If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Cardiothoracic and Vascular Surgery, Deutsches Herzzentrum der Charité (DHZC), Berlin, GermanyCharité-Universitätsmedizin Berlin, corporate member of Freie Universität Berlin and Humboldt-Universität zu Berlin, Berlin, Germany
Department of Cardiothoracic and Vascular Surgery, Deutsches Herzzentrum der Charité (DHZC), Berlin, GermanyCharité-Universitätsmedizin Berlin, corporate member of Freie Universität Berlin and Humboldt-Universität zu Berlin, Berlin, Germany
Myeloid-derived suppressor cells (MDSCs) are naturally occurring leukocytes that develop from immature myeloid cells under inflammatory conditions that were discovered initially in the context of tumor immunity. Because of their robust immune inhibitory activities, there has been growing interest in MDSC-based cellular therapies for transplant tolerance induction. Indeed, various pre-clinical studies have introduced in vivo expansion or adoptive transfer of MDSC as a promising therapeutic strategy leading to a profound extension of allograft survival due to suppression of alloreactive T cells. However, several limitations of cellular therapies using MDSCs remain to be addressed, including their heterogeneous nature and limited expansion capacity. Metabolic reprogramming plays a crucial role for differentiation, proliferation and effector function of immune cells. Notably, recent reports have focused on a distinct metabolic phenotype underlying the differentiation of MDSCs in an inflammatory microenvironment representing a regulatory target. A better understanding of the metabolic reprogramming of MDSCs may thus provide novel insights for MDSC-based treatment approaches in transplantation. In this review, we will summarize recent, interdisciplinary findings on MDSCs metabolic reprogramming, dissect the underlying molecular mechanisms and discuss the relevance for potential treatment approaches in solid-organ transplantation.
Introduction
Although clinical solid-organ transplantation has improved short-term outcomes, the procedure still faces limitations due to substantial side effects of immunosuppression and compromised long-term organ survival [
]. Calcineurin inhibitors, as the backbone of immunosuppression, are associated with nephrotoxicity, cardiovascular and metabolic side effects, as well as increased risk of opportunistic infections and de novo malignancies [
]. Moreover, current first-line immunosuppressive drugs have shown limited effects on chronic rejection, impeding long-term graft survival even under strict follow-up and maintenance immunosuppression [
]. To further improve the quality and expectance of life of transplant recipients, innovative therapeutic solutions are required. Of note, inducing transplant tolerance has been considered as a desirable solution, as it enables organ acceptance without the need for long-term immunosuppression [
Myeloid-derived suppressor cells (MDSCs) are a heterogeneous population of immature myeloid progenitor cells with potent immunosuppressive activity, which has led the focus on investigating their therapeutic potential for allogeneic transplantation [
]. At the same time, MDSCs were shown to facilitate allograft tolerance through promoting the development and recruitment of regulatory T cells (Treg) [
]. Moreover, targeted depletion of MDSCs leads to cardiac allograft rejection in recipient mice that had been treated with donor splenocyte transfusion and anti-CD40L, known to prevent transplant rejection [
], underscoring the essential role of MDSCs for induction and maintenance of transplant tolerance. In human studies, expansion of MDSCs has been observed in peripheral blood of kidney transplant recipients [
Monocytic and promyelocytic myeloid-derived suppressor cells may contribute to G-CSF–induced immune tolerance in haplo-identical allogeneic hematopoietic stem cell transplantation.
Myeloid-derived suppressor cells increase and inhibit donor-reactive T cell responses to graft intestinal epithelium in intestinal transplant patients.
American journal of transplantation:.2018; 18: 2544-2558
]. Expanding the number of MDSCs and modifying their functional capacities to enhance their inhibitory effects on alloimmune responses is a promising approach for organ transplantation. The transduction of human monocytes with FADD-like interleukin (IL)-1β–converting enzyme-inhibitory protein (c-FLIP)-expressing lentivirus vectors, for instance, increased the suppressive activity on stimulated T cells and abrogated graft-versus-host disease in a xenograft model upon adoptive transfer, leading to improved long-term survival of recipient mice when compared with those receiving luciferase-expressing (control) monocytes. Strikingly, c-FLIP–transduced monocytes exerted more prominent suppression on graft-versus-host disease when compared with thawed human Tregs [
Monocytic and promyelocytic myeloid-derived suppressor cells may contribute to G-CSF–induced immune tolerance in haplo-identical allogeneic hematopoietic stem cell transplantation.
Myeloid-derived suppressor cells increase and inhibit donor-reactive T cell responses to graft intestinal epithelium in intestinal transplant patients.
American journal of transplantation:.2018; 18: 2544-2558
In addition to challenges due to their heterogeneous nature and limited expansion capacity, there are also transplant-specific challenges that need to be addressed. Common immunosuppressive drugs such as rapamycin (RPM), cyclosporine A (CsA) and tacrolimus interfere with complex metabolic and immunological signaling pathways affecting different cells, and MDSCs are no exception. Although CsA appears to promote the proliferation of MDSCs and thus enhance immunosuppression [
]. A detailed understanding of these interactions is needed for clinical application, but it seems worthwhile to investigate. MDSCs proliferate after transplantation and inhibit graft rejection, but durable graft tolerance has not been achieved in preclinical studies with MDSCs [
Given the advancements made in cancer research, cellular metabolic reprogramming represents a promising target for immunoregulation in transplantation [
]. Upon activation, immune cells reprogram the expression of critical metabolic enzymes and rewire metabolic pathways to overcome the bioenergetic and biosynthetic demand [
]. Notably, distinct features of different immune cells in metabolic reprograming provide the opportunity to target specific subsets, allowing tailored immunosuppression [
], which constitutes a major goal of current transplant research in order to lower systemic side effects and provide long-term immune tolerance. Indeed, several studies have reported on targeted immune cell reprogramming towards an anti-inflammatory phenotype, setting the cornerstones for further investigations of this therapeutic approach. M1 macrophages, for instance, have been skewed towards M2 macrophages using dimethyl fumarate, metformin or TEPP-46 [
Metformin inhibits the production of reactive oxygen species from NADH:Ubiquinone oxidoreductase to limit induction of interleukin-1β (IL-1β) and boosts interleukin-10 (IL-10) in lipopolysaccharide (LPS)-activated macrophages.
In transplantation, preclinical research has demonstrated the feasibility of metabolic inhibitors in prevention of transplant rejection and tolerance induction [
]. Most studies of these studies have focused solely on T-cell immunity as major drivers of allograft rejection. However, other critical immune cell components such has MDSCs remained largely unexplored. Although the metabolic phenotype and adaptations of MDSCs to hypoxic and inflammatory environments have been thoroughly characterized in cancer research, little effort has been made to translate this knowledge to the field of transplantation, where metabolic treatments are gaining increasing interest. Therefore, in the present review, we introduce MDSCs and their metabolic characteristics, dissect the underlying molecular mechanisms in MDSCs metabolic reprogramming and delineate the relevance for potential treatment approaches in solid organ transplantation.
Phenotype and Function of MDSCs
MDSCs in both human and mice are heterogeneous and can be classified into two major groups according to their origin from the granulocytic or monocytic myeloid cell lineages, monocytic MDSCs (M-MDSCs) and polymorphonuclear MDSCs (PMN-MDSCs) [
]. Both MDSC subsets have been characterized by expression of CD11b, whereas their phenotype can be distinguished according to the expression of distinct surface markers. Hence, in mice M-MDSCs display a CD11b+Gr-1midLy6ChiLy6G−CD49d+ phenotype whereas PMN-MDSCs are defined as CD11b+Gr1hiLy6ClowLy6G+CD49d−4,27. In humans, in turn, M-MDSCs are characterized as CD11b+CD14+HLA-DRlow/−CD15− and PMN-MDSCs as CD11b+CD14−CD15+(or CD66b+) [
Monocytic and promyelocytic myeloid-derived suppressor cells may contribute to G-CSF–induced immune tolerance in haplo-identical allogeneic hematopoietic stem cell transplantation.
Of note, human MDSCs also comprise another distinct group of myeloid precursor cells that are defined as Lin-HLA-DR-CD33+ early-stage MDSCs which, however, only account for up to 5% of the total MDSCs population [
] (Table 2). In transplantation, alterations in MDSC subsets are rarely reported. In a murine cardiac transplant model, most posttransplant splenic MDSCs had been CD11b+Ly6ClowLy6G+ PMN-MDSCs [
Table 2Phenotype and function of MDSC subsets in humans and mice.
M-MDSC
PMN-MDSC
e-MDSC
Mice
CD11b+Gr-1midLy6ChiLy6G−CD49d+
CD11b+Gr1hiLy6ClowLy6G+CD49d−
–
Human
CD11b+CD14+HLA-DRlow/−CD15−
CD11b+CD14−CD15+(or CD66b+)
Lin−(CD3/14/15/19/56)/HLA−DR−/CD33+
Cell-specific function
Generation of NO, Production of immunosuppressive cytokines (IL-10, TGF-β), Inhibitory checkpoint (PD-1)
Generation of ROS, Depletion of L-arginine and cysteine levels from the environment
Poorly understood
General function
Blocking of T cell homing, Promotion of Treg development, ectoenzyme-induced adenosine production, Fas/FasL-dependent apoptosis induction of cytotoxic T cells, Methylglyoxal-mediated effector T-cell paralyzation
Surface markers defining the phenotype of M-MDSCs, PMN-MDSCs and e-MDSCs in human and mice are shown. In addition, cell-specific function of each subset as well as general functions common in distinct subsets are outlined.
One of the key functions of MDSCs is T-cell suppression. Data from cancer research suggest that M-MDSCs and PMN-MDSCs share common mechanisms of immunosuppression but are somewhat different in their preference [
]. Hence, M-MDSCs have been shown to exhibit their major immunosuppressive effects through inducible nitric oxide synthase (iNOS), immunosuppressive cytokines and cell-surface molecules [
]. iNOS converts L-arginine to nitric oxide (NO), which impairs T-cell proliferation by inducing apoptosis, suppressing T-cell mitogenic responses or inhibiting major histocompatibility complex class II expression [
]. In addition, M-MDSCs up-regulate the expression of immunosuppressive cytokines such as IL-10 and transforming growth factor-β, as well as immune regulatory cell-surface molecules like membrane-bound programmed death receptor ligand 1 (PD-L1) [
]. PMN-MDSCs, in turn, have been characterized to exert their suppressive effects on immune responses promoting the generation of reactive oxygen species (ROS) and arginase 1. In detail, MDSCs express NADPH oxidase 2 (Nox2) catalyzing the production of ROS, whereas increased expression of arginase-1 causes a reduction of local L-arginine and cysteine levels, which constitute a key nutrition for T-cell proliferation and function [
]. Noteworthy, the described dichotomy does not imply an exclusive gene expression of either iNOS or arginase-1. Synergic expression of iNOS and arginase-1 has namely been characterized to constitute a cardinal feature of the immunosuppressive function of MDSCs, supporting the generation of reactive nitrogen species [
]. Reactive nitrogen species in turn impede protein–protein interactions and functions including chemotaxis, antigen recognition and activation of T cells [
]. In addition, MDSCs appear to mediate the development of Treg that suppress T-cell responses against transplanted grafts via interferon (IFN)-γ–dependent pathways [
]. IFN-γ seems to play a crucial role for tolerance induction in transplantation, as tolerance is not induced in IFN-γ–deficient recipient mice due to active CD28 and CD40 ligand T-cell co-stimulation pathways, resulting in the proliferation of alloreactive effector T lymphocytes [
]. Other mechanisms of MDSC-mediated immunosuppression that are mediated by both M-MDSCs and PMN-MDSCs include blockade of T-cell homing through reducing L-selectin expression on T cells [
]. Notably, a recent study showed a novel mechanism in which MDSC-derived methylglyoxal can be transferred to T cells via cell–cell transfer. Subsequently, methylglyoxal then paralyzed T-cell effector function through metabolic inhibition of these cells [
Metabolic Reprogramming Orchestrating the Immunosuppressive Effects of MDSCs on T-Cell Immunity After Transplantation
Metabolic reprogramming is a general term initially derived from cancer research for the response of cells to crucial changes in the surrounding microenvironment, involving altered expression in critical metabolic enzymes, metabolites and metabolic pathways, which in turn leads to altered functions [
] can be observed deriving from ischemia–reperfusion injury (IRI), shear forces and allo-immune responses, which in turn affect the metabolic phenotype and function of immune cells, including MDSCs.
MDSCs rapidly expand after transplantation exerting immunosuppressive effects through mechanisms described previously [
]. Triggered by external stimuli, metabolic reprogramming of MDSCs hereby alters the acquisition and processing of nutrients in order to meet the requirement of augmented energy supply and biosynthetic intermediates [
]. Oxidate stress for instance, can be recognized by general control nonderepressible 2 (GCN2) acting as an environmental sensor crucially involved in the instigation of MDSC metabolic reprogramming through inducing the upregulation of ATF4, a transcription factor that regulates the expression of genes involved in amino acid metabolism, oxidative stress and the unfolded protein response [
Potent GCN2 inhibitor capable of reversing MDSC-driven T cell suppression demonstrates in vivo efficacy as a single agent and in combination with anti-angiogenesis therapy.
]. During metabolic reprogramming glucose, lipid and amino acid metabolism undergo significant alteration tightly regulating MDSCs effector pathways (Table 3).
Table 3Overview of metabolic pathways involved in MDSC-derived immunosuppressive function on T cells.
Metabolic pathway/product
Regulated effector function
A
Glucose metabolism
HIF-1a
Differentiation and proliferation under hypoxia
PD-L1 expression
Inhibition of CD4 proliferation
Inhibition of CD8 IFN-γ production
Sirt1/HIF-1α
M2 polarization
Il-10 and TGF-β expression
Phosphoenolpyruvate
Restraining ROS self-damage
B
Lipid metabolism
FAO
FA, AA, rbose
IL-10, PD-1, iNOS expression
FAT 2/CD36
CD8+ suppression
C
Amino acid metabolism
Arginase/iNOS
Arg depletion
Decreasing CD3 Expression
NO release
Inhibition of T-cell proliferation
Blocking T-cell Trafficking
IDO/Kyn
Trp depletion
Cell cycle arrest of T cells in G1
Inhibition of T-cell–derived IFN-γ expression
Increasing Treg frequencies
Kyn, 3-HK, 3-HAA
Suppression of T-cell proliferation
QUIN
TH2 polarization
Inhibition of T-cell IFN-γ and TNF-α production
3005.66
3005.61
2150
2260
2003.74
6011.22
6011.22
4300
2260
2003.74
33011.19
Overview of metabolic pathways and their downstream metabolite products of (A) glucose metabolism (B) lipid metabolism and (C) amino acid metabolism. Every subtable lists the regulated effector function of MDSCs on T cells regulated by the reflective metabolic pathway.
]. Oxygen-rich environments cause immune cells including quiescent naïve T cells, B cells and monocytes to prioritize glucose flux for maximum ATP production through mitochondrial oxidation [
] recruit aerobic glycolysis to generate energy and biosynthetic intermediates within a shorter time period when compared with oxidative phosphorylation.
Notably, dynamic metabolic flux analysis revealed that MDSCs recruit glycolysis as the major metabolic pathway of glucose regardless of oxygen availability and activation status [
]. Thus, MDSCs were found to upregulate glycolytic genes, increase glucose uptake and reduce oxygen consumption during differentiation and activation [
]. Thereby, hypoxia-inducible-factor-1-alpha (HIF-1α) constitutes a key transcriptional factor that regulates the expression of glycolytic enzymes and promotes anaerobic glycolysis in MDSCs [
]. In addition, the mammalian target of rapamycin (mTOR) pathway has also been identified to increase glucose uptake and glycolytic enzyme expression [
] and during organ transplantation, limited arterial blood supply causes an ischemic state with tissue hypoxia that induces an anaerobic metabolism with increasing lactate levels and ROS production [
]. HIF-1α, upregulated in response to tissue hypoxia, has thereby been delineated as a cardinal factor dampening the pathological consequences of IRI inevitably accompanying organ transplantation [
]. HIF-1α has namely been shown to improve mitochondrial respiratory function through the phosphoinositide 3-kinase/AKT and Janus kinase 2/STAT3 pathways [
]. Notably, HIF-1α has also been characterized as a key transcriptional factor that regulates the expression of glycolytic enzymes as well as lactate transporters and promotes glycolysis in MDSCs [
], thus accelerating differentiation and proliferation under organ hypoxia. In addition, HIF-1α has been shown to increase the expression of PD-L1, an immune checkpoint receptor, which mediates the inhibition of T cells by binding to the PD-L1 receptor of T cells. Moreover, acidic microenvironment caused by hypoxia has been characterized to increase the proliferation of MDSCs and strengthen their suppressive activity on T cells, compromising CD4+ T-cell proliferation and CD8+ T-cell–derived IFN-γ production via the HIF-1α pathway [
]. Furthermore, local tissue hypoxia activates SIRT1 in MDSCs, which in turn exerts agonistic effects with HIF-1α deacetylating several of its downstream targets, thus mediating the differentiation of MDSCs towards a regulatory and immunosuppressive M2-phenotype. Of note, SIRT1 deficiency was found to impair M2-MDSCs with compromised glycolytic activity and lower arginase activity as well as IL-10 and TGF-β1 expression. These studies indicate that hypoxia/HIF-1α signaling is a potent inducer of immunosuppression by enhancing glycolysis in MDSCs. Of note, aerobic glycolysis does not only work as an energy generation pathway for MDSCs, but also constitutes a critical regulator of redox homeostasis. Thus, phosphoenolpyruvate generated through augmented aerobic glycolysis, for instance, has been identified as glycolytic intermediate with antioxidant activity that restrains self-damage by ROS generated during the effector response of MDSCs [
], which may also interfere with the impact of exogenous ROS, deriving for instance from IRI following transplantation, on MDSCs (Table 2, A).
Lipid Metabolism
Lipid metabolism constitutes a cardinal pathway for the suppressive function of MDSCs on the local microenvironment. The major pathway of cellular lipid metabolism is fatty acid β-oxidation (FAO). Studies in cancer showed that MDSCs, upon infiltrating the tumor and becoming activated, increased the uptake of lipids by the upregulation of the scavenger receptor CD36 through STAT3 and STAT5 signaling [
]. Peroxisome proliferator–activator receptors signaling pathway sense the increased level of fatty acids and regulate downstream lipid metabolic pathways [
]. Thus, mitochondrial electron transfer complex and the key enzymes in tricarboxylic acid cycle are upregulated, promoting FAO and TCA. These metabolic reprogramming events shift the primary source of energy generation from glycolysis to FAO. Indeed, compared with peripheral MDSCs, tumor-infiltrating MDSCs showed increased mitochondrial mass, oxygen consumption and FAO-derived ATP generation [
]. Notably, in MDSC, FAO also displays the primary source of biosynthetic intermediates for the generation of FA, AA and ribose, which are essential for further synthesis of anti-inflammatory molecules such as IL-10, PD-1 and iNOS. IL-10, in turn, has been dissected to exert strong immunosuppressive effects on allo-immune responses following transplantation [
]. Of note, FAO also serves as the cardinal resource for the expression of pro-inflammatory cytokines such as IL-2. However, MDCSs have been delineated to display a primarily anti-inflammatory secretome in transplantation with IL-10 representing a cardinal cytokine [
]. The lipid metabolism of MDSCs has furthermore been associated with a suppressive function on dendritic cells (DCs). Accumulated lipids in PMN-MDSCs are hereby oxidized through ROS and MPO and subsequently transferred to DCs inhibiting their antigen cross-presentation and their orientation on major histocompatibility complex class II in cancer [
]. This, in turn, blocks antigen-mediated cross-presentation and inhibits T-cell stimulation, which is of translational relevance for transplantation as antigen cross-presentation represents a crucial step in the initiation of T-cell allo-immune responses [
]. Of note, extracellular oxidized lipids can also be taken up by T cells through CD36, subsequently leading to T-cell exhaustion mediated by p38 signaling [
]. Clearance of cellular oxidized lipids in T cells by overexpression of glutathione peroxidase 4, in turn, has been shown to restore the effector function in vivo [
]. Supporting the importance of lipid metabolism for the immunosuppressive function of MDSCs was the observation that an inhibition of FA uptake blocking fatty acid transporter 2 [
] through pharmacological inhibition or genetic deletion, abrogated the suppressive effects of MDSCs on CD8+ T-cells (Table 2, B).
Amino Acid (AA) Metabolism
AA metabolism represents a critical mechanism of MDSC effector function mediating T-cell immunosuppression since arginine, tryptophan and cysteine are essential amino acids for T-cell function. MDSCs can deplete these AA from the microenvironment and thus starve T cells, leading to impaired T-cell survival, proliferation and function. Moreover, several products of AA metabolism inhibit T-cell immune responses through different signal pathways.
MDSCs metabolize arginine though arginase-I and iNOS. Upon activation through IFN-γ, MDSCs have been shown to take up large amounts of Arg by inducing the cationic AA transporter 2, Arg-1 and iNOS. The reduction of local Arg levels in turn leads to decreased T-cell–derived CD3ζ expression, thus compromising T-cell antigen-specific proliferation [
]. Similarly, adoptive transfer of tumor necrosis factor-α–induced MDSCs prevented allograft rejection in male-to-female skin transplants through inhibiting T-cell proliferation that was abrogated in the presence of an NO-inhibitor or following genetic-depletion of NO [
]. Notably, PMN-MDSCs and M-MDSCs display divergent features of Arg metabolism with PMN-MDSCs expressing greater levels of Arg-1, and M-MDSCs exhibiting higher levels of iNOS [
Indoleamine 2, 3-dioxygenase (IDO) is the key enzyme of tryptophan metabolism in the kynurenine (Kyn) pathway. MDSCs have been shown to exhibit an increased IDO expression and augmented Kyn pathway causing a metabolization of Trp from the surrounding microenvironment associated with decreased Trp levels [
Myeloid-derived suppressor cells suppress antitumor immune responses through IDO expression and correlate with lymph node metastasis in patients with breast cancer.
], thus compromising T-cell–derived immune responses. Indeed, IDO expression was found to be critical, inhibiting T-cell–driven allo-immune responses with abrogated long-term cardiac allograft survival in CTLA4-Ig–treated mice deficient for IDO [
] in mice through inhibiting T-cell–derived IFN-γ expression and increased Treg frequencies.
Noteworthy, recent studies have demonstrated an entwined pathway of arginine and tryptophan immunometabolism, leading to immunosuppressive function of DCs, which can be mediated by Arg1+-MDSCs. Hereby, IDO1 phosphorylation and consequent activation of immunosuppressive, non-enzymatic IDO1 signaling in DCs was dependent on previous expression of Arg1 and Arg1-dependent production of polyamines which in turn can be released by MDSCs [
]. Thus, joint regulation of arginine and tryptophan metabolism could represent an important target for regulating immunomodulatory effects of MDSCs.
Of additional relevance, downstream products of the Trp metabolism display bioactive compounds that further restrains T-cell derived immune response. Thus, Kyn, 3-hydroxykynurenine and 3-hydroxyanthranilic acid (3-HAA) were shown to strongly suppress T-cell proliferation in mixed leukocyte reactions leading to a prolonged rat skin allograft survival [
Studying the immunosuppressive role of indoleamine 2,3-dioxygenase: tryptophan metabolites suppress rat allogeneic T-cell responses in vitro and in vivo.
]. Similarly, 3-HAA also prevented the rejection of rat cardiac allografts diminishing the proliferation of T cells that could not be restimulated by donor-specific DCs [
], Moreover, 3-hydroxykynurenine and 3-HAA were shown to compromise T-cell proliferation, thus preventing T-cell–induced kidney allograft rejection and protect the tubular epithelial cell from injury in pigs [
], has been shown to shift the human immune balance towards a regulatory Th2-dominated phenotype while inhibiting Th1-cell–derived INF-γ and tumor necrosis factor-α production through agonistic effects on D-methyl-D-aspartate receptors [
Effects of Immunosuppressants on MDSC Metabolic Reprogramming
The calcineurin inhibitors, CsA and tacrolimus, constitute the backbone of the current first-line immunosuppressive regimen. Experimentally, in vitro and in vivo data from murine transplant models showed that CsA promoted MDSCs proliferation and enhanced their immunosuppressive function on alloreactive T cells [
]. Hereby, the calcineurin-NFATc pathway is the primary target of CsA regulating AA metabolism of MDSCs through regulating IDO and iNOS. Notably, inhibition of NFAT by CsA in MDSCs up-regulated IDO enhancing Trp metabolism as well as production the downstream immunosuppressive metabolite KYNA (Figure 1A). Thus, CsA treated MDSCs compromised both CD4+- as well as CD8+-T-cell–derived INF-γ production, increased Treg differentiation and promoted allograft survival when compared to PBS treated MDSCs in a model of murine skin transplantation [
Figure 1Treatment approaches and targets of metabolic reprogramming of MDSCs in transplantation. The outlined three metabolic pathways involved in the immunosuppressive function on T cells (TCR signaling, proliferation and IFN-y) and proliferation/differentiation of MDSCs can be upregulated through diverse approaches in the context of transplantation while immunosuppressive drugs may cause significant interference. AA metabolism: (A) Cyclosporin inhibits NFATc, which revokes its inhibitory effect on AA metabolism and promotes IDO and iNOS expression. (B) Rapamycin inhibits IDO expression and mTOR signaling, thus restraining glycolysis. (C) Glucocorticoids promote iNOS expression. Glycolysis: (D) Treatment with metformin exerts antagonistic effects on mTOR as rapamycin, promoting glycolysis both in vitro and in vivo. (E) Exogenous lactate increases MDSCs glycolysis and proliferation in vitro. (F) Following transplantation normoglycemia could ensure sufficient glycolysis of MDSCs. (G) Hypoxic preconditioning of MDSCs during cell culture or organs during ex vivo organ perfusion may promote HIF-α–mediated glycolysis. (H) HIF-α agonists can be recruited during in vitro cell culture of MDSCs or during ex vivo organ perfusion. Fatty acid oxidation: (I) Subjecting MDSCs to very low-density lipoproteins (VLDL) in vitro could augment FAO, leading to increased immunosuppressive effects on T cells. (J) Targeting STAT3 through optimizing culture conditions with different cytokines could augment FAO via increased lipid uptake. (K) Inhibiting fatty acid synthesis for instance with soraphen A may cause a compensatory upregulation of fatty acid uptake. (L) Viral transduction to artificially increase CD36 expression may promote lipid uptake. (M) Supplementing lipids as a diet to transplant recipients or during ex vivo organ perfusion may promote lipid uptake in MDSCs.
Moreover, CsA treatment of MDSCs increased levels of iNOS which convert Arg to immunosuppressive NO, thus restraining CD8+- and CD4+ T-cell differentiation while prolonging survival of murine skin grafts [
] (Figure 1A). Regarding tacrolimus, MDSCs isolated from tacrolimus-treated lung transplantation recipients were able to inhibit T-cell proliferation in vitro [
], RPM has been found to regulate innate immune homeostasis also impacting MDSCs in transplantation. Thereby, RPM inhibited MDSC differentiation, proliferation and immunosuppressive effects on T cells through directly impeding glycolysis with down-regulated glycolytic enzymes hexokinase 2, phosphofructokinase 1, pyruvate kinase muscle as well as the transporter Glut1, both in vitro in sorted granulocyte-macrophage (GM)-colony-stimulating factor (CSF)–induced M-MDSCs and in vivo in induced M-MDSCs of AlloSkin-grafted mice [
] (Figure 1B). Similarly, genetically altered mice deficient for mTOR also showed decreased levels of M-MDSCs in skin-grafts as well as draining lymph nodes in addition to compromised inhibitory effects on CD4+- as well as CD8+-T-cell proliferation. Moreover, RPM also inhibited iNOS and Arg-1 as well as HO-1, IDO and NOX2 expression underlying the impaired immunosuppressive function of MDSCs on T-cells [
] (Figure 1B). Consistent with these experimental data, a clinical study reported that MDSCs isolated from kidney transplant recipients that had been treated with RPM failed to inhibit T-cell proliferation through decreased iNOS activity [
Myeloid-derived suppressor cells increase and inhibit donor-reactive T cell responses to graft intestinal epithelium in intestinal transplant patients.
American journal of transplantation:.2018; 18: 2544-2558
] (Figure 1C). In a recent study, this effect was further attributed to an inhibition of HIF-1α–driven glycolysis upon activation of MDSCs. However further insights on metabolic changes such as the lipid metabolism that is known to display the major energy source upon MDSCs activation had not been explored [
Costimulatory receptors are a class of molecules expressed by T cells regulating activation and generation of effector T-cell responses. Blockade of costimulatory signals represents a promising strategy to control T-cell responses. As the most recognized molecule, CTLA4-Ig has been shown to dampen activation of naïve T cells by blocking costimulation via CD28 [
]. Notably, in a randomized controlled trial in patients with advanced-stage melanoma, CTLA4-Ig treatment was found to decrease PMN-MDSC levels and compromise Arg1expresssion [
Ipilimumab treatment results in an early decrease in the frequency of circulating granulocytic myeloid-derived suppressor cells as well as their Arginase1 production.
]. However, the underlying mechanism of this effect remained unclear.
Metabolic Optimization of MDSCs for Transplantation
Besides inhibiting the rejection of allografts following transplantation, mechanistic animal studies have further delineated MDSCs as essential component for the development and maintenance of immune tolerance [
]. Thus, MDSCs have emerged as a novel, cellular treatment opportunity to improve outcomes of transplant recipients. However, preclinical studies with MDSCs failed to achieve durable transplant tolerance [
] and haven't been any reports concerning MDSCs infusion in human transplant recipients.
Since metabolic reprogramming has been identified to crucially mediate differentiation, proliferative expansion and the immunosuppressive effects of MDSCs, manipulating or supplementing cardinal metabolic pathways may augment MDSCs function and improve treatment approaches. As a proof of concept, targeting metabolic reprogramming in other cellular therapies could improve the survival and expansion of in vitro–cultured cells [
Unlike cancer research, which aims to inhibit the suppressive function of MDSCs and enhance T-cell–derived anti-tumor immune responses through metabolic reprogramming, transplantation requires the opposite approach. Targeting MDSCs metabolism during transplantation can be achieved in vitro before adoptive transfer by altering the culture conditions or by genetic modification. Moreover, in vivo drug treatments and diet alterations could be recruited while ex vivo organ perfusion offers a treatment approach between ex- and implantation.
As described previously, MDSCs are highly dependent on increasing their glycolytic activity to differentiate and expand. In a tumor derived microenvironment hypoxia and excessive lactate levels force adjacent cells to adapt their metabolism with MDSCs upregulating their aerobic glycolysis. Thus, promoting MDSCs derived glycolysis in vitro may increase proliferation and survival before cellular therapy. Indeed, subjecting MDSCs to metformin that enhances the glycolytic pathway through augmenting glucose uptake while compromising mitochondrial respiration chain activity through an 5' AMP-activated protein kinase-dependent and -independent pathway [
]. This mechanism has also been targeted in vivo with translational relevance for clinical application. Namely, allograft-transplanted mice subjected to RPM displayed reduced MDSC frequencies with impeded immunosuppressive effects on T cells through compromised glycolysis, which was reversed upon treatment with metformin [
] (Figure 1D). Of note, metformin has been shown to improve immune homeostasis through reduced Th17-cell and increased Treg frequencies in kidney transplant recipients treated with tacrolimus [
]. Exogenous lactate as a product of glycolysis has been shown to increase the percentage of MDSCS derived from in vitro–cultured mouse bone marrow cells stimulated with GM-CSF and IL-6, which may improve culture conditions to expand MDSCs generation (Figure 1E). At the same time, MDSCs have also been shown to depend on sufficient glucose supply in order to exert glycolytic activity in vivo. Namely, depletion of glucose levels using a ketogenic diet to lower lactate production by glycolytic tumors resulted in smaller tumors, decreased MDSC frequency, and improved T-cell immune responses [
]. Diets for patients following transplantation should therefore ensure sufficient glucose supply with normoglycemia (Figure 1F). Notably, further studies have delineated stronger immunosuppressive capacity of MDSCs when stimulated with lactate via the HIF-1α pathway [
]. HIF-1α can be induced in vitro through hypoxic preconditioning of cells, thus providing another opportunity to boost MDSCs differentiation and function (Figure 1G). Preconditioning of mesenchymal stroma cells in hypoxic conditions prior to transplantation for instance, has been shown to improve cell survival through a metabolic switch towards a more glycolytic state mediated by HIF-1α activation [
]. Moreover, augmenting MDSCs derived HIF-1α–mediated glycolysis could also be performed through pharmacologically stabilization. In support, treating mice subjected to IRI with GSK360A that inhibits the prolyl hydroxylase domain-containing-enzyme led to an increased expression of HIF-1α–regulated genes such as pyruvate dehydrogenase kinase-1 and hexokinase II, thus comprising oxidative stress (Figure 1H). Equitable results were achieved in vitro treating a cardiac cell line [
]. These drugs could either be applied in vitro to MDSCs before cellular therapy to boost differentiation and proliferation; to transplant recipients potentially exerting dichotomous, beneficial effects on both, IRI and MDSC function; or during ex vivo organ perfusion. Thereby, allograft resident MDSCs could be manipulated either through decreasing local oxygen delivery or treating organs with HIF-1α agonistic drugs and via increasing HIF-1α levels (Figure 1G and 1H).
At least, increasing FAO of MDSCs through altering culture conditions or genetic modification of MDSCs during in vitro expansion could display another approach to improve their immunosuppressive function. Strikingly, subjecting murine and human MDSCs to very low-density lipoproteins promoted iNOS expression and augmented the suppressive effects on T-cell proliferation [
]. Thus, augmenting MDSCs FAO could be orchestrated by expanding MDSCs in the presence of very low-density lipoprotein, which may also promote the immunosuppressive function of these cells (Figure 1I). Moreover, culture conditions could be further optimized through supplementing a combination of GM-CSF, G-CSF, IL-4, IL-6, IL-10 and vascular endothelial growth factor that are recognized as the most important inducers of STAT3 in MDSCs [
], thus inducing CD36 expression and improving lipid uptake (Figure 1J). Furthermore, inhibiting the de novo synthesis of fatty acids, conversely increasing FAO, with small molecules such as soraphen A that impedes the acetyl-CoA carboxylase 1 and 2 may present an additional option (Figure 1K). Of note, disruption of fatty acid synthesis has been shown to prevent acute graft-versus-host diseases with diminished T-cell proliferation and augmented Treg frequencies [
]. At least genetic manipulation of MDSCs through lentiviral transduction could be employed to induce durable CD36 expression that may improve lipid uptake in vivo after adoptive transfer (Figure 1L). Supplementation of fatty acids to either transplant recipients or allografts during ex vivo organ perfusion may constitute a clinically implementable therapeutic approach to increase FA levels and improve recipient or donor derived MDSCs function, respectively (Figure 1M).
Concluding Remarks
Metabolic reprogramming constitutes a mechanistic underpinning of MDSCs immunosuppressive effector functions in the microenvironment of solid malignant tumors. From a clinical perspective, the detrimental immunosuppressive effects of MDSCs in cancer translate into beneficial effects in transplantation regulating allo-immune responses and prolonging allograft survival. Metabolic reprogramming of MDSCs may thus also be crucially affected in the inflammatory microenvironment of allografts, IRI and immunosuppressive drugs. However, thus far experimental studies investigating the relevance of metabolic reprogramming of MDSCs in the context of transplantation have been limited, and further investigations are required to confirm the outlined mechanisms in both pre-clinical experimental models and human studies. Moreover, despite axiomatic similarities, the tumor microenvironment differs from that of a transplanted organ undergoing IRI or acute rejections episodes, which needs to be considered when evaluating the transferability of mechanistic insights deriving from cancer studies.
With a view on future perspectives, the metabolic characteristics of MDSCs in transplantation require further investigation to evaluate the potential of targeting those pathways for tolerance induction. Therefore, experimental animal transplant models with integrative omics combining transcriptomics, proteomics and metabolomics could serve to evaluate the expression of metabolic enzymes as well as metabolite levels and evaluate the proportional interplay of metabolic pathways during metabolic reprogramming in the context of transplantation.
Targeting metabolic reprogramming of MDSCs either in vitro, in vivo or during ex vivo organ perfusion may ultimately provide novel treatment approaches to improve MDSCs derived tolerance induction or MDSCs based cell therapy approaches.
Funding
Jasper Iske was supported by the Junior Clinician Scientist Program of the Berlin Institute of Health. Yeqi Nian was supported by a grant (82101874) from the National Natural Science Foundation of China and an internal grant from Tianjin First Central Hospital.
Declaration of Competing Interest
The authors have no commercial, proprietary or financial interest in the products or companies described in this article.
Author Contributions
Conception and design of the study: JI and YN. Drafting and revising manuscript: YC, MR, SZS. All authors have approved the final article.
References
Leventhal J.R.
Mathew J.M.
Outstanding questions in transplantation: tolerance.
American journal of transplantation.2020; 20: 348-354
Monocytic and promyelocytic myeloid-derived suppressor cells may contribute to G-CSF–induced immune tolerance in haplo-identical allogeneic hematopoietic stem cell transplantation.
Myeloid-derived suppressor cells increase and inhibit donor-reactive T cell responses to graft intestinal epithelium in intestinal transplant patients.
American journal of transplantation:.2018; 18: 2544-2558
Metformin inhibits the production of reactive oxygen species from NADH:Ubiquinone oxidoreductase to limit induction of interleukin-1β (IL-1β) and boosts interleukin-10 (IL-10) in lipopolysaccharide (LPS)-activated macrophages.
Potent GCN2 inhibitor capable of reversing MDSC-driven T cell suppression demonstrates in vivo efficacy as a single agent and in combination with anti-angiogenesis therapy.
Myeloid-derived suppressor cells suppress antitumor immune responses through IDO expression and correlate with lymph node metastasis in patients with breast cancer.
Studying the immunosuppressive role of indoleamine 2,3-dioxygenase: tryptophan metabolites suppress rat allogeneic T-cell responses in vitro and in vivo.
Ipilimumab treatment results in an early decrease in the frequency of circulating granulocytic myeloid-derived suppressor cells as well as their Arginase1 production.
00104-4&id=gr1.jpg){kind=link}