Metabolic Checkpoints in Neurodegenerative T Helper 17 (TH17) and Neuroregenerative Regulatory T (treg) Cells As New Therapeutic Targets for Multiple Sclerosis.
Hongxing Shen,James A Bonner,Lewis Zhichang Shi
DOI: https://doi.org/10.4103/1673-5374.265552
2020-01-01
Neural Regeneration Research
Abstract:The central nervous system (CNS) is an immune-privileged site with tightly-regulated immune responses, a concept proposed by Nobel Laureate Sir Peter Medawar in 1960. Under physiological conditions, only a few T lymphocytes conducting immunosurveillance can infiltrate the CNS. However, in neurodegenerative pathology such as multiple sclerosis (MS), a devastating inflammatory demyelinating disease, transmigration of encephalitogenic T cells from the periphery to the CNS is evident, contributing to the etiology of MS. Among the encephalitogenic T cells, in recent years, interleukin-17-producing T helper 17 (TH17) cells have attracted intensive research interests because of their superior ability to induce MS, as compared to the interferon-γ-producing T helper 1 (TH1) cells that had long been regarded as the primary culprit in the pathogenesis of MS, prior to the discovery of TH17 cells in 2005. An early study showed that C-C chemokine receptor 6 (CCR6)+ TH17 cells are the first encephalitogenic T cells to infiltrate the CNS, which leads to the second wave of infiltration by other neuroinflammatory immune cells including TH1 cells (Reboldi et al., 2009). These T cells drive the development of clinical signs of MS. Conversely, forkhead box P3 (Foxp3)+ regulatory T (Treg) cells suppress TH17 and TH1 cells, mitigating neuroinflammation in MS. Intriguingly, a recent study showed that Treg cells can actually promote remyelination, a key event in neural regeneration (Dombrowski et al., 2017). Analogous to the well-established dichotomy of TH1 and interleukin-4-producing T helper 2 cells, TH17 and Treg cells are considered dichotomous cell fates of activated CD4+ T cells, because of the intimate physical and functional interactions of their master transcriptional factors (retineic-acid-receptor-related orphan nuclear receptor-γ (RORγt) and Foxp3) and the dynamic two-way transdifferentiation between them in various pathogenic conditions. While immune signals (TCR ligation, CD28-mediated co-stimulation, and cytokine) seemingly orchestrate the differentiation of these T cell subsets, an emerging frontier in immunology research is the realization that at a fundamental level, it is the cellular metabolism that dictates the T cell fate decisions including TH17 vs. Treg. To this end, TH17 cells primarily engage glycolysis and fatty acid synthesis (FAS), while Treg cells mainly use oxidative phosphorylation and fatty acid oxidation (FAO) to satisfy their bioenergetic and biosynthetic demands. This suggests that targeting these metabolic pathways may offer a legitimate approach to cure MS by suppressing neurodegenerative TH17 and concomitantly promoting neuroregenerative Treg cells. Thus, in this perspective, we focus on metabolic checkpoints capable of tipping the TH17/Treg balance in favor of Treg formation. Using mice with T cell-specific deletion of hypoxia-inducible factor 1α subunit (HIF1α), we were among the first to show that loss of HIF1α, a downstream target of the mechanistic target of rapamycin (mTOR), dampens TH17 development and concurrently augments Treg induction (Shi et al., 2011). Consistent with a prominent role of HIF1α in regulating glycolysis, polarizing HIF1α–/– TH17 cells have significantly reduced glycolytic activity, coupled with profound downregulation of a spectrum of essential glycolytic molecules including Glut1 (the dominant glucose transporter on T cells) and hexokinase 2 (the first rate-limiting enzyme in glycolysis). Moreover, inhibition of glycolysis with 2-DG (an inhibitor of all hexokinases) phenocopied effects of HIF1α deficiency, including attenuated progression of experimental autoimmune encephalomyelitis (EAE), a well-established murine model for MS (Shi et al., 2011). Pyruvate dehydrogenase (PDH) is the key bifurcation enzyme in the glycolytic pathway, under tight regulation of PDH kinases (PDHKs) and PDH phosphatases. In its active dephosphorylated form mediated by the activity of PDH phosphatases, PDH drives pyruvate into mitochondria for further oxidation and adenosine triphosphate generation through the tricarboxylic acid cycle. In its inactive phosphorylated form controlled by the action of PDHKs, it prevents pyruvate entry into the mitochondria and instead feeds pyruvate to lactate fermentation to sustain the high-flow of the glycolytic pathway. As such, inhibition of PDHKs by dichloroacetate and consequently reduced glycolysis, led to suppression of TH17 differentiation and promotion of Treg induction, as well as amelioration of EAE (Gerriets et al., 2015), in line with our findings (Shi et al., 2011). More recently, Li et al. (2019) found that CG-5, an inhibitor of glucose transporters including Glut1, also suppressed TH17 differentiation and enhanced Treg induction. Collectively, these studies indicate that suppressing the mTOR-HIF1α-glycolysis axis, by favoring the formation of Treg cells, may represent an effective approach to cure MS. In support of this notion, administration of metformin, which indirectly inhibits activation of mTOR and HIF1α, led to reduced TH17 and increased Treg, attenuating the clinical signs of EAE (Sun et al., 2016). Mechanistically, metformin activates the adenosine monophosphate activated protein kinase (AMPK), a downstream target of the liver kinase B1, which then phosphorylates and activates the TSC1/2 complexes that negatively regulate mTORC1. In addition, AMPK activation by metformin boosts FAO, a catabolic process that favors Treg induction and impairs TH17 differentiation. Notably, metformin is a commonly-prescribed drug to treat type 2 diabetes mellitus, providing an easy-to-adopt translational approach to test this idea in clinical trials of patients with MS. Unlike Treg cells catabolizing already-available extracellular fatty acids to fuel their metabolic needs, TH17 cells utilize de novo FAS to sustain their proliferative and functional properties. The de novo FAS is under a delicate regulation of sterol response element binding proteins and liver X receptors, with peroxisome proliferator-activated receptors acting as coactivators of liver X receptors. Of note, the mTOR signaling can activate sterol response element binding proteins, linking the mTOR-HIF1α-glycolysis axis to de novo FAS to construct the glycolytic-lipogenic network. Acetyl-CoA carboxylase 1 (ACC1) catalyzes carboxylation reaction of acetyl-CoA to malonyl-CoA, which is the first step in FAS. Compellingly, deletion of ACC1 in T cells and treatment with an ACC1 pharmacological inhibitor soraphen A drastically blocked TH17 development and drove differentiation towards a functional Treg phenotype (Berod et al., 2014), supported by Treg-favoring uptake of exogenous fatty acids and FAO. Other than increased glucose uptake, activated T cells also markedly increase utilization of glutamine, the most abundant amino acid in serum. Once transported into the cells, glutamine is first hydrolyzed to glutamate by glutaminase, followed by conversion to α-ketoglutarate and concurrent release of the amino group. In differentiating TH17 and Treg, the amino group primarily fuels biosynthesis of aspartate in a reductive amination reaction that is catalyzed by glutamate oxaloacetate transaminase 1 and 2 (GOT1 and GOT2) (Xu et al., 2017). Using a GOT1 selective inhibitor (aminooxy)acetic acid (AOA), Xu et al. (2017) elegantly demonstrated that suppression of GOT1 led to reduced formation of α-ketoglutarate, which can be metabolized to 2-hydroxyglutarate in an error-prone dehydrogenase-mediated reaction. Consequently, accumulation of 2-hydroxyglutarate is significantly decreased in AOA-treated TH17 cells. This in turn curtails hypermethylation of Foxp3 gene locus and augments Foxp3 expression, exerting greater antagonistic effects on RORγt and repressing TH17 differentiation. A net outcome would be a tilted TH17/Treg balance towards Treg that delays disease progression in AOA-treated EAE mice (Xu et al., 2017). Furthermore, AOA induces phosphorylation of AMPK in polarizing TH17 cells, a trademark of Treg cells, pointing to an interconnection between the glutamate-α-ketoglutarate-2-hydroxyglutarate axis and the AMPK-FAO pathway. These identified metabolic regulators with opposing effects on TH17 and Treg are summarized in Figure 1. Although they are depicted in isolation, as discussed here, intimate interactions exist among them.Figure 1: The four metabolic checkpoints in the TH17/Treg balance.Blue triangles indicate favoring effects towards TH17 cells; red triangle indicates promoting effects towards Treg. 2-DG: 2-Deoxy-D-glucose; 2-HG: 2-hydroxyglutarate; 2-KG: 2-ketoglutarate; ACC1: acetyl-CoA carboxylase 1; AMPK: adenosine monophosphate activated protein kinase; AOA: (aminooxy)acetic acid; CG-5: a novel glucose transporter inhibitor; DCA: dichloroacetate; FAO: fatty acid oxidation; FAS: fatty acid synthesis; Gln: glutamine; Glut: glucose transporter; LKB1: liver kinase B1; LXRs: liver X receptors; mTOR: mechanistic target of rapamycin; PPARs: peroxisome proliferator-activated receptors; SorA: soraphen A; SREBPs: sterol response element binding proteins; TH17 cells: T helper 17 cells; Treg cells: regulatory T cells.The “immune privilege” status of the CNS is maintained by brain barriers such as the blood-brain barrier (BBB) and the blood-cerebrospinal fluid barrier (BCB) (Figure 2). The tight junctions between adjacent endothelial cells in BBB and neighboring choroidal epithelial cells in BCB (located in the choroid plexus) limit the entry of solutes and ions, as well as circulating immune cells into the CNS. How do encephalitogenic TH17 cells cross these barriers and enter the CNS? One of the key players in this process is the interaction between very late antigen-4 (VLA-4) with vascular cell adhesion molecule-1 expressed on endothelial cells (particularly in BBB) in the initial rolling and adhesion of T cells. On the basis of promising outcomes from phase III clinical trials (Polman et al., 2006), natalizumab, a humanized blocking antibody against VLA-4, was approved in 2006 to treat patients with MS. With respect of T cell transmigration through BCB, additional players contribute to the rolling and subsequent firm adhesion, including the melanoma cell adhesion molecule-laminin 411 axis and the leukocyte function-associated molecule-1-intercellular adhesion molecule-1 axis (Figure 2) (Rothhammer et al., 2011). Following arrest and adhesion, as aforementioned, diapedesis of TH17 into the CNS is mediated by the interaction between CCR6 expressed on TH17 cells and chemokine (C-C motif) ligand (CCL)-20 constitutively expressed on the choroidal epithelia of choroid plexus (Reboldi et al., 2009). However, little information on how Treg cells penetrate the CNS is available, although CCL2- and CCL17-mediated migration has been suggested (Keil et al., 2016). Given the safeguarding BBB and BCB structures in the CNS and known mechanisms underlying transmigration of TH17/Treg cells into the CNS, a pertinent question to ask is how the aforementioned metabolic checkpoints impact the intactness of the BBB and BCB and hinder these transmigrating processes to prevent the development of MS. While this can be explained, to a certain extent, by direct debilitating effects of interleukin-17 on cerebral endothelium through the Rho kinase-dependent inhibitory phosphorylation of nitric oxide synthase and reduced nitric oxide production (Faraco et al., 2018), intrinsic alterations associated with metabolic reprogramming in TH17 and Treg remain to be explored, such as changes of their migratory ability and expression of essential molecules involved in the transmigration process (CCR6, melanoma cell adhesion molecule, leukocyte function-associated molecule-1, VLA-4, CCL-2, CCL-17, etc.). While natalizumab remains to be the most commonly-used therapeutic reagent in relapsing-remitting MS, concerning issues with long-term use of natalizumab (≥ 24 months) include development of life-threatening progressive multifocal leukoencephalopathy and recurrence in some patients. Since metabolic modulators (e.g., AOA) can reprogram TH17 to become Treg by either epigenetically modifying methylation of the Foxp3 gene locus or increasing phosphorylation of AMPK (Xu et al., 2017), we reason that targeting these “metabolic checkpoints” has distinct advantages over conventional therapeutic reagents including natalizumab and fingolimod, a sphingosine-1-phosphate receptor modulator that sequesters lymphocytes in lymph nodes and prevents them from migrating to the CNS. With further systematic evaluations (prophylactic and therapeutic) with different dosing and timing, these metabolic modulators may be developed as standalone therapies or used in combination with other already-approved reagents (e.g., natalizumab) to mitigate toxicities and better therapeutic efficacy. The fact that some of them are already being used in the clinic would expedite such clinical transitions.Figure 2: The underlying mechanisms of TH17 crossing the BBB and BCB.The VLA-4-VCAM-1 axis represents the primary mechanism of TH17 transmigration through the BBB. In BCB, MCAM-laminin 411 and LFA-ICAM-1 axes also contribute to the penetration of TH17 cells into the CNS. BBB: Blood-brain barrier; BCB: blood-cerebrospinal fluid barrier; ICAM-1: intercellular adhesion molecule-1; LFA-1: leukocyte function-associated molecule-1; MCAM: melanoma cell adhesion molecule; TH17 cells: T helper 17 cells; VCAM-1: vascular cell adhesion molecule-1; VLA-4: very late antigen-4.Work in our lab has been funded by the V Foundation Scholar Award (V2018-023), ACS-IRG (91-022-19), and R21 (1R21CA230475-01A1) to LZS.