In patients with UC, there is a higher prevalence of hyperglycemia than in controls (68). Interestingly, patients who have both psoriasis and IBD have a significantly higher prevalence rate of diabetes than those who have sole psoriasis (69). Moreover, patients with IBD and diabetes mellitus exhibit a significantly higher levels of C-reactive protein, erythrocyte sedimentation rate, eosinophil counts, monocyte counts in blood (70). Furthermore, use of metformin, a medication for the treatment of type 2 diabetes, known as an 5’ adenosine monophosphate-activated protein kinase (AMPK) activator, significantly reduces the hazardous ratio of new-onset IBD by almost a half (71), and positively prevents dextran sulfate sodium (DSS)-induced colitis in mice (38). These clinical observations may suggest that systemic glucose metabolism and intolerance is notably associated with IBD.

Hyper-glycolysis in cells often leads to mitochondrial dysfunctions by a hyper-polarizing the mitochondrial membrane potential, and oversupply of mitochondrial ROS (mROS). A high glucose concentration induces more mROS, and drives Th17 cell generation. Accordingly, the treatment of the ROS scavenger, N-acetyl-L-cysteine (NAC), or the mitochondria-targeted anti-oxidant, mitoquinone (MitoQ), in T cells significantly inhibits Th17 cell differentiation even in the presence of high glucose (23).

In animal models of colitis, high glucose availability has been shown to exacerbate the development of inflammation. Colonic expression of glycolysis-associated proteins such as HK2, lactate dehydrogenase A (LDHA), phosphate fructose kinase, and c-MYC are evidently enriched in inflamed tissues of DSS-induced colitis mouse model (72). Likewise, a high-sugar diet clearly deteriorates lymphocyte infiltration, epithelial damage, and cytokine expressions in the gut after colitic DSS challenges in mice (24). High glucose consumption (10% in a drinking water) also exacerbates T cell transfer colitis model in mice by increasing proinflammatory Th17 populations in the gut (23). Conversely, overexpressing a transgenic glucose transporter (Glut) 1 receptor in regulatory T cells fails to achieve disease remission in an adoptive T cell transfer colitis model (25) ( Figure 3A ; Table 1 ).

Glucose restriction achieves an effective prevention of colitis disease. Treatment of potential Glut blocker, Ritonavir, has been reported to ameliorate the disease severity in NOD-scid IL-2Rγ^null colitis animal model (26). Glut1-depleted CD4⁺ T cells fail to trigger intestinal inflammation in nonsteroidal anti-inflammatory drug-induced T cell transfer colitis model in mice (27). Adoptive transfer of Glut3-deficient T cells fail to induce intestinal inflammation (28). T cell specific-HK2 deficiency partially recovers intestinal inflammation in a spontaneous colitis model of IL-10 knockout (KO) mice, although HK2-deficient CD4⁺ T cells appear to have normal proliferation, viability, activation and differentiation in vitro (29) ( Figure 3A ; Table 1 ).

This metabolic shifts from OXPHOS to glycolysis reliance often accompanies induction of the mammalian target of rapamycin (mTOR), c-Myc, and Hif1α pathways (73). The importance of these factors in regulating T cell metabolism and effector functions is well described in the previous literature and review (73, 74).

Tuberous sclerosis 1 (TSC1) is a negative regulator of mTORC1. Constitutive activation of mTORC1 by CD4⁺ specific deletion of TSC1 promotes Th1 and Th17 cell differentiation and suppresses immunosuppressive activity of Treg, resulting in an enhanced disease severity in an adoptive T cell transfer colitis model (30). Conversely, mTOR deficient T cells fail to differentiate into effector T cells, including Th1, Th2, and Th17 (75). The therapeutic efficacy of rapamycin, a classic pharmacological inhibitor of mTORC1, has been substantially demonstrated to ameliorate intestinal inflammations in experimental animal colitis models in a diverse perspective (31–36). Another pharmacological mTORC1 inhibitor, arctigenin, has been reported to decrease Th1 and Th17 cell differentiation, leading to amelioration of disease severity in DSS-induced colitis model (37) ( Figure 3A ).

The activation of mTOR signaling pathway further stimulates the activity of Hif1α, a master regulator of oxygen homeostasis. Although Hif1α-deficient T cells have better cellular growth and proliferation in response to TCR engagement with IL-3 (76), HIF1α–mediated metabolic reprogramming toward ‘aerobic glycolysis’ (76, 77) promotes Th17 cell differentiation and declines Treg cell differentiation in both in vitro experiment, and in vivo animal model of MS (77, 78) ( Figure 3A ). In patients with CD, and UC, Hif1α expression is highly augmented in the inflamed colonic mucosa (39), especially in Th17 cells (79). These experimental evidences may suggest that HIF-signaling pathway is hypothetically to be a fascinating therapeutic target. Despite this, the modulation of HIF levels for IBD has several limitations to induce clinical responses. First, the role of Hif1α in regulatory T cells seems to be debatable in an animal model of colitis (39, 40, 80, 81). That is presumably because Hif1α under hypoxic condition not only facilitates glycolysis but also paradoxically protects mitochondria from ROS by reducing complex I activity, pyruvate oxidation, autophagy, and mtDNA encoded mRNA levels (82). Certainly, T cell specific deletion of Hif1α significantly downregulates Foxp3 over IL-17 expression, leading to the uncontrolled immune responses in DSS-induced colitis model (39). In contrast, Von Hippel-Lindau (VHL)-deficient (i.e. Hif1α-stabilized) Treg cells favor Th1 differentiation over Treg differentiation, resulting in colitis development in adoptive T cell transfer model (40) ( Figure 3A ). Besides, HIF signaling pathways plays an important role in the maintenance of the epithelial barrier functions. In an animal colitis model, the systemic administration of Hif1α inhibitor, PX-478, attenuates the protective effects of exogenous H₂S and epithelial regeneration in dinitrobenzene sulfonic acid-treated rat colitis model (41), whereas Hif1α activation by DMOG (42, 43), FG-4497 (44) or CG-598 (45) results in enhancement of epithelial barrier functions against animal colitis models. Collectively, Hif1α, a master transcription factor driving aerobic glycolysis, regulates T cell differentiation as well as intestinal barrier functions ( Table 1 ).

Myc is a proto-oncogene that acts as a transcriptional factor downstream of the mTOR signaling pathway. It is strongly upregulated in both CD4⁺ and CD8⁺ T cells after TCR activation (83, 84). Myc regulates the expression of glycolytic enzymes at both mRNA (84) and protein (83) levels, and affects the expression of lactate transporter (Slc16a1), glycolytic flux (83), and PPP flux (84). Myc also supports the de novo pyrimidine/purines synthesis, which are essential for nucleotide production and proliferation in activated T cells (83, 84). Moreover, Myc enhances the expression of amino acid transporter and glutaminolysis in the early phase of TCR activation (4 to 8 hours). Thus, Myc modulates metabolic pathways in CD4⁺ T cells during activation.

Regulatory T cells have a different metabolic signature from conventional effector T cells, and Myc plays a different role in them. Foxp3, the key transcription factor for Treg cells, inhibits c-Myc expression by binding directly to the TATA box of the c-Myc gene (85). In addition, overexpression of Glut1 in Treg cells impairs their functions and exacerbates colitis in a colitis model with adoptive transfer (25) ( Table 1 ). However, c-Myc expression is critical for effector Treg generation during early neonatal Treg development (46). Thus, deleting Myc specifically in Foxp3⁺ cells causes early-onset inflammation in multiple organs such as skin, pancreas, liver, lung, and colon in mice (46) ( Figure 3A ). Therefore, the use of Myc inhibitors against proinflammatory T cells should be carefully taken into account for optimal therapeutic effects.

Mitochondrial pyruvate oxidation is a metabolic checkpoint for cell cycle and cellular functions during T cells development. T cells progenitor originate from hematopoietic stem cells (HSC) in the bone marrow. Under physiological conditions in the bone marrow, long-term hematopoietic stem cells (LT-HSC) show a high glycolytic-dependent metabolic profile and impaired oxygen consumption. LT-HSC can preserve their stemness and quiescence by increasing Hif1α-dependent expression of PDKs and reducing mitochondrial mass (89), indicating that pyruvate oxidation is not essential for T cell development at BM. Consistently, defects in PDHA1, PDK2/PDK4 or mitochondrial pyruvate carrier 1 (MPC1) do not affect T cell development at BM (89–91). However, PDK2 or PDK4 knock-in can rescue Hif1α deletion-induced mROS and cell death. Interestingly, only LT-HSC can survive in the in vitro cell culture condition with PDH inhibitor, 1-aminoethylphosphinic acid (1-AA), while short-term HSC and multipotent progenitors comes to cell death (89). Thus, pyruvate oxidation is not a requirement for T cell progenitor development at BM ( Figure 4 ).

In the later stage, thymocytes mature and differentiate into double negative (DN) T, double positive (DP) T and finally CD4⁺ or CD8⁺ single positive (SP) T cells. During this maturation, T cells metabolically need glucose-derived pyruvate oxidation (90, 91). Blocking pyruvate flux into mitochondria by MPC1 deletion reduces OXPHOS gene expression, cellular oxygen consumption rate, and thymic T cell development (90). PDHE1α deficiency in the thymocytes also affects redox balance and carnitine metabolism (91). Loss of PDHE1α remarkably compromise thymic DP T cell survival and PTEN deletion-induced malignant proliferation in thymocytes (91) ( Figure 4 ).

Fully differentiated T cells during homeostasis are largely quiescent and demand a relatively low level of cellular activity. Therefore, they require somewhat higher mitochondrial glucose-derived pyruvate oxidation (65). Stimulation of TCR with CD28 drives expedited metabolic reprogramming through upregulating glycolysis, the pentose phosphate pathway, and glutaminolysis, presumably in the cytosol, yet downregulating pyruvate oxidation (65) in a mitochondrion (84). Certainly, the levels of p-PDHE1 and PDK4, which signify inhibition of pyruvate oxidation, were raised in T cells after early TCR/CD28 activation (13, 49), demonstrating that mitochondrial pyruvate oxidation is essential ( Figure 3B ).

In addition to the T cell activation, cellular pyruvate availability also affects T cell differentiation. Ethyl pyruvate (EP) supplementation notably enhances Treg cell proliferation and differentiation in both in vivo and in vitro (92, 93), whereas Th1 and Th17 differentiation are not affected by EP supplementation in vitro (93). Furthermore, the role of pyruvate oxidation in activated CD4⁺ T cells is more important because in vivo activated CD8⁺ T cells favor carbon sources from PDH flux (pyruvate oxidation) over pyruvate carboxylase (PC) flux, while in vitro activated CD8⁺ T cells prefer PC over PDH flux (94). In addition, CD4-specific deletion of PDHE1α (95) or Glut3 (28) in CD4⁺ T cells compromises glucose-derived acetyl-CoA availability, leading to reduction in the histone acetylation and Th17 polarization (28, 95).

Defeats in mitochondrial pyruvate influx induced by MPC inhibition disrupts mitochondrial calcium homeostasis. This is evident by decreasing carbonyl cyanide 3-chlorophenylhydrazone-induced mitochondrial Ca²⁺ release as well as ATP-induced mitochondrial Ca²⁺ uptake (96). Consequently, nutrient stress by restricted pyruvate oxidation substantially compromises mitochondrial respiration (oxygen consumption rate, OCR) and ATP production and compensatorily enhances autophagy flux (96). This may suggest that mitochondrial pyruvate oxidation sustains mitochondrial health. MPC1 deletion leads to abnormal activation in splenic T cell and results in expansion of auto-reactive T cells in the response to CD3/CD28 stimulation (90) ( Figure 3B ).

In colonic mucosal biopsies from patients with IBD, phosphorylation of PDHE1α, a substrate of PDK4, has been strongly correlated with CD4⁺ T cells ( Figure 3B ). In line with this, the experimental DSS-induced colitis model also displays the significantly upregulated expressions of PDK4 and p-PDHE1α after DSS challenge. Interestingly, CD45⁺ hematopoietic cells including CD4⁺ T cells, neutrophils, macrophages and dendritic cells appears to be a central player in charge with PDK and PDHE1α expression (49) ( Table 2 ). Collectively, the restoration of pyruvate oxidation via supplementation of pyruvate or inhibition of PDKs may be directly connected to therapeutic strategy for IBD.

Dichloroacetate (DCA) is a structural analogue of pyruvate that inhibits PDKs activity (PDK1 at most among PDKs). Previously, its ability to modulate mitochondrial respiration, ROS generation, and metabolic reprograming has been demonstrated in respect of metabolic syndrome (97, 98), osteoporosis (99) and cancer metabolism (100, 101) as well as immune response (65, 88, 102). In human alloreactive-human peripheral blood mononuclear cell (PBMC), DCA treatment significantly decrease glucose uptake, lactate production, and protein expressions of aerobic glycolysis related enzymes such as Glut1, HK2, LDHA, p-PDH and Myc (103, 104). Furthermore, DCA markedly increased regulatory T cell signature (IL-10 secretion and Foxp3 protein expression), and yet decreased effector T cell signature (such as T-bet, GATA3, and RORγT expression) in alloreactive human PBMC. These data suggests that PDK blocker, DCA, mitigate aerobic glycolysis and favor differentiation toward immunosuppressive regulatory T cell rather than proinflammatory effector T cells (103, 104) ( Figure 3B ). In line with this observation, activated CD4⁺ T cells (isolated from BALB/c normal mice) by CD3/CD28-mediated TCR stimulation had decreased lactate production and effector cytokine productions (IFN-γ, IL-5, and IL-17) in response to DCA for 48 hours (105) ( Figure 3B ).

Under hypoxia, both glycolysis and pyruvate oxidation are critical for the survival of effector memory (EM) T cells Glucose deprivation or 2-DG treatment significantly reduces mitochondrial membrane potentials and consequently induces apoptosis in EM T cells. However, the addition of sodium pyruvate or DCA dramatically restores 2DG-induced mitochondrial membrane potential declines and enhances survival under hypoxic stress (106). This finding suggests that glucose oxidation via pyruvate metabolism stabilizes mitochondrial membrane potentials and thereby increases survivability under hypoxic conditions.

PDK also modulates CD4⁺ T cell differentiation. First, its pharmacological inhibitor, DCA, has been clearly demonstrated to induce Treg differentiation and suppress Th17 differentiation (48, 107, 108) in part through ROS generation ( Figure 3B ). However, the current understanding of the mechanisms by which PDK inhibitors regulate T cell differentiation is contradictory. The predominant isoforms of PDKs in naïve, Th17 and regulatory T cells are PDK1 and PDK3 (48, 107). One paper shows that PDK1 deficiency, the strongest target isoform among PDKs that is possibly inhibited by DCA, compromises Th17 differentiation and accelerates Treg differentiation (48), which suggests that the effects of DCA are dependent on PDK –presumably PDK1- activity. On the other hand, it has been shown that the immunosuppressive effects of PDK inhibitors were ironically independent of PDKs because the effect of DCA on manipulating T cell differentiation is persistent even under the knockdown of PDK1 and PDK3 (107). This may suggest that the anti-inflammatory effects of DCA are likely followed through the inhibition of PDK2 and PDK4, or unidentified non-canonical pathways such as ROS production (107). Despite the differences in methods of isolation and differentiations of naïve CD4⁺ T cells that were used in those papers, PDK activity is undoubtedly pivotal in T cell differentiation. In addition, PDK1 (60) or PDK4 (49) knockdown significantly reduces IL-17 secretion in in vitro-polarized Th17 cells ( Figure 3B ).

The effects of PDK inhibition on Th1 and Th2 differentiation have not been thoroughly investigated. DCA treatment significantly enhances IFN-γ secretion and production in splenocytes (109) as well as Th1-polarized CD4⁺ T cells (108), indicating enhanced Th1 differentiation. In contrast, one report showed that DCA treatment fails to reduce Th1 differentiation and effector function (48).

PDK1 has been implicated in the early stages of TCR signaling pathways during T cell activation, in addition to its role in the mitochondrial matrix ( Figure 3B ). In PA-R CD8⁺ T cells, TCR activation by CD3 results in the downstream signaling complex, including ZAP70 and Lck, which binds and activates PDK1 within a few minutes (65). Consequently, due to the impediment of mitochondrial pyruvate influx, the cumulative cytosolic pyruvate allows an augmentation in lactate biosynthesis. In this way, the TCR complex rapidly utilizes aerobic glycolysis during the activation process (65, 110). In summary, these data suggest that inhibition of PDKs reduces anaerobic glycolysis, improves mitochondrial function, promotes ROS production, decreases Th17 differentiation, and increases immunosuppressive Treg function. Therefore, targeting pyruvate oxidation flux to manipulate T cell functions is a promising strategy in the treatment of inflammatory bowel disease (IBD).

Notably, EP has been shown to significantly increase the population of CD4⁺CD25⁺Foxp3⁺ regulatory T cells in both peritoneal cells, lamina propria, and the Peyer’s patches (92) ( Figure 3B ). Furthermore, EP treatment attenuates Th17 cell infiltration in the intestine and prevents trinitrobenzene sulphonic acid (TNBS)-induced colitis in mice (47). In vivo DCA treatment (2g/L, ad libitum) significantly reduces CD4⁺ T cell accumulation, especially Th17 cells, in the spleen and mesenteric lymph nodes in a naive T cell adoptive transfer colitis model (48). Unfortunately, DCA treatment fails to prevent intestinal inflammation due to the non-responsive Th1 cells to DCA. However, PDK inhibition by DCA successfully attenuates EAE progression, which is another Th17-mediated neuronal disease (48). Although DCA treatment is an efficient way to restrain the development of Th17-mediated inflammation, other approaches inhibiting PDKs may be warranted.

Recently, we have reported the pathological role of PDK4 in CD4⁺ T cell (49, 91). PDK4 is highly expressed during early T cell activation, and its deletion leads to the suppression of aerobic glycolysis. Certainly, PDK4 KO mice were found to be more resistant to DSS-induced colitis. In addition, CD4⁺ T cells deficient in PDK4 induce less intestinal inflammation in both DSS-induced colitis and naïve T cell adoptive transfer colitis. Furthermore, treatment with the pharmacological PDK4 inhibitor, GM-10395, compromises CD4⁺ T cell activation and attenuates DSS-induced colitis (49) ( Table 2 ).

Pyruvate dehydrogenase phosphatase (PDP) is a mitochondrial matrix enzyme that sustains PDC activity. PDP1 is predominantly expressed in mitochondria from skeletal muscle, whereas the PDP catalytic subunit 2 (PDP2) is expressed in the liver and many other cells, including white blood cells. In agreement with the effects of DCA on T cell differentiation, PDP2 overexpression, which enhances PDC and pyruvate oxidation, inhibits glycolysis and Th17 differentiation (111) ( Figure 3B ).

In activated T cells, glutaminolysis plays a vital role in their proliferation and effector functions. In mouse splenocytes or T cells, removal of glutamine completely blocks their effector functions (i.e., proliferation and secretion of IL-2 and IFN-γ (112)) in spite of comparable expression of activation cell surface marker (CD25, CD69, and CD98 (112)). This dependence of glutamine for proliferation of activated splenocytes or PBMC cannot be replaced by supplementation of glutamate, α-KG, asparagine, or proline, which are substrates that can substitute for glutamine (112, 113).

In addition, activated human CD4⁺ T cell for 24 hours do not utilize glucose or glutamine as a carbon source to fuel mitochondrial respiration, as the pharmacological treatment using MPC inhibitor (UK5099), GLS inhibitor (CB839 or BPTES), or a combination of both fail to decrease mitochondrial oxygen consumption (52, 114). However, glutamine deprivation reduces mitochondrial oxygen consumption and ATP production (51) ( Figure 3C ).

Moreover, glutamine availability is still critical in the differentiation of naïve CD4⁺ T cells. Glutamine deprivation leads to the differentiation of naïve CD4⁺ T cells into Foxp3⁺ regulatory T cells with robust in vivo proliferative potentials and immune suppressive effects (51, 113) and a reduction in effector cytokine production such as IFN-γ and IL-17A (52) ( Figure 3C ). Strikingly, the skewing of naïve CD4⁺ T cells to Foxp3⁺ Tregs with glutamine restriction cannot be reversed by glutamate or α-KG supplementation, indicating that glutamine is not required for carbon source (113).

Moreover, naïve CD4⁺ T cells grown with low glutamine availability, even under Th1 or Th17-polarizing conditions, favor the differentiation into Foxp3⁺ Tregs while blocking the differentiation into Th1 and Th17 cells (51, 52), respectively ( Figure 3C ). These observations suggest that glutamine itself, not those intermediates of glutaminolysis such as glutamate and α-KG, plays an essential role in modulating effector functions, mitochondrial ATP synthesis, and differentiation in CD4⁺ T cells.

De novo nucleotide synthesis requires glutamine-driven aspartate as a source of nitrogen atoms for the formation of the purine/pyridine ring (115). It is well established that aspartate biosynthesis from glutamate via GOT1/2 is required for proliferation in mammalian cells (116, 117). In CD4⁺ T cells, inhibition of de novo purine/pyrimidine synthesis by N-phosphonacetyl-l-aspartate (PALA) and mycophenolic acid (MPA) promotes the differentiation of Foxp3⁺ Tregs, which is consistent with the effects of glutamine deprivation ( Figure 3C ). In addition, the generation of Foxp3⁺ Tregs is abolished under low glutamine conditions when glutamine synthetase (GS), which synthesizes endogenous glutamine from glutamate, is inhibited by Methionine sulfoximine (MSO) (113) ( Figure 3C ). These findings suggest that the nitrogen atoms in the amide group of glutamine are primarily used for nucleotide synthesis, rather than the carbon backbone of glutamine, in proliferating CD4⁺ T cells.

Despite that glutamine supplementation attenuates intestinal inflammation in the DSS-induced colitis animal model (50) ( Table 2 ), Foxp3⁺ T cells induced by glutamine restriction have shown superior immunosuppressive capacity in vivo to prevent IBD using the adoptive T cell transfer mouse model. Injection of either CD4⁺Foxp3⁺ natural Tregs or Foxp3⁺ T cells grown under glutamine-limited conditions fully protects against colonic infiltration of immune cells, weight loss, and activation of effector T cells. Moreover, glutamine withdrawal has been shown to enhance the proliferation of Foxp3⁺ T cell in vivo (51) ( Figure 3C ).

GLS enzyme activity and mRNA expression are highly induced during early T cell activation (112, 118). However, GLS activity is considerably lower than the activity of GOT or GDH in T cells (112). This indicates that GLS, which is the first enzymes in the process of deamidation of glutamine, is likely to be a limiting factor in T cell activation and presumably T cell differentiation. Under Th0 activating condition by CD3 and CD28 stimulation, GLS inhibition by compound 968 or BPTES significantly impairs the expression of T cell activation marker (CD25 and CD226), secretion of effector cytokines such as IFN-γ, TNF-α, IL-2, and IL-17 and chemokine receptors (CCR6 and CXCR3) (118). Nevertheless, the mode of action by GLS inhibition is quite astonishingly different from that by glutamine deprivation on T cell differentiation. Treatment with GLS inhibitors (CB839 or BPTES) significantly decreases proliferation, IL-17 cytokine production (52, 119) and ATP-coupled OCR (120) in Th17 cells but increases proliferation and IFN-γ production in Th1 cells with an enhanced exhausted phenotype (i.e. PD-1, Lag3, and Tim3) (52) ( Figure 3C ). Moreover, the genetic deletion of GLS leads to the attenuation of RORγt expression in Th17 cells but the accumulation of t-bet expression in Th1 cells while not affecting Foxp3 expression in Treg ( Figure 3C ). In addition, the generation of α-KG decreases in CB839-treated Th1 but not Th17 cells, whereas 2-HG increased in both Th1 and Th17 ( Figure 3C ). Given that α-KG is an important cofactor for both histone and DNA demethylation, GLS deficiency may influence T cell differentiation mechanistically through not only TCA anaplerotic intermediate but also alteration of epigenetic modifications. CB839-treated Th1 cells display decreased global H3K27 trimethylation and more Th1 related genes such as Ifng with opened chromatin accessibility, and supplementation of a-KG reverses the decreased global methylation and opened chromatin status in CB839-treated Th1 cells (52) ( Figure 3C ). Of note, α-KG has been previously reported to regulate IL-2-sensitive effector gene expression in Th1 cells through the association of CCCTC-binding factor in part (121). Taken together, it is suggested that GLS deficiency has distinct mechanisms of differentiation of Th1 and Th17 cells. Patients with CD possess significantly higher numbers of GLS1-positive cells in the inflamed regions of the lamina propria than the control patients (53). Collectively, targeting GLS may impair the immune responses of the infiltrated T cell, particularly Th17 in vivo for treating IBD. Adoptive transfer of GLS-deficient naive CD4⁺ T cell to Rag1 KO mice indeed fails to induce weight loss and intestinal inflammation (52) ( Figure 3C ). Treatment of BPTES also effectively ameliorates spontaneous intestinal inflammation in IL-10 deficient mice by restoring Th17/Treg balance (53) ( Figure 3C ; Table 2 ).

Metabolic adaptations has emerged as a critical process in T cell activation, survival, and differentiation. Manipulating the metabolism of T cells can be promising strategies for treating IBD. However, metabolic modulators may have unwanted effects on not just only pathogenic T cells, but also intestinal epithelial cells, endocrine system, bile acid production, and intestinal microbial metabolism to some extent. Thus, further studies are required.

Deficiency of TFAM in CD4⁺ T cells leads to mitochondrial dysfunction characterized by reduced mtDNA copy number, decreased transcription of mitochondrial proteins, and impaired fatty acid oxidation, despite compensatory increases in mitochondrial mass. Strikingly, Tfam-deficient CD4⁺ T cells also have dysfunctions in transcription factor EB (TFEB)-induced endolysosomal biogenesis, impaired autophagic flux, and abnormal accumulation of lipids such as sphingomyelin and triglycerides (14). Furthermore, the mitochondrial dysfunctions induced by Tfam deficiency lead to effector T cell differentiation favoring Th1 cell differentiation under Th1 or Th2 polarizing conditions, as well as more pathogenic IFN-γ-secreting Th17 cells (14). Consistent with this, Tfam deletion in regulatory T cells induces mROS and switches metabolism towards glycolysis (54). Tregs in the absence of Tfam have significantly reduced Foxp3 expression, which is indispensable for sustaining immune-tolerance (54). Moreover, Tfam deficiency leads to the development of more proinflammatory IFN-γ (54), IL-13 or IL-17a secreting Tregs (54) ( Figure 5A ). Collectively, mitochondrial Tfam in CD4⁺ T cells governs mitochondrial fitness, the balance between effector and regulatory T cells, and autoimmunity. Therefore, it is suggested that Tfam deficiency in CD4⁺ T cells is likely to prompt the development of autoimmune diseases such as IBD. Undoubtedly, CD4-specific Tfam KO mice are more susceptible to chemical-induced experimental colitis model using 3% DSS (14) ( Table 3 ). Of note, CD4-specific Tfam-deficient mice over 2 months of age exhibit premature aging syndromes with multiple organ dysfunctions associated with aging (130). Moreover, the infusions of Tfam-deficient Tregs functionally fail to resolve the accumulation of pathogenic CD4⁺ effector T cells in the adoptive T cell transfer colitis model and accordingly advance disease severity (54) ( Figure 5A ; Table 3 ).

In addition to TFAM, there are more gene sets related to mitochondrial biogenesis regarding mitochondrial DNA replication (e.g., Polymerase G and Twinkle) and mitochondrial transcription (e.g., RNase H1 and mitochondrial aminoacyl-tRNA synthetase) (131, 132). Yet, none of them have been thoroughly elucidated in mitochondrial functions and T cell responses. Recently, one paper shows that T cells with sti mutation (defects in alanyl-tRNA synthetase) have compromised TCR signal initiation machinery in T cells (133). Therefore, targeting mitochondrial biogenesis is still considered a viable option for IBD therapy and new developments in this field are expected to appear in the near future.

IL-15 is a cytokine known for T cell homeostasis, generation of memory T cells and prevention from cell death (134–136). Nevertheless, IL-15 has been recognized as a critical cytokine in mitochondrial biogenesis in T cells. IL-15-treated CD8⁺ T cells have elongated hyper-fused mitochondria with increased mitochondrial mass (137, 138) and TFAM expression (139) compared to unstimulated or IL-2 treated CD8⁺ T cells. In addition, IL-15 has also been shown to play a central role in CD4⁺ T cell proliferation (140), Th17 effector function (141), T helper cell differentiation (142), and neuronal-autoimmunity (141). Moreover, IL-15 remarkably restored augmented mitochondrial biogenesis (TFAM and PGC1α expression) ( Figure 5A ), diminished mitochondrial mass (mitotracker-green staining), and impaired proliferation in cycling Tregs of immune non-responder subjects after HIV infection (143). In addition, deprivation of the IL-15 cytokine presumably shifts immunosuppressive Foxp3-positive cells to proinflammatory RORγt-positive cells (55). Collectively, it is highly affirmative that IL-15 has a possible implication in controlling mitochondrial biogenesis in CD4⁺ T cells and its therapeutic applications in IBD. Although genetic deletion of IL-15 (i.e. using IL-15 whole-body KO mice) is not successful in treating DSS-induced colitis model in mice (144) ( Table 3 ), adoptive transfer of B6 CD4⁺ T cells to IL-15-deficient Rag KO mice induces colonic inflammation compared to the control, and adoptive transfer of IL-15 receptor-deficient Treg fails to suppress severe colonic inflammations with extensive mucosa damage in CD4⁺ T cell transfer colitis model (55) ( Table 3 ), suggesting a direct role of IL-15 in colitogenic CD4⁺ T cells.

Mitochondria contain a number of calcium transport channels, including the mitochondrial calcium uniporter (MCU), which is sub-localized in the inner membrane of mitochondria (181). Mitochondrial-mediated Ca²⁺ uptake plays an important role in cytosolic Ca²⁺ buffering in mast cells. Studies have shown that in MCU-deficient mast cells, antigen-induced beta-hexominidase release was suppressed, suggesting a role for MCU-mediated mitochondrial Ca²⁺ flux in mast cell degranulation (182). Furthermore, knockdown of mitochondrial MCU has been shown to reduce cytoplasmic calcium fluctuation and SOCE in mast cells (183). However, the pathological role of both MCU and MICU1 (a gatekeeper of MCU-mediated mitochondrial Ca²⁺ uptake) in T cells remains poorly understood. Loss of MICU1 has been shown to generate excessive ROS, delay proliferation, and impair migration in HeLa cells (184), Thus, blocking MCU or MICU1 in CD4⁺ T cells is likely to result in mitochondrial Ca²⁺ instability and ROS generation, leading to suppression of cell proliferation and gut penetration in pathogenic CD4⁺ T cells ( Figure 7 ).

Voltage-dependent anion channel (VDAC) is a mitochondrial outer membrane transporter, and has been associated with mitochondrial calcium transport, cell death, and lupus-like autoimmunity (185). Upon 24-hour T cell activation, VDAC1 expression increases along with the Glut1 and HK2 in PBMC (186), which are activation signatures of T cells. In addition, PBMC from patients with coronavirus-A displays mitochondrial dysfunctions characterized by fragmented mitochondrial morphology and apoptotic signaling (186).

The efficacy of VBIT-4 and VBIT-12, VDAC inhibitors, has been documented in treating IBD using DSS- or TNBS-induced colitis animal models (57–59) ( Figure 7 ; Table 3 ). Although the detailed mechanism behind VDAC inhibition has been shown to be associated with MAVS and inflammasome activation on intestinal epithelial cells (58), it may normalize mitochondrial calcium imbalance and attenuate effector functions in pro-inflammatory T cells (186).

Mitochondria and ER are physically connected to form a junction called mitochondria-associated membrane (MAM), which provide a platform for various cellular processes including calcium homeostasis, autophagy, lipid metabolism, and apoptosis. The MAMs contain several tethering molecules such as IP3R, VDAC, and GRP75. In 2018, Bantug and colleagues demonstrated the critical role of MAMs in early activation of effector/memory CD8⁺ T cells ( Figure 7 ), which controls effector functions in memory CD8⁺ T cells through an immunometabolic reprogramming that involves mitochondrial respiration, HK binding to VDAC, mTOR/AKT/GSK3b signaling. Inhibition of MAM formation using nocodazole significantly reduces mitochondrial respiration, HK expression, mTOR/AKT/GSK3b signaling pathway, and cytokine secretion ( Figure 7 ). However, inhibition of mitochondrial calcium does not affect cytokine production in CD8⁺ T cells (187).

PDK4 has been identified as a novel modulator of MAM integrity and mitochondrial quality control (97). Genetic deletion of PDK4 in activated CD4⁺ T cells significantly diminishes MAM formation, SOCE, and mitochondrial calcium transfer compared to control cells. A novel PDK4 inhibitor, GM-10395, also ameliorates SOCE, mitochondrial calcium, and T cell activation. The anti-inflammatory effects of PDK4 suppression have been demonstrated in DSS-induced colitis animal models and naïve T cell transfer colitis models in vivo (49) ( Figure 7 ; Table 3 ).