Mitochondrial metabolism in macrophages.

Mitochondria are considered to be the powerhouse of the cell. Normal functioning of the mitochondria is not only essential for cellular energy production but also for several immunomodulatory processes. Macrophages operate in metabolic niches and rely on rapid adaptation to specific metabolic conditions such as hypoxia, nutrient limitations, or reactive oxygen species to neutralize pathogens. In this regard, the fast reprogramming of mitochondrial metabolism is indispensable to provide the cells with the necessary energy and intermediates to efficiently mount the inflammatory response. Moreover, mitochondria act as a physical scaffold for several proteins involved in immune signaling cascades and their dysfunction is immediately associated with a dampened immune response. In this review, we put special focus on mitochondrial function in macrophages and highlight how mitochondrial metabolism is involved in macrophage activation.

INTRODUCTION

Mitochondria are double membrane organelles that are assumed to be of bacterial origin according to the endosymbiotic theory. This theory suggests that mitochondria evolved from an aerobic prokaryote which was engulfed by a host nucleated cell and from then on started to take advantage, depending on the prokaryotic cell for energy production; this led to the development of modern mitochondria over time (1). Although mitochondria have dramatically developed since then, their prokaryotic origin is still evident and includes the presence of their own circular genome, mitochondrial DNA (mtDNA), which encodes some parts of the respiratory chain as well as mitochondrial tRNAs and rRNAs (2). Moreover, mitochondria synthesize their own membrane and reproduce by partitioning and pinching in the middle, a process used by bacteria and is known as fission (3, 4).

After establishment of the mitochondrion within the early eukaryotic cell, this organelle has become responsible for several cellular key functions such as energy generation, reactive oxygen species (ROS) production, calcium homeostasis, iron sulfur cluster biogenesis, and the regulation of apoptosis (5, 6). Mitochondria supply the major portion of cellular adenosine triphosphate (ATP) through oxidative phosphorylation (OXPHOS) (6, 7) and are a central hub for several metabolic activities such as the tricarboxylic acid cycle (TCA), β-oxidation of fatty acids, parts of the urea cycle, and amino acid synthesis (Fig. 1; 8, 9). Many intermediates of these pathways, including ROS are in`volved in the control of cellular gene expression (10).

Figure 1.

Metabolic pathways associated with mitochondria. 3-PG, 3-phosphoglyceric acid; ADP, adenosine diphosphate; ALT, alanine aminotransferase; Arg, arginase; ASL, argininosuccinate lyase; ASS1, argininosuccinate synthetase 1; ATP, adenosine triphosphate; CAD, cis-aconitate decarboxylase; CoA-SH, coenzyme A; CPT-1, carnitine palmitoyl transferase-1; DHAP, dihydroxyacetone phosphate; F1,6BP, fructose 1,6-bisphosphate; G3P, glycerol 3-phosphate; G6P, glucose 6-phosphate; γ-GLU-Cys, γ-glutamyl-cysteine; GAP, glyceraldehyde 3-phosphate; Gly, glycine; GOT1, cytoplasmic aspartate aminotransferase; GOT2, mitochondrial aspartate aminotransferase; GPD1, cytosolic glycerol 3-phosphate dehydrogenase; GPD2, mitochondrial glycerol 3-phosphate dehydrogenase; GPx, glutathione peroxidase; GR, glutathione reductase; GSH, glutathione; GSSH, glutathione disulfide; IDH, isocitrate dehydrogenase; iNOS, inducible nitric oxide synthase; LDH, lactate dehydrogenase; MDH1, cytosolic malate dehydrogenase; MDH2, mitochondrial malate dehydrogenase; MPC, mitochondrial pyruvate carrier; NADP+, nicotinamide adenine dinucleotide phosphate; NADPH, reduced nicotinamide adenine dinucleotide phosphate; PDC, pyruvate dehydrogenase complex; Ser, serine; ROS, reactive oxygen species; SHMT1, serine hydroxymethyltransferase-1; SDH, succinate dehydrogenase.

Because of their central cellular role, it is not surprising that mitochondrial function is crucial for the normal activity of mammalian cells, and dysfunction of these organelles has been linked to several neurological and metabolic disorders, aging, and cancer (8). Mitochondrial dysfunction might arise from ROS-induced mutations of the mtDNA due to its close proximity to the mitochondrial electron flow (8, 11). Moreover, mtDNA has a nucleotide imbalance or asymmetry that decreases the fidelity of the DNA polymerase, POLG, which is responsible for replicating the mitochondrial genome (12). Depending on the tissue type, a single mammalian cell contains up to 1,000 mitochondria, each of which maintains several copies (2–10) of its circular mtDNA (13, 14). As a consequence, every cell contains up to several thousand mtDNA copies, rather than just one copy of nuclear DNA in postmitotic cells. This makes the mtDNA more susceptible to heterogeneity as this could lead to heterogenous copies of mtDNA within the same cell or even within the same mitochondrion (15). Although mitochondria still maintain their own DNA, they rely heavily on nuclear genes for the production of mitochondrial proteins (16). This heterogeneity and increased mutation rate may be one of the reasons why mitochondria have lost most of their genomic information in favor of the nuclear genome (16).

The fundamental contribution of nuclear DNA and cellular protein biosynthesis to basic mitochondrial function renders a precise cross talk between nuclear and mtDNA essential for regular functioning of the cell (17). To convey the bidirectional cross talk between nuclear and mtDNA, the cells rely on epigenetic and post-transcriptional mechanisms to modulate gene expression (17). These mechanisms include DNA methylation, the regulation of gene expression by microRNA (miRNA), and post-translational histone modifications (17). There are many miRNAs which have been reported to regulate mitochondrial transcripts (17, 18). Of particular interest in the context of this review are miRNAs involved in the regulation of mitochondrial metabolism in macrophages. For example, miR-15a/16 has been shown to play a role in inflammatory macrophages, as Moon et al. (19) demonstrated that the deletion of miR-15a/16 increased phagocytosis and the production of mitochondrial reactive oxygen species (mROS). Moreover, a deficiency of miR-15a/16 resulted in an increase of proinflammatory cytokines (19). Another example is miR-33; this miRNA targets PGC-1α in macrophages, resulting in the inhibition of mitochondrial respiration and ATP production (20, 21). Taken together, miRNAs represent a novel therapeutic target that can be exploited for the treatment of immune disorders (21).

Among the diseases associated with mitochondrial dysfunction are neurological disorders such as neurodegenerative diseases, ischemia, and hypoxia-induced brain injury (22). Dysfunction of mitochondrial biogenesis, fusion and fission have been considered responsible for these pathogeneses (22). Mitochondrial dysfunction has also been implicated in the pathogenesis of immune cells in immunological diseases such as systemic lupus erythematosus (SLE), rheumatoid arthritis, and type I diabetes (23). Furthermore, several classes of pharmaceutical drugs have been reported to exert mitochondrial toxicity, such as pain medications [nonsteroidal anti-inflammatory drugs (NSAIDs)], cholesterol lowering drugs (statins), anticancer drugs (kinase inhibitors and anthracyclines), antidiabetic drugs (thiazolidinediones, fibrates, and biguanides), and antibiotics such as fluroquinolones and macrolide (24).

Mitochondria are known to be involved in the immune response against pathogens as they supply the necessary energy and required metabolic intermediates for proper immune cell function and are even involved in sensing danger or stress signals (25, 26). This has been shown to be essential for host defense and tissue homeostasis (26). Hence, mitochondrial dysfunction has been reported to be associated with defective macrophage phagocytosis in chronic obstructive pulmonary disease (COPD) (27). Furthermore, it was demonstrated to prevent the repolarization of inflammatory macrophages (28).

Macrophages are the effector cells of the innate immune system (29). They are responsible for phagocytosing pathogens and releasing proinflammatory and antimicrobial mediators. Moreover, macrophages are also responsible for maintaining tissue homeostasis by eliminating dead cells, debris, and foreign materials and exerting regulatory and repair functions (29). In this review, we discuss the details of mitochondrial metabolism and how it is involved in the inflammatory response of macrophages. In addition, we will briefly touch upon the role of mitochondria in immunological signal transduction processes.

ROLE OF MITOCHONDRIAL METABOLISM IN IMMUNE SIGNALING

Upon activation, proinflammatory macrophages undergo considerable metabolic transformations to meet the energetic and biochemical requirements of pathogen defense. Due to their central metabolic role, mitochondria are obviously an integral part in the governance of this procedure. In the following, we cover metabolic changes in proinflammatory macrophages with specific focus on mitochondria (Table 1).

Table 1.

Summary of the most important metabolic pathways in macrophages

Metabolic Pathway	Specific Features	References
Glycolysis	Glycolysis is a cytosolic process but it relies on mitochondrial function for NADH/NAD+ cofactor balancing Induced glycolysis is one of the hallmarks of classically activated macrophages It provides the cells with energy at a much higher rate than OXPHOS	(30, 31)
Glycerol 3-phosphate shuttle	GPS is one of the two transport mechanisms for glucose derived electrons to the ETC GPS shuttle was found to be important in proinflammatory macrophages and regulates the inflammatory response	(32)
Malate aspartate shuttle	MAS is the second most efficient mechanism for transporting electrons produced during glycolysis across the inner mitochondrial membrane to the ETC MAS is required in activated macrophages to maintain redox balance between mitochondria and cytoplasm and supports a high glycolytic flux	(33)
One carbon metabolism	1 C metabolism refers to connected metabolic pathways which include the folate and methionine cycles LPS was found to induce 1 C metabolism in proinflammatory macrophages and it was found to work synergistically with the induced pentose phosphate pathway and serine synthesis pathway to achieve epigenetic regulation and to induce IL1β expression	(34, 35)
TCA cycle	TCA cycle is central to mitochondrial metabolism and fuels the respiration chain with reduced co-factors in form of NADH and FADH2 TCA cycle activity is a provider of critical intermediates for cell and macrophage function such as aspartate and itaconate	(36)
Itaconate	Itaconate is produced in proinflammatory macrophages from cisaconitate Irg1 is the gene encoding ACOD1, the enzyme producing itaconate. This gene is one of the most upregulated genes during proinflammatory conditions in murine macrophages	(37)
The urea cycle and the aspartate-argininosuccinate shunt	The aspartate-argininosuccinate shunt interconnects the urea- and the TCA- cycle It is an alternative source for fumarate synthesis and can replenish reduced fumarate synthesis in pro-inflammatory macrophages It is crucial for arginine synthesis which is critical for antimicrobial NO production through NOS2	(5)

ACOD1, cis-aconitate decarboxylase; GPS, glycerol-3-phosphate shuttle; LPS, lipopolysaccharides; MAS, malate-aspartate shuttle; NO, nitric oxide; NOS2, nitric oxide synthase II; OXPHOS, oxidative phosphorylation; TCA, tricarboxylic acid cycle; 1 C, one carbon.

Glycolysis

Although glycolysis is a cytosolic process, it relies on mitochondrial function for NADH/NAD+ cofactor balancing through the malate-aspartate shuttle (MAS) or the glycerol-3-phosphate shuttle (GPS). Moreover, a significant fraction of glycolytic pyruvate is transported into mitochondria for oxidation and ATP production (Fig. 1; 31). Induced glycolysis is one of the hallmarks of classically activated macrophages and several factors are involved in its regulation, including hexokinase (HK), phosphofructokinase (PFK), hypoxia-inducible factor (HIF)-1α, glucose transporter 1 (GLUT1), carbohydrate kinase-like protein (CARKL), and pyruvate kinase (PKM; 31, 38). All of these factors operate in a well-concerted manner to achieve the rapid supply of energy and metabolic precursors for proper macrophage function. Although glycolysis is not economic in terms of energy production per glucose molecule, it provides cells with energy at a much higher rate than OXPHOS (39). Indeed, glycolysis is considered to be 100 times faster in energy production than OXPHOS and a high glycolytic flux ensures a sufficient supply with intermediates of branching pathways required for nucleotide, amino acid, and lipid biosynthesis (30). All of these are essential for the metabolic adaptation of the cell (39).

Glycerol 3-phoshphate Shuttle

The glycerol 3-phosphate shuttle (GPS) is one of the transport mechanisms for glucose-derived electrons to the electron transport chain (ETC) as glycolytic NADH itself cannot pass through the mitochondrial membrane. The GPS starts in the cytosol with a reduction of dihydroxyacetone phosphate (DHAP) to glycerol 3-phosphate (G3P) by cytosolic glycerol 3-phosphate dehydrogenase (GPD1) (Fig. 1). G3P is further transferred to the inner mitochondrial membrane, where it is oxidized back to DHAP by glycerol 3-phosphate dehydrogenase (GPD2) while the electrons are passed to the ETC by the reduction of FAD to FADH2, eventually resulting in the reduction of ubiquinone (Q) to ubiquinol (QH2; 32). The G3P shuttle was found to be important in proinflammatory macrophages and regulates their inflammatory response (32). GPD2 has been demonstrated to increase glucose oxidation, to induce expression of inflammatory mediators, and to fuel the production of acetyl coenzyme A (acetyl-CoA) required for the subsequent acetylation of histones in proinflammatory macrophages (32).

Malate-Aspartate Shuttle

The malate-aspartate shuttle (MAS) is a second mechanism for transferring electrons produced during glycolysis across the inner mitochondrial membrane to the ETC. The pathway starts in the cytosol with the reduction of oxaloacetate to malate by cytosolic malate dehydrogenase (MDH1). Malate is then transported into the mitochondria via the malate-α-ketoglutarate antiporter. Once inside, malate is oxidized back to oxaloacetate by mitochondrial malate dehydrogenase (MDH2). During this reaction, electrons are transferred to mitochondrial NADH to fuel the ETC via complex 1; it is thus more efficient than the GPS. To close the cycle, oxaloacetate and glutamate are transaminated to aspartate and α-ketoglutarate by the mitochondrial aspartate aminotransferase (GOT2). Aspartate is then transported to the cytosol by the glutamate-aspartate antiporter. Finally, it gets converted to oxaloacetate by the cytoplasmic aspartate aminotransferase (GOT1) to allow for the next electron transport process (Fig. 1; 40). Like the GPS, the MAS is required to maintain redox balance between mitochondria and cytoplasm and to support a high glycolytic flux in activated macrophages (33).

One Carbon Metabolism in Activated Macrophages

Folate-mediated one carbon (1 C) metabolism refers to connected metabolic pathways which include the folate and methionine cycles (34). 1 C pathways are located in the cytosol, the mitochondria and the nucleus (41). This pathway is essential to provide 1 C or methyl groups for the synthesis of DNA, amino acids, polyamines, creatine, and phospholipids (41). Moreover, 1 C metabolism plays a role in epigenetic regulations and redox defense (42). 1 C metabolism is particularly vital in highly metabolic active cells such as neuronal and immune cells (42, 43). In this regard, there is evidence which suggests that 1 C metabolism contributes to the energy demand of highly metabolically active cells. This can be demonstrated by reactions of 1 C metabolism which produce ATP and NADPH. For example, serine catabolism has been found to generate ATP with the activity of reverse 10-CHO-THF synthase (FTHFS; 44). Furthermore, serine catabolism is associated with NADPH generation through the activity of cytosolic and mitochondrial methylenetetrahydrofolate dehydrogenase (MTHFD) and 10-CHO-THF dehydrogenase (FTHFD; 44).

Besides their role in energy production, lipopolysaccharides (LPS) were found to induce 1 C metabolism in proinflammatory macrophages (35). Furthermore, 1 C metabolism was found to work synergistically with the induced pentose phosphate pathway and serine synthesis pathway to achieve epigenetic regulation and to induce interleukin (IL)-1β expression (35). S-adenosylmethionine (SAM), an end product of 1 C metabolism, has been demonstrated to be the driver for histone H3 lysine 36 trimethylation and IL-1β induction (35). The inhibition of SAM generation through impairment of the three aforementioned metabolic pathways leads to an anti-inflammatory effect, confirming the role of SAM in driving the inflammatory response (35).

Pyruvate Import into Mitochondria and Its Oxidative Decarboxylation

Mitochondrial pyruvate carrier 1 (MPC1) transfers pyruvate from the cytosol into the mitochondrial matrix where it can be oxidized and decarboxylated by the pyruvate dehydrogenase complex (PDC) to acetyl-CoA, the starting metabolite of the TCA or Krebs cycle (Fig. 1; 45). MPC1 plays a critical role in activated macrophages as inhibition of MPC1 resulted in reduced levels of the immune metabolite itaconate and eventually decreased the expression of immunoresponsive gene 1 (Irg1), inducible nitric oxide synthase (iNOS), and tumor necrosis factor α (Tnfα; 38, 46).

βOxidation and Carnitine Palmitoyl Transferase-1 System

β Oxidation of long chain fatty acids has been proposed to be important for IL-4-induced alternative polarization of macrophages which feature enhanced fatty acid oxidation (47). Carnitine palmitoyl transferase-1 (CPT-1), a mitochondrial outer membrane protein that facilitates long chain fatty acid uptake into the mitochondrial matrix for oxidation, was assumed to play an important role in macrophage polarization (Fig. 1; 47). Experiments with the CPT-1 inhibitor etomoxir highlighted an inhibiting effect on anti-inflammatory macrophage polarization (47–49). However, a recent study by Divakaruni et al. (47) has shown that long chain fatty acid oxidation was not a critical part of this process. In this context, they showed that previously applied etomoxir levels were high enough to deplete the intracellular pool of free coenzyme A, possibly via the conversion of prodrug etomoxir into active etomoxiryl CoA (47).

TCA Cycle

The Krebs or TCA cycle is central to mitochondrial metabolism and fuels the respiratory chain with reduced co-factors in the form of NADH and FADH2. It is a cycle of nine reactions that starts with the irreversible condensation of acetyl-CoA and oxaloacetate to citrate and ends with the synthesis of the next round’s precursor oxaloacetate (Fig. 1). Due to the two oxidative decarboxylation reactions catalyzed by isocitrate dehydrogenase (IDH) and 2-oxoglutarate dehydrogenase (2OGDH) in mammalian metabolism, mammalian cells cannot gain net biomass from acetyl-CoA through this cycle and the main role of the mitochondrial TCA cycle activity is the supply of energy. Nevertheless, TCA cycle activity is also a provider of critical intermediates for cell and macrophage function, such as aspartate and itaconate (36). This is the reason why TCA cycle fluxes are rerouted in pro-inflammatory macrophages to support macrophage function. The first major modification is the reduction of IDH activity. IDH is an enzyme which converts isocitrate to α-ketoglutarate and the downregulation of IDH in proinflammatory macrophages results in citrate accumulation. The accumulated citrate serves as a precursor for the synthesis of itaconate and cytosolic acetyl-CoA (Fig. 1; 37, 50). The second adjustment of TCA cycle activity in activated macrophages is the inhibition of succinate dehydrogenase (SDH) by accumulating itaconate (46, 51). SDH oxidizes succinate to fumarate and its inhibition ultimately yields to succinate accumulation. Tannahill et al. (52, 53) suggested that accumulated succinate induces IL-1β expression through HIF-1α stabilization (Fig. 2).

Figure 2.

Immune signaling of pattern recognition receptors (PRRs) and how mitochondria contribute to immune cell activation. RIG-I, retinoic acid inducible gene-I; MDA5, melanoma differentiation associated gene 5; MAVS, mitochondrial-antiviral-signaling-protein; TANK, TRAF family member-associated NF-κB activator; TRAF, tumor necrosis factor receptor (TNFR) associated factor; IRF3, interferon regulatory factor 3; IFNs, interferons; NF-κB, inducing nuclear factor κ-light-chain-enhancer of activated B cells; TLR, toll-like receptor; TRIF, TIR domain containing adapter-inducing interferon-β; TRAM, TRIF-related adapter molecule; MyD88, myeloid differentiation factor 88; IRAK, IL-1 receptor-associated kinase; IRG1, immunoresponsive gene 1; NRF2, nuclear factor erythroid 2-related factor 2; IκBζ, nuclear factor κ B zeta; HIF-1α, hypoxia-inducible factor; CREB, cAMP response element-binding protein; IL-6, interleukin-6; IL-1β, interleukin-1β; BAX, Bcl-2-associated X protein; BCL2L11, BCL2 like 11; MARCH5, protein E3 ubiquitin protein ligase; NLR, nuclear oligomerization domain (NOD)-like receptor; NLRP3, NLR family pyrin domain containing 3; ROS, reactive oxygen species; CASP1, caspase 1; IL-18, interleukin-18.

Contrary to proinflammatory macrophages, TCA cycle activity is induced in anti-inflammatory macrophages (5). Upon the induction of anti-inflammatory macrophages with IL-4, signal transducer activator of transcription 6 (STAT6), and peroxisome proliferator activator receptor γ (PPARγ)-coactivator-1β (PGC-1β) are induced, contributing to the development of the anti-inflammatory macrophage phenotype, metabolically including the upregulation of fatty acid oxidation, mitochondrial respiration, and the stimulation of mitochondrial biogenesis (5, 54).

Itaconate

Besides energy supply, TCA cycle metabolism is directly involved in antimicrobial defense mechanisms and immune modulation through the compound itaconic acid. Itaconic acid is an organic unsaturated dicarboxylic acid also known as methylene succinic acid (50). Michelucci et al. (37) demonstrated that this compound is produced in proinflammatory macrophages from cis-aconitate. Irg1 has been revealed as the gene encoding cis-aconitate decarboxylase (ACOD1), the enzyme producing itaconic acid (Fig. 1; 37). Irg1 is one of the most upregulated genes during proinflammatory conditions in murine macrophages and microglial cells (Fig. 2; 37, 55–57).

Itaconate acts as an antimicrobial compound because it is a potent inhibitor of isocitrate lyase (ICL), a key enzyme of the glyoxylate shunt, which is a biosynthetic pathway present in many microorganisms but absent in animals (58, 59). ICL in pathogens catalyzes the important conversion of isocitrate to glyoxylate and succinate, thereby circumventing the two decarboxylation reactions of the TCA cycle. The glyoxylate shunt and ICL activity are essential for microbial biomass production when growing on acetate or fatty acids as carbon sources (60). For example, the growth of Mycobacterium avium and M. tuberculosis depends on ICL activity when surviving inside macrophages, because they mainly employ host cell cholesterol as a carbon source, which is degraded to acetyl-CoA (58, 61). Another enzyme that is inhibited by itaconate is propionyl-CoA carboxylase. This enzyme catalyzes the conversion of propionyl-CoA to methylmalonyl-CoA, an important step for the detoxification of propionate via the citramalate cycle (62).

Many pathogens such as Yersinia pestis or Pseudomonas aeruginosa have been associated with developing countermeasures against itaconate intoxication (63). Three enzymes have been identified to degrade itaconate to pyruvate and acetyl-CoA: itaconate CoA transferase (Ict), itaconyl CoA hydratase (Ich) and (S)-citramalyl-CoA lyase (Ccl) (64). The genes encoding for these enzymes cluster in a genomic region called “required for intracellular proliferation” (ripABC) and represent a bacterial barrier against the antimicrobial effects of itaconate, even exploiting this anti-microbial metabolite as a carbon source (65). Inhibiting the itaconate degradation pathway in bacteria might represent a promising target for future anti-bacterial treatments (50, 65).

Apart from its antimicrobial properties, itaconate has also been demonstrated to possess anti-inflammatory properties. Part of the anti-inflammatory effect of itaconate has been suggested to be rooted in its inhibition of SDH (64). That is because of the role of SDH in driving the inflammatory response by reverse electron flow to complex I and ROS-mediated HIF-1α stabilization in activated macrophages (66). Mills et al. (67) have investigated the effect of the itaconate ester 4-octyl itaconate (4-OI) on mouse and human macrophages and revealed that this itaconate derivative increases levels of the transcription factor nuclear factor erythroid 2-related factor 2 (NRF2) via the inhibition of Kelch-like ECH-associated protein 1 (KEAP1). In turn, this results in an increased expression of NRF2 downstream target genes such as the anti-inflammatory protein heme oxygenase 1 (HMOX1). The increase of NRF2 expression was explained by an alkylation of a cysteine residue in KEAP1 through 4-OI. Under normal conditions, KEAP1 promotes degradation of NRF2, however, alkylation of cysteine residues in KEAP1 protein resulted in accumulation and activation of newly synthesized NRF2 (67).

In addition to the anti-inflammatory effect mediated by NRF2, itaconate has also been found to impair the glycolytic flux by inhibiting fructose-bisphosphate aldolase A (ALDOA) and glyceraldehyde 3-phosphate dehydrogenase (GAPDH), subsequently attenuating the inflammatory response in activated macrophages (68, 69). Furthermore, Bambouskova et al. (70) investigated the impact of the itaconate derivative dimethyl itaconate (DI) on mouse macrophages. DI was found to induce electrophilic stress and regulate the NF-kappa-B inhibitor zeta (IκBζ) protein via the key mediator activating transcription factor 3 (ATF3). Moreover, DI was found to inhibit IL-17 mediated-IκBζ induction in keratinocytes and ameliorate the pathology of psoriasis in a psoriasis mouse model which underlines the importance of this pathway as a new target for the treatment of autoimmune disease (70). However, recently ElAzzouny et al. (71) and Swain et al. (72) have demonstrated that none of these derivatives were converted to itaconate intracellularly, putting the proposed anti-inflammatory mechanisms of itaconate into question.

Finally, itaconate has been proposed to work synergistically with iNOS to achieve NLR Family Pyrin Domain Containing 3 (NLRP3) tolerance and prevent full caspase-1 activation (73). In another study, 4-OI was found to block NLRP3 activation by blocking NLRP3-NEK7 interaction, whereas Irg1−/− BMDMs exhibited increased NLRP3 inflammasome activation (74). This indicates a potential role of itaconate as a regulator of the NLRP3 inflammasome. Even though many studies have addressed itaconate’s function in immunomodulation, several aspects of related mechanisms are still not fully understood. Unraveling the details of the itaconate mode of action in mammals and bacteria could help to identify new promising antibacterial targets (65).

UREA CYCLE AND THE ASPARTATE-ARGININOSUCCINATE SHUNT

The aspartate-argininosuccinate shunt interconnects the urea and the TCA cycle. Argininosuccinate synthetase catalyzes the conversion of citrulline and aspartate to arginosuccinate which is then hydrolyzed to arginine and the TCA intermediate fumarate by argininosuccinate lyase (5). After hydration to malate by fumarase activity, malate can then pass the mitochondrial membrane. Malate is further involved in the TCA cycle, where it can be converted into either oxaloacetate or fumarate (Fig. 1). The action of the arginosuccinate shunt is twofold: First, it is an alternative source for fumarate synthesis and can replenish reduced fumarate production in proinflammatory macrophages in which SDH is inhibited by endogenous itaconate (75). Second, it is crucial for arginine synthesis which is critical for antimicrobial nitric oxide (NO) production through nitric oxide synthase II (NOS2; 5).

ATP-CITRATE LYASE AND HISTONE ACETYLATION

Upon Toll-like receptor (TLR)4 activation with LPS, glucose uptake and mitochondrial citrate synthesis are induced, as highlighted earlier. The accumulated citrate is partially exported into the cytosol via the citrate transporter protein (CTP) and then further converted to acetyl-CoA and oxaloacetate by ATP-dependent citrate lyase (ACLY), an enzyme which is upregulated under pro-inflammatory conditions. Lauterbach et al. (76) recently demonstrated that ACLY inhibition in activated macrophages leads to the downregulation of inflammatory genes such as Il6, Il12β, Il18, Il27, Cxcl9, and Cxcl10. The authors showed that reduced cytosolic acetyl-CoA caused by the inhibition of ACLY directly attenuates histone acetylation which is a prerequisite for the induction of specific LPS-inducible gene sets (76).

Even though ACLY is a major regulator of cytosolic acetyl-CoA production, this metabolite can also be produced from acetate through acetyl-CoA synthetase short-chain family member 2 (ACSS2) (77). This could explain why ACLY inhibition only reduces histone acetylation to a basal level and does not completely abolish it (76).

PROFILING MITOCHONDRIAL METABOLISM

In the previous sections, we highlighted the essential role of mitochondrial metabolism to mount the inflammatory response. In this regard, profiling mitochondrial metabolism is of high relevance for the understanding of immunometabolic mechanisms. However, the isolation of mitochondrial metabolites can be challenging as mitochondria are heavily tied to and intermingled with their cell. Several techniques have been developed to address this issue and aim to fractionate mitochondria from their cellular context. In 2007, Frezza et al. (78) reported a protocol for the isolation of mitochondria and performed a subsequent metabolic analysis. While this procedure is very efficient in terms of purity, it requires 1–2 h for the isolation, which drastically impacts the stability of the mitochondrial metabolome as metabolic reactions operate in the range of seconds. Later on, Chen et al. introduced a protocol to access mitochondrial metabolites based on high-affinity magnetic immunocapture and successfully profiled mitochondrial metabolites after a period of only 12 min (79). More recently, Nonnenmacher et al. developed a protocol to enrich mitochondrial metabolites via the selective permeabilization of the cytosolic membrane using the detergent digitonin (80, 81). This protocol has been demonstrated to be the fastest since selective permeabilization requires only one minute for removing the cytosolic contents of the cell while maintaining functional mitochondria for either direct metabolomic measurements or for further stable isotope-based targeted and non-targeted metabolic flux analyses (82, 83).

ROLE OF MITOCHONDRIA IN MOUNTING AN INFLAMMATION

Macrophages are activated by sensing sets of molecular targets known as pathogen-associated molecular patterns (PAMPs) or damage-associated molecular patterns (DAMPs) by pattern recognition receptors (PRRs) (84). PRRs include receptors such as retinoic acid inducible gene (RIG-I)-like receptors (RLRs), TLRs and nucleotide-binding oligomerization domain (NOD)-like receptors (NLRs) (25). In the next paragraphs, we will highlight how mitochondria support the signal transmission of PRRs after their activation.

RIG-I-LIKE RECEPTORS

RLR receptors such as RIG-I and melanoma differentiation associated gene 5 (MDA-5) are responsible for recognizing viral double stranded (ds) RNA. Activation of these receptors results in the induction of nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) and interferon regulatory factor 3 (IRF-3). This, in turn, causes the upregulation of type I interferons (IFNs) and other pro-inflammatory cytokines which promote antiviral immunity (85–88). Mitochondria have been demonstrated to be an integral part of this mechanism as they express the outer mitochondrial membrane (OMM) protein mitochondrial-antiviral-signaling-protein (MAVS). MAVS is also known as IFNβ promoter stimulator 1 (IPS1) or CARD adapter inducing IFNβ (CARDIF) or virus-induced signaling adaptor (VISA) (89–93). Upon activation with viral dsRNA, MAVS interacts with RIG-1 which results in the recruitment of tumor necrosis factor receptor (TNFR) associated factor (TRAF) 2, 5 and 6 (92). This promotes the antiviral response, including the activation of NF-κB and IRF signaling pathways (Fig. 2; 25, 86, 87). In this regard, localization of MAVS to the mitochondrial membrane has been demonstrated to be critical for its function (91). Seth et al. demonstrated that the mislocalization of MAVS either to the plasma membrane or the endoplasmic reticulum (ER) significantly reduced its activity (91). Additionally, an intact mitochondrial membrane potential as well as mitochondrial fusion and fission are a prerequisite for the RLR-induced antiviral response (25, 94).

TOLL-LIKE RECEPTORS

Another set of receptors involved in immune cell activation are TLRs. TLRs are transmembrane receptors expressed on the plasma membrane, the endosome or ER (25, 95). There are 13 members of the TLR family that have been identified in mammals so far, which are activated by PAMPs from bacteria, fungi, parasites or viruses (96–98). The activation of TLRs results in the upregulation and processing of several inflammatory cytokines such as IL-1β, IL-6 and TNFα (95). In this regard, mitochondria were found to be involved in TLR7 signaling (99). This receptor senses viral single stranded (ss)RNA and its activation induces increased expression of pro-inflammatory cytokines and type I interferons. The OMM protein E3 ubiquitin-protein ligase MARCH5 is important for TLR7 signal transmission to NF-κB (99). For this purpose, it interacts with TRAF family member-associated NF-κB-activator (TANK) and impairs its ability to inhibit TRAF6 (99). This is required for the induction of pro-inflammatory cytokines and immune cell activation (Fig. 2) (99). Furthermore, MARCH5 localization to the outer membrane of the mitochondrion is critical for its function as mislocalization has been shown to diminish its function (99). Another receptor thought to be associated with mitochondria is TLR3. TLR3 is activated by both, viral dsRNA and host mRNA released from damaged cells (100). Triggering TLR3 has been reported to induce apoptosis via a downstream regulator known as TAp63α (101). TAp63α induces the mitochondrial apoptosis pathway through an upregulation of B cell lymphoma 2 (Bcl-2) family proteins such as Bcl-2-associated X protein (Bax) and BCL2-like 11 (BCL2L11) (Fig. 2; 102).

Lastly, TLR2/4 also have an impact on mitochondria as they control mitochondrial biogenesis through several transcription factors such as cAMP response element-binding protein (CREB), NF-κB, IRF3 and NRF2 (Fig. 2; 25, 103). TLR4 is possibly the most studied TLR receptor in nucleated cells and recognizes LPS, part of the outer membrane of gram-negative bacteria (104). Upon LPS recognition, two major signaling pathways are activated. The first pathway is activated by myeloid differentiation factor 88 (Myd88) and IL-1 receptor-associated kinase 4 (IRAK4), which leads to NF-κB activation and the upregulation of inflammatory cytokines such as TNFα and IL-1β (104). The second pathway is activated via TIR domain-containing adapter-inducing interferon-β (TRIF) and TRIF-related adapter molecule (TRAM), which leads to IRF3 activation and the induction of type I interferons (104). LPS activation of macrophages leads to several modifications of mitochondrial metabolism, finally resulting in the polarization of macrophages into a pro-inflammatory subtype.

NOD-LIKE RECEPTORS

NLRs are the third family of receptors which are involved in innate immune cell activation. In addition to other intracellular danger signals, NLRs sense microbial products (105). Upon triggering NLRs, the innate immune system is activated through the induction of NF-κB and inflammatory caspases (105, 106). There are currently 22 known members of the NLR family (105), with NLRP3 possibly being the most studied among those (107, 108). It is an integral component of the inflammasome, a multiprotein complex acting as a danger signaling platform (25, 109), whereby it regulates the activity of caspase 1 and thus controls the processing and secretion of IL-1β and IL-18 (109–111).

The NLRP3 inflammasome is activated by a two-step procedure which includes several signals triggered by infection and metabolic dysregulation (108). To activate the NLRP3 Inflammasome, macrophages have to be primed with microbial molecules or endogenous cytokines to first induce NLRP3 and pro-IL-1β expression via NF-κB activation (112, 113). The second step is not fully understood and involves the sensing of either of ATP, viral RNA or pore-forming toxins, eventually inducing inflammasome assembly and the activation of pro-caspase 1 which then catalyzes the proteolytic cleavage of IL-1ß and IL-18 (112, 114).

Upon inflammasome activation, NLRP3 translocates from its resting location at the ER to mitochondria and mitochondria-associated ER membranes (25, 115). MAVS promotes NLRP3 recruitment to mitochondria and the induction of IL-1β processing (Fig. 2; 108). NLRP3 inflammasome activation is highly dependent on mitochondrial function, hence mitochondrial dysfunction was found to impair NLRP3 inflammasome activation (25, 115).

Independent of DAMPs and PAMPs, there are other factors which play a role in modulating mitochondrial metabolism and macrophage response (116). Other factors include oxygen and nutrient availability and metabolic cargo from cell debris (116). For example, decreased oxygen supply in hypoxic conditions lead to a shift in macrophage metabolism to anerobic glycolysis (46). Moreover, it was reported that hypoxic and inflammatory signals share certain transcriptional events such as the activation of HIF-1α and NF-κB families (117). This demonstrates that the microenvironment could modulate the metabolic profile of macrophages (117). This is especially important in cases such as ischemia and cancer, where the microenvironment changes, imposing metabolic alterations on macrophages (117). Studying such metabolic changes may help to identify novel targets which could have therapeutic applications in the future.

MITOCHONDRIAL REACTIVE OXYGEN SPECIES

Mitochondrial reactive oxygen species are produced as a byproduct during OXPHOS or in response to external stimuli (118–120). Too high levels of ROS are considered to be toxic as they can cause severe damage to DNA, lipids and proteins (120, 121). To a certain extent, cells can counteract this toxicity with the production of several ROS scavenging enzymes such as superoxide dismutase, catalase, and glutathione peroxidase (120).

Despite the harmful effects of ROS, several studies have revealed that moderate levels of ROS are essential for proper cell signaling (119, 122–124). As an example, H2O2 is produced when cells are stimulated by cytokines such as TNF-α, transforming growth factor beta-1 (TGF-β1) or interleukin-1 (119, 125–127). Upon induction, H2O2 functions as a second messenger in NF-κB activation (128). Furthermore, mROS produced by complex I is involved in NF-κB induction through the stabilization of HIF-1α and c-Src mediated phosphorylation of IκB-α (129). This activation could be prevented by treating cells with an antioxidant such as N-acetyl-l-cysteine (128). ROS also functions as an activator of the mitogen-activated protein kinase (MAPK) pathway and triggers its activation either by the inactivation of MAPK phosphatases (MKPs) (a group of protein phosphatases, responsible for dephosphorylation and inactivation of MAPKs) or by the oxidative modification of MAPK signaling proteins such as receptor tyrosine kinase (RTK) and MAP3Ks (130, 131). The activation of MAPKs and NF-κB pathways by ROS highlights their role during inflammation.

SUMMARY AND PERSPECTIVE

During evolution, mitochondria have developed to an integral player of macrophage function. This is not only because of their role in energy supply but also due to their metabolic support during inflammatory conditions. Moreover, these organelles provide macrophages with anti-microbial intermediates such as NO and itaconate, which are needed to combat pathogens. Our understanding of mitochondrial metabolism provides a key aspect for our knowledge of innate immune system signaling.

Proinflammatory and anti-inflammatory macrophages have distinct metabolic needs. Whereas pro-inflammatory macrophages depend more on glycolysis to achieve rapid energy supply at the inflammatory site, anti-inflammatory macrophages depend on OXPHOS to achieve sustainability for long-term processes (5, 26, 132).

Moreover, mitochondria are an important source of ROS which play a regulatory function during inflammation. Besides the stabilization of HIF-1α, ROS triggers the activation of NF-κB, MAPKs pathways and the NLRP3 inflammasome.

A key metabolite produced in activated macrophages is itaconate. In addition to its anti-microbial function, it is involved in immunomodulatory processes. Although its anti-inflammatory function is under debate, the specific mechanism is still not clear. Moreover, neither the subcellular location of its synthesis nor its degradation pathway in macrophages is completely explained.

With this review, we highlight the importance of mitochondria in macrophages. Furthermore, we want to motivate to study these organelles in more depth to gain a better understanding of their contribution to inflammatory processes and thus to the pathogenesis of infectious, inflammatory and autoimmune diseases.

GRANTS

This work was supported by Federal State of Lower Saxony, Niedersaechsisches Vorab CDInfect project, the University of Bonn and integrative data analytics for respiratory syncytial virus risk assessment (INDIRA).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.E.H. prepared figures; M.N. drafted manuscript; M.Z.N. and K.H. edited and revised manuscript; M.Z.N., J.E.H., and K.H. approved final version of manuscript.