Mitochondria are the major energy producers within a cell that couple metabolite oxidation to aerobic respiration (10). Glucose and fatty acids, after being catabolized through glycolysis and β-oxidation, respectively, fuel the TCA cycle in the form of acetyl-CoA. Acetyl-CoA is further oxidized into carbon dioxide to generate NADH and FADH₂, the main sources of electrons for the ETC. The ETC ultimately transfers electrons provided by NADH and FADH₂ to oxygen, while concurrently generating ΔΨm across the mitochondrial inner membrane. This proton gradient is further utilized to produce ATP (9). In addition to generating ATP, mitochondria also contribute to lipid and amino acid synthesis to build macromolecules. Acetyl-CoA can be transported out of the mitochondria into the cytosol where it is used for protein acetylation as well as de novo fatty acid synthesis (11). Citrate can also be exported into the cytosol for de novo fatty acid synthesis and into the nucleus to serve as an acetyl-CoA donor for histone modification (12). Furthermore, the metabolic intermediates of the TCA cycle can also be used for production of cholesterol, nucleotides, and amino acids. These processes provide building blocks for cell proliferation and growth at a cost of TCA cycle metabolite depletion. Altogether, mitochondria bridge nutrient metabolism to fulfill the bioenergetic demands of cell through the coordination of TCA cycle and ETC.

Mitochondrial quality and activities are maintained and adjusted through mitochondrial dynamics. In response to the type, strength, and duration of metabolic or genomic stress, mitochondrial dynamics regulate mitochondrial fusion, fission, mobility, biogenesis, and degradation. Mitochondrial mobility regulates the subcellular localization of mitochondria, whereas mitochondrial fusion and fission controls mitochondria mass and metabolism by fusing (fusion) or splitting (fission) the inner and outer membranes and matrix components (13, 14). Nutrient deprivation induces the formation of a tubular network of mitochondria by promoting mitochondrial fusion and suppressing mitophagy (a mitochondrial clearance process) (15). By contrast, severe and prolonged DNA damage leads to mitochondrial fission and further facilitates the clearance of damaged mitochondria via mitophagy (16). Thus, mitochondrial fusion and fission provide a method to efficiently and flexibly adjust mitochondrial quality and quantity. Importantly, most of the mitochondrial fusion and fission machineries are conserved from yeast to mammals, further implicating the importance of these processes (17). Mechanistically, mitofusin and optic atrophy 1 (OPA1) are two dynamin-like GTPases that control fusion of the mitochondrial outer and inner membranes, respectively (18). These proteins are regulated by ubiquitination and proteolytic cleavage. When mitophagy is induced, mitofusin 1 and 2 are ubiquitinated in a PTEN-induced putative kinase 1/Parkin-mediated manner (19). Moreover, OPA1 is constitutively cleaved by protease Yme1L in the intermembrane space in order to shape proper cristae structures (20). Furthermore, loss of mitochondrial ΔΨm induces OPA1 cleavage by protease OMA1, a process that further dampens mitochondrial fusion (21, 22). Conversely, mitochondrial fission is triggered by phosphorylation of dynamin-related protein 1 (Drp1) on serine 616 by protein kinase C. This event promotes Drp1 translocation to the mitochondrial outer membrane and facilitates the association between Drp1 with other adaptor proteins, including Fis1 (mitochondrial fission 1 protein), Mff (mitochondrial fission factor), and MiD49/51 (23–25). By contrast, phosphorylation of Drp1 on serine 637 by protein kinase A (PKA) leads to Drp1 inactivation (26). Additionally, mitofusin 2 is regulated via JNK phosphorylation, which when coordinated with Huwe1-regulated ubiquitination promotes stress-induced mitochondrial fragmentation and apoptotic cell death (27). These regulatory steps tightly control the balance of mitochondrial fusion and fission to actively fine-tune the mitochondria’s metabolic activity. Mitochondrial fusion can increase cristae formation and respiratory complex formation as well as increasing the substrate uptake fueling oxidative phosphorylation (OXPHOS). Moreover, fusion also promotes fatty acid oxidation (FAO), which is important for the formation and survival of memory T cells (discussed in Section “Metabolic Reprogramming Fuels T Cell Activation and Differentiation”) (28). On the other hand, mitochondrial fission not only acts to eliminate dysfunctional mitochondria but is also an adaptation that occurs in response to increased aerobic glycolysis (29).

Unlike the short-lived effector T cells that rely on aerobic glycolysis, memory T cells display increased mitochondrial mass, spare respiratory capacity, and utilize FAO to meet their metabolic demands (35, 41). Reduced glucose uptake and glycolysis has been shown to not only suppress effector T cell functions but also promote T cell differentiation toward memory-like precursor cells that have an elevated dependence on mitochondrial metabolism (42). Additionally, it has been shown that IL-15, an important cytokine for the survival of memory T cells, increases mitochondrial mass and copy numbers in memory T cells (35, 43). Furthermore, the elevated oxidative metabolism and mitochondrial activity in memory T cells could positively impact survival and the rapid transition to glycolysis after re-stimulation (36, 41, 44). These findings suggest that mitochondrial mass and activity play critical roles in guiding the generation of memory T cells. Interestingly, mitochondrial dynamics have also been reported to modulate the differentiation process in effector and memory T cells. Effector T cells primarily possess fragmented mitochondria, which might facilitate aerobic glycolysis. During TCR activation by antigen-presenting cells, Drp1-mediated mitochondrial fission and translocation of mitochondria to the immune synapse could be promoted by suppression of PKA-induced Drp1 phosphorylation at serine 637 by mTOR activation and the calcium (Ca²⁺) flux-mediated phosphatase calcineurin (45–48). By contrast, Opa1-dependent mitochondrial fusion promotes memory T cell formation and oxidative metabolism by maintaining the tightly associated cristae structures and ETC complexes (Figure 1). Notably, OXPHOS can stimulate mitochondrial inner membrane fusion through Yme1L-mediated cleavage of Opa1 (49). In addition, oxidative stress-induced glutathione oxidation results in elevated mitochondrial fusion by promoting mitofusin dimerization (50). Moreover, mitochondrial deacetylase sirtuin-3 (SIRT3) deacetylates OPA1, which activates OPA1’s GTPase activity promoting mitochondrial networking and preventing cell death (51). Thus, TCR signaling and metabolites derived from oxidative metabolism contribute to the regulation of T cell mitochondrial dynamics.

In addition to regulating mitochondrial structure, metabolites derived from mitochondrial metabolic pathways may also regulate T cell differentiation by affecting the epigenetic and transcriptomic landscape. One recent study demonstrated that S-2-hydroxyglutarate (S-2HG) can act as an immunometabolite altering the gene expression profile with changes in the epigenetic landscape toward memory-like CD8⁺ T cells (52). Since S-2HG is a chemical analog of α-ketoglutarate and succinate, we postulate that production of α-ketoglutarate and succinate could regulate T cell differentiation by modulating Jumonji D3 (Jmjd3)- and ten eleven translocation (Tet)-dependent histone and DNA methylation, respectively (53, 54). These mitochondria-derived metabolites may provide signals that impact differentiation, exhaustion, and the secretome of T cells through mitochondria–nucleus communication. Thus, future investigations to determine if mitochondrial dynamics and activity can modulate T cell differentiation through as yet undefined mitochondria–nucleus communication will provide exciting new insights for immunometabolic regulation.

T cell exhaustion describes a state when T cells are incapable of proliferating or producing effector molecules, which often occurs as a result of chronic antigen exposure in diseases such as viral infections and tumors (55). Although changes in the transcription program of T cells can contribute to T cell exhaustion, recent studies have revealed that metabolic insufficiency and deregulation of nutrient-sensing pathways also contribute to T cell exhaustion (56). During chronic lymphocytic choriomeningitis virus infection, Bengsch et al. reported that glycolytic and mitochondrial metabolism in early effector CD8⁺ T cells are repressed by programmed cell death protein 1 (PD-1) signaling. PD-1 signals also suppress the expression of peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC1α) that inhibits mitochondrial biogenesis in viral-specific CD8⁺ T cells leading to declined effector function in exhausted T cells (57). Intriguingly, PGC1α overexpression in exhausted T cells is able to enhance their effector functions and improve mitochondrial biogenesis. By contrast, it has been reported that HIV-specific CD8⁺ T cells display increased mitochondrial mass, resulting in higher cluster of differentiation 95 (CD95)/CD95-ligand induced apoptosis (28, 58, 59). Thus, it is important to further investigate the contributions of mitochondrial biogenesis on T cell exhaustion and how we can target mitochondrial metabolism of T cells when treating chronic viral infection.

The tumor microenvironment can also cause metabolic insufficiency in T cells by reducing PGC1α-mediated mitochondrial biogenesis (60). Similar to chronic viral infection, overexpressing PGC1α in tumor antigen-specific CD8⁺ T cells sustains their metabolic fitness and improves their antitumor responses in the tumor microenvironment. Of note, PD-1 signal is dispensable in tumor-induced impairment of mitochondrial biogenesis in T cells. Since hypoxia and nutrient deprivation caused by elevated tumor metabolism could drastically affect mitochondrial metabolism and dynamics (61), it is likely that T cells fail to sustain their mitochondrial activity and biogenesis due to a metabolic status in the tumor microenvironment that does not exist in the chronic viral infection condition. Interestingly, PD-1 signaling has been shown to increase FAO and mitochondrial spare respiratory capacity in effector T cells (62). Therefore, we postulate that the impairment in mitochondrial biogenesis while receiving PD-1 signaling might result in cell death and senescence in tumor-infiltrating T cell due to metabolic catastrophe (Figure 2). Altogether, these recent studies in chronic viral infection and tumors suggest that mitochondrial metabolism in T cells could critically regulate T cell immune responses in disease progression. Further investigations are needed to reveal how mitochondrial activity is connected with T cell immune responses and whether regulation of mitochondrial dynamics also controls T cells behavior. These investigations will provide novel insights for immunometabolic regulation and facilitate the design of immunotherapies.

Accumulation of dysfunctional mitochondria is frequently observed in aged cells (63). The mitochondrial free radical theory of aging has been recognized for decades (64); oxidation of cellular macromolecules and mitochondrial DNA damages are thought to be mainly caused by reactive oxygen species (ROS) produced from dysfunctional mitochondria (65, 66). Aged T cells display a reduction in respiratory metabolism and ETC activity; this is accompanied by reduced Ca²⁺ influx into mitochondria and elevated mitochondrial ROS production. These changes could lead to decreased ΔΨm and ATP production, as well as activation of nuclear factor of activated T-cells (67). Aged T cells may also upregulate the expression levels of coinhibitory receptors such as PD-1, TIGIT (T-cell immunoreceptor with Ig and ITIM domains), Lag-3 (lymphocyte-activation protein 3), and Tim-3 (T-cell immunoglobulin and mucin-domain containing-3). Therefore, two attractive strategies to rejuvenate the immune responses of aged T cells are (1) activate mitochondrial biogenesis and (2) improve mitophagy-mediated mitochondrial quality control. Among the proposed strategies, nicotinamide adenosine dinucleotide (NAD)-dependent regulation has delivered promising outcomes in different aging studies (68). Manipulating cellular metabolism or treating with nicotinamide riboside increases NAD and may prevent aging by restoring mitochondrial activity. The restoration of mitochondrial activity occurs mainly because of mitochondrial biogenesis activation induced by a sirtuin 1–PGC1α pathway and SIRT3-induced mitophagy (69, 70). However, it remains unclear if these strategies can also be applied to aged T cells. Detailed analysis of the contributions of deregulated mitochondrial dynamics in T cell aging will provide a springboard for developing immunometabolic boosters to rejuvenate aged T cells.

A well-functioning immune system is reliant on the selection of a lymphocyte repertoire that is adequately diverse in response to numerous foreign antigens yet sufficiently self-tolerant to avoid the development of autoimmunity. As described above, production of ROS and ATP synthesis, which are tightly associated with the ΔΨm, regulate the activation, proliferation, and selection of the cell death pathway in T cells. Recent evidence has proposed that the regulation of programmed cell death in T cells is impaired in both human and murine systemic lupus erythematosus (SLE) contributing to other disease pathogenesis (71). Meanwhile, ΔΨm and ROS levels are elevated in patients with SLE in comparison with healthy controls (72). When mitochondrial ROS was blocked in lupus T cells, mice showed reduced interferon pathway signaling and had decreased signs of autoimmune disease (73). Furthermore, coordinated mitochondrial hyperpolarization and ATP depletion play key roles in abnormal T cell death in lupus patients (74). Hence, persistent mitochondrial hyperpolarization, which leads to increased ROS production and ATP depletion, might be responsible for the unusual increase in spontaneous apoptosis in SLE patients and other relative autoimmune diseases. However, the relative impact and metabolic signaling pathways involved still require further investigation. This will likely allow us to identify novel targets for pharmacological intervention in patients with autoimmune diseases.