Immunometabolism of Myeloid-Derived Suppressor Cells in Oncology

Immunometabolism of Myeloid-Derived Suppressor Cells in Oncology: Comparison

Please note this is a comparison between Version 1 by Brian SM Munansangu and Version 2 by Lindsay Dong.

The field of immunometabolism seeks to decipher the complex interplay between the immune system and the associated metabolic pathways. The role of small molecules that can target specific metabolic pathways and subsequently alter the immune landscape provides a desirable platform for new therapeutic interventions. Immunotherapeutic targeting of suppressive cell populations, such as myeloid-derived suppressor cells (MDSC), by small molecules has shown promise in pathologies such as cancer and support testing of similar host-directed therapeutic approaches in MDSC-inducing conditions such as tuberculosis (TB). MDSC exhibit a remarkable ability to suppress T-cell responses in those with TB disease. In tumors, MDSC exhibit considerable plasticity and can undergo metabolic reprogramming from glycolysis to fatty acid oxidation (FAO) and oxidative phosphorylation (OXPHOS) to facilitate their immunosuppressive functions.

MDSC
tuberculosis
metabolic reprogramming

1. MDSC Classification

1.1. Origin and Identification

Myeloid-derived suppressor cells (MDSC) originate in the bone marrow during the development of myeloid progenitor cells, i.e., monocytes and neutrophils ^[1][20]. MDSC are divided into two major populations in both humans and mice, namely, polymorphonuclear MDSC (PMN-MDSC) and monocytic MDSC (M-MDSC), as described previously in humans ^[2][3][21,22] and mice ^[4][5][23,24]. Recently, the markers lectin-type oxidized LDL receptor 1 (LOX1) ^[6][25] and fatty acid transporter 2 (FATP2) ^[7][26] have been shown to discriminate MDSC in tumor biology, although these have not yet been confirmed to be expressed by MDSC in infectious diseases such as active TB disease. Although M-MDSC and PMN-MDSC are considered the two main subsets of this suppressive population, recent evidence from human studies points towards a precursor subset termed early-stage MDSC ^[8][27]. Agrawal et al. also described a subverted DC phenotype characterized by low expression of MHC class II and CD80, and expression of CD14 that lacked CD1a molecule with the presence of CD83 and CD86 ^[9][28]. Additionally, our research group has also shown that CD14⁺ MDSC produces IL-10 and Il-6 with low levels of HLA-DR and CD80 ^[9][10][28,29]. Another recently described subset of MDSC has been termed eosinophilic (Eo)-MDSC and was identified in a mouse model infected with Staphylococcus aureus ^[11][30].

1.2. Recruitment and Expansion of MDSC

MDSC are recruited from the bone marrow in response to a variety of growth factors, hormones, and transcription factors, along with antigens linked to these pro-inflammatory conditions, thereby regulating myelopoiesis and inducing MDSC expansion. This process is enhanced by alarmins such as prostaglandin E2 (PGE2) and high mobility box group 1 (HMGB1) secreted by host tissues ^[12][13][31,32], and pro-inflammatory cytokines such as tumor necrosis factor (TNF-α), interleukin-1β (IL-1β), IL-6, IL-13, and S100A8/A9 ^[14][33]. Hypoxia and nutrient starvation may also affect the epigenetic status of myeloid cells ^[15][34], which is critical to the accumulation and reprogramming of MDSC in response to inflammatory or pathogenic insults ^[16][35]. Additionally, chemokines such as CCL2 (MCP-1, monocyte chemoattractant protein-1), CXCL2 (MIP-2, macrophage inflammatory protein 2), and CXCL8 have been identified to be vital in MDSC migration ^[17][18][36,37].

1.3. Immunosuppressive Mechanisms of MDSC

Although MDSC are widely considered regulators of the immune response, contributing to the homeostatic balance between pro- and anti-inflammatory responses, under aberrant conditions of chronic infection, MDSC are mainly considered detrimental to efficient host control of the pathogen. Several studies have shown how metabolic alterations, as well as those induced by infections, can influence the function, reprogramming, and differentiation of MDSC ^[19][38]. The mechanisms by which these cells suppress the immune response include:

(i)

(i)

Sequestration of cardinal amino acids (AA) such as L-arginine, L-cysteine (Cys), and L-tryptophan by the activity of inducible enzymes such as arginase 1 (ARG1), inducible nitric oxide synthase (iNOS/NOS2) and indolamine dioxygenase (IDO) ^[8][20][27,39]. The reduction of these AA leads to the inhibition of T-cell activation and proliferation and reduced expression of the T-cell receptor TCR-CD3 ζ chain, inducing the cell to undergo proliferative arrest ^[21][22][23][40,41,42];

(ii)

(ii)

The production of reactive oxygen species (ROS) and reactive nitrogen species (RNS) by MDSC skews the polarization of monocytes, T cells, and macrophages towards anti-inflammatory and regulatory phenotypes ^[24][25][43,44];

(iii)

(iii)

Indirect suppression of T and effector B cells through the induction of tolerogenic immune cells such as de novo generation of Fox-P3⁺ regulatory T cells (Tregs) ^[26][45], regulatory B cells, and tumor-associated macrophages (TAMs) ^[22][27][41,46]. Recently, a novel mechanism was elucidated through which MDSC-dependent metabolic and functional paralysis of CD8⁺ T cells occurs ^[28][47]. The mechanism involves methylglyoxal-derived glycation of L-arginine products such as argpyrimidine and hydroimidazolone, thereby depleting cytosolic amino acids such as L-arginine, resulting in T-cell paralysis ^[28][29][47,48]. The transfer of methylglyoxal from MDSC to T cells is dependent on cell–-cell contact, resulting in more T-cell suppression at sites where MDSC may accumulate, such as tumor tissue ^[28][47];

(iv)

(iv)

Secretion of suppressive cytokines such as TGF-β and IL-10 that exert direct suppressive effects on T-cell responses ^[26][45];

(v)

(v)

Induction of T-cell apoptosis through the induction of the B7 family of immune-regulatory ligands, a co-signaling network superfamily that plays an essential role in the modification of T-cell activation and tolerance ^[30][49], such as B7-H1 (programmed cell death ligand 1 (PD-L1)), B7-H3, and B7-H4 ^[31][50], and impairment of T-cell migration through the reduction of CD62L expression ^[10][29].

For the past ~10 years, most immunosuppressive mechanisms have been elucidated from cellular interactions within the tumor microenvironment (TME), driving inferences for MDSC immunosuppressive functions in other conditions.

2. Immunometabolism of MDSC in Oncology

Metabolic dysregulation is a hallmark of numerous cancer types ^[32][107]. Altered metabolism has been observed in aggressive tumors such as glioblastoma (GBM) and ovarian and gastric cancer. MDSC have been shown to sense or exhibit plasticity in their environment and respond by selecting the most efficient metabolic pathways to sustain their suppressive and pro-tumorigenic functions ^[33][108]. In the TME, MDSC increase uptake of FA and activate the switch from glycolysis to FAO ^[34][109].

2.1. Metabolic Reprogramming of MDSC

As previously noted, MDSC exhibit distinct functions and phenotypes in various disease settings ^[23][42]. They display a certain degree of plasticity where they can assume a pro- or anti-inflammatory phenotype to support tumor proliferation ^[35][110]. Although the underlying mechanism leading to MDSC function and activation in TB is not fully known, metabolic reprogramming of MDSC underpins many of the suppressive functions in cancer ^[23][42]. As with cancer cells, available evidence indicates that MDSC can undergo anaerobic glycolysis and OXPHOS, which are influenced by substrate availability and by signaling pathways elicited by metabolites or pathogen-derived inflammatory signals ^[23][36][42,111]. During maturation, MDSC exhibit elevated utilization of central carbon metabolism, the PPP, glycolysis, and the TCA cycle ^[37][112]. Various signaling pathways, such as the phosphatidylinositol 3-kinase (PI3K)–serine threonine protein kinase (AKT)–mTOR pathway, operate in concert to control metabolic reprogramming/activity in immune cells ^[38][113]. Under conditions of oxygen deprivation, mTOR signaling stimulates the HIF-1α pathway, resulting in upregulation of glycolytic enzymes, glucose, and lactate transporters—all of which result in the Warburg effect ^[28][47]. The phenotypic heterogeneity of MDSC could cause MDSC to compete with other immune cells for carbon acquisition in the TME and fall under the control of energy metabolic pathways, such as FA metabolism ^[39][114], which has a direct effect on the regulation of OXPHOS and glycolysis ^[39][114]. Furthermore, the TME may dysregulate some key genes associated with hallmark cancer pathways ^[40][41][115,116], such as HIF-1α (de-regulating cellular energetics), telomerase activation (enabling replicative immortality) ^[41][116], and activation of the NF-κβ and TGF-β pathways (antagonizing growth suppressor activity of p53 pathways) ^[42][117]. The pulmonary granuloma from M. tb infection may similarly dysregulate some key genes as seen in cancer. HIF-1α binds to the promoter of lactate dehydrogenase A (LDHA), which catalyzes the conversion of pyruvate to lactate, and in some cancers, isoforms of LDHA subunits have the highest efficiency of converting pyruvate to lactate and are linked to increased HIF-1α and VEGF expression, increased tumor size, enhanced metastatic potential, and a poor prognosis ^[3][43][44][22,118,119]. Non-host immune response therapies such as radiation therapy have also been associated with activation of MDSC and were dependent on enhanced lactate secretion and mediated by HIF-1α activity, resulting in the lactate-regulating MDSC function to reprogram the TME into a more immunosuppressive phenotype ^[45][120]. LDH has been found in active pulmonary TB patients’ bronchoalveolar lavage (BAL) fluid at high levels, and BAL LDH corresponds with increased serum LDH (Emad and Rezaian, 1999). Sputum-positive TB patients had higher serum levels of LDH1, LDH2, and LDH3, which have four, three, or two B subunits, respectively ^[46][121]. Increases in cerebrospinal fluid (CSF) lactate has also been linked to increased severity of tuberculous meningitis clinical stage ^[47][122]. This suggests that LDH levels and isoform specificity, as well as lactate levels, might be used as diagnostic or prognostic markers for pulmonary tuberculosis.

2.2. Metabolism of Glucose, Lipids, and Amino Acids by Tumor-Derived MDSC

Hossain et al. found that tumor-infiltrating MDSC (T-MDSC) increased fatty acid uptake and activated FAO, including upregulation of key FAO enzymes ^[48][49][5,123]. These findings exposed new avenues as targets for the development of host-directed therapies to enhance treatment outcomes; inhibiting MDSC has also been shown to improve prospects for successful immunotherapy and/or radiation and chemotherapy ^[2][50][21,124]. Additionally, other metabolic pathways could impact the reconfiguration of host immune responsiveness, such as glycolysis, pentose phosphate pathways (PPP), tricarboxylic cycle (TCA), fatty acid synthesis (FAS), and amino acid synthesis (AAS) ^[51][52][125,126]. T-MDSC have been shown to metabolically shift from glycolysis to OXPHOS and upregulate lipid-associated markers such as CD36 and Msr1 ^[53][127]. Studies by Al-Khami et al. showed that the intracellular accumulation of lipids increases oxidative metabolism and activates the immunosuppressive mechanisms. Inhibition of STAT3 or STAT5 signaling or genetic depletion of the fatty acid translocase CD36 inhibits the activation of oxidative metabolism and the induction of immunosuppressive function in tumor-infiltrating MDSC and results in a CD8⁺ T-cell-dependent delay in tumor growth ^[53][54][127,128]. The TME is characterized by a complex network of blood vessels, tumor cells, and host immune cells, featuring extensive crosstalk of chemokines, cytokines, immune regulatory molecules, and transcription factors that shape the phenotype of MDSC and cancer cells ^[55][56][129,130]. The TME favors a hypoxic environment with a low pH that is thought to arise from the extracellular accumulation of lactate ^[57][131] as a result of the interaction of HIF-1α and MYC proto-oncogene transcription factor (Myc) during the upregulation of glycolytic enzymes such as GLUT and the influx of glucose and lactate production ^[58][132]. MDSC in tumors have been shown to utilize both glycolysis and OXPHOS, as observed in nasopharyngeal cancer (NPC) ^[59][133]. Liu et al. demonstrated that glycolytic activation via the mTOR-pathway was essential for M-MDSC differentiation into M1 and M2 macrophage phenotypes and was dependent on SIRT1 (a negative regulator of mTOR) ^[60][61][134,135]. Many similarities exist between TME and TB granulomas. Mature TB granulomas are heterogeneous, and several distinct TB lesions can coexist within the same patient ^[62][140]. Histological restationing of the lung tissues from pulmonary TB, sarcoidosis, and lung adenocarcinoma (LUAD) patients showed that TB-infected lung tissue shares a set of potential pathogenic mediators with LUAD. MDSC are considered a major cellular contributor to the suppressive TME ^[63][64][141,142], and tumor-associated growth factors such as granulocyte-macrophage colony-stimulating factor (GM-CSF), granulocyte colony-stimulating factor (G-CSF), and signaling pathways such as STAT3 and STAT5 are known to upregulate lipid transport receptors (such as CD36 and Msr1) in MDSC and enhance lipid accumulation ^[49][123]. Similarly, the TB granuloma microenvironment is known to be enriched with lipid content, fatty acids (FAs), cholesterol, and “foamy” macrophages ^[65][66][143,144]. Considering the numerous immune and physical parallels that exist between the TME in cancer and the TB granuloma microenvironment, a similar outcome is anticipated for MDSC in tuberculous granulomas. Additional support comes from studies demonstrating M2 macrophage dominance during M. tb infection. These rely heavily on OXPHOS for ATP production. Once monocytes enter the interstitial tissue, they differentiate into macrophages and align to an M2 phenotype, although reports suggest that M. tb can induce aerobic glycolysis upon infection ^[67][92]. On the contrary, other researchers have shown that M. tb inhibits aerobic glycolysis in infected macrophages ^[68][69][93,145]. Foamy macrophages in TB granulomas have been suggested to participate in sustaining persistent bacteria and tissue pathology that leads to cavitation and the release of infectious bacilli ^[66][144]; this may be true for MDSC, which may internalize M. tb, allowing it to flourish in the lipid-rich niche of the granuloma ^[70][71][146,147]. This was demonstrated in cancer, whereby MDSC overloaded with lipids suppressed CD8⁺ T cells, whereas MDSC with normal lipid content did not ^[71][147].

2.3. ROS regulates MDSC-Mediated Immune Suppression

Signals such as ROS and NO have been well researched and documented to form signatures of human immune responses in oncology ^[72][149]. ROS are by-products of cellular oxidative metabolism ^[72][149]; in a healthy organism they help sustain an internal redox environment that balances free radicals produced by cellular antioxidants and enzyme systems ^[72][149], but during cancer they are produced by MDSC as a mechanism of suppressive function ^[73][74][150,151]. Studies by Nagaraj et al. highlighted that ROS and peroxynitrite production by MDSC has the potential to modify CD8⁺ T cells so that they lose their capacity to bind major histocompatibility complex (MHC) molecules and induce antigen-specific tolerance of peripheral CD8⁺ T cells ^[75][152]. By-products such as H₂O₂ formed from MDSC interaction with superoxide attenuate T-cell CD3ζ expression and consequently inactivate T cells and reduce IFN-γ expression ^[73][150]. In the lung, the initial encounter of alveolar macrophages and M. tb bacilli induces an oxidative burst ^[75][76][152,153] leading to the production of ROS that offers resistance to the growth of bacteria or other invading microorganisms ^[76][77][153,154]. Energy-generating pathways are a critical source of ROS, and Jian et al. showed that MDSC counteract OXPHOS via upregulation of glycolysis. They also identified the glycolytic metabolite, phosphoenolpyruvate (PEP), as a pivotal antioxidant that prevents excess ROS production by MDSC, resulting in protection from apoptosis ^[78][155]. Other signaling molecules such as HIF-1α and nuclear respiratory factor (NRF) are critical in conditions of hypoxia and pathways that modulate MDSC differentiation, allowing them to take a dual M1/M2 phenotype and mTOR-induced glycolytic activities ^[13][60][79][32,134,156] and contribute to immune cell reprogramming. In addition, Nrf2 promotes the expression of NAD(P)H-quinone oxidoreductase 1 (NQO1), the enzyme responsible for the reduction and detoxification of reactive quinones, and the enzyme responsible for the cleavage of heme into biliverdin, heme oxygenase of expression-1 (HMOX1) ^[80][157]. Interestingly, HMOX1 functions in part as an antioxidant but has also been shown to be anti-inflammatory, reducing IL-12 p40, IL-16, and TNF-α levels in dendritic cells and in allergic airways ^[81][158]. In contrast to HIF-1α, AMP-activated protein kinase (AMPK) exerts immunosuppressive function by inhibiting glycolysis through the PI3K–AKT–mTOR pathway. AMPK drives glycolysis toward OXPHOS during glucose metabolism. MDSC thrive in high-ROS environments and upregulate ROS production, which is one of the key ways they confer immuno-regulatory effects on immune cells ^[73][74][82][150,151,159]. It is possible, considering the similarities between the TME and the TB granuloma microenvironments, that the same mechanism may be present/active in the TB granuloma, where MDSC can accumulate and thrive in a ROS environment and enhance the suppressive effect of MDSC.