that keratinocytes respond to C. acnes by increasing lipid biosynthesis (Fig. 1) [40]. Propionate was sufficient to induce lipid synthesis, and no other SCFAs, such as butyric acid, acetic acid, and isovaleric acid, altered lipid metabolism [40]. Furthermore, treatment of keratinocytes with propionate led to increased expression of lipid biosynthesis genes [40]. Overall, the evidence clearly illustrates how the unique capacity of C. acnes to assimilate lipids from the skin and produce propionate can dramatically perturb innate immunity and suggest that bacterial pathways associated with the synthesis of SCFAs could be targeted for antimicrobial therapy in the context of acne (Fig. 1) [35].

The skin can also sustain the growth of other bacteria with significant pathogenic potential, such as S. aureus [17]. S. aureus is part of the microbiota of 30% of the human population, where it colonizes the skin, nasopharynx, and gastrointestinal tract [41]. However, this bacterium becomes a significant health risk when it breaches epithelial barriers, leading to several pathologies ranging from skin infection to bacteremia [42]. The skin has characteristically low levels of glucose, making it a limiting growth factor in this environment [17]. Notably, S. aureus encodes four highly efficient glucose transporters that are absent in nonpathogenic species of Staphylococci [43]. This observation suggests a potential association between glucose assimilation and pathogenesis [44]. In support of this possibility, recent studies found that high glucose levels can exacerbate infection outcomes in mouse models of streptozotocin-induced hyperglycemia, which resembles some aspects of diabetes [44]. When infected intradermally with S. aureus, these mice have more severe dermonecrotic lesions, higher bacterial loads in abscesses, and greater dissemination into the kidneys than nondiabetic mice. In addition, diabetic animals showed elevated blood glucose levels and reduced tissue production of NO, supporting the idea that diabetes increases the susceptibility to bacterial infections [44].

In a recent study by Thurlow et al. [44], the increased pathogenesis of S. aureus in the skin with high circulating glucose was found to be dependent on the accessory gene regulatory (Agr) system. The Agr system of S. aureus regulates the expression of virulence factors, including PSMs, lytic toxins, and other immunomodulatory virulence factors [45–47]. This system comprises four genes: agrD, agrB, agrC, and agrA. agrD encodes the autoinducing peptide (AIP), while agrB encodes a peptidase that matures AIP. AgrC is a receptor histidine kinase, and AgrA is a response regulator. The activation of AgrA occurs when the extracellular concentrations of AIP reach a threshold concentration, leading to the binding of the AIP to AgrC and triggering the phosphorylation of ArgA and subsequent transcription of virulence factors [23]. Recent findings suggest that AgrC requires high ATP concentrations to achieve optimal activity on account of its low affinity for ATP in its histidine kinase domain [48, 49]. This observation led to the hypothesis that Agr requires high glucose to generate sufficient ATP for AgrC activation [44]. In agreement with this hypothesis, the study found that mutants in nonessential glycolytic genes of S. aureus showed low toxin production and ATP levels [44]. The reduced toxin production of the glycolytic mutants coincided with reduced dermonecrotic lesion severity, although bacterial load was not affected. This study supports the notion that glucose is a critical nutrient that promotes barrier disruption by regulating toxin production.

S. aureus can also activate glycolysis in keratinocytes to evade immune clearance by forming a persister subpopulation termed small colony variants (SCVs) [50, 51]. SCVs are important for chronic and relapsing infections as they are highly refractory to antibiotic treatment [52]. SCVs promote glycolysis in keratinocytes by breaking down fumarate, a metabolite that is known to inhibit glycolysis, from the extracellular environment (Fig. 1) [50]. Low fumarate concentrations and generation of mitochondrial ROS by keratinocytes lead to necroptosis, a form of programmed cell death characterized by inflammation and epithelial damage in the absence of significant recruitment of macrophages and neutrophils as seen in pyroptosis (Fig. 1) [53, 54]. The tissue damage and reduced recruitment of macrophages and neutrophils allow the internalization and passage of SCVs across the skin barrier [50]. This mechanism is supported by several pieces of evidence, including reduced fumarate in SCV-infected keratinocytes, increased expression of fumC (a gene involved in fumarate catabolism) in SCVs, and induction of necroptosis in keratinocytes infected with SCVs (Fig. 1) [50]. Additionally, induction of necroptosis in mice impairs bacterial clearance. Mice defective in activating necroptosis (mlkl^−/−) or treated with glycolysis inhibitors and ROS scavengers have improved clearance [50]. Altogether, these studies argue that SCVs induce glycolysis in keratinocytes by depleting fumarate and initiating a feedback loop that triggers necroptosis and fails to eliminate S. aureus from the skin [50].

It is clear that glycolysis plays a role in S. aureus survival on the skin, with direct effects on innate immunity [50, 51]. This point is further supported by genome analyses of clinical strains of S. aureus isolated from the skin, which determined that colonization drives the emergence of single nucleotide polymorphisms in hotspots like the quorum sensing gene, agrC, and single nucleotide polymorphisms in the metabolic pathways known to promote the emergence of SCVs such as genes involved in the TCA cycle and electron transport chain [51, 55]. Several representative isolates were found to stimulate glycolysis in keratinocytes, suggesting that the skin environment may promote the selection of strains that induce glycolysis to promote survival [51]. However, the impact of these mutations on the inflammatory response has yet to be fully assessed, and there is no experimental validation of the mutations and their effect on metabolic enzyme activity, metabolite levels, immune cell activity, and survival during infection [51].

In this section, we have provided two pertinent examples of how bacterial pathogens harness metabolism to regulate the immune response of the skin and promote pathogenesis. Although not exhaustive, these studies highlight how (i) metabolism directly drives compromised barrier integrity; (ii) accumulation of pro-inflammatory metabolites via nutrient catabolism promotes infection; (iii) reprogramming of host metabolism promotes a persister state at epithelial surfaces; and (iv) accumulation of host nutrients such as glucose regulate expression of factors that break down physical barriers and compromise innate cell function. These examples touch on the wide variety of potential ways bacterial metabolism could drive innate immune responses of the skin.

Bacterial Metabolism and Phagocytic Leukocytes

Macrophages and neutrophils are major effector cells of the innate immune system. They act as a first line of defense against foreign agents that penetrate the body’s protective barriers [20]. Maintaining adequate levels of these cells in the body is crucial, as individuals with low counts are at a significantly higher risk of infection [20]. Neutrophils continuously surveil the human body to detect microbial infection and respond quickly by entrapping and killing the invading pathogens [56]. Meanwhile, upon leaving the bloodstream, monocytes rapidly differentiate into macrophages, which can kill pathogens, produce a broad range of cytokines, eliminate damaged host cells, and promote tissue repair after the insult is eradicated [57]. During infection, cytokines released from epithelial cells recruit neutrophils and macrophages into the infected tissue [21]. Neutrophils are most abundant within the first 24–48 h of infection [6, 7, 20]. However, macrophages progressively replace neutrophils due to their short lifespan [56]. Macrophages can survive longer, self-renew at the infection site, and produce cytokines to sustain innate cell recruitment. They also drive several facets of adaptive immunity, including the amplification of antimicrobial signals via the presentation of antigens to helper T cells [20, 57]. Both types of cells can engulf pathogens and produce antimicrobial compounds, including free radicals, AMPs, and lytic enzymes that destroy the invading pathogen [6, 7, 20].

Ensuring complete elimination of a bacterial infection requires macrophages and neutrophils to adjust their metabolism to facilitate the production of antimicrobial molecules, synthesize soluble factors to activate the immune response while repairing cellular damage, and restore homeostasis after the infection has been resolved [58]. The metabolic status of these immune cells plays a crucial role in preventing infections and has been extensively reviewed elsewhere [14, 58–60]. However, bacterial pathogens also harness metabolism to evade killing by phagocytes. This section will discuss how bacterial pathogens use metabolism to resist free radicals produced by phagocytes, how lipid derivatives drive the immune response in macrophages, and how bacterial metabolism can maximize intracellular survival. While there are several other critical cells of the innate immune system (dendritic cells, NK cells, eosinophils, etc.), we will focus primarily on macrophages as most of the current literature in this area considers these cells.

Bacterial Metabolism Promotes Resistance to Free Radicals Produced by Phagocytes

Macrophages and neutrophils produce free radicals such as ROS in a process called respiratory burst to eliminate pathogens [61]. NADPH oxidase catalyzes the formation of superoxide, the precursor of hydrogen peroxide (H₂O₂), hydroxyl radical, hypochlorous acid (HOCI), and peroxynitrite. In contrast, myeloperoxidase synthesizes HOCI from H₂O₂ and chloride [62]. Despite the effectiveness of oxidative burst, macrophages and neutrophils are unable to clear several pathogens due in part to the production of antioxidants that transform these free radicals into less harmful byproducts [5, 62, 63]. Several small molecules like vitamin E, glutathione, bacillithiol, and coenzyme A, as well as enzymes like superoxide dismutase, catalase, glutathione peroxidase, and peroxiredoxins, all neutralize ROS to prevent cellular damage [62–64]. These antioxidants often neutralize most ROS; however, prolonged exposure can weaken these defenses, requiring alternative strategies to combat oxidative stress [63]. Recent research has shown that several bacterial pathogens can adapt to the infectious environment by tightly regulating their metabolism to minimize the impact of free radicals [15, 63, 65, 66]. This metabolic adaptation often involves switching from respiration to fermentation and coincides with the upregulation of the NAD kinases, the reduction of NADH synthesis, and the generation of ATP through substrate-level phosphorylation (Fig. 2) [15, 65–70].

Fig. 2.

The regulation of aerobic respiration and fermentation controls levels of internal ROS to promote bacterial survival in the presence of external ROS produced by phagocytic leukocytes. S. aureus and S. Typhimurium prioritize glycolysis and fermentation upon exposure to ROS from phagocytes to alleviate oxidative stress. The shift from respiration to fermentation reduces the generation of intracellular ROS while maintaining redox balance and ensuring ATP synthesis through substrate-level phosphorylation. Similarly, persister cells are tolerant to antibiotics and ROS on account of a metabolic transition characterized by reduced TCA cycle and OXPHOS activity.

In S. aureus and S. enterica serovar Typhimurium (S. Typhimurium), it is well established that fermentation significantly reduces the adverse effects of free radicals generated by immune cells. Aerobic respiration is detrimental for bacterial cells under oxidative stress due to the overproduction of intracellular ROS and the dissipation of the membrane potential (Fig. 2) [66]. For example, in S. aureus, resistance to NO is achieved by activating lactate fermentation through a fermentable carbon source such as glucose [71]. Bacteria that grow in media supplemented with nonfermentable carbon sources are more vulnerable to NO [71]. Additional studies showed that the high resistance of S. aureus to NO stress is attributed to the expansion of the Rex regulon [72]. The inhibition of respiration via NO leads to the accumulation of NADH, which in turn binds to Rex. This binding results in the derepression of genes involved in fermentation [72]. The Rex regulator in S. aureus controls additional genes and confers a higher level of NO resistance, with faster activation and better respiratory capacity than coagulase-negative staphylococci [72]. These studies suggest that the Rex regulon in S. aureus has evolved to respond efficiently to free radicals like NO, allowing a rapid switch between respiration and fermentation [71, 73–75].

S. Typhimurium also undergoes fermentation to prevent damage caused by free radicals [66]. By screening transposon mutants for increased sensitivity to H₂O₂, it was discovered that mutants in several glycolysis genes, including phosphoglycerate mutase (gpmA) and pyruvate kinase (pykF), had increased sensitivity to H₂O₂ [66]. The H₂O₂ sensitivity of a ΔgpmA mutant was dependent on respiratory activity, as the phenotype was abrogated under anaerobic conditions [66]. The study further validated the connection between H₂O₂ resistance and fermentation by testing H₂O₂ sensitivity in a mutant with defects in acetogenesis (ΔackA Δpta), a fermentation pathway that produces acetate from glucose catabolism while simultaneously regenerating ATP via substrate-level phosphorylation (Fig. 2) [76]. The ΔackA Δpta double mutant was more susceptible to H₂O₂ compared to the WT strain, confirming the link between fermentation and tolerance to free radicals (Fig. 2) [66].

In S. Typhimurium, fermentation is also required to maintain the membrane potential upon H₂O₂ exposure [66]. Dissipation of the membrane potential increases S. Typhimurium susceptibility to H₂O₂, an effect that was exacerbated in the ΔgpmA mutant. Determination of several parameters of respiratory activity in the ΔgpmA mutant uncovered a higher NADH/NAD+ ratio, reduced intracellular pH, and increased NADH dehydrogenase activity [66]. Additionally, the ΔgpmA mutant had a lower abundance of DsbAB proteins, which are part of a thiol-disulfide oxidoreductase system [66]. The DsbAB system catalyzes the formation of disulfide bonds within proteins using the respiratory chain as an electron acceptor [77]. This study emphasizes the significance of glycolysis in activating fermentation and maintaining membrane potential, which is required for the function of oxidoreductase systems like DsbAB that contribute to protein folding [66, 73, 74, 78].

Thus far, the composite evidence points toward a model where S. aureus and S. Typhimurium increase glycolytic flux and rely on fermentation to counteract oxidative stress caused by free radicals produced by phagocytic leukocytes (Fig. 2) [66, 71, 79]. This alternative strategy to produce energy during oxidative stress is beneficial because it limits the production of intracellular ROS and alleviates the stress imposed on respiratory enzymes, which are highly prone to oxidation (Fig. 2) [80–82]. Furthermore, the severe virulence defects of mutants in fermentative pathways suggest that S. aureus and S. Typhimurium do not always undergo aerobic respiration and instead employ fermentation to exploit the nutrients available in the infection site. Therefore, the shift between respiration and fermentation is a metabolic adaptation that offers an effective solution to free radical stress and ensures energy conservation during infection. Given the redundancy of fermentation pathways in S. aureus and S. Typhimurium, the identification of which pathways best promote resistance to free radicals and how the nutrients in the infection environment control the expression of these pathways will be an insightful area of future study.

Lipids and Fatty Acid Derivatives as Bacterial Mediators of Macrophage Responses

Pathogenic bacteria can adapt to the nutritional environment of the host by altering metabolism and acquiring host nutrients. One such nutrient is lipoic acid, an essential cofactor for several metabolic enzymes, including pyruvate dehydrogenase (PDH), oxoglutarate dehydrogenase, branched-chain α-ketoacid dehydrogenase (BCODH), and the glycine cleavage system [12]. Not only is lipoic acid required for the activity of these metabolic enzymes, but it is also implicated in oxidative stress resistance in M. tuberculosis on account of its redox-sensitive terminal disulfide [83]. Given its critical importance to cell viability, bacteria have evolved to ensure lipoic acid availability through de novo biosynthesis or salvage pathways [12, 84]. Disruption of the lipoic acid biosynthesis and salvage pathways in some pathogens results in attenuation during infection and altered sensitivity to oxidative stress [12, 85]. Recently, the relevance of lipoic acid to oxidative stress resistance and immunological defense was expanded. In a screen for S. aureus transposon mutants that confer defects in macrophage activation, a mutation in the gene encoding the lipoic acid synthetase (LipA), an enzyme involved in the last step of de novo synthesis, was identified that led to a TLR2-dependent hyperinflammatory response when added to macrophages [86]. A ΔlipA mutant lacks lipoylation on E2 subunits of metabolic enzyme complexes, including PDH, which decarboxylates pyruvate to generate acetyl-CoA [86]. E2-PDH is abundant within bacterial cells but is also released by S. aureus throughout growth, leading to the hypothesis that lipoyl-E2-PDH might directly blunt macrophage activation [86]. Experiments designed to test this hypothesis showed that purified lipoyl-E2-PDH can directly block TLR2 activation by interfering with the recognition of native ligands [86]. This effect requires lipoic acid, as its removal from E2-PDH leads to normal TLR2 responses [86]. Subsequent infection experiments confirmed that a ΔlipA mutant elicits greater pro-inflammatory cytokines and promotes the recruitment of pro-inflammatory macrophages, with improved restrictive capacity, to the site of infection [86, 87]. Moreover, a ΔlipA mutant is more susceptible to ROS produced by macrophages [86, 87]. These studies suggest that the attachment of lipoic acid to E2-PDH promotes immune evasion by blocking TLR2-mediated macrophage activation [12, 86, 87].

Interestingly, it is also known that PDH mediates resistance to macrophages in other Gram-positive pathogens, such as Streptococcus pyogenes [88]. S. pyogenes uses PDH to enable mixed-acid fermentation, and deletion of PDH in S. pyogenes leads to decreased CFU and reduced dermonecrosis during skin infection [88]. Furthermore, upon challenging macrophages, a Δpdh mutant was more sensitive to killing, which was in part attributed to an inability of the mutant to interfere with the acidification of the phagolysosome [88]. Additionally, macrophages infected with the Δpdh mutant produced a greater amount of IL-10, although the significance of this observation as it relates to enhanced sensitivity to macrophage killing is unclear at this time [88]. Nevertheless, these findings suggest that PDH hinders the ability to prevent phagolysosome acidification and sustain residence within macrophages. Consequently, PDH is critical in enabling both S. pyogenes and S. aureus evasion of macrophage defenses, although via starkly different mechanisms [88].

Lipoic acid-containing enzymes such as PDH and BCODH generate acetyl-CoA and the CoA precursors of branched-chain fatty acids, which are critical metabolites for fatty acid synthesis in S. aureus [89]. Despite their essential role in fatty acid synthesis, the dependency on PDH and BCODH for survival can be bypassed in environments where sufficient host fatty acids are available for incorporation into the membrane [89, 90]. Most lipids found in the host contain unsaturated fatty acids. S. aureus uses the lipase, glycerol ester hydrolase (Geh), to break down these lipids and release free fatty acids. Unsaturated free fatty acids are then assimilated into bacterial membrane phospholipids and lipoproteins using the fatty acid kinase, FakA, and fatty acid-binding proteins FakB1 and FakB2 [89–94]. The incorporation of unsaturated fatty acids into bacterial phospholipids and lipoproteins is an unusual occurrence since S. aureus does not synthesize unsaturated fatty acids de novo. Rather, it relies on the synthesis of branched-chain fatty acids to confer membrane fluidity and viability [90]. Lipoproteins are critical extracellular PAMPs derived from Gram-positive bacteria that are sensed through TLR2 [54]. S. aureus lipoproteins engage TLR2 receptors via esterified fatty acids, the composition of which could alter immunostimulatory capacity [93]. Indeed, unsaturated fatty acid incorporation into S. aureus phospholipids and lipoproteins stimulates robust TLR2-dependent production of IL-6 and keratinocyte chemoattractant in infected macrophages [92].

The incorporation of unsaturated fatty acids into the S. aureus membrane also drives excessive inflammation in mouse infection models. Several strains of S. aureus were generated to test the possibility that unsaturated fatty acid incorporation during infection augments immune activation [91, 92]. Indeed, levels of TNF, IL-6, and keratinocyte chemoattractant were highly elevated in the livers of mice infected with S. aureus strains that require host fatty acids for survival, causing dramatic changes in infection dynamics [89, 92]. These studies highlight the potential relevance of the acyl chain composition of the bacterial cell envelope in dictating the tone and duration of the innate immune response [93]. By balancing de novo biosynthesis of branched-chain lipids (less inflammatory) with scavenging unsaturated fatty acids from the host (pro-inflammatory), S. aureus calibrates immunological responses and infection dynamics [89, 91–93]. It is worth noting that BCODH is also relevant for the pathogenesis of other Gram-positive organisms, but in different ways. For example, the BCODH mutant in L. monocytogenes has significantly decreased bacterial burdens during systemic infection and impaired growth in minimal media [95]. However, the mechanisms behind this attenuation require further research.

As previously mentioned, bacteria have evolved intricate pathways for metabolizing lipids, which can ultimately calibrate the immune response in favor of bacterial survival [36, 91]. Diverse bacterial species can secrete enzymes, such as lipases and phospholipases, to break down host lipids [96, 97]. While many of these enzymes provide the bacteria with a carbon source or promote the recycling of released fatty acids, lipases can also promote pathogenesis and modulate immunological responses [96, 97]. For example, S. aureus secretes two lipases, Sal1 and Geh, which hydrolyze host and bacterial lipids. It was recently shown that the activity of Geh blunts the innate immune response by hydrolyzing bacterial lipoproteins to release free fatty acids and prevent engagement of TLR2 [91]. Interestingly, mice infected with a Δgeh mutant had a TLR2-dependent increase in proinflammatory cytokine production and accelerated clearance from tissues, indicating that the absence of Geh improves innate immune detection and bacterial clearance [91]. Together, these studies highlight the relevance of the lipid environment to the host response to infection. There are likely to be several variations on these themes in different microbes dependent on the numbers and types of secreted lipases they encode [96, 98].

Indeed, the importance of bacterial lipid metabolism in the interaction with host macrophages has been reported in several other pathogens [91, 99, 100]. In vivo experiments with S. Typhimurium established that mutants of the outer membrane translocase FadL and its activator protein, FadD, are attenuated during infection [99]. When bacteria were administered via oral gavage, the ΔfadL and ΔfadD strains had lower bacterial counts in tissue than the WT strain. Similar results were observed for other mutants in lipid metabolism, such as YafH, FabB, and FadA. Furthermore, ΔfadL, ΔfadD ΔyafH, ΔfabA, and ΔfabB strains were rendered more sensitive to macrophage mediated-killing by interferon-γ treated macrophages (M1 polarized) [99]. In contrast, there was no difference in killing among the bacterial strains when macrophages were pre-stimulated with interleukin-4 (M2 polarization) [99]. This study confirms the relevance of lipid metabolism in macrophage survival and its potential role in modulating immune signaling. However, it remains unclear whether the byproducts of lipid metabolism have immunomodulatory properties, and the molecular mechanism behind resistance to macrophage killing remains unclear.

Overall, these studies underscore the significance of bacterial lipids and their derivatives in the modulation of immune responses. Pathogenic bacteria use essential lipid derivatives like lipoic acid to support growth and suppress inflammation and modulate lipid composition in the membrane to redirect the inflammatory response and promote infection. Identification of the repertoire of bacterial and host lipids with immunomodulatory properties and further delineation of the molecular mechanisms by which lipases and their byproducts shape innate immunity will be fruitful areas for future investigation.

Bacterial Metabolism Promotes Resistance to the Intracellular Environment of Phagocytes

Bacteria encounter a harsh environment upon phagocytosis [101]. One of the most common metabolic pathways modulated by host cells in the presence of bacteria is glycolysis, which drives the acidification of the phagolysosome (Fig. 3) [102]. A recent study by Gutiérrez et al. [102] confirmed that glycolysis is critical for eliminating S. Typhimurium. The authors used metabolomics to establish that aerobic glycolysis increases when macrophages are infected with S. Typhimurium. Conversely, chemical inhibition of glycolysis in infected macrophages promotes intracellular replication. Based on these studies, the authors surmised that low glycolytic activity leads to poor acidification of the macrophage phagolysosome, which promotes bacterial survival. They subsequently determined that aldolase A, an enzyme involved in glycolysis, is required to regulate phagosome acidification (Fig. 3). It does so by coordinating the assembly of the vacuolar-type ATPase, an enzyme involved in pumping protons across the membrane of several intracellular organelles, including the phagolysosome [102].

Fig. 3.

S. Typhimurium links metabolite sensing to activation of virulence factors. Left: (1) To eliminate S. Typhimurium, macrophages activate glycolysis for phagosomal acidification. (2) The enzyme aldolase A promotes proper assembly of the v-ATPase complex in the phagosomal membrane. (3) The v-ATPase complex is responsible for driving protons into the phagosome, which results in rapid acidification. Right: (1) S. Typhimurium survives in phagosomes by sensing metabolites from central carbon metabolism and inducing the expression and translocation of SPI-2 effectors (2). The expression of these virulence effectors manipulates the macrophage response, creating a less hostile environment (3).

S. Typhimurium also directly exploits glycolysis to promote its pathogenicity [101, 103]. S. Typhimurium uses several glycolytic intermediates as carbon sources, including 3-phosphoglycerate, whose catabolism generates pyruvate and lactate, known activators of Salmonella pathogenicity islands (SPI)-1 and SPI-2 [103]. SPI-1 induction suppresses macrophage serine synthesis through the T3SS effector, SopE2, leading to the accumulation of 3-phosphoglycerate to maintain SPI-1 expression [103]. A series of carefully conducted experiments support this model, including the observation that macrophages infected with S. Typhimurium have increased expression of several glycolytic genes, along with the accumulation of glycolytic intermediates. Additionally, the concentrations of serine are significantly reduced in infected macrophages. This effect is only observed with live bacteria, implying the bacteria must be metabolically active [103]. Further, deletion of the phosphoglycerate transport protein (PgtP) led to attenuation in mice and reduced infection of the spleen and liver. However, when infected mice were administered glucose, the infectivity of the ΔpgtP mutant was partially restored, implying that S. Typhimurium relies on glycolysis to cause infection. The last piece of evidence that linked the metabolism of 3-phosphoglycerate with pathogenesis was the evaluation of ΔSPI-1 and ΔsopE2 mutant strains. These mutants could not suppress serine metabolism or promote 3-phosphoglycerate production in macrophages, suggesting that SopE2 is responsible for inhibiting macrophage serine synthesis [103]. Overall, this study demonstrates a remarkable virulence mechanism that couples virulence factor expression with glycolysis and serine metabolism to promote the survival of S. Typhimurium (Fig. 3) [103].

Macrophages infected with S. Typhimurium also accumulate the TCA intermediate, succinate, which is known to induce the production of IL-1β [102]. Like SPI-1 regulation via 3-phosphoglycerate, S. Typhimurium promotes antimicrobial resistance and T3SS activity by sensing succinate derived from macrophages [104]. Specifically, S. Typhimurium detects and internalizes succinate through the DcuB transporter, triggering the expression of SPI-2, which encodes the T3SS and several antimicrobial resistance genes. Activation of the SPI-2 regulon enhances intracellular survival and pathogenesis via expression of the T3SS effector protein, SteC, and the two-component systems PhoPQ and PmrAB, which confer resistance to polymyxin (Fig. 3) [104]. Deleting the dcuB and phoP genes in S. Typhimurium abolishes intracellular survival in macrophages and the ability to infect mice, suggesting that succinate is essential to induce the expression of the SPI-2 and cause disease [104]. These studies provide compelling examples of how S. Typhimurium adapts in response to macrophage metabolites to induce virulence genes and survive in the intracellular environment. However, there remain open questions regarding how hijacking macrophage metabolism impacts the recruitment or function of other immune cells and whether these metabolic shifts occur in these other cells [102–104].

M. tuberculosis is also highly resistant to macrophage killing [105]. A major pathologic feature of infection with M. tuberculosis is the formation of a granuloma, an organized structure of infiltrated macrophages, epithelioid, and foamy cells surrounded by a rim of lymphocytes [106]. M. tuberculosis can survive for long periods in a granuloma by adopting a nonreplicating but energy-producing lifestyle [105]. This metabolic state involves using fatty acids from macrophages, which promote the transition from a replicating to a nonreplicating or dormant state [100, 105]. Upon stimulation with M. tuberculosis, macrophages increase saturated and monounsaturated fatty acid synthesis [100, 107]. Lipidomic data obtained from infected macrophages with M. tuberculosis showed substantially increased intracellular levels of saturated and monounsaturated fatty acids due to the upregulation of the fatty acid synthase FASN-1 gene in macrophages [100]. These macrophages also increased the intracellular content of polyunsaturated lipids, which are catabolized to produce sub-metabolic products involved in cell signaling, including the activation of macrophages via TLR2/4 [107]. Interestingly, M. tuberculosis imports higher amounts of polyunsaturated lipids, including arachidonic acid, via the Mce1 transporter than other fatty acids, implying that polyunsaturated lipids could play a role in bacterial survival [100, 107].

M. tuberculosis possesses genomic regions of difference that are absent in nonpathogenic species and harbor factors that aid in promoting infection [108]. One such regions of difference factor, Rv1768, promoted survival in macrophages and during infection [109]. Rv1768 is a secreted protein that interacts with S100A9, a calcium- and zinc-binding protein that regulates inflammatory processes and immune responses [110]. This interaction promotes bacterial survival in macrophages by regulating nuclear factor-kappa B and TNF-α signaling pathways, as well as disrupting arachidonic acid metabolism [109]. Arachidonic acid metabolism in macrophages generates leukotrienes and lipoxins via 5-lipoxygenase (5-LO), which are important immunomodulators [100, 109]. Macrophages infected with a Δrv1768 mutant had reduced levels of 5-LO, an effect that was not observed in S100A9 knockout macrophages [109]. Macrophages infected with the WT strain produced leukotrienes but had reduced the generation of lipoxins, leading to an imbalance in macrophage activation, which promoted M. tuberculosis survival. Further studies suggested that Rv1768 may interact with S100A9 to suppress the nuclear factor-kappa B-TNF-α signaling pathways to control arachidonic acid metabolism and promote M. tuberculosis survival within cells; however, additional studies are needed to substantiate this possibility [109].

In contrast to S. Typhimurium, macrophages infected with M. tuberculosis have decreased glycolysis and OXPHOS activity, ultimately leading to a decline in ATP production [111]. This metabolic shift results in a quiescent state in which macrophages primarily rely on lipid metabolism to conserve energy. Notably, this quiescent state is exclusively triggered by pathogenic M. tuberculosis, as exposure of macrophages to the nonpathogenic species Mycobacterium bovis does not reduce glycolysis or OXPHOS activity [112]. A recent study confirmed that lipid metabolism in infected macrophages is linked to the transition of M. tuberculosis to a nonreplicative state [111]. The study used an engineered M. tuberculosis strain containing three fluorescent proteins, each under the control of a different promoter: (i) the GroEL promoter, which is constitutively expressed; (ii) the hspX promoter, which is expressed in the nonreplicative state, and (iii) the single-stranded DNA-binding protein promoter, which is expressed during active replication. By infecting macrophages with the reporter strain, the authors could isolate nonreplicative and replicative subpopulations of M. tuberculosis via cell sorting [111]. Upon reinfecting macrophages with the different bacterial subpopulations, transcriptomic analyses revealed that nonreplicative M. tuberculosis caused an increase in fatty acid metabolism compared to macrophages reinfected with the replicative subpopulation [111]. A genome-wide CRISPR screen subsequently identified important host factors that contribute to the nonreplicating behavior of M. tuberculosis, with lipid metabolism being significantly enriched [100, 111, 113]. These studies confirmed that fatty acids, triacylglycerols, and lipid metabolism from the host are associated with the emergence of the nonreplicative subpopulation of M. tuberculosis. Furthermore, the studies suggest that M. tuberculosis induces the synthesis of lipids in the host to trigger nonreplicative behaviors [111].

The highlighted studies suggest that several pathogens rely on detecting and responding to the metabolic state of macrophages. It is noteworthy that different pathogenic bacteria can initiate diverse metabolic responses that are tailored to the pathogenic lifestyle of that microbe. For instance, S. Typhimurium and M. tuberculosis have evolved sophisticated mechanisms to sense and respond to specific metabolic intermediates produced by macrophages in response to infection. The detection of these metabolic intermediates triggers changes in bacterial physiology, particularly the activation of virulence factors or the transition into a metabolic state in which bacterial cells are more resistant to macrophage-mediated killing. While we have provided examples of how these pathogenic bacteria survive within the intracellular environment of macrophages, further research is necessary to confirm whether these mechanisms apply to other bacterial species and other immune cells.

Bacterial Metabolism and Soluble Components of the Innate Immune System

The complement cascade is the principal soluble component of the innate immune system responsible for inducing inflammation and combating pathogens. Several pathogenic bacteria show impressive resistance to complement. This resistance is mediated by several mechanisms, including the secretion of proteases that degrade complement proteins and alterations of the microbial surface to prevent antibody binding [114]. The complement system consists of soluble proteins that remain inactive in the bloodstream until they encounter a pathogen. Once activated, complement proteins initiate a cascade of proteolytic events that result in protein cleavage and oligomerization. The complement proteins act on microbial surfaces to form the membrane attack complex (MAC) that causes cell lysis and release of cytoplasmic content. Moreover, the binding of the C3 complement protein on the bacterial surface improves recognition and subsequent uptake by phagocytes [114, 115].

There are three different complement pathways, each with their own activation pathway [114]. The classical pathway was the first to be discovered and is a critical component of the humoral immune response to infection [114, 116]. The classical pathway is triggered by the Fc portions of microbe-specific antibodies, where the plasma protein, C1q, recognizes antibodies attached to the surface of microbes [117]. Once bound to the Fc portion of the antibodies, C1q activates two associated serine proteases, C1r and C1s. Activation of C1r and C1s triggers a proteolytic cascade, which recruits C4 and its products, C4a and C4b, to form a complex with activated C2 [114]. This complex catalyzes the generation of C3 convertase, which cleaves free C3 into C3a and C3b [114]. C3b interacts with C5 convertase, cleaving C5 fragments into C5b, which forms a complex with C6, C7, C8, and C9 proteins to generate the MAC [115, 118]. The mannan-binding lectin pathway is activated when lectins or carbohydrate-binding molecules detect specific sugar-containing structures on bacterial surfaces, such as the capsule [114]. Finally, the alternative pathway is triggered by direct binding of C3 on the bacterial surface, leading to opsonization and direct killing of pathogens by recruitment of phagocytes [20, 114]. While several mechanisms used by bacteria to evade complement activation are well understood, comparatively less is known about the connection between bacterial metabolism and resistance to complement-mediated killing. This section will consider how some bacterial pathogens shift metabolism upon encountering components of the complement system and how metabolic proteins can interfere with the complement cascade.

Soluble Factors and Bacterial Metabolism

Transcriptomic data indicate that bacterial metabolism might be involved in complement resistance, as certain metabolic pathways related to central carbon metabolism, aerobic respiration, metals, and purine uptake are differentially expressed upon serum exposure [117, 119–121]. For instance, in E. coli, glycine, serine, and threonine catabolism are downregulated in the presence of serum [116]. Interestingly, when glycine was supplemented, E. coli became more susceptible to serum killing, which may suggest that bacteria need to decrease their glycine metabolism to thrive in the presence of complement factors [116]. Exogenous glycine is believed to be converted to serine and threonine through GlyA and Kbl, respectively. Serine is then broken down to α-ketoglutarate in the TCA cycle, increasing the membrane potential [116]. At the same time, the increased GlyA activity downregulates purine metabolism, which reduces the AMP and ADP pools in the cell, impacting ATP regeneration [116]. Low ATP regeneration and high membrane potential promote the expression of membrane-associated proteins, including HtrE, NfrA, and YhcD, which have a high affinity for complement factors, potentially conferring susceptibility [116]. While this study suggests that bacterial metabolism may be crucial in determining susceptibility to complement factors, it is important to note that further research is necessary to determine whether the upregulated pathways during serum incubation are directly responsible for mediating resistance to complement factors [116].

In addition to the metabolic changes in the presence of serum, bacteria can evade complement-mediated killing by assimilating sub-metabolic products from the host [122]. For example, Neisseria meningitidis takes up extracellular lactate via the LctP transporter to promote infection [122]. A ΔlctP mutant has a survival disadvantage relative to WT N. meningitidis, which correlates with increased serum sensitivity [122]. The increased sensitivity to serum was associated with greater C3 deposition on the surface of the ΔlctP strain [122]. A ΔlctP mutant does not have capsule defects but has reduced sialylation of its lipopolysaccharide (LPS). This study suggests that lactate metabolism might contribute to the sialylation of LPS that subsequently prevents C3 binding, although more work is needed to establish exactly how lactate metabolism causes changes to LPS and whether the change in sialylation is responsible for serum resistance [122].

A recent study with Helicobacter pylori provided some new insight into how lactate metabolism inhibits complement activity [115]. The study determined that H. pylori colonized the gastrointestinal tract of C3 knockout mice more efficiently than WT mice, highlighting the relevance of the complement system in protecting the host from bacterial colonization [115]. The authors found that the classical pathway was a primary mediator of H. pylori killing that could be abrogated by adding lactate to the serum. The study revealed that lactate consumption halts the accumulation of C4b on the cell envelope, thereby blocking C3 and C5 convertases and inhibiting MAC formation [115]. Although the mechanism by which H. pylori reduces C4b deposition remains unclear, it is hypothesized that lactate metabolism might modify the cell envelope composition or promote rapid elimination of C4b on the cell surface by activating an unidentified protease [115]. It is also possible that lactate induces the expression of the sialic acid-binding adhesin in H. pylori, which interacts closely with sialic acid and attracts factor H to inhibit C3 convertase [115, 123, 124]. It is worth mentioning that the addition of lactate to serum also promoted resistance in Haemophilus influenzae [125]. It appears that lactate production/assimilation may be a broad strategy used by several bacterial pathogens to resist complement-mediated killing, though the molecular details remain unclear [122, 124, 125].

The presence of serum also dramatically affects cell wall and lipid metabolism [114, 126]. These changes to the cell envelope play a crucial role in mediating serum resistance by impeding the accessibility of complement factors to the bacterial surface [114]. In addition to complement resistance, serum-induced changes to lipid and peptidoglycan metabolism also drive AMP resistance. S. aureus induces the production of cardiolipin and peptidoglycan after treatment with serum, which renders the bacteria less vulnerable to AMPs as well as the antibiotic daptomycin. The changes in cell wall thickness were mediated by the GraRS two-component system, which reduces the production of peptidoglycan hydrolases [126].

The presented studies implicate metabolism in resistance to complement-mediated killing. Several bacteria may modulate their metabolism to camouflage the cell surface and reduce the binding of complement proteins. Notably, this metabolite-dependent modification of the cell surface can be achieved through the acquisition of metabolites produced by the host during infection. However, there remains uncertainty about the molecular mechanisms of complement resistance in response to host metabolites, and future studies are warranted.

Metabolic Proteins That Directly Interfere with Complement-Mediated Killing

While the relationship between metabolism and complement resistance has principally been studied in Gram-negative bacteria, there is comparatively less research on this topic in Gram-positive bacteria. One reason is that most Gram-positive organisms are considered naturally resistant to the MAC due to the thick cell wall, which interferes with the insertion of C5b9 into the membrane [127]. However, genetic defects in the complement system increase the likelihood of colonization by both Gram-negative and Gram-positive organisms [128]. Furthermore, complement effectively eliminates Gram-positive bacteria through opsonization and enhanced phagocytosis [114]. Gram-positive bacteria possess several traits that help avoid complement-mediated opsonization, including inhibiting surface ligand binding, interfering with complement receptor recognition, and directly cleaving complement proteins [127]. However, it is unclear how bacterial metabolism affects the resistance of Gram-positive organisms to complement.

A study with S. pyogenes found that the transcripts for fructose-bisphosphate aldolase and enolase, both enzymes of glycolysis, were highly upregulated in the presence of blood [129]. Although indirect, the study indicated glycolysis is altered in the presence of blood and blood components, which include serum and serum proteins. Notably, a later study found that S. pyogenes uses proteins involved in glycolysis as an unconventional strategy to avoid complement attack [130]. One such protein is the phosphoglycerate kinase (PGK), which converts 1,3-bisphosphoglycerate to 3-phosphoglycerate [130]. This 41.9-kDa protein is found in the cytoplasm, membrane, and culture supernatant of S. pyogenes. It has been suggested that membrane-associated and released PGK interfere with complement by inhibiting MAC formation on the cell surface [130]. Biochemical experiments support this conclusion by demonstrating that recombinant PGK interacts with several complement factors, including C5, C7, and C9. This interaction inhibits the activation of complement by sequestering the C5, C7, and C9 proteins and interfering with C9 polymerization [130]. Importantly, recombinant PGK does not activate the complement pathway, and MAC inhibition was independent of its enzymatic activity [130]. Given the relative resistance of Gram-positive bacteria to lysis by MAC, it will be interesting to further evaluate the relevance of PGK in infection models.

In a similar vein, S. pyogenes can dampen the recruitment of neutrophils by binding the anaphylatoxin, C5a, a central chemoattractant for neutrophils and macrophages [131]. S. pyogenes inhibits C5a activity through a multiprotein complex comprising surface dehydrogenase (SDH) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH). SDH-GAPDH has a soluble cytoplasmic form but is also found on the bacterial surface upon activating the immune response. Experimental evidence suggests that SDH-GAPDH forms a complex with C5a, reducing neutrophil chemotaxis and H₂O₂ production [132]. The genes encoding SDH and GAPDH are essential in S. pyogenes, and their biological function was not established until Jin et al. [131] generated a truncated version of SDH in which a hydrophobic tail was added at the C-terminus, preventing its export. Further, transcriptional analysis of this mutant strain showed downregulation of genes involved in virulence, carbohydrate and amino acid metabolism, and lipid metabolism. In addition, the mutant was attenuated during intraperitoneal infection. These studies support a model where the export of the SDH-GAPDH complex is critical for virulence, reduces the intracellular concentrations of SDH, and protects the bacterium against complement [131].

Enolase is another example of a glycolytic protein that exhibits nonclassical cell surface display. Enolase appears to promote recruitment of the host factor C4b-binding protein (C4BP) on the surface of cells. This interaction inhibits the classical and lectin pathways of complement activation in response to Streptococcus pneumoniae by accelerating the decay of C3 convertase, thereby interfering with the MAC formation and chemotaxis of neutrophils [133]. Enolase is thought to interact with another protein called pneumococcal surface protein C, which promotes resistance to complement [133].

Altogether, the studies introduced in the section highlight the relevance of glycolytic enzymes in mediating resistance to complement in S. pyogenes and S. pneumoniae. Although direct interactions with complement proteins appear to be involved, it remains unclear whether changes in glucose levels affect sensitivity to complement proteins. There is still much to discover about how bacteria modify their metabolism in the presence of serum and how these changes may manifest in resistance to complement proteins and other soluble factors present in serum [129–131, 133].

Bacterial Metabolism, the Innate Immune System, and Antibiotics

Antibiotic resistance in pathogenic bacteria is a global issue [134]. Experts predict that multidrug-resistant (MDR) bacterial infections will soon become one of the leading causes of death worldwide [134]. Several factors contribute to resistance, including the overuse of antibiotics in livestock, international travel, inadequate hygiene practices, and a lack of knowledge regarding proper antibiotic administration [134–136]. Studies have shown that antibiotic resistance is primarily driven by genetic mutations and the acquisition of resistance genes through horizontal gene transfer [137]. However, recent evidence suggests bacterial metabolism can also directly promote antibiotic resistance through reduced metabolic activity that causes slow growth or growth arrest (Fig. 2) [138, 139]. Thus, the metabolic status of bacteria seems to play a crucial role in determining susceptibility to antibiotics [113].

A recent study with S. aureus serves as an intriguing example of the significant impact bacterial metabolism can have on antimicrobial susceptibility [140]. S. aureus is more resistant to β-lactam antibiotics when the nonessential gene, pgl, which encodes the pentose phosphate pathway enzyme 6-phosphogluconolactonase, is deleted [140]. The Δpgl mutant does not induce the alternative penicillin-binding (PBP2a) that confers resistance to β-lactams [140, 141]; rather, the resistance phenotype is associated with a change in metabolism. The Δpgl mutant redirects carbon flow to the production of cell wall precursors, ultimately changing cell envelope composition and increasing resistance to β-lactams [140]. The change in the cell envelope composition in the Δpgl strain was mainly associated with reduced production of lipoteichoic acid (LTA). This conclusion was supported by Western blot and wheat germ agglutinin agglutination assays, which showed lower LTA abundance and wheat germ agglutinin binding. The Δpgl mutant was also more susceptible to Congo red, which inhibits the LTA synthase, LtaS, but was more resistant to D-alanylation inhibitors, suggesting an imbalance between LTA and wall teichoic acids [140].

One important observation from this study was that the choice of growth medium can significantly impact β-lactam sensitivity [140]. Bacteria grown in Mueller-Hinton had the highest levels of resistance, while growth in chemically defined media conferred intermediate resistance, and growth in chemically defined media supplemented with glucose had the least resistance [140]. These results validated the prevailing hypothesis that the metabolic state of bacteria governs sensitivity to antibiotics [113]. Moreover, these results suggest that the cultivation methods often fail to align with the actual antibiotic susceptibility in vivo because they do not replicate the nutritional conditions in the host [113]. For example, MDR Gram-negative bacteria, such as E. coli, Pseudomonas aeruginosa, and Klebsiella pneumoniae, show high resistance to carbapenems due to the presence of excess zinc in Mueller-Hinton media as Zinc is a metal cofactor required for carbapenemase activity [142–144]. In validation studies, it was found that K. pneumoniae, E. coli, and P. aeruginosa were also highly resistant to carbapenems when grown in Mueller-Hinton broth [142, 143]. However, when the Mueller-Hinton broth was supplemented with a metal chelator, the bacteria had significantly increased carbapenem sensitivity. The effect of the metal chelator was only observed when assessing carbapenems but not other antibiotics, such as quinolones [143]. Mice infected with MDR K. pneumoniae and E. coli and treated with meropenem using 10% and 2.5% of the human-simulated regimen demonstrated that this carbapenem has good antibacterial activity against all MDR strains tested [143]. This increased antibacterial activity of carbapenems against the MDR strains coincides with low levels of zinc in mice, supporting the notion that the nutritional conditions of the host can drive the antimicrobial resistance profile [143].

The immune system can also influence antibiotic efficacy. A common example centers on the concept that inhibition of bacterial growth by antibiotics facilitates more effective immune clearance [138, 145]. However, the relationship between antibiotics and the immune system can also have harmful effects on the host [146]. For instance, a recent study by Ronneau et al. [147] discovered that the internalization of S. Typhimurium by macrophages treated with antibiotics drives the formation of persister cells [139]. Reactive nitrogen species (RNS) produced by macrophages in response to cefotaxime halts the bacterial TCA cycle and OXPHOS, resulting in growth arrest. The study provided evidence to support the idea that the production of RNS is essential for persister formation, as knockout macrophages in the NO synthase, Nos2, that were infected with S. Typhimurium generated fewer persister cells. RNS has bacteriostatic activity, affecting lipids, nucleic acids, and proteins, particularly metabolic enzymes containing thiols, iron-sulfur clusters, or heme [15, 147]. Thus, RNS has the potential to target several metabolic complexes of the TCA cycle and OXPHOS in S. Typhimurium, thereby reducing respiration and ATP levels. As a result of the lower respiratory activity, S. Typhimurium enters a nonreplicative state characterized by high tolerance to antibiotic treatment (Fig. 2). Curiously, by chemically inhibiting the production of RNS in macrophages or promoting bacterial growth by supplementing the media with glycerol, the authors found that S. Typhimurium cannot enter into a persister state and remains sensitive to antibiotics [147].

The same principle is true for E. coli and S. aureus, as stressors in the host environment, such as ROS, low pH, and antibiotics, cause the generation of persister populations through damage to TCA and OXPHOS enzymes (Fig. 2) [148–150]. Prolonged antimicrobial treatment in patients with S. aureus infections often leads to the emergence of SCVs [52, 148]. Clinical strains of S. aureus produce more SCVs than domesticated strains when exposed to low pH or incubated with neutrophils in combination with some antibiotics [51]. This phenotype is also observed in mice infected with S. aureus and treated with antibiotics, indicating that factors present in the host encourage the selection of persister populations [148]. In E. coli, redirecting carbon flux to the glyoxylate shunt increases the number of persister cells. In contrast, activation of the TCA cycle reduces the number of persister cells [150, 151]. Together, these studies highlight how host factors, like free radicals from phagocytes or the low pH in the phagosome, can contribute to the generation of persister cells by reducing OXHPOS and TCA cycle activity and validate the connection between low energy states and antibiotic tolerance [150, 151].

Besides tolerance to antibiotics, persister cells can also interfere with immune signaling and cause innate immune cells to adopt an anti-inflammatory state, which allows bacteria to survive [146]. A recent study showed that persister cells of P. aeruginosa are remarkably resistant to killing by macrophages and complement factors [152]. These cells have reduced C5b deposition and appear to promote the polarization of infected macrophages toward an M2 phenotype [152]. The M2-like macrophages exhibit anti-inflammatory characteristics and secrete anti-inflammatory cytokines such as IL-10 and CCL5 [152]. Interestingly, persister cells of E. coli show similar resistance to macrophage-mediated killing, whereas actively replicating cells are susceptible, indicating that the transition to the non-replicative state could be considered a selective advantage to thrive in the presence of antibiotics and innate immune cells [152]. Polarization of macrophages toward an M2 phenotype by persister cells also occurs with S. Typhimurium [153]. When infected macrophages are treated with antibiotics, S. Typhimurium generates persister cells, characterized by a nonreplicative but metabolically active state that permits the expression of the SPI-2 regulon and mediates macrophage polarization through the effector protein, SteE (Fig. 3) [153]. Notably, the deletion of steE no longer triggers M2 polarization in macrophages [153].

How antibiotics impact persister cell physiology and the downstream effect on macrophages can be complex due to the range of antibiotics used in clinical settings [154, 155]. For example, in a recent study, the treatment of S. aureus with trimethoprim-sulfamethoxazole, a dihydrofolate reductase (DHFR) inhibitor, was found to promote the emergence of a specific type of SCV designated as thymidine-dependent SCV (TD-SCV) [155]. TD-SCV triggers the stimulator of interferon genes (STING) signaling pathway in murine macrophages [156]. STING is an endoplasmic reticulum-bound protein that recognizes cyclic dinucleotides such as c-di-AMP and promotes the transcription of interferon-β and several pro-inflammatory cytokines [157]. The mechanism by which trimethoprim-sulfamethoxazole activates STING appears to be a consequence of inhibiting DHFR in S. aureus. DHFR converts dihydrofolate to tetrahydrofolate, an essential metabolic intermediate involved in purine and thymidylate synthesis [155]. Trimethoprim-sulfamethoxazole treatment depletes DHFR from bacteria, resulting in a state known as thymineless death, characterized by increased production of the second messenger, c-di-AMP. High levels of c-di-AMP trigger the STING signaling pathway in macrophages [155, 157, 158]. DHFR inhibitors also trigger the STING pathway in other Firmicutes, including L. monocytogenes and Enterococcus faecalis, which are both known to synthesize c-di-AMP [155]. However, the c-di-AMP-deficient Proteobacteria (S. Typhimurium and Francisella novicida) do not induce the expression of Ifnb1 in macrophages [155, 159]. Infection studies furthered these observations by demonstrating that administration of DHFR inhibitors to infected mice leads to the emergence of TD-SCV and leads to the overproduction of pro-inflammatory cytokines via STING. This, in turn, leads to increased neutrophil recruitment, causing tissue damage and creating an ideal environment for bacteria to avoid clearance [155, 160, 161].

The studies described in this section highlight several effects of antibiotic usage on bacterial metabolism, the consequences of which can have significant impacts on the immune response to infection and persistence. While our understanding of the important relationships between antibiotics, metabolism, and immune defenses is developing, it is crucial to recognize that each antibiotic has the potential to induce a distinct response. Further research in this area will undoubtedly provide new insights.

Conclusion

The studies outlined in this review highlight how several bacterial pathogens have adapted to exploit the host as a source of nutrients and bypass innate immunity. These pathogens effectively sense the metabolic status of their host, produce metabolites with immunomodulatory properties, shift metabolism to promote survival against antibacterial defenses, control energy balance, directly block soluble immune mediators, and shift immune cell metabolism to support bacterial survival. While themes are beginning to emerge, there is still much to learn about the intersections between bacterial metabolism and evasion of immune defenses. Just as pathogenesis is driven by adaptive traits that are often unique to the infecting organism, the roles of metabolism in driving resistance against host defenses are likely to be affected by unique traits of the infecting organism, site of infection, and replicative niche. We look forward to seeing considerable growth in this area of study as researchers continue to delve into the complexities of bacterial metabolism and its impact on host biology and immune defenses.

Acknowledgments

We thank Reginald Woods, Andrew Albers, and Dr. Liwei Fang for constructive feedback on this manuscript. Portions of figures were created with Biorender.com.

Conflict of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

This study was supported by grants NIH R01 AI120994, NIH R01 AI153059, and a Burroughs Wellcome Fund Investigators in the Pathogenesis of Infectious Disease Award to Francis Alonzo. The funders had no role in the manuscript conception, planning, writing, or decision to publish.

Author Contributions

Dr. Ivan C. Acosta and Dr. Francis Alonzo contributed equally to this work. All authors have read and agreed to the published version of the manuscript.

Funding Statement

This study was supported by grants NIH R01 AI120994, NIH R01 AI153059, and a Burroughs Wellcome Fund Investigators in the Pathogenesis of Infectious Disease Award to Francis Alonzo. The funders had no role in the manuscript conception, planning, writing, or decision to publish.