Immunometabolism: From basic mechanisms to translation

Abstract

Immunometabolism has emerged as a major mechanism central to adaptive and innate immune regulation. From early observations that inflammatory cytokines were induced in obese adipose tissue and that these cytokines contributed to metabolic disease, it was clear that metabolism and the immunological state are inextricably linked. With a second research wave arising from studies in cancer metabolism to also study the intrinsic metabolic pathways of immune cells themselves and how those pathways influence cell fate and function, immunometabolism is a rapidly maturing area of research. Several key themes and goals drive the field. There is abundant evidence that metabolic pathways are closely tied to cell signaling and differentiation which leads different subsets of immune cells to adopt unique metabolic programs specific to their state and environment. In this way, metabolic signaling drives cell fate. It is also apparent that microenvironment greatly influences cell metabolism. Immune cells adopt programs specific for the tissues where they infiltrate and reside. Ultimately, a central goal of the field is to apply immunometabolism findings to the discovery of novel therapeutic strategies in a wide range of diseases, including cancer, autoimmunity, and metabolic syndrome. This review summarizes these facets of immunometabolism and highlights opportunities for clinical translation.

1 |. INTRODUCTION

Two of the most fundamental features of multicellular life are the need to effectively distribute nutrients across cells and tissues and to protect from infection and injury. It is not surprising, therefore, that links between metabolism and immunity have become apparent in the past twenty years. Indeed, inflammation is now known to mediate many of the pathologies of metabolic syndrome. Conversely, the metabolic programs of individual cells of the immune system are under tight control and strongly influence cell function and differentiation. While malnutrition has long been known as widely immune suppressive,¹ Hotamisligil et al showed in a landmark paper in 1993 that adipose tissue produced TNFα in murine models of obesity and diabetes² to demonstrate that overnutrition can promote inflammation. Inflammation drives insulin resistance and features of diabetes and metabolic syndrome.³ In addition to metabolic tissues regulating immune cells, the metabolism of immune cells themselves is highly regulated. Warburg described metabolic changes and aerobic glycolysis in cancer cells early in the 20th century⁴ that are now known to be driven by activated oncogenic signaling pathways.⁵ Those same signaling pathways are activated to promote aerobic glycolysis in stimulated immune cells^6–8 and play key roles to reprogram metabolism from catabolic oxidative pathways to anabolic pathways.

Each hematopoietic cell type has clearly defined morphological, expression, and functional characteristics with defined subtypes that are continuing to be defined as technological approaches enabling increased sensitivity arise. However, these varied lineages and subsets of each cell type also reside in or travel to distinct tissue depots which impact function. In this light, it is not surprising that each population of immune cells has a distinct metabolism and nutrient usage. First demonstrated by Vats et al⁹ that macrophage metabolism can influence production of inflammatory cytokines as alternatively activated macrophages rely on a different metabolic program than classically activated macrophages, this principle has since been shown in T cells,¹⁰ myeloid derived suppressor cells (MDSCs),¹¹ and dendritic cells (DCs).¹² The cytokines and signaling pathways that guide distinct immune responses also promote these specific programs that support bioenergetics and biosynthesis. The intersection of metabolism and immunity at the systemic and cellular levels now forms the rapidly evolving field of immunometabolism. The most up-to-date concepts on the control of cell metabolism through cell intrinsic signaling mechanisms, the effect of the tissue microenvironment and nutrient availability on metabolism, and the potential for these pathways to be targeted or modified in disease are reviewed here and this volume (Figure 1).

FIGURE 1.

Immunometabolism: From Basic Mechanisms to Translation—Graphical summary of reviews in this collection. This figure summarizes the series of reviews provided by experts in the field demonstrating the mechanisms of immunometabolism, impacts on certain tissues such as adipose tissue, and disease states such as inflammation, cancer, and lupus. ATM: adipose tissue macrophage; VAT: visceral adipose tissue; OXPHOS: oxidative phosphorylation

2 |. BASIC MECHANISMS THAT REGULATE IMMUNE METABOLISM

The metabolism of T cells has been an area of intense interest and has many potential biological and therapeutic implications that illustrate the potential of immunometabolism. The majority of research has focused on uptake and metabolism of glucose, amino acids primarily glutamine, and certain fatty acids, although other metabolites impact metabolism. When T cells are stimulated through the antigen receptor, CD28 costimulation can increase glucose transporter GLUT1 expression, glucose uptake and subsequent glycolysis, and mitochondrial capacity.^6,13,14 Conversely, CD28 family inhibitory receptors cytotoxic T-lymphocyte-associated protein 4 (CTLA4) and programmed cell death-1 (PD-1) can suppress this metabolic transition.^15,16 T cells that fail to receive CD28 costimulation become anergic and are metabolically suppressed (ie, have low uptake of substrates and little mitochondrial metabolism).¹⁷ The metabolic program of T cells as they differentiate into different subsets and fates, however, is not shared. While effector T cells induce aerobic glycolysis, it was shown that memory T cells instead rely on mitochondrial metabolism and lipid oxidation^18,19 but can rapidly revert to glycolysis upon restimulation through endoplasmic reticulum-mitochondrial direct interactions that serve as metabolic hubs.²⁰ Unlike effector T cell subsets, regulatory T cells (Tregs) do not require GLUT1 or high levels of glutamine uptake through the amino acid transporter ASCT2^21,22 and instead predominantly rely on mitochondrial lipid, pyruvate, and lactate oxidation.^10,23–25 Tregs can be highly glycolytic, but the primary Treg transcription factor, FoxP3, has also been shown to repress glycolysis and high rates of glucose metabolism can impair Treg suppressive capacity.^24,26,27 Individual effector T cell populations also differ, as glutamine metabolism is essential for CD4 Th17 cells, but effector differentiation of CD4 Th1 and CD8 cytotoxic T cells as well as that of classically activated M1-like macrophages is restrained by this pathway.^28–30 These differences likely reflect varied modes of regulation, roles, and requirements of specific tissue sites, and offer distinct opportunities to selectively modulate immunity (Figure 2).

FIGURE 2.

Metabolism regulates immune cell activation. Glycolytic metabolism in immune cells typically leads to activation of an effector phenotype (eg, effector T, M1-like macrophages, DC, NK, and B cells), whereas oxidative metabolism of substrates such as fatty acids (lipid) and amino acids including glutamine leads to a regulatory or memory phenotype. Immunometabolism is necessary for not only cellular energetics, but also contributes to biosynthetic intermediates and signaling

Multiple signaling mechanisms coordinate after immune cells are stimulated to mediate metabolic reprogramming and the specific differences of each cell type and subset. In addition to transcriptional regulation through Myc and hypoxia-inducible factor (HIF1α) that drive induction of anabolic genes for glycolysis and mitochondrial metabolism in T cells^8,31,32 and macrophages,³³ the mechanistic target of rapamycin (mTOR) pathway plays a critical role in immunometabolism and cell fate. mTOR acts as part of two protein complexes, with the predominant immunometabolic regulation occurring through the actions of mTOR complex I (mTORC1) that is activated downstream of receptor stimulation of phosphatidylinositol 3-kinase (PI3K) and Akt. As described by Huang et al³⁴ the PI3K/Akt/mTORC1 pathway serves as a central mechanism to sense and integrate nutrient availability and signaling to promote T cell metabolism and function. mTORC1 is activated at lysosomes through the combined actions of T cell activation and signaling through PI3K and Akt that lead to activation of the Rheb GTPase and the coordinated sensing of essential amino acids through the sestrin/GATOR complex. The lysosomal localization plays a key role in this process as the amino acid transporter SLC38a9 transports arginine from the lysosome allow the ability of the sestrin/GATOR complex to sense leucine and isoleucine and activate mTORC1.^35,36 When activated, mTORC1 regulates a host of cellular functions to promote the transition from quiescence to anabolic metabolism and proliferation while suppressing catabolic metabolism. A key target for mTORC1 is the suppression of the AMP-activated protein kinase (AMPK) by phosphorylation of ULK1. AMPK is activated when cells experience energy stress to maximize oxidative pathways and promote autophagy, but also can have a role in normal physiology to balance mTORC1 activity.^37,38 The generally shared aerobic glycolysis processes across effector T cells may suggest similar dependence on mTORC1. Yet, it is clear that the mTORC1 pathway has distinct roles in each effector T cell subset and that a balanced and dynamically regulated metabolism may be the key to these differential roles. For instance, while CD4 Th1 cells rely on the mTORC1 activator protein Rheb, Th2 has been shown capable of Rheb-independent activation. This may, however, reflect graded dependencies on the levels of mTORC1 activity for each effector T cell subtype. Indeed, Tregs are dependent on a graded level of mTORC1 activity. In the absence of mTORC1 function, Tregs fail to suppress effector T cells and severe autoimmunity can ensue.^39,40 However, AMPK can promote Treg accumulation in vivo¹⁰ and may do so in part by moderating mTORC1 activity, as the mTORC1 inhibitor rapamycin can promote development of Treg.⁴¹ These findings about regulator and effector T cells demonstrate that factors that balance and fine-tune mTORC1 signaling function to integrate microenvironmental nutrients and signals and are fundamental elements of immunometabolic signaling.

B cell metabolism is also highly regulated and undergoes reprogramming as cells progress through development and differentiation. B cell activation and cytokine stimulation promote glycolysis^42,43 much like T cells, but other pathways can differ. Indeed, flux through the pentose phosphate pathway may be limiting in B cells⁴⁴ and mitochondrial activity is tightly regulated that may influence susceptibility to apoptosis.⁴⁵ Germinal centers also provide a dynamic microenvironment which may be hypoxic to drive glycolysis and alter B cell proliferation and fate,³¹ although increased fatty acid metabolism in germinal center localized B cells has also been reported.⁴⁶ How these changes relate to the particular activation and differentiation states of B cells as they transition through activation, germinal centers, and plasmablast differentiation is discussed in detail by Jellusova.⁴⁷ Because germinal center B cells undergo class switch DNA recombination and somatic hypermutation, errors leading to mutations can occur. Hence, many of the factors that drive B cell activation and metabolism can also lead to malignant transformation. The mismatch between the oncogene-driven metabolic demands and the metabolic pathways and flux meant to support those demands, however, can lead to B cell apoptosis. Thus, Jellusova et al emphasize that understanding mechanisms that regulate B cell metabolism, therefore, informs normal B cell biology as well as B cell transformation to leukemia and lymphoma. Ultimately, antibody-producing B cells have clear metabolic demands to support high rates of protein synthesis; however, the metabolic requirements of plasmablasts and long-lived plasma cells remain uncertain.

Innate immune cells also undergo metabolic programming and use metabolic pathways to control cell fate. Bakker and Pearce provide a detailed review on the past decade of progress in understanding immunometabolism with a focus on DC and macrophage metabolism.⁴⁸ Mononuclear phagocytes are important in sampling the environment and crosstalk with neighboring cells, effectively bridging innate and adaptive immunity. Appropriate control over cell metabolism is necessary for these interactions and acts to fine-tune inflammatory responses. After lipopolysaccharide (LPS) stimulation, DCs exhibit a brief burst of oxidative phosphorylation (OXPHOS) that is downregulated after several hours to allow cells to engage aerobic glycolysis to fuel the need for metabolic intermediates. Macrophages have similar metabolic reprogramming after danger signal activation such as LPS. Much of the foundational work for this myeloid immunometabolism arose from work in macrophages showing two breaks in the tricarboxylic acid (TCA) cycle after pro-inflammatory stimuli. A main function of inhibition of the TCA cycle is that this allows excess citrate to feed into other anabolic pathways and succinate to promote inflammation.^49,50 Also, production of the bioactive metabolite itaconate is stimulated by pro-inflammatory stimuli. Itaconate plays an anti-microbial role but more critically it blunts succinate dehydrogenase (SDH) enzymatic function,⁵¹ leading to a build-up of succinate. The block at SDH leads to the second break in the TCA cycle which promotes mitochondrial ROS production, HIF1α stabilization, and a continued commitment to glycolysis in the activated cell. In addition to inhibition of SDH, itaconate leads to activation of a potent anti-inflammatory and anti-oxidant transcription factor nuclear factor erythroid 2—related factor 2 (Nrf2).⁵² There is great interest in itaconate or its derivatives in therapeutic approaches due to its ability to alkylate cysteines on metabolically sensitive enzymes. In contrast, alternative M2-like activation of macrophages is reliant upon TCA cycle—mediated metabolism of glutamine, glucose, and fatty acids.^9,53 Indeed, transgenic expression of Glut1 increased classically activated macrophage inflammatory function, while genetic deletion of Glut1 specifically repressed those functions.^54,55

Lastly, Bakker and Pearce touch upon the role of mitochondrial metabolism downstream of a little studied pathway called hypusination through the polyamine spermidine derived from arginine, an amino acid well studied in macrophage metabolism.⁵⁶ Inhibition of hypusination of eukaryotic initiation factor 5 (eIF5A) led to the reduction of alternatively activated macrophage markers, TCA cycle flux, fewer electron transport chain (ETC) components, and an overall reduced OXPHOS metabolism. Surprisingly, inhibition of this pathway had no effect on pro-inflammatory macrophage activation. These disparate findings emphasize the fact that mitochondrial metabolism in pro-inflammatory macrophages is not to fuel energy production but to generate metabolic intermediates, whereas in alternatively activated macrophages, OXPHOS is necessary for the M2 phenotype.

Another innate mechanism of immunometabolic regulation is through the complement system. The complement system is a pattern recognition receptor (PRR) system that has newly identified roles to regulate basic T cell metabolism and physiology.⁵⁷ In addition to secreted complement that arises from the liver, West et al review recent findings showing that T cells produce small amounts of intracellular forms of complement.⁵⁸ C3 and C5 in this setting are each cleaved to C3a and C3b and C5a and C5b by intracellular proteases. A subsequent set of intracellular receptors then recognize these ligands that promote glycolysis and lipid oxidation. Indeed, West et al speculate that complement arose in single-celled organisms as a metabolic and stress-sensing mechanism that subsequently evolved to recognize and help eliminate pathogens. This intracellular system may be essential for cell viability, as individuals with complement deficiency syndromes nevertheless have been shown to retain expression of intracellular complement components. In addition to direct actions of complement, the complement regulatory cofactor CD46 is a human-specific complement receptor which can promote T cell glycolysis in response to engagement with extracellular complement.⁵⁷ While the CD46 proximal signals are not certain, the Notch and PI3K pathways are likely candidates to then promote T cell glycolysis and inflammation. Importantly, individuals deficient in CD46 have T cell immunodeficiencies. Overall, complement has been previously recognized to promote defense from invading pathogens, but current data support an additional role to maintain cell survival and tissue homeostasis.

One approach to identify novel immunometabolic regulators is to take a human genetics strategy. In particular, Mendelian genetic metabolic disorders provide insight into genes and pathways essential for normal metabolic function and influence on immunity, as reviewed in detail by McGuire.⁵⁹ While the majority of these diverse disorders have not presented apparent immunological phenotypes, immune cells have not yet been extensively explored in these settings. With clear examples, such as glycogen storage disease-1b that leads to lymphopenia and diminished ability of T cells to upregulate glycolysis, it is apparent that these human genetic diseases can provide a wide-ranging opportunity for immunometabolic discovery. Indeed, screening patients with immune disorders for metabolic phenotypes recently identified gain-of-function mutations in SDH,⁶⁰ whereas polymorphisms in ETC proteins were found to lead to increased incidences of infection.⁶¹ In sum, taking this genetic approach based on human disease coupled with biochemical analyses of T cells may reveal new immunometabolic regulators and pathways.

3 |. IMMUNOMETABOLISM IN TISSUES

Much of the work on immunometabolism has been performed by necessity in vitro and with immune cells from secondary lymphoid organs, such as spleen and lymph nodes. This is due to technical challenges caused by the need for large numbers of cells for many biochemical assays and of the rarity and heterogeneity of specific immune cell populations in other tissues. These barriers are rapidly being overcome due to technical advances, and it is clear that while metabolism based on in vitro assays can teach us a great deal about immunometabolism, pathways in vivo can be regulated differently. For example, it was recently shown using ¹³C-glucose tracing that effector T cells use glucose differently in vivo, with a greater proportion directed to biosynthetic pathways derived from glycolysis.⁶² Stable isotope metabolomics and other in vivo and tissue-based assays are now increasingly accessible and allow for innovative directions in our approaches to understand immunometabolism.

One tissue that has been a focus of immunological research is adipose tissue. A wide range of changes in the adipose tissue—resident immune cell populations occur with obesity, which can vary by depot. Macrophages, in particular, change in content and phenotype from lean to obese to weight-cycled states and influence the development of insulin resistance, diabetes, cardiovascular disease, and cancer. Since back-to-back reports of adipose tissue infiltration by macrophages in 2003,^63–65 the field has shown the importance of essentially every immune cell in adipose tissue with unique depot specific differences. Caslin et al⁶⁶ review the contribution of the predominant immune cell in fat tissue and adipose tissue macrophages (ATM) where extrinsic effects on local and systemic immunometabolism, as well as intrinsic immunometabolism and effects on phenotype, are discussed. From the first report of differences in ATM characteristics between lean and obese,⁶⁷ the evolution of the ATM phenotype has varied. In short, in vitro differentiation studies to generate pro-inflammatory M1-like macrophages vs. anti-inflammatory immunomodulatory M2-like macrophages failed to hold true in vivo. Advances in flow cytometry, proteomics, metabolomics, and single cell sequencing of macrophages have shown varied phenotypes that blend both classical M1-like and alternative M2-like markers depending on the age of mice, duration on diet, and extent of obesity or weight loss. The interaction of ATM with oxidized lipids and other metabolites secreted from adipocytes or varied stromal cells in adipose impacts inflammation and insulin resistance.⁶⁸ Thus, it is increasingly important to assess the metabolic and cellular interactions in the adipose tissue.

In addition to macrophages, a unique visceral fat—specific Treg population was recently defined that functions to keep this vast depot of adipose tissue quiescent in terms of inflammation to maximally sensitize insulin signaling.^69,70 A review by Li and Spallanzani et al details a critical role of visceral adipose tissue (VAT) mesenchymal stromal cells (VmSCs) in regulating VAT Tregs, local, and systemic inflammation in obesity.⁷¹ VAT Tregs were first identified in 2009 and comprise 40%−80% of CD4+ T cells. VAT Tregs are distinct from lymphoid tissue—associated Tregs which comprise only 5%−15% of CD4T cells in that they are derived from the thymus, become more “VAT like” in the spleen (a poorly understood step), and then seed VAT early in life. VAT Tregs proliferate and survive in VAT, rather than result from conversion or recruitment. VAT Tregs are typically more clonal in a MHCII VAT antigen presenting cell-dependent manner and less heterogeneous than lymphoid-resident Tregs. VAT Tregs function to improve insulin sensitivity, reduce pro-inflammatory macrophages, and increase anti-inflammatory macrophages and reduce inflammation in adipocytes, among other functions. VAT Tregs are highly dependent upon the nuclear hormone receptor PPARγ (distal to TCR signaling) and have a unique transcriptome with many lipid metabolism genes upregulated compared to non-VAT Tregs. Tregs in human studies have been shown to be both higher and lower in obesity, but the nature of these comparative studies is difficult to interpret without controlled interventions.

VAT Tregs are highly dependent on the IL-33-ST2 axis, which is typically associated with allergic Th2 response, but is increasingly appreciated for potent wound healing capacity, promotion of tumor immunosuppression, and activation of increased metabolism of white and brown fat, potentially through VmSC-dependent IL-33 production.^72,73 They describe a fibroblast-like population of mSC with several phenotypic subsets, three of which produce IL-33 (ie, immunocyte promoting) compared to IL-33-negative adipogenic cells with markers of pre-adipocytes. High-fat diet feeding transiently reduces IL-33-producing VmSC cells followed by a reduction in VAT Tregs. Interestingly, sex hormones appear to regulate subsets of VmSC populations and production of IL-33, which could account for differences in Tregs in basal or obese state between males and females. Like other immune cells in diet-induced obese models, short-term high-fat diet—induced obesity versus long-term and severe obesity affect this axis differently. While short-term high-fat diet VmSC-IL33 + cells are associated with Tregs, this positive association is lost in severe obesity. Morbidly obese patient adipose tissue contains VmSCs (ie, are CD9hi) that associate with diabetes, fibrosis, and inflammation. On the other hand, the CD9lo population that is enriched for PPARγ and is reduced with diabetes. However, the role of VmSCs in IL-33-dependent Treg homeostasis in obesity is unclear. Controlled time course single cell sequencing studies in the future are necessary to inform on the specifics of cell kinetics that are lacking in single endpoint obesity studies. More research is necessary to understand the bi-directional crosstalk between stromal cells and Tregs, commonalities between human and mouse, and the kinetics and phenotype of these cells in obesity and other diseases. Tregs also play key roles to maintain homeostasis in other peripheral tissues including liver, skin, and tumors and are reviewed by Wang et al⁷⁴.

4 |. IMMUNOMETABOLISM IN DISEASE AND TRANSLATION

Like activation and differentiation of lymphocytes and myeloid cells that lead to metabolic reprogramming, viral infections have been shown to reshape cell metabolism in ways that are central to the antiviral immune response. Viral infection can alter cholesterol synthesis,⁷⁵ and Bahadoran et al focus on influenza-induced metabolomic reprogramming of immune cells in the lung airway.⁷⁶ The lung microenvironment comprises barrier cells and mechanisms as well as immune regulatory mechanisms that maintain homeostasis and defend the body against pathogens. Influenza infection induces metabolic reprogramming of both immune and epithelial cell populations to co-opt the host’s cells. In epithelial cells, one example of influenza-induced metabolic reprogramming is by the induction of an aberrant activation of the PI3K/mTOR/Akt pathway. Additionally, influenza induces immune cell activation and differentiation, although the impact of influenza A viral infection on immune cells is largely uncharacterized. The authors discussed metabolism in both steady state and upon influenza viral infection relevant to immune cell populations such as DCs, macrophages, natural killer (NK) cells, and T cells. For example, influenza viral infection causes DCs to enter a state of hyperglycolysis with relatively low oxygen utilization. Macrophages in the respiratory tract largely play a role in phagocytosing the virus and infected cells. Yet, M1 macrophages activate a strong antiviral response including macrophage-derived reactive oxygen species (ROS) that is deleterious to the host, thus representing potential targets to reduce disease severity. Influenza infection of T cells induces a switch from OXPHOS to T effector utilization of glucose and glutamine. In sum, effector functions of immune cells such as DCs, NK cells, macrophages, and effector T cells rely upon glycolysis, while T regulatory cells use OXPHOS. Also, systemic glucose or other microenvironmental foci of influenza infection are discussed because they shape the immune response. Like tumors, influenza infection may induce immune suppression and infected epithelial cells become glucose- and glutamine-dependent. Therefore, a delicate balance is needed to fuel the immune response without exacerbating damage to the host.

An ultimate goal of immunometabolism is to exploit the specific metabolic programs of distinct immune cell populations to treat disease. Systemic lupus erythematosus (SLE) has been studied extensively in this regard and is reviewed by Teng et al⁷⁷. The metabolism of T cells in SLE has been recognized to be altered for nearly twenty years.⁷⁸ The wide range of T cell subsets, B cells, and DCs that contribute to SLE have been shown to have differential metabolic requirements. These are impacted by altered systemic nutrition, diet, and cell intrinsic changes to metabolism. Obesity and metabolic syndrome are associated with SLE and may contribute to a variety of SLE-associated pathologies.⁷⁹ This may occur due to increased nutrient acquisition by inflammatory cells in settings of hyperglycemia and hyperlipidemia, but may also be regulated through changes in hormones or adipokines, such as leptin. A variety of metabolic inhibitors have been tested in animal models of SLE.^80–82 Inhibitors of mTORC1 have shown some promise, and direct inhibition of glycolysis may also protect from disease. This effect was strengthened, however, by addition of the AMPK activator, metformin. A concern in these metabolic inhibitor treatments has been the potential for widespread toxicity. This concern, however, has not materialized, and treatments have been found to be remarkably well tolerated. Even blockade of the fundamental process of glucose uptake was not found to be widely cytotoxic and provided protection in animal models of SLE.⁸¹ Combinations of metabolic inhibitors for glycolysis and alternate metabolic pathways now offer new directions with successful preclinical proof of principle studies, and it is now important to move to human clinical trials to combat this auto-immune disease.

Tumor microenvironments can also alter the metabolism of T cells and macrophages to impair anti-tumor immunity.⁸³ Critically, treatments such as checkpoint blockade therapies act in part through relieving this metabolic inhibition.^15–16,84 Siska et al focus on the “Warburg effect” which is a key feature of tumor cells and leads to acidification and the accumulation of lactate in the tumor microenvironment.⁸⁵ As a consequence of localized lactic acidosis, immune cell populations within the tumor microenvironment exert an immunosuppressive effect and mediate an effective immune escape by tumor cells. The authors suggested a metabolic-tumor-stroma score (MeTS) as a predictor of the efficacy of anti-tumor immune responses. MeTS are broken down into four sub-categories: MeTS 1 which consists of respiring tumor cells (OXPHOS) with high T cell infiltration (“Hot” tumors); MeTS 2 with a respiring tumor and glycolytic (Warburg-like) stromal cells; MeTS 3 with a glycolytic tumor and respiring stromal cells; and MeTS 4 with a glycolytic tumor and low T cell infiltration (“Cold” tumors). First, the authors provided a historical timeline of the Warburg effect, and then discussed the Warburg effect in immune cells and its underlying mechanisms in thorough detail. For example, HIF activation is not solely dependent on hypoxic conditions, but can also be achieved under normoxia conditions and the presence of lactate or pyruvate. The authors also discussed how tumor-derived lactate has a broad spectrum of effects on tumor growth and metastasis such as the promotion of cell motility, modulation of tumor-associated fibroblasts, impairment of anti-tumor immune cells (NK cells, CD8 T cells, M1 macrophages, and DCs), and promotion of immunosuppressive immune cells (MDSC, M2 macrophages, Tregs). Finally, the authors focused on strategies targeting the Warburg effect for tumor therapy and immune activation. For example, preclinical and clinical strategies aimed at targeting LDH-A, LDH-B, MCT1, and MCT4 blocked lactate secretion which improved anti-tumor immune responses.

Both systemic and local influences can also affect metabolism and anti-tumor immunity.

Turbitt et al focus on the effect of obesity on T cell metabolism and function in the presence and absence of solid tumors.⁸⁶ Obesity alters not just Treg, but also CD4 and CD8 T cells that play roles in anti-tumor immunity.^87,88 Adipokines, such as leptin, are elevated in obese individuals and can promote T cell glycolysis and polarization toward an anti-tumor Th1 phenotype.⁸⁹ Thus, the authors provide an explanation of why immune checkpoint blockade provides superior anti-tumor immune responses in both obese preclinical and clinical models compared to lean models. Leptin reverses metabolic defects in T cells and increases the expression of PD-1 on their surface. Therefore, immunotherapy such as anti-PD-1 is more efficient in high-PD-1-expressing T cells that are present in obese individuals. Moreover, the authors discuss the effects of glycolytic tumors on the impairment of T cell functions. Highly glycolytic tumor cells induce diminished glucose uptake by T cells and promote immunosuppressive myeloid cells such as MDSCs, therefore rendering T cells less effective. Additionally, the authors discussed obesity-driven intracellular lipid accumulation in T cells and DCs as a mechanism by which tumors inhibit anti-tumor immune cell functions. Finally, the authors highlight novel therapeutic approaches focused on improving T cell metabolic fitness. Few examples include mTOR inhibition and the use of anti-diabetic drugs such as metformin in combination with current immunotherapeutic approaches. These strategies aim at improving T cell OXPHOS and cytokine production to increase the efficacy of anti-PD-1 blockade.

Sipe and Chaib et al tackle the role of host-microbial interactions and the impact of microbially derived metabolites on immunometabolism in anti-tumor immunity and other chronic diseases. Microbes are altered by many interventions and lifestyle factors, and the metabolic sequelae are just beginning to be uncovered. While most work on the microbiome has focused on the gut, extra-intestinal microbes are detected in various tissues, including tumors, and are susceptible to modification by dietary intake or changes in body composition such as weight gain or loss. How gut or extra-intestinal microbes alter their own and the host’s metabolism is actively being researched because the magnitude of our microbiome’s non-mammalian genetic capacity to generate or modulate metabolites is vast. The impact of host bile acids and modification of bile acids by microbes is discussed in detail for cell types including tumor-associated macrophages, DCs, MDSC, T cells including Tregs, Th17, mucosal-associated invariant T (MAIT), NKT, and B cells. Current literature suggests that bile acids or agonism of bile acid receptors such as FXR or TGR5 in innate and adaptive immune cells leads to a general downregulation of inflammation in most cell types. However, microbial modification of bile acids and signaling through epithelial cells led to activation of bile acid targeted recruitment of NKT cells and improved anti-tumor immunity.⁹⁰ Thus, BA composition may influence diseases such as cancer through reprogramming of the immune microenvironment. Interestingly, microbial dysbiosis and specific microbes including bacteria and fungi correlate with several cancers, notably tumor subtypes, stage, or race in breast cancer.⁹¹ Furthermore, certain microbes have been shown to impact response to immune checkpoint blockade in anti-tumor therapy in mice and correlate with therapeutic response in humans.^92–94 Based on these and other related studies, it is possible that immune perturbations resulting from microbiome composition and microbially modified metabolites can be targeted for therapeutic benefit. There is increasing interest in “bugs as drugs.” The discovery of novel pathways driven by microbial metabolism that could be exploited in the treatment of cancer including the clinical efficacy of immune checkpoint blockade, obesity, and metabolic syndrome, as well as auto-immune diseases, would be impactful to understand risk and impact of personalized therapy and improve therapeutic outcomes.

5 |. CONCLUSION AND PERSPECTIVE

The field of immunometabolism has emerged as a major and attractive area of research in immunology and metabolic disease. The focus on systemic and cellular metabolism and the diversity of metabolic programs of specific cell populations have enhanced our understanding of both basic biology and metabolic dysfunction, as well as identified novel targets for anti-inflammatory and anti-tumor immune-based therapies. This issue aims to highlight key areas of current research, but many important questions and challenges remain especially with regard to commonalities or differences between humans and preclinical models, sex, or race. It is clear that the tissue location is critical to immune cell intrinsic metabolism through changes such as cell-cell crosstalk, viral foci, competition for nutrients, and extent of hypoxia. Future studies must incorporate the effect of the microenvironment on immune cell metabolism, and vice versa. Finding new approaches to understand the diversity of metabolic programs in specific tissue sites will open new avenues for discovery and immune modulation. Ultimately, identifying therapeutic targets is a key goal for the field. While standard-of-care treatments, such as methotrexate, demonstrate the potential of targeting metabolic pathways, better understanding of cell and tissue biology will certainly reveal new targets with increased specificity and reduced toxicity. As a rapidly maturing field, immunometabolism is now poised to move forward to address these needs.