A Metabolic Regulation of Immune Response: Immunometabolism during Viral Infection

1. Introduction

Metabolism is generally referred to as the sum of all biochemical reactions within an organism designed to produce or consume energy and, by extension, allow the maintenance of life. Metabolism thus includes all the anabolic or catabolic reactions taking place at the cellular or systemic level. On the one hand, it provides the energy necessary for the cell to ensure its activities, whether in the form of ATP or reduced substrates such as NADH or NADPH. On the other hand, via anabolic pathways, metabolism participates in the synthesis of biomass (nucleic acids, amino acids and proteins, lipids, and carbohydrates) useful for the storage of energy, and information, or for the formation of functional units of the cell (i.e., proteins). The balance between anabolism and catabolism is finely regulated to preserve energy homeostasis. This regulation is dependent on enzymes that are very sensitive to changes in energy status (i.e., glucose deprivation vs. high glucose level). Indeed, the cell seeks to maintain an adequate ATP/ADP ratio with either cellular functions or growth.

As metabolism is the keystone of cellular activity, the mechanisms by which it is governed deserve attention. It should be noted that the consequences of metabolic disorders on cell functions are studied in the specific case of human pathologies, such as diabetes, obesity, or cancer. Immune responses (i.e., innate and adaptive) as well as pathogen sensing are among the issues for which a close link with metabolism has also been established [1]. During viral infections, particular metabolic states, namely the metabolic pathways active at the time of infection, seem to be favorable to the virus multiplication. This may be through a direct action of some metabolites in the early stages of the viral cycle or the establishment of antiviral responses (notably interferon-dependent) [2].

To maintain its homeostasis in a fluctuating environment, the cell must have a consistent metabolic status so that a shift in the dynamics of cell metabolism is possible depending on the cell's fate. Schematically, a quiescent cell will tend to favor the use of carbohydrate substrates, degraded to pyruvate to fuel the tricarboxylic acid cycle (TCA) and provide large amounts of energy in the form of ATP. This mechanism is driven via oxidative phosphorylation – OXPHOS, which enables cells to maintain their activities (e.g., enzymatic activities, ATP-dependent pump, ATP/ADP ratio-dependent signaling). On the contrary, a proliferating cell will favor the rapid production of energy through ‘aerobic glycolysis’, leading to the utilization of pyruvate to produce lactate even under aerobic conditions (Figure 1). This mechanism, known as the Warburg effect, was initially observed in cancer cells and tends to facilitate the abundant synthesis of biomass at the expense of efficient energy production [3]. More recently, the ability of some viruses to modulate the metabolism in this way has been established, reminiscent of the Warburg effect [4,5]. During infection, the virus exploits the cellular machinery to its advantage, aiming to rapidly produce and release an important quantity of virions. This intense viral replication within the infected cell necessitates a significant amount of nucleotides, amino acids, and lipids, all supplied in a manner similar to that in a cancer cell, through metabolic reprogramming [4]. This ability to reprogram or redirect the host's metabolism seems to be shared by many viruses, spanning different families and groups, such as the Human Immunodeficiency Virus (HIV), Dengue virus (DENV), Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), or Herpes Simplex Virus (HSV) [4,6]. Moreover, it is plausible that the tendency of viruses to redirect metabolism could elucidate the preferential targeting of cells engaged in aerobic glycolysis, such as tumor cells. These cells, with an existing metabolism suitable for infection and viral replication, emerge as a prime target. This preference may confer oncolytic potential to the infecting virus, capable of reinstating immune responses within tumors.

Another putative reason for modulating the host's metabolism during infection is to impede the establishment of the antiviral cellular response, thereby promoting viral expansion. The hypothesis that certain transient or specific metabolic states due to chronic diseases favor viral infection takes on full significance in this context, partly explaining the differences in systemic reactivity to the same virus.

Figure 1. Schematic representation of the Warburg effect. According to this concept created by Otto Warburg, proliferating cells, including tumor cells, supply their biomass requirements by an adaptating of their metabolism. This adaptation involves the production of lactate from pyruvate even under aerobic conditions and has therefore been established as aerobic glycolysis. The intermediates of glycolysis are then utilized to provide lipids, nucleotides, and amino acids essential for high-rate cellular proliferation.

Figure 1. Schematic representation of the Warburg effect. According to this concept created by Otto Warburg, proliferating cells, including tumor cells, supply their biomass requirements by an adaptating of their metabolism. This adaptation involves the production of lactate from pyruvate even under aerobic conditions and has therefore been established as aerobic glycolysis. The intermediates of glycolysis are then utilized to provide lipids, nucleotides, and amino acids essential for high-rate cellular proliferation.

Indeed, the interactions between metabolism and immunity are increasingly discussed with the emergence of the term ‘immunometabolism’. This facet of immunology has been the subject of numerous advances in the context of cancer (see section 2.a. Immunometabolism and cancer) and has already been extensively presented elsewhere. The aim of this review is rather to discuss, in the light of the latest available literature, how metabolic disorders (diabetes, obesity, and others) might affect the outcome of viral infections. Then, we will discuss the advances made regarding the interplays between metabolism and immunity in the context of viral infections, with an emphasis on the antiviral innate immunity of the infected host cell.

2. Immunometabolism

The immune system is an intricate network of cells and molecules that safeguard the body against threats. Recently, increasing attention has been directed towards the regulation of this network by metabolism, giving rise to the concept of immunometabolism. The study of immunometabolism could provide insight into how obesity fosters inflammation in conditions such as arthritis [7]. It also offers a way of understanding how drugs acts on metabolism or metabolites, such as methotrexate or 25-hydroxycholesterol, to find utility in the treatment of infectious diseases [8,9].

a. Immunometabolism and cancer

A proper introduction to immunometabolism is its contribution to the understanding and treatment of pathologies of cancer origin. Indeed, cancer has been one of the most prolific subjects of study for immunometabolism. Cancer is a group of pathologies that can occur in virtually all organs and have in common the anarchic division of cells, possibly leading in the case of malignant tumors to the invasion of distant tissues. The metabolism of cancer cells is of crucial importance for the development of these pathologies, as cancer cells have a particular metabolism allowing them to support their development. In addition, the tumor metabolic environment also influences the immune system by participating in immune tolerance and immune system failures in the clearance of tumor cells [10].

Indeed, tumors have been shown to modulate the concentration of nutrients critical to immune cell function. Infiltration of immune cells into the tumor tissue is generally associated with a better prognosis [11]. In particular, T cell infiltration is well documented, and the promotion of their activity is even the subject of new therapeutic approaches [12]. The tumor cells microenvironment is specific and results from the metabolism of tumor cells. The latter is based primarily on high-rate of glycolysis to allow the large production of biomass (e.g., amino acids, nucleotides and lipids) required for their sustained division. This leads to a depletion of the available intratumoral glucose pool [13]. Of note, glucose depletion has been associated with decreased infiltration and reactivity of the immune system, including T cells on the tumor lcoation [14]. For example, a deficit in the production of IFN-γ, a key effector secreted by tumor-infiltrating CD8⁺ T cells, was observed after glucose deprivation [15]. This decrease is induced by the binding of glyceraldehyde phosphate dehydrogenase (GAPDH) to the 3’ untranslated region (3’UTR) of the IFN-γ mRNA when the glycolytic enzyme is not devoted to its metabolic function, namely, at low levels of glucose [15]. Alternatively, phosphoenolpyruvate (PEP), an intermediate glycolysis metabolite, whose level decreases following glucose deprivation, is essential for T cell anti-tumor responses H [16]. It supports T cell receptor (TCR) and nuclear factor of activated T cells (NFAT)-induced signaling, whose activation depends on inositol trisphosphate receptors (IP3R) located on the mitochondrial membrane [16]. Ca²⁺ then acts as an integrator of TCR signaling and PEP by inhibiting sarco/ER Ca²⁺−ATPase (SERCA) activity. This increased cytosolic Ca²⁺ concentration then promotes the anti-tumor activities of TCR-dependent CD4⁺ and CD8⁺ T cells [16].

b. A case of metabolic-dependent activation in immune cells: monocyte and macrophage

Macrophages are CD11b⁺CD68⁺ leukocytes historically described for their anti-microbial and phagocytic activities. However, many other functions have been associated with them over the years [17]. Indeed, their participation in tissue homeostasis, as well as tissue repair in case of damage, is no longer to be proven [17]. To fulfill these essential but diverse functions, macrophages can adopt a variety of phenotypes. Given the complexity of these phenotypes, it is generally accepted to classify macrophages according to whether they have pro-inflammatory (M1) or anti-inflammatory (M2) activities [17]. Nevertheless, it is important to keep in mind that this classification greatly simplifies the repolarization capabilities of these cells, which rely on transcriptionally dynamic capacities to adopt the phenotype and functions best suited to their tissue microenvironment [17]. Moreover, macrophages are among the only cells to have so many tissue-specific counterparts (i.e., Kupffer cells in the liver, microglia in the brain, and alveolar macrophages in the lung) [17]. All these tissue-specific macrophages have an embryonic origin, unlike the monocyte lineage-macrophages, which derive from the differentiation of monocytes following their tissue infiltration. In the following, we will discuss the metabolic characteristics of macrophages according to whether they are M1 or M2 polarized in vitro and in vivo. Some differences that may occur depending on their tissue of residence, have been reviewed very recently by Wculek et al. [18].

M1 macrophages (CD11b⁺CD68⁺CD64⁺) have a predominantly pro-inflammatory activity whose activation in vitro occurs either through TLR stimulation by LPS or by cytokine stimulation with TNF or IFNγ. This activation reprograms cell metabolism in a way that supports the establishment of an oxidative burst by the production of reactive oxygen species (ROS) and nitrogen species (RNS), as well as the production of antimicrobial effectors. M1 macrophages are characterized by predominant glycolytic metabolism (Figure 2) as evidenced by their increased consumption of glucose, enabled in particular by an increase in the Glucose Type 1 receptor (GLUT1) [19,20]. Hypoxia Inducible Factor-1α (HIF-1α) was shown to be a transcription factor essential for the onset of this metabolic reprogramming [21,22]. Stabilization of HIF-1α is mediated in M1 macrophages by blocking TCA at two distinct levels. First, the inhibition of isocitrate dehydrogenase (IDH-1) leads to citrate accumulation, as well as itaconate synthesis [23]. On the one hand, the accumulated citrate is converted to acetyl-CoA, which is used for histone acetylation and inflammatory gene transcription [24]. On the other hand, itaconate inhibits the activity of the succinate dehydrogenase complex (SDH), leading to the accumulation of succinate [25]. Ultimately, the succinate generated in this way stabilizes HIF-1α [22]. Notably, this metabolic reprogramming has been shown to be essential for IL-1β production in response to LPS stimulation [22]. Mitochondrial activity in M1 macrophages is then diverted from ATP production to mitochondrial ROS production. The latter will serve within phagosomes, participating in antigen processing and presentation [26]. At this stage, ATP production is therefore mainly dependent on aerobic glycolysis, similarly to the effect described by Warburg for cancer cells [5,22]. This also supplies the necessary intermediates for the pentose phosphate pathway (PPP) leading in the end to an increase in the NADPH pool, as well as to the production of biomass (here as nucleotides) and ROS [27].

Figure 2. General metabolism involved in pro-inflammatory M1 macrophages orientation. M1 macrophages develop a glycolytic metabolism enabled by the overexpression of the glucose transporter, GLUT1, in which HIF-1α is involved. Stabilization of this factor is, on the one hand, linked to the inhibition of isocitrate dehydrogenase (IDH-1), leading to citrate accumulation and itaconate synthesis. Accumulated citrate is converted into acetyl-CoA which is used for histone acetylation and thus promotes the transcription of inflammatory genes. On the other hand, itaconate inhibits the succinate dehydrogenase complex (SDH) activity, resulting in succinate accumulation. The succinate thus produced stabilizes HIF-1α. The mitochondrial activity of M1 macrophages is dedicated to the production of mitochondrial reactive oxygen species, to be used within phagosomes. ATP production therefore relies primarily on aerobic glycolysis.

Figure 2. General metabolism involved in pro-inflammatory M1 macrophages orientation. M1 macrophages develop a glycolytic metabolism enabled by the overexpression of the glucose transporter, GLUT1, in which HIF-1α is involved. Stabilization of this factor is, on the one hand, linked to the inhibition of isocitrate dehydrogenase (IDH-1), leading to citrate accumulation and itaconate synthesis. Accumulated citrate is converted into acetyl-CoA which is used for histone acetylation and thus promotes the transcription of inflammatory genes. On the other hand, itaconate inhibits the succinate dehydrogenase complex (SDH) activity, resulting in succinate accumulation. The succinate thus produced stabilizes HIF-1α. The mitochondrial activity of M1 macrophages is dedicated to the production of mitochondrial reactive oxygen species, to be used within phagosomes. ATP production therefore relies primarily on aerobic glycolysis.

Unlike M1 macrophages, macrophages with a M2 phenotype (CD11b⁺CD68⁺CD163⁺) are obtained in vitro in response to IL-4 or IL-13 and display anti-inflammatory properties as well as involvement in tissue repair [17]. In contrast to M1, M2 macrophages have a cellular metabolism based on increased OXPHOS and mitochondrial respiration (Figure 3), as evidenced by the elevation of OCR in macrophages polarized towards an anti-inflammatory phenotype [28]. Transcriptional shifts leading to this metabolism are controlled in M2 macrophages by activation of several factors including PPARδ, PPARγ, PGC-1β, and IRF4. These factors activation depends on both STAT6 and the mTORC2 complex [28,29]. Some of these factors are associated with lipid metabolism and lead, in particular here, to the upregulation of the scavenger membrane protein CD36 that allows internalization of lipids used for fatty acid oxidation (FAO) [29,30]. While CD36 has been shown to be critical for polarization in M2 [30], the role of FAO remains ambiguous. Indeed, the use of etoxomir to inhibit FAO or genetic models that prevent it, do not appear to affect M2 polarization [31,32], but rather the ability of M2 macrophages to produce ATP. Thus, one might conclude that FAO does not appear to be essential for M2 macrophages to perform their inflammatory functions, with fatty acids serving instead as a substrate for energy production through OXPHOS [31]. In contrast, glutaminergic metabolism appears to be of paramount importance with glutamine providing one-third of the carbon derivatives of TCA [33]. Glutamine leads to the accumulation of α-ketoglutarate, which orchestrates M2 polarization by activating Jumonji domain-containing protein D3 (JMJD3) demethylases [34]. They mediate histone demethylation of M2-specific gene promoters [34]. As a result, these epigenetic modifications are partly responsible for M2 polarization.

Figure 3. General metabolism involved in anti-inflammatory M2 macrophages orientation. In M2 macrophages, activation of the PPARδ, PPARγ, PGC-1β and IRF4 factors enables transcriptional changes directing the cellular metabolism towards OXPHOS. The activation of these factors depends on both STAT6 and mTORC2. In particular, overexpression of the scavenger CD36 provides lipidic substrates required for fatty acid oxidation. Fatty acid oxidation is essential for ATP production via OXPHOS, rather than for M2 macrophage functions. On the other hand, glutamine, which supplies the Krebs cycle, is essential since it orchestrates M2 polarization through the activation of JMJD3 demethylases.

Figure 3. General metabolism involved in anti-inflammatory M2 macrophages orientation. In M2 macrophages, activation of the PPARδ, PPARγ, PGC-1β and IRF4 factors enables transcriptional changes directing the cellular metabolism towards OXPHOS. The activation of these factors depends on both STAT6 and mTORC2. In particular, overexpression of the scavenger CD36 provides lipidic substrates required for fatty acid oxidation. Fatty acid oxidation is essential for ATP production via OXPHOS, rather than for M2 macrophage functions. On the other hand, glutamine, which supplies the Krebs cycle, is essential since it orchestrates M2 polarization through the activation of JMJD3 demethylases.

As with other cells of the immune system, deficits in macrophage metabolism can occur pathologically, as in the case of atheromatous intraplaque foamy macrophages or in obesity [18].

c. The metabolic-dependent activation of tissue sentinels: dendritic cells

Dendritic cells (DCs, CD11c⁺) are antigen-presenting cells (APCs), which like macrophages, have variable origins [35]. In fact, several subtypes of DCs can be distinguished according to their origin. Conventional DCs (cDC1 and cDC2) and plasmacytoid DCs (pDCs) derived from specific progenitors, and others derived from monocytes (mo-DCs) [35]. As APCs, DCs are found in a resting state in many tissues (e.g., Langerhans cells in the dermis) where their role is to sense their environment, and then activate in case of danger. Activation occurs in response to pathogen associated molecular patterns (PAMPs) or damage associated molecular patterns (DAMPs) and results in initiating the immune response by the secretion of cytokines, as well as by the presentation of antigen in lymphoid tissues [35].

The metabolism of resting DCs has already been studied both in vitro and in vivo. Thus, it is known that the mTORC1 complex has a preponderant activity here again. Indeed, in mice, the use of rapamycin leads to the inhibition of the ex vivo growth of pDCs and cDCs from the bone marrow (BM); in vivo the pool of DCs is considerably reduced [36]. Confirming these results obtained in mouse models, the use of rapamycin on human mo-DCs results in impaired survival [37]. Conditional expression models were used to further explore the importance of mTORC1 in DCs survival. Phosphatase and TENsin homolog (Pten) deletion using a Cd11c-Cre system leads to expansion of DCs in mice. This result is consistent with the previous one, PTEN being a negative regulator of mTOR [36]. Deletion of Tsc1, another negative regulator of mTOR, using a similar system, led to an increased growth of cultured BM-DCs [38]. As one of the main factors that regulate cellular metabolism, mTORC1 directs quiescent DCs to use OXPHOS to satisfy their low energy requirement [39].

As for the other immune cells presented so far, the activation of DCs following stimulation of one of their pattern recognition receptors (PRRs) leads to a shift in metabolism with glucose becoming the main fuel for cellular metabolism. This process takes place in two steps. Indeed, shortly after the activation of DCs in a TLR-dependent manner, a rapid increase in glycolytic activity and entry of intermediates into the PPP was observed [40]. These changes are initiated by activation of AKT via TANK-binding kinase 1 (TBK1) and IκB kinase ε (IKKε) [40]. AKT phosphorylates and subsequently activates hexokinase 2 (HK2), the glycolysis-limiting enzyme whose activity increases after association with the mitochondrial surface [40]. Lately, maintenance of glycolysis within DCs relies on activation by the mTOR complex of HIF-1α [41]. Interestingly, if inhibition of HIF-1α or mTOR occurs early after TLRs stimulation, it has no impact on DCs activation, in contrast to early inhibition of AKT. This illustrates the importance of the timing of different signaling pathways in DCs activation [40]. In contrast, activation of the AMP- activated protein kinase (AMPK) pathway appears to control DCs activation, by promoting OXPHOS, in particular, through PGC1α activation, which favors mitochondrial biogenesis. In fact, AMPK knockdown has been associated with a potentiation of DCs activation [42]. Whereas its activation or that of PGC1α, leads to tolerogenic DCs [42].

Figure 4. Dendritic cell metabolism. Maintenance of dendritic cells quiescence depends on an overriding activity of mTORC1. PTEN and TSC1, negative regulators of mTOR limit dendritic cell growth. mTORC1 orientates quiescent cells towards OXPHOS utilization, with protein catabolism and triacylglycerol providing amino acids and fatty acids to supply ATP production via the Krebs cycle. Activated dendritic cells take up glucose as the main fuel for a metabolism geared towards aerobic glycolysis. Activation of mTORC1and HIF-1α, in this case, is enabled by TBK1 and IKKε activity on AKT downstream of dendritic cells PRR.

Figure 4. Dendritic cell metabolism. Maintenance of dendritic cells quiescence depends on an overriding activity of mTORC1. PTEN and TSC1, negative regulators of mTOR limit dendritic cell growth. mTORC1 orientates quiescent cells towards OXPHOS utilization, with protein catabolism and triacylglycerol providing amino acids and fatty acids to supply ATP production via the Krebs cycle. Activated dendritic cells take up glucose as the main fuel for a metabolism geared towards aerobic glycolysis. Activation of mTORC1and HIF-1α, in this case, is enabled by TBK1 and IKKε activity on AKT downstream of dendritic cells PRR.

Therefore, we have shown here that essential changes in immune cell metabolism are required during their activation to enable them to support their function, whether as a support or effector. We will now discuss the contributions of immunometabolism in certain pathological situations.

3. At the systemic level: Interplays between host metabolic status and infectious diseases

It is gaining increasing awareness that the resolution of an infection depends on multiple intrinsic physiological factors. In this light, the host metabolic status and the related comorbidities are now taken into account in understanding immune responses, whoch are often more or less efficient during the course of an infectious disease, conditioning resolution or subsequent pathophysiological processes.

a. Diabetes, obesity and other metabolic disorders

Metabolic diseases (e.g., type 2 diabetes, obesity) represent a growing risk for the world’s population, particularly through their cardiovascular (myocardial infarction, stroke) and neurological (diabetic neuropathy) complications. Previously restricted to high-income countries and western lifestyles, these diseases are now tending to develop in low- and middle-income countries. As depicted in this section, these metabolic disorders may present an associated risk of infection. It was also recently demonstrated during the COVID-19 epidemic that one of the complications of SARS-CoV-2 infection is the development of post-infection diabetes.

i. Diabetes

Diabetes is a chronic disease characterized by hyperglycemia and chronic low-grade inflammation. Two main types of diabetes can be distinguished. Type 1 or insulin-dependent diabetes (T1D) is an autoimmune disease targeting the β-cells of the islets of Langerhans in the pancreas which secrete insulin, a key hormone in the regulation of circulating glucose. In contrast, the pathophysiology of type 2 or insulin-resistant diabetes (T2D) involves hyperinsulinemia, due to a progressive resistance of cells to insulin, following the installation of a low-grade inflammation [1]. Diabetes accounts for a total of more than 460 million cases in 2019 with a projection to 700 million cases in 2045, the majority of cases being T2D [43]. Beyond its cardiovascular and neurological complications, diabetes could lead to impaired host defenses and an increased risk of infection.

In fact, diabetes has been found to cause immune deficiencies in several essential compartments to establish an adequate immune response. Chronic hyperglycemia is accompanied by the production of advanced glycation end products and ROS. This pro-oxidant status promotes chronic low-grade inflammation, which is detrimental to the immune response. More specifically, T2D is accompanied by the recruitment of M1 polarized macrophages within the adipose tissue where they will secrete significant amounts of pro-inflammatory mediators (Tumor necrosis factor α - TNF-α, C-reactive protein - CRP, interleukin-1β - IL-1β, CCL2, CCL3, CXCL8, IL-6, and IL-12) [44,45]. However, under bacterial stimulation, peripheral blood mononuclear cells (PBMCs) from diabetic patients have been found to have a decrease in functional capabilities compared to healthy individuals. Thus, a decrease in the secretion of cytokines such as IL-1β, IL-6 and IL-2 is observed [46,47,48]. Of note, IL-6 plays an important role in protection against pathogens and in the adaptive immune response by inducing antibody production and the development of effector T cells [49]. Therefore, these studies reveal that inhibition of these cytokines in hyperglycemia can affect the immune response development against pathogens. In addition, a suppression of type I interferon (IFN) production is reported in PBMC cultured under high glucose conditions and stimulated with viral RNA mimetic, namely polyinosinic:polycytidylic acid - poly:IC [50]. It was also demonstrated that leukocyte recruitment is inhibited in a diabetic context, notably due to an alteration in the expression of adhesion molecules [51]. More specifically, for macrophages, there is a decrease in Fcγ receptor expression, as well as in the antimicrobial and phagocytosis capacities [52,53]. Similarly, for neutrophils, a reduction in antimicrobial capacities was noted. This is due to an alteration of the pro-oxidant and degranulation capacities [54,55,56]. Furthermore, the production of neutrophil extracellular traps (NETs) or NETose is decreased under hyperglycemic conditions, leading to increased susceptibility to infections [57]. Deficiencies in degranulation have also been demonstrated in NK cells from T2D subjects, resulting from a defect in NKG2D and NKp46 activating receptors [58]. Of note, complement activation is impacted in hyperglycemia, due to dysfunction of opsonization by C4 fragments of complement [59].

In addition to effector cells, T2D is accompanied by altered presentation and thus pathogen detection. Therefore, diabetic mice showed reduced expression of the adaptor protein containing the Toll-like receptor 2 (TLR) and the Toll / IL-1R domain (TIRAP) [60], involved in the detection of pathogens. However, contradictory data exist since it has been shown that the expression of TLRs by monocytes and neutrophils from diabetic subjects was increased [61]. In fact, it is now known that the quality of blood glucose control has an impact on TLRs expression. A subject with hyperglycemia and poor control will have decreased TLRs expression [61]. While a subject with controlled hyperglycemia will have higher TLRs expression [61]. Dendritic cells, the main antigen-presenting cells (APCs), showed delayed recruitment to the site of infection in diabetic mice in the course of M. tuberculosis infection [62]. This delay could be attributed, as previously for monocytes/macrophages, to reduced levels of CCL2 and CCL5 [62]. In addition, in the context of intracellular bacterial infection, this delay in recruitment has been shown to be accompanied by impaired phagocytosis [63].

Diabetes has been repeatedly shown to be a risk factor for developing bacterial and fungal infections, and this has already been reviewed [64,65]. However, viral infections have less been reported. Nevertheless, a study including nearly 190,000 people in China showed that diabetic individuals have a 1.5 times higher relative risk of developing hepatitis B virus infection than non-diabetic individuals [66]. Associated with the immunological disorders discussed above, Middle East Respiratory Syndrome coronavirus (MERS-CoV) infection in diabetic mice is accompanied by unresolved lung inflammation leading to exacerbated pathology [67]. These mice were also shown to have fewer monocytes/macrophages and CD4⁺ T cells, without affecting viral clearance [67]. For SARS-CoV-2, chronic inflammation found in T2D appears to be a risk factor for developing severe forms of COVID-19 [68]. Furthermore, strengthening the link between TD2 and viral infection, it is now known that SARS-CoV-2 can infect adipocytes, leading to a decrease in adiponectin secretion. This contributes to the establishment of an inflammatory state conducive to the development of insulin resistance [69], explaining the increased risk of transitioning from a prediabetic state to diabetes in COVID-19 patients. Finally, complications of diabetes include cardiovascular damage (atherothrombosis, stroke, myocardial infarction). Many of these cardiovascular complications are related to endothelial damage. Moreover, severe forms of DENV are known to be associated with damage to the endothelial barrier. Therefore, it is hypothesized in the literature that severe forms of DENV infection are more likely to occur in diabetic individuals due to pre-existing endothelial damage [70,71].

ii. Obesity

Overweight and obesity are defined as an excessive accumulation of fat in an individual. Abnormal accumulation of visceral adipose tissue (VAT) is a risk factor for developing cardiovascular events and is often associated with the development of insulin resistance [72]. Previously limited to high-income countries, overweight and obesity are now also increasing rapidly in low- and middle-income countries. According to the WHO, today more people are suffering from obesity than from malnutrition [73]. In 2016, there were 1.9 billion overweight people, of whom over 650 million were obese [73].

Obesity is a multifactorial metabolic disorder combining genetic, environmental and developmental factors [74], whose simplest cause corresponds to a caloric intake that is not consistent with the individual's energy expenditure. However, more complex mechanisms can be underlined, such as an impairment on the neurons of the arcuate nucleus of the hypothalamus involved in the control of food intake. This is mediated by the expression of orexigenic factors (neuropeptide Y, agouti-related protein) or anorexigenic factors (proopiomelanocortin, Cocaine and amphetamine regulated transcript - CART) under the hormonal control of leptin and insulin secreted by adipocytes and β-cells of the islets of Langerhans, respectively [74]. More recently, the impact of dysbiosis of the intestinal microbiota on the risk of obesity has been discussed. Indeed, the abundance of firmicutes bacteria seems to be linked to the development of obesity, due to the ability of these bacteria to extract energy more efficiently, thus promoting calorie absorption and weight gain [75].

Beyond the cardiovascular and endocrine complications associated with obesity, obese individuals exhibit immune disorders. In addition to its important storage role, VAT is a tissue that is strongly linked to immunity, in particular through the secretion of adipokines (leptin, adiponectin), cytokines (TNF-α, lL-6) and chemokines (CCL2). They exhibit activity not only in the inflammatory state, but also in the immune system [76]. Thus, hyperplasia and hypertrophy in obesity are accompanied by a deregulation of the immune system, which we will describe here. Adipose tissue contains different cell types including, in addition to adipocytes, endothelial cells, fibroblasts, pre-adipocytes, but especially leukocytes [77]. First, in obese subjects, there is an alteration in the recruitment, proliferation and polarization of macrophages within the adipose tissue [78]. Increased adipocyte size has been shown to trigger a stress response and the release of chemokines (CCL2 and CCL5) [79], leading to the recruitment and accumulation of monocytes and macrophages in adipose tissue [78]. On the other hand, the increased presence of free and non-esterified fatty acids in obesity results in the activation of TLR4 and TLR2, leading to the production of pro-inflammatory cytokines (e.g., IL-6) and reduced expression of anti-inflammatory cytokines (e.g., IL-10) [80,81]. This pro-inflammatory state may ultimately be responsible for the development of insulin resistance and T2D, as previously explained [81]. Finally, while most M2 polarized macrophages are found in healthy subjects' adipose tissue, macrophage polarization in obese individuals is oriented toward the M1 phenotype [77]. This accumulation is, on the one hand, due to the recruitment of blood monocytes, but also to the proliferation and polarization shift of resident macrophages in adipose tissue [82]. Leukocytosis in adipose tissue in obesity also involves neutrophils. Similarly, to monocytes, neutrophils from obese subjects show increased migration within adipose tissue, as well as increased basal superoxide production, thus participating in low-grade inflammation [83]. Finally, NK cells follow this trend, with increased activation, proliferation, and IFN-γ secretion [84]. Cells involved in adaptive immunity are also not spared in obesity. Indeed, adipose tissue from obese subjects showed a higher presence of CD4⁺ helper 1 and CD8⁺ cytotoxic T cells [85]. Although this pro-inflammatory status is the one found in case of infection, its chronicity is accompanied by an enhanced susceptibility to infections and a lower reactivity to pathogens in obese subjects [83,86].

The immune alterations in obesity are responsible for a heightened risk of developing infections. These infections often occur opportunistically or nosocomially in tissues harboring a commensal microbial flora, at the interface between the internal and external medium. Consequently, obese mice have shown insufficient antimicrobial activity to combat Staphylococcus aureus during a cutaneous infection [87]. Fungal infections with Candida albicans are also more prevalent in obese individuals [88]. Furthermore, there is an increased risk of developing Escherichia coli infections in the urinary tract of these subjects [89]. However, it is in the respiratory system that the most data have been reported. Indeed, obese subjects are at a higher risk of contracting upper or lower respiratory system infections [90]. Mice^ob/ob have shown alterations in their immune response when exposed to pathogens of bacterial origin, including Mycobacterium tuberculosis, Listeria monocytogenes, and Klebsiella pneumoniae [91,92,93]. Due to poor control of these infections by the immune system, they are more likely to breach the endothelial barrier, leading to septicemia [91]. The inability to control infection under these conditions has been partly attributed to defects in antigen-specific T cell immune responses [93], as well as deficiencies in the phagocytic activity of neutrophils and macrophages [92]. The severity of infections in obese subjects is also heightened in the case of viral infection. Consequently, successive epidemics of H1N1 influenza and SARS-CoV-2 have been correlated with an increased risk of mortality and morbidity in obese patients [94,95]. In the case of Influenza virus infection, it has been demonstrated in a murine model that the increased susceptibility of obese individuals is linked to a defect in antigen presentation to T lymphocytes by dendritic cells, leading to deficient cytotoxic and memory responses [96,97]. It is noteworthy that obesity has recently been associated with greater severity of West Nile virus (WNV) infection in a murine model [98]. This increase in severity was found in females and is linked to dysfunctions of the adaptive immunity in the acute phase, as well as in the late phase with a decrease in the production of neutralizing antibodies in these individuals [98]. Correspondingly, with this inability to produce a humoral response, it is known that obesity has a negative impact on immune response in the case of vaccination. Thus, the immunogenicity induced by vaccines targeting the Hepatitis B and H1N1 influenza viruses is reduced in obese individuals [99,100].

b. Exercise

Much like nutrition, physical activity has an impact on an individual's metabolic status; however, its influence on immune function remains a subject of debate (Figure 5). Epidemiological evidence suggests that regular physical activity reduces the incidence of chronic diseases and the risk of viral and bacterial infections [101]. In fact, frequent physical exercise may enhance immune competence. A prospective cohort study of 1,509 Swedish men and women aged between 20 to 60 years demonstrated that higher levels of physical activity were associated with a lower incidence of self-reported upper respiratory tract infections [102]. Acute exercise sessions of less than 60 minutes have been shown to result in a transient alteration in leukocyte counts and lymphocyte proliferation capabilities [103]. Although these changes remained modest, they could be considered as an enhancement of immunosurveillance during moderate physical activity. This type of exercise serves as a crucial adjunct to the immune system, facilitating the continuous turnover of leukocytes between circulation and tissues [104]. Moreover, physical exercise has been suggested as a valuable tool for COVID-19 prevention by improving immune function, particularly innate and adaptive immunity, and modulating the anti-inflammatory state [105].

However, prolonged and intensive exercise in athletes is associated with immune dysfunction, inflammation, and oxidative stress [106,107]. Exercise may transiently impair immune protection, thereby increasing the risk of infection [108]. Consequently, athletes may experience a temporary or sustained reactivation of viruses, such as the Epstein-Barr virus, due to the suppression of immune parameters during years of intense training [109,110,111].

Figure 5. Benefice and risk of physical activity in the context of infection. The role of physical activity on immunity in case of viral infection remains a subject of debate in the literature. The data available to date seem to point towards the description of a hormesis phenomenon. Physical activity has a beneficial effect, illustrated by a reduction in the incidence of infection and an increase in immune-surveillance, in the case of regular and moderated activity. On the other hand, when the “dose” of exercise becomes intense and prolonged, deleterious effects and a context conducive to infection are found, namely an increase of inflammation, oxidative stress and alteration of the immune system

Figure 5. Benefice and risk of physical activity in the context of infection. The role of physical activity on immunity in case of viral infection remains a subject of debate in the literature. The data available to date seem to point towards the description of a hormesis phenomenon. Physical activity has a beneficial effect, illustrated by a reduction in the incidence of infection and an increase in immune-surveillance, in the case of regular and moderated activity. On the other hand, when the “dose” of exercise becomes intense and prolonged, deleterious effects and a context conducive to infection are found, namely an increase of inflammation, oxidative stress and alteration of the immune system

Collectively, these studies demonstrate that physical exercise is beneficial for maintaining functional immunity. However, when practiced at a very high intensity, it is suggested to alter immune function and increase susceptibility to infections (Figure 5). It is also crucial to bear in mind that the connection between exercise and immune function is undoubtedly a multifactorial concept involving other aspects such as nutrition and changes in the microbiome, for example [112]. To this end, further studies using a multifactorial approach will be necessary to enhance our understanding of the immune changes induced by exercise and their underlying mechanisms.

4. Concluding remarks

In conclusion, although viral control of cellular metabolism for its own benefit appears to be a common consequence of infection that affects the antiviral response, our work tends to show that immune evasion through metabolic control appears to have been thwarted by the host. Regardless of the metabolic context, whether glycolytic or OXPHOS-dependent, the host cell is able to shape its response and produce antiviral effectors; on the one hand, viperin, and on the other, ISGs such as ISG15, 54 or 56. However, as witnessed in the course of several successive epidemics, viruses consistently employ strategies to disrupt the antiviral response during replication, in a frenetic evolutionary race. This crosstalk between virus and immunometabolism, once again, illustrates the red queen effect in infectious processes [172].