Targeting mitochondria: restoring the antitumor efficacy of exhausted T cells

Abstract

Immune checkpoint blockade therapy has revolutionized cancer treatment, but resistance remains prevalent, often due to dysfunctional tumor-infiltrating lymphocytes. A key contributor to this dysfunction is mitochondrial dysfunction, characterized by defective oxidative phosphorylation, impaired adaptation, and depolarization, which promotes T cell exhaustion and severely compromises antitumor efficacy. This review summarizes recent advances in restoring the function of exhausted T cells through mitochondria-targeted strategies, such as metabolic remodeling, enhanced biogenesis, and regulation of antioxidant and reactive oxygen species, with the aim of reversing the state of T cell exhaustion and improving the response to immunotherapy. A deeper understanding of the role of mitochondria in T cell exhaustion lays the foundation for the development of novel mitochondria-targeted therapies and opens a new chapter in cancer immunotherapy.

Introduction

In recent years, tumor immunotherapy strategies, particularly immune checkpoint blockades, have achieved remarkable breakthroughs. However, only a subset of patients has benefited from these treatments, with the majority facing the challenge of primary or secondary immune resistance. The functional exhaustion of tumor-infiltrating lymphocytes (TILs) is one of the key factors limiting the response to immunotherapy, resulting in ineffective T cells recognizing and attacking tumor cells, failing to effectively control tumor growth, and contributing to disease progression [1]. In the tumor microenvironment (TME), T cells exhibit reduced cytokine secretion, increased chemokine expression, and upregulation of inhibitory receptors (e.g., PD-1, CTLA-4, LAG-3, TIM-3) [2,3,4]. Upon binding to the appropriate ligands, T cell proliferation is impaired, effector functions are attenuated, and they eventually enter an exhausted state [5]. Exhausted CD8⁺ T cells (T_ex) are not a single subpopulation, but consist of multiple subpopulations with distinct transcriptional, metabolic, and epigenetic properties. Under prolonged antigenic stimulation, TCF-1⁺ CXCR5⁺ progenitor-like precursor exhausted T cells (T_pex) progressively differentiate into terminally exhausted T cells that express high levels of inhibitory receptors, such as PD-1 [6]. Although PD-1 inhibitors partially reprogram T_pex cells, inducing clonal expansion and migration to the tumor site, they have a limited effect on already deeply exhausted T cells [7]. Therefore, restoring the function of T_ex cells or even converting them into T_pex or effector T cells (T_eff) is crucial for enhancing T cell killing and increasing the efficacy of immunotherapy.

Mitochondria, as the core of metabolism and signaling, have a crucial influence on T cell activation, proliferation, differentiation, and effector functions. Resting T cells undergo metabolic reprogramming upon antigen stimulation and differentiate into effector or memory CD8⁺ T cells (T_m) to adapt to different energy requirements [8]. Upon antigen activation, resting T cells, originally dependent on oxidative phosphorylation (OXPHOS), are rapidly activated and transformed into glycolysis-dependent T_eff cells. At the end of the immune response, they are transformed into fatty acid β-oxidation (FAO)-active CD8⁺ T_m cells by activation of 5′-adenosine monophosphate-activated protein kinase (AMPK) or inhibition of the mTOR pathway [9, 10]. Mitochondria play a key role in this metabolic transformation process [11], generating ATP through OXPHOS to provide energy for T cell activation and proliferation [12]. Mitochondrial dynamics, such as dynamin-related protein 1 (Drp1)-mediated fission and fusion, influence metabolism and signaling and regulate T cell activation [13]. Inhibition of Drp1-mediated fission and promotion of mitochondrial fusion leads to a reduction in glycolysis and a shift towards a memory-like phenotype in T_eff cells [14]. Overexpression of peroxisome proliferator-activated receptor-gamma (PPARγ) coactivator-1-alpha (PGC-1α), a master regulator of mitochondrial biogenesis, increases mitochondrial biogenesis, restores T cell function, and enhances antitumor immunity [15]. In addition, mitochondrial membrane potential (MMP) and reactive oxygen species (ROS) levels directly affect T cell function. T cells with high MMP have high ROS levels, a fast glycolysis rate, and increased expression of effector-related genes, whereas T cells with low MMP have upregulated expression of memory genes and enhanced antitumor activity [16]. Moderate ROS levels promote the activation of the nuclear factor of activated T cells (NFAT) transcription factor and transcription of IL-2 genes, which are conducive to T cell proliferation and differentiation [17]. Conversely, high levels of ROS inhibit T cell receptor (TCR) signaling and transcription factor activity, weakening T cell function and promoting tumor immune escape [18].

In summary, physiological processes such as mitochondrial energy metabolism, kinetics, and biogenesis are critical for CD8⁺ T cell activation, proliferation, and anti-tumor function. Recent studies have shown that abnormal mitochondrial function can contribute to T cell exhaustion and tumor development through multiple pathways. Approaches to restore the antitumor function of exhausted CD8⁺ T cells by remodeling mitochondrial metabolism, dynamics, and biogenesis have been shown to be effective, suggesting that mitochondria may be a potential target for future cancer therapies. The aim of this review is to explore the key role of mitochondrial dysfunction in T cell exhaustion and its mechanisms, to summarize the strategies for targeting mitochondria to restore the function of T_ex cells, and to provide a theoretical basis for the development of novel mitochondria-targeted therapies.

Mitochondrial dysfunction and T cell exhaustion

Changes in mitochondrial characteristics and T cell exhaustion

T cells in the TME are subjected to sustained antigenic stimulation and metabolic stress, which together drive mitochondrial dysfunction [15], manifested by defective OXPHOS, impaired mitochondrial adaptations, and mitochondrial depolarization [19, 20]. These changes promote T cell exhaustion at the level of T cell energy metabolism, epigenetics, transcription, and signaling. Specifically, T_ex cells exhibit reduced metabolic efficiency and mitochondrial dysfunction. Long-term antigenic stimulation triggers structural and functional changes in mitochondria, impairing T cell glycolysis and OXPHOS and accelerating the process of exhaustion.

The metabolic pathways of T_ex cell subpopulations are distinct; T_pex cells tend to be catabolic and are characterized by mitochondrial FAO and OXPHOS, whereas terminal T_ex cells favor the glycolytic pathway. Mitochondria play a central role in this metabolic reprogramming. Furthermore, elevated ROS levels are an important hallmark of T cell exhaustion [20,21,22]. Diminished mitochondrial adaptation or morphological fragmentation leads to aberrant ROS production and impaired OXPHOS function, exacerbating the dysfunction of T cell proliferation, differentiation, and effector function, thereby deepening the exhausted state.

In addition, mitochondrial depolarization and MMP damage have been observed in T_ex cells, reflecting an imbalance in mitochondrial dynamics. This imbalance may affect the activity and efficiency of OXPHOS and interfere with T cell metabolic reprogramming [23]. In summary, T_ex cells are characterized by impaired metabolism, defective OXPHOS, and abnormal mitochondrial dynamics. These mitochondrial dysfunctions drive the exhaustion process from multiple perspectives, including energy metabolism, epigenetics, transcriptional regulation, and signaling (Fig. 1). Here, we review the specific roles and mechanisms of mitochondrial dysfunction in T cell exhaustion.

Fig. 1

Mitochondrial alterations in exhausted T cells. In the TME, under continuous antigen stimulation, the following events occur: (1) Inhibition of PGC-1α-mediated mitochondrial biogenesis: Continuous antigen exposure inhibits PGC-1α, a key regulator of mitochondrial biogenesis, leading to a reduction in the number and function of mitochondria; (2) Impaired mitochondrial metabolism: Healthy mitochondria undergo metabolic stress, resulting in impaired mitochondrial oxidative phosphorylation (OXPHOS), decreased ATP production, and impaired mitochondrial adaptability; (3) Structural and functional changes: Sustained antigen stimulation induces structural and functional changes in mitochondria, including a decline in mitochondrial division and increased fusion, which prevents PINK1-mediated mitochondrial autophagy; (4) Accumulation of depolarized mitochondria: Impaired membrane potential and accumulation of depolarized mitochondria occur, making exhausted T cells unable to effectively use OXPHOS to provide energy; (5) Increased reactive oxygen species (ROS): The accumulation of dysfunctional mitochondria leads to abnormal production of ROS, which further exacerbates T cell exhaustion; these mitochondrial alterations increase the transcription of genes associated with exhaustion, leading to epigenetic alterations in T cells; eventually, there is an increase in inhibitory receptors such as PD-1, a decrease in cytokine production, and impairment of T cell proliferation and differentiation, all of which contribute to T cell exhaustion

Dysregulated mitochondrial metabolism promotes T cell exhaustion

Antigen stimulation is associated with impaired mitochondrial OXPHOS function

Under prolonged antigenic stimulation, T cells exhibit impaired mitochondrial OXPHOS function, leading to increased reliance on glycolytic processes for ATP production [24]. Glucose transporter protein 1 (GLUT1), which is induced by hypoxia, was found to be upregulated in exhausted CD8⁺ T cells, reflecting the high dependence of T cells on the glycolytic pathway [24]. However, in the TME, under conditions of glucose deprivation, a progressive decrease in the ATP/AMP ratio was observed in terminally exhausted T cells, despite their enrichment for genes related to glycolysis. The impaired mitochondrial OXPHOS function was unable to maintain the DNA and protein synthesis processes necessary for a continuous supply of ATP, which not only suppressed the proliferative capacity of T cells but also progressively weakened their cytokine production, while upregulating the expression of genes associated with T cell exhaustion [22].

In addition, T cells often experience severe hypoxia in the TME. Sustained stimulation in a hypoxic environment can rapidly cause T cells to enter a state of exhaustion. The underlying mechanism shows that hypoxia promotes the expression of the transcriptional repressor Blimp-1 (encoded by Prdm1), which leads to the inability of T cells to adapt to the hypoxic microenvironment and ultimately to the fate of exhaustion by inhibiting mitochondrial metabolic reprogramming mediated by PGC-1α [23]. Mitochondrial dysfunction induces oxidative stress and ROS accumulation, which in turn inhibits the proteasomal degradation pathway of hypoxia-inducible factor-1 alpha (HIF-1α). This cascade of reactions affects the expression of glycolysis-related enzymes, such as glyceraldehyde-3-phosphate dehydrogenase (GAPDH), phosphofructokinase (PFKL), enolase 1 (ENO1), and pyruvate kinase M (PKM), which further promote the transcriptional and metabolic reprogramming of T cells towards their terminal state of exhaustion [25].

Regulation of T cell metabolism by immunosuppressive signals in the TME

Immunosuppressive signals in the TME severely disrupt the oxidative metabolism of T cells, preventing them from meeting the metabolic demands of their effector functions (Fig. 2). When the TCR/CD3 complex interacts with the co-stimulatory molecule CD28, calmodulin-dependent protein kinase 2 (CaMKK2) is activated, which in turn triggers the initiation of AMPK. This calcium-dependent AMPK activation mechanism is a key driver in promoting mitochondrial biogenesis and metabolic reprogramming [10, 26]. Meanwhile, TCR co-stimulatory signaling activates phosphatidylinositol 3-kinase (PI3K), which subsequently initiates a signaling cascade involving protein kinase B (AKT) and mammalian target of rapamycin (mTOR), which further activates a number of key transcription factors, such as c-Myc and HIF-1α, to regulate the glycolytic program of T cells [27, 28]. The specific metabolic pathway induced by mTOR activation is essential for metabolic reprogramming of T cells to support cell growth, proliferation, differentiation, and cytokine production, among other immune functions.

Fig. 2

Mitochondrial dysfunction and T cell exhaustion. Under continuous antigen stimulation, T cell mitochondrial oxidative phosphorylation (OXPHOS) is impaired, leading to an inability to meet the metabolic demands of the T cells; this results in the upregulation of glucose transporter 1 (Glut 1), reflecting a shift towards glycolytic ATP production. (1) The T cell receptor (TCR) initiates downstream signaling pathways: Under TCR and CD28 co-stimulation, phosphorylated PI3K activates the PI3K-AKT-mTOR pathway, mTOR plays a pivotal role in regulating mitochondrial metabolism, including the upregulation of glycolysis-related genes such as HIF-1α and c-Myc, promoting glycolysis and PGC-1α-mediated mitochondrial biogenesis and other essential metabolic processes; meanwhile, the AMPK pathway, activated by calcium/calmodulin-dependent protein kinase 2 (CaMK2), is also a significant driver for mitochondrial biogenesis and metabolic reprogramming. (2) PD-1 inhibition of the PI3K axis and downstream signaling: PD-1 signaling alters T cell metabolic reprogramming by restraining the PI3K-AKT-mTOR signaling axis via dephosphorylation of PI3K, thereby inhibiting T cell glycolysis and promoting fatty acid β-oxidation (FAO); additionally, PD-1 signaling decreases the activity of the mitochondrial fission protein Drp1 by directly affecting mTOR and ERK1/2 activity, disrupting mitochondrial dynamics and contributing to mitochondrial dysfunction and OXPHOS, ultimately driving T cell exhaustion. (3) Mitochondrial dysfunction under hypoxic signaling: Hypoxic signaling within the TME promotes the activation of the transcriptional repressor Blimp-1, which suppresses PGC-1α-mediated mitochondrial metabolic reprogramming, leading to T cell exhaustion. (4) Increased reactive oxygen species (ROS) generation causes altered transcription: mitochondrial dysfunction leads to redox stress and ROS accumulation; by inhibiting the activation of T cell NF-κB, this promotes NFAT activation and increases the expression of the transcription factor TOX, thereby enhancing the expression of exhaustion-related genes such as Pdcd1 and Lag3, increasing the expression of inhibitory receptors, and resulting in T cell exhaustion

Fig. 3

Differentiation of T cells in tumor microenvironment. During acute infection, naive T cells are stimulated by TCR engagement, leading to enhanced glycolysis, reduced oxidative phosphorylation (OXPHOS), and transformation into effector cells (T_eff) or directly into stem-like memory cells (T_scm). Upon antigen clearance, AMP-activated protein kinase (AMPK) pathway activation, mTOR pathway inhibition, and increased fatty acid oxidation (FAO) promote the majority of effector cells to undergo apoptosis or differentiate into memory cells (T_m), while T_scm also differentiates into T_m. T_m contributes to long-term immunity against tumors. (2) In the presence of transforming growth factor-β (TGF-β), naive T cells can differentiate into regulatory T cells (T_reg), which inhibit the activation and function of other T cells through the secretion of TGF-β and interleukin-10 (IL-10), thereby promoting immune escape. (3) Chronic antigen stimulation (such as in tumors) and mitochondrial dysfunction within the tumor microenvironment, characterized by decreased glycolysis and OXPHOS, inhibited mitochondrial fission, and elevated reactive oxygen species (ROS) levels, drive T cells towards an exhausted state rather than memory formation. This leads to reduced cytokine production and increased expression of inhibitory receptors such as programmed cell death protein 1 (PD-1) and T-cell immunoglobulin and mucin-domain containing-3 (TIM-3), along with unique transcriptional and epigenetic regulation that suppresses the immune response. (4) Exhausted T cells consist of precursor exhausted T cells (T_pex) and terminally exhausted T cells (T_ex). T_pex possess self-renewal capacity and respond to PD-1 blockade therapy. Enhancing glycolysis and OXPHOS, promoting mitochondrial fission, and decreasing ROS levels can facilitate the conversion of T_pex into T_eff or T_m. (5) T_pex experiencing sustained increases in ROS levels and damage to mitochondrial membrane potential further differentiate into T_ex, which highly express inhibitory receptors such as PD-1 and TIM-3, lose self-renewal and effector functions, and do not respond to PD-1 blockade therapy

However, the interaction between PD-1 and its ligand PD-L1 regulates key signaling pathways, such as TCR-activated PI3K/Akt/mTOR and MAPK/ERK, thereby inhibiting the activity of downstream signaling molecules and affecting the expression of metabolism-related transcription factors. This further suppresses metabolic processes closely related to immune function, leading to a reduction in cytokine secretion, impairing T cell proliferation and differentiation, and ultimately leading to a decline in immune function [3, 29].

Specifically, PD-1/PD-L1 binding induces phosphorylation of the immunoreceptor tyrosine inhibitory motif (ITIM) and immunoreceptor tyrosine switch motif (ITSM) regions of PD-1, which attract and activate tyrosine phosphatases Src homology phosphatase 1 (SHP-1) and Src homology phosphatase 2 (SHP-2). These phosphatases contribute to the dephosphorylation of the downstream effector molecule PI3K, attenuate the activation of AKT, and inhibit the PI3K-AKT-mTOR signaling axis, promoting the expression of the adipose triglyceride lipase (ATGL) and the key enzyme for FAO, carnitine palmitoyl transferase (CPT1A), which inhibits T cell glycolysis and instead promotes FAO, thus reprogramming the metabolic pathways of T cells [30]. PD-1 signaling also directly inhibits mTOR activity, inducing mitochondrial dysfunction and defective OXPHOS, driving T cells into an exhausted state [29]. In addition, AKT hyperactivation induced by sustained antigenic stimulation and its subsequent inhibition of the activity of the transcription factor forkhead box O (FOXO) led to a sharp decrease in the expression of PGC-1α, which impeded the normal course of mitochondrial biogenesis, and this series of knock-on effects was a key link contributing to the metabolic imbalance and ultimately the terminal exhaustion phenotype [15].

Taken together, the TME interferes with the oxidative metabolism of T cells through multiple pathways, preventing them from performing their effector functions and leading to dysfunction. This finding highlights the central impact of metabolic pathways on T cell function and predicts that by regulating metabolic reprogramming, oxidative stress, and hypoxia adaptation, new strategies may be developed in the future to restore T cell efficacy and enhance antitumor immunity. These findings open new avenues of research and provide potential breakthroughs for cancer immunotherapy.

Abnormal mitochondrial ROS accumulation promotes T cell exhaustion

ROS generation and mitochondrial dysfunction

ROS are a class of highly reactive oxidative molecules or free radicals, which mainly include superoxide anions (O₂^–·), hydroxyl radicals (OH·), and hydrogen peroxide (H₂O₂). Excessive accumulation of ROS is closely associated with oxidative stress, and mitochondria serve as the major site of ROS generation, with complexes I and III of the electron transport chain considered to be the major sources of mitochondrial ROS (mROS) [31].

In the electron transport chain, complex I, a ubiquinone oxidoreductase, is responsible for the transfer of electrons from NADH to flavin mononucleotide (FMN), which is then transferred to ubiquinone via iron-sulfur clusters, driving the transfer of electrons along the mitochondrial respiratory chain. During this process, oxygen molecules intercept high-energy electrons at the FMN site of complex I, resulting in the generation of superoxide radicals. Superoxide radicals promote ROS generation as ROS precursors [31].

Correspondingly, an efficient antioxidant enzyme system exists in the mitochondria, including manganese superoxide dismutase (MnSOD), glutathione peroxidase (GPX), and catalase (CAT), which can effectively scavenge excess ROS and maintain the balance of intracellular ROS levels.

Dual role of ROS in T cell function within the TME

Within the TME, various cell types exhibit distinct dynamics of ROS, which influence their physiological functions and interactions. First, tumor cells often upregulate ROS production through metabolic reprogramming, activation of oncogenes, and mitochondrial dysfunction [32]. This increased ROS can promote tumor cell proliferation and survival by activating certain signaling pathways such as nuclear factor kappa B (NF-κB) [32]. Additionally, ROS can induce the expression of angiogenic factors, promoting the formation of new blood vessels that supply extra nutrients to the tumor [32]. Second, stromal cells within the TME, such as endothelial cells, mesenchymal stem cells (MSCs), and cancer-associated fibroblasts (CAFs), contribute to tumor growth and metastasis while also affecting anti-tumor immune responses [33]. Under high ROS levels produced by tumor cells, fibroblasts are activated into myofibroblasts, which express mesenchymal markers like α-smooth muscle actin (α-SMA), calponin, and vimentin, facilitating tumor formation and metastasis, and produce large amounts of hydrogen peroxide, increasing oxidative stress in the microenvironment [32].

It has been shown that TCR stimulation can induce rapid production of ROS through multiple pathways (Fas/FasL, NADPH oxidase [NOX]) and regulate T cell signaling, gene expression, and immune effects [34]. Under normal conditions, small amounts of ROS produced by mitochondria can promote T cell activation and proliferation to gain effector functions [35]. For example, mROS triggers the activation of NFAT transcription factors and the transcription of IL-2 cytokine genes, which supports the maturation and proliferation of initial T cells [17].

However, tumor-infiltrating T cells often exhibit impaired mitochondrial adaptation or fragmentation in response to prolonged antigenic stimulation and tumor microenvironmental stressors, including hypoxia and glucose deprivation, and induce abnormally elevated mROS and impaired OXPHOS [20]. Prolonged exposure to high levels of ROS impairs T cell proliferation and effector functions, driving them into a state of exhaustion [20] (Fig. 3). Specific mechanisms include: ROS interfering with TCR and MHC peptide complex recognition, attenuating T cell immune responses [18, 36]; Persistent ROS inhibiting NF-κB phosphorylation and activation, affecting IFN-γ and CD39 expression and inducing TIL hypo-responsiveness [37]; Mitochondrial dysfunction leading to overproduction of ROS, which triggers oxidative stress and promotes upregulation of NFAT [21, 22]; overexpression of NFAT impairs the function of tumor-infiltrating CD8⁺ T cells and contributes to their exhaustion by acting on genes encoding exhaustion-associated factors [38,39,40].

Moreover, tumor-associated macrophages (TAMs) and myeloid-derived suppressor cells (MDSCs) in the TME also contribute to ROS accumulation. These cells release ROS and other suppressive molecules, inhibiting T-cell activity and maintaining an immunosuppressive state in the TME. TAMs are one of the most abundant immune cells in the TME and are primarily classified into M1 and M2 types. M1 macrophages exhibit antitumor activity, while M2 macrophages display immunosuppressive properties, promoting tumor growth and angiogenesis [41]. TAMs produce large amounts of ROS via NOX, which can directly impair T cell function, leading to T cell exhaustion [42]. Additionally, ROS produced by TAMs can influence T cell differentiation and function through epigenetic modifications, such as DNA methylation and histone deacetylation [43].

MDSCs are a heterogeneous population of myeloid cells that primarily suppress T cell immune function by producing ROS and other inhibitory molecules. MDSCs can competitively deplete cysteine in the body, upregulate inducible nitric oxide synthase (iNOS) and arginase-1 (Arg-1) activity, thus depleting L-arginine and inhibiting T cell function [44]. ROS produced by MDSCs not only directly damage T cells but also affect T cell metabolic states, such as inhibiting aerobic glycolysis and the TCA cycle, reducing ATP production, and further suppressing T cell proliferation and effector function [45].

TAMs and MDSCs exhibit synergistic effects in the TME, collectively promoting T cell exhaustion and immune evasion. ROS produced by TAMs can activate MDSCs, leading to further production of ROS and other inhibitory molecules, creating a positive feedback loop that exacerbates T cell dysfunction [46]. Furthermore, TAMs and MDSCs can secrete various cytokines and chemokines, such as IL-6, IL-10, and TGF-β, further suppressing T cell function [47].

In addition, macrophage-derived ROS has been shown to enhance the function of regulatory T cells (T_regs), which are key immunosuppressive cells in the TME [48]. T_regs, through the secretion of TGF-β, activate NOX to generate ROS, suppressing the activation and proliferation of effector T cells and thus weakening the immune response [49]. Furthermore, studies have indicated the effect of ROS on CD4⁺ T cells, where NOX-2-derived ROS can promote the differentiation of T cells from Th17 to Th2 phenotype [50]. Other inflammatory cells, such as activated phagocytes (including neutrophils, eosinophils, and mononuclear phagocytes), also produce ROS, further exacerbating oxidative stress in the TME and influencing the interactions among different cell types [51, 52].

In conclusion, in the tumor environment, excessive ROS accumulation due to mitochondrial abnormalities is a key factor contributing to T cell exhaustion, which not only directly impairs T cell function but also exacerbates T cell exhaustion by interfering with signaling, gene expression, and oxidative stress, forming a vicious cycle and significantly weakening the efficacy of anti-tumor immune responses. In the face of T cell exhaustion caused by mitochondrial dysfunction and excessive ROS accumulation, potential solutions can be explored by drawing on various strategies from existing studies. For example: Antioxidants can neutralize ROS and protect T cells; Specific drugs such as MitoQ have demonstrated potential therapeutic value by improving mitochondrial function and reducing ROS generation; Modulation of signaling pathways, particularly inhibition of the NF-κB/NFAT pathway, may help restore T cell viability; Gene editing techniques such as CRISPR-Cas9, although still in the experimental stage, have shown great potential in correcting mitochondrial DNA (mtDNA); Immunotherapies combining the use of PD-1/PD-L1 or CTLA-4 inhibitors have been shown in clinical trials to enhance the anti-tumor capacity of T cells. Ultimately, the personalized integration of these strategies into treatment regimens tailored to each patient may be an important direction for future treatment.

Mitochondrial dynamics and T cell exhaustion: from kinetic abnormalities to dysfunction

Importance of maintaining mitochondrial dynamic properties

Mitochondria, the energy factories of the cell, exhibit highly dynamic properties and their morphology and function are finely regulated. This dynamism is primarily driven by fusion and fission processes, both of which are mediated by a group of nuclear-encoded GTPases that perform mechanical work on biological membranes using the energy released by GTP hydrolysis [53]. Mitochondria not only carry their own genetic material, mtDNA, but also require constant repair and renewal to maintain a healthy state. The mitochondrial fission mechanism ensures the quality of the mitochondrial population by removing damaged or disabled members in a timely manner, whereas mitochondrial autophagy further removes damaged parts and prevents deeper cellular injury when the cell is under severe stress [54]. At the same time, mitochondrial fusion creates an interconnected network structure that facilitates the sharing of internal mitochondrial resources, including proteins, lipids, and other small molecules, a process that is essential for maintaining mitochondrial OXPHOS function and preventing damage caused by the accumulation of mtDNA mutations that accompany the aging process [55]. In a healthy state, fusion and fission maintain a delicate balance, with the remodeling of mitochondrial cristae adapting to different metabolic demands. Disruption of this balance leads to mutations and dysfunction in mitochondrial morphology.

Imbalance of mitochondrial dynamics with T cell exhaustion

When comparing normal T cells with T_ex cells, the latter exhibit mitochondrial depolarization and decreased MMP, accompanied by shortened and fewer mitochondrial cristae, suggesting that the mitochondrial autophagy pathway is less efficient in removing damaged and depolarized mitochondria. This demonstrates that impaired mitochondrial dynamics exacerbate T cell exhaustion characteristics, as well as the degree of epigenetic reprogramming [20]. Notably, the upregulated expression of PD-1 in T_ex cells indirectly affects T cell function by altering the structure of the mitochondrial cristae, specifically impairing mitochondrial OXPHOS [56]. An increase in mitochondrial mass and decrease in MMP were observed in terminally exhausted T cells, which limits the efficient use of OXPHOS energy by T cells [20]. Recent studies have shown that in T_ex cells, the activity of Drp1, a GTPase that plays a central role in mediating mitochondrial fission and maintaining mitochondrial dynamic homeostasis, is inhibited, and its function has been shown to be critical for T cell activation, proliferation, and effector functions [57] (Fig. 3). The decrease in Drp1 activity is directly linked to the degradation of mitochondrial morphology and function, which in turn affects the energy supply of T cells, rendering them unable to maintain their basic activities and functions and weakening the body’s overall immune response [23]. Studies have confirmed that PD-1 signaling inhibits mitochondrial fission by downregulating the phosphorylation level of Drp1 at Ser616 through the regulation of the Extracellular signal-regulated kinases 1/2 (ERK1/2) and mTOR pathways, which in turn promotes T cell exhaustion [58]. In addition, recent studies have shown that Drp1-deficient T cells exhibit reduced TCR expression, which impairs TCR signaling and reduces the viability of T cell activation and proliferation, as well as significant differences in the expression of TCF1 in the control group. TCF1 is a key transcription factor in T_pex cells and is essential for the maintenance of T cell self-renewal; together, these findings confirm that reduced Drp1 activity is likely to be a key driver of T cell exhaustion and dysfunction [59].

Interplay between mitochondrial metabolism and epigenetics in T cell exhaustion

Mitochondrial abnormalities and their impact on T cell exhaustion

Mitochondrial abnormalities in promoting T cell exhaustion are not only manifested at the level of energy metabolism but also profoundly affect the transcriptional and epigenetic status of T cells. In the TME, mitochondrial depolarization of CD8⁺ T cells is tightly linked to exhaustion-associated epigenetic modifications, revealing a complex interplay between mitochondrial health and T cell functional status [20]. T cell exhaustion is driven by a unique set of transcriptional programs that confer a phenotype specific to T_ex cells, including specific transcription factor activity, inhibitory receptor expression, effector functions, metabolic state, and epigenetic features [6, 60,61,62,63]. As cells enter the exhausted state, their epigenetics are remodeled, allowing them to acquire a stable exhausted phenotype. This means that PD-1 blockade therapy alone makes it difficult to adequately activate late-stage exhausted T cells or induce their conversion to T_m cells to generate a durable antitumor response [1].

Role of transcription factors in T cell exhaustion

The expression of the TOX transcription factor is key to the process of T cell exhaustion, which is activated by the TCR pathway in response to prolonged antigenic stimulation and is induced by calmodulin phosphatase-mediated activation of NFAT [64, 65]. Activation of TOX then initiates a series of cascading responses, including the activation of transcription factors such as TCF7 (encoding TCF-1) and Nr4a2, promoting the expression of inhibitory receptors such as PD-1, LAG-3, and HAVCR2, and accelerating the process of T cell exhaustion [66]. Abnormal mitochondrial function further drives TOX expression by promoting the activity of transcription factors, such as NFAT1/2, while inhibiting the production of cytokines, such as IFN-γ, TNF, and IL-2, exacerbating the state of T cell exhaustion. In addition, metabolic dysregulation and reduced adaptive mitochondria in T_ex cells further drive epigenetic changes, such as histone modification and DNA methylation, which in turn affect gene expression and cellular function [67].

Metabolic regulation and epigenetic changes

Inhibition of aerobic glycolysis in T cells has been shown to reduce the rate of the tricarboxylic acid (TCA) cycle and decrease the catabolic substrates of acetyl coenzyme A, such as acetate, citrate, or pyruvate, thereby reducing acetyl coenzyme A levels and the degree of acetylation at the histone H3 site [67]. This phenomenon was confirmed in chimeric antigen receptor T (CAR-T) cells, where studies by Si X and Jaccard et al. showed that mitochondrial isocitrate dehydrogenase 2 (IDH2) inhibits CAR-T cell function by inhibiting antioxidant metabolism and histone acetylation [68, 69].

IDH2, a key enzyme in the TCA cycle, plays a crucial role in cellular metabolism and epigenetic regulation. In CAR-T cells, IDH2 can impede the metabolic adaptation of CAR-T cells by inhibiting the pentose phosphate pathway, reducing glucose utilization, and limiting cytoplasmic levels of acetyl-coenzyme A. This, in turn, inhibits the histone acetylation process that promotes memory cell formation. IDH2 inhibitors can reverse these effects by promoting the conversion of citrate to acetyl-coenzyme A in the cytoplasm, thereby activating the epigenetic mechanism of memory differentiation in CAR-T cells [68, 69].

IDH2 mutations frequently occur in various cancer types, particularly in gliomas and acute myeloid leukemia. These mutations lead to abnormal IDH2 enzyme activity, resulting in the production of the oncometabolite 2-hydroxyglutarate (2-HG), which affects cellular metabolism and epigenetic modifications, promoting tumorigenesis and progression [70, 71]. Recent studies have shown that IDH2 mutations not only influence the metabolism and function of tumor cells but also affect the function of immune cells, particularly T cell exhaustion and immune evasion [72].

The 2-HG produced by IDH2 mutations can interfere with T cell metabolic pathways, especially glycolysis and OXPHOS. Research indicates that 2-HG can inhibit glycolysis in T cells, reducing ATP production and leading to insufficient energy supply [72, 73]. Additionally, 2-HG can disrupt glutamine metabolism, and dysregulated glutamine metabolism has been shown to promote T cell exhaustion [74]. Inhibition of 2-HG reduces T cell exhaustion in IDH-mutant gliomas and promotes the generation of memory CD8⁺ T cells [75]. Furthermore, 2-HG can interfere with the function of other immune cells, further promoting immune evasion. Specifically, 2-HG can suppress the antigen-presenting capacity of dendritic cells (DCs) and the pro-inflammatory functions of macrophages, thereby facilitating immune evasion [72, 73].

Given the negative impact of IDH2 mutations on T cell metabolism and function, combining IDH2 inhibitors with immune checkpoint inhibitors (ICIs) may offer a new approach in cancer immunotherapy. The use of IDH2 inhibitors can reduce 2-HG production, restore T cell metabolic balance, and enhance T cell effector function and persistence [69, 76]. Combining IDH2 inhibitors with ICIs, such as PD-1/PD-L1 inhibitors, can synergistically enhance anti-tumor immune responses and improve the efficacy of immunotherapy [77].

Based on the phase III INDIGO trial, the IDH1/2 inhibitor vorasidenib was approved by the U.S. Food and Drug Administration (FDA) for treating adult patients with recurrent or progressive low-grade glioma with an IDH1 or IDH2 mutation. The trial showed that vorasidenib significantly extended the progression-free survival (PFS) compared to placebo in these patients [78]. Multiple clinical trials are currently evaluating the safety and efficacy of combining IDH2 inhibitors with ICIs. The results will provide important insights for future therapeutic strategies.

ROS-Induced epigenetic modifications and their impact on T cell exhaustion

ROS not only directly affect T cell function but also lead to extensive reprogramming of T cells through the induction of epigenetic modifications, which can promote their exhaustion. These epigenetic modifications include DNA methylation, histone modifications, and regulation of non-coding RNAs, all of which can alter the expression patterns of target genes, influencing T cell differentiation and function [61, 79].

DNA methylation

DNA methylation plays a crucial role in epigenetic modification, and ROS can enhance the activity of DNA methyltransferases (DNMTs), leading to an upregulation of DNA methylation levels [80]. DNA methylation typically occurs at CpG islands, affecting the promoter regions of genes and thereby regulating gene expression. In T cells, increased DNA methylation induced by ROS can lead to the upregulation of genes associated with T cell exhaustion, such as PD-1 and TIM-3 [1]. α-Ketoglutarate (α-KG), as a precursor of several key amino acids, is involved in the TCA cycle by supporting OXPHOS while counteracting the inhibitory effects of glutamine restriction on T cell exhaustion by demethylating lysine at position 27 of H3 [81,82,83].

Histone modifications

Furthermore, ROS can cause changes in histone modifications by activating histone-modifying enzymes, such as histone deacetylases (HDACs) and histone methyltransferases (HMTs) [84]. These modifications influence chromatin structure and accessibility, thereby controlling gene expression. Previous studies have confirmed that ROS-induced histone deacetylation can suppress the expression of anti-apoptotic genes, promoting apoptosis and exhaustion in T cells [85, 86].

Non-coding RNAs

Moreover, ROS can also affect the expression and function of non-coding RNAs, such as microRNAs (miRNAs) and long non-coding RNAs (lncRNAs). miRNAs regulate the expression of target genes by binding to mRNAs. Changes in miRNA expression induced by ROS can influence T cell differentiation and function. For example, upregulation of miR-155 is associated with T cell exhaustion, while downregulation of miR-155 can restore T cell function [87]. lncRNAs can also affect T cell function through various mechanisms, such as regulating gene expression, chromatin structure, and protein stability [88].

In conclusion, mitochondria, as centers of energy metabolism, play a critical role in the transcriptional and epigenetic regulation of T cell exhaustion by linking metabolic adaptation to epigenetic regulation (Fig. 3). The link between metabolism and epigenetics cannot be ignored in the relationship between mitochondrial abnormalities and T cell exhaustion, providing important insights for the development of new epigenetic strategies to reverse T cell exhaustion and suggesting that the intrinsic effects of mitochondrial metabolism on T cell functional status should be fully considered when designing therapeutic regimens.

Targeting mitochondria to restore exhausted T cell antitumor function

By targeting T cell mitochondria metabolism, enhancing mitochondrial biogenesis, and improving mitochondrial structure and function to restore the energy supply and function of exhausted T cells, the strategies that have been applied up to date are as follows (Table 1).

Table 1 Targeting mitochondrial strategies to restore exhausted T cell function

Strategies to regulate mitochondrial metabolism

Mitochondrial metabolic remodeling promotes functional recovery of exhausted T cells

Recent studies have shown that the functional status of exhausted T cells can be effectively restored and reversed by fine-tuning T cell metabolic pathways, specifically by targeting mitochondrial metabolic remodeling. Specifically, optimizing metabolic pathways, adjusting hypoxic and nutrient conditions in the cellular microenvironment, and using metabolism-modulating drugs have all been shown to improve T cell activity and efficacy.

Role of co-stimulatory molecules

The co-stimulatory molecule CD28 not only regulates the metabolic activity of T cells but also promotes their proliferation by activating the PI3K/Akt/mTOR signaling cascade, a process that is essential for improving T cell fate [89]. In the context of renal cell carcinoma, the role of CD28 is particularly important. It reprograms the metabolic pattern of CD8⁺ TILs by upregulating GLUT3, a glucose transporter protein that increases glycolysis and mitochondrial oxidative capacity, thereby restoring T cell effector function [90].

Role of ICIs in metabolic reprogramming and T cell function

ICIs have revolutionized cancer immunotherapy by enhancing T cell activation and function. These inhibitors target various inhibitory receptors on T cells, such as PD-1, CTLA-4, and more recently, TIGIT. While PD-1 and CTLA-4 inhibitors have shown significant clinical success, there is growing interest in targeting additional immune checkpoints to further improve therapeutic outcomes.

TIGIT, an inhibitory receptor found on various immune cells, is emerging as a key target in the regulation of metabolic processes that influence immune responses. Several studies have reported that TIGIT inhibits the activation and function of T cells and natural killer (NK) cells by interacting with its ligand CD155 on DCs [91,92,93]. Additionally, TIGIT is expressed in terminally exhausted CD8⁺ T cell subsets, and when co-expressed with other co-inhibitory receptors, it directly suppresses T cell proliferation and function. Recent research has shown that TIGIT activation can inhibit glycolysis in T cells, leading to insufficient energy supply and promoting T cell exhaustion [94, 95]. Inhibitors of the immune checkpoint TIGIT/CD155 contribute to the reprogramming of glucose metabolism by activating the PI3K/AKT/mTOR signaling axis, which in turn activates the secretion of the cytokines IFNγ and TNFα by CD8⁺ T cells in triple-negative breast cancer, effectively counteracting the inhibitory effects of tumors on T cells [91]. It is currently unclear whether TIGIT/CD155 inhibitors exert their effects by directly impacting mitochondrial function or through indirect metabolic pathways [94].

The promising preclinical and mechanistic data on TIGIT/CD155 inhibitors have led to their incorporation into clinical trials. Dual PD-(L)1/TIGIT blockade is considered a promising cancer immunotherapy combination and has been studied in various solid tumors, including lung, liver, and cervical cancers [96, 97]. In the phase II CITYSCAPE study, dual PD-L1/TIGIT blockade (atezolizumab/tiragolumab) showed promising clinical benefits as a first-line treatment for PD-L1-positive non-small cell lung cancer (NSCLC) patients [98]. Similarly, in another phase II study, tiragolumab in combination with atezolizumab and bevacizumab improved anti-tumor responses compared to atezolizumab plus bevacizumab alone [99]. However, in two key phase III clinical trials in lung cancer(SKYSCRAPER-01 and SKYSCRAPER-02), the combination therapy did not demonstrate strong synergistic effects [100]. Ongoing Phase III clinical studies are evaluating the efficacy of TIGIT inhibitors in hepatocellular carcinoma and other solid tumors.

Modulation of glycolytic enzymes

Modulation of the expression of specific metabolic genes and the activity of glycolytic enzymes, such as lactate dehydrogenase (LDH), can significantly inhibit T cell exhaustion and enhance the antitumor effect of CD8⁺ T cells.

Recent studies have shown that LDH inhibitors, in synergy with IL-21, promote the generation of stem cell-like memory T cells (T_scm) and enhance mitochondrial adaptation, thereby effectively inhibiting T cell exhaustion [101]. These inhibitors facilitate the entry of pyruvate into the TCA cycle for OXPHOS without affecting IL-21-induced metabolic quiescence, which increases the proportion of T_scm cells and stimulates a more potent and prolonged antitumor response. LDH-A, a key isoform of LDH, catalyzes the conversion of pyruvate to lactate, which is the final step in glycolysis. Under hypoxic conditions, T cells rely on LDH-A to produce lactate to maintain energy supply. High expression of LDH-A is associated with T cell metabolic reprogramming, shifting from OXPHOS to glycolysis, thereby accelerating T cell exhaustion [102, 103]. Additionally, LDH-A regulates the HIF signaling pathway in the TME, influencing the hypoxia-induced expression of PD-L1 mediated by HIF1-α, which affects the immune response [104]. LDH inhibitors, such as GNE-140, have been shown to effectively inhibit T cell exhaustion and enhance the anti-tumor efficacy of immune checkpoint blockade (ICB) [102, 105, 106].

SLC2A1, also known as GLUT1, is a critical glucose transporter responsible for the uptake of glucose from the extracellular environment into the cell. The expression of SLC2A1 is upregulated during T cell activation to meet high energy demands. Studies have shown that high expression of SLC2A1 is associated with T cell exhaustion in chronic viral infections and the TME. In conditions of chronic antigen stimulation and adverse microenvironments, sustained high expression of SLC2A1 can lead to T cell dependence on glycolysis rather than OXPHOS, accelerating T cell exhaustion [107]. Inhibition of SLC2A1 using small-molecule inhibitors like STF-31 can reduce glucose uptake and alleviate T cell exhaustion [108, 109].

Phosphoglycerate Kinase 1 (PGK1) is a key enzyme in the glycolytic pathway, catalyzing the formation of ATP. PGK1 has also been shown to play a significant role in immune regulation in various tumors. PGK1 drives tumor progression by regulating glycolysis and mitochondrial metabolism, and its high expression is associated with poor prognosis in many types of cancers [110, 111]. During T cell exhaustion, the expression of PGK1 is upregulated, affecting T cell energy production and effector function [111]. In breast cancer, enhanced glycolytic activity regulated by PGK1 contributes to the modulation of the TME by regulating the expression of immune suppressive factors such as PD-L1 and CTLA4 [112]. PGK1 inhibitors, such as NG52, have shown inhibitory effects on tumor cell proliferation [113]. Overall, PGK1 plays a crucial role in T cell function and metabolic regulation, and changes in its activity levels may be involved in the mechanisms of T cell exhaustion. Modulating PGK1 activity could be a potential strategy to improve T cell function and enhance anti-tumor immunity.

IL-2 induces the expression of key genes involved in glycolysis, including SLC2A1, PGK1, and LDHA. This accelerates glycolysis and lactate production, which are essential for T cell activation and function. However, chronic exposure to IL-2 can also contribute to T cell exhaustion by promoting an overreliance on glycolysis [101].

In summary, modulating the activity of glycolytic enzymes such as LDH, SLC2A1, and PGK1 can significantly impact T cell metabolism and function. These enzymes play crucial roles in T cell exhaustion and the development of effective antitumor responses. Targeting these enzymes, either alone or in combination with other immunotherapies, holds promise for improving the outcomes of cancer immunotherapy.

Role of IL-10

IL-10 is a multifunctional cytokine that has been shown to promote anti-tumor immunity in various mouse tumor models [114, 115]. It not only reduces pro-tumorigenic inflammation but also plays a role in rejuvenating exhausted TILs. As previously discussed, exhausted T cells exhibit impaired mitochondrial respiration and/or glycolysis, which can exacerbate T cell exhaustion due to poor metabolic adaptability. Research has found that an extended half-life interleukin (IL)-10/Fc fusion protein reprograms the metabolism of terminally exhausted T cells by upregulating pyruvate carrier-dependent OXPHOS and downregulating key glycolytic genes, thereby improving their mitochondrial function, proliferation, and effector functions, and endowing them with renewed vigor and enhanced anti-tumor effects [116]. Sustained elevation of IL-10 serum concentrations achieved through pegylated IL-10 (pegilodecakin) also enhances cytotoxicity, expansion of tumor-specific CD8⁺ T cells, and their tumor-killing effects [117]. In several Phase I/Ib studies involving solid tumors, pegilodecakin, whether used as a monotherapy or in combination with chemotherapy or PD-1 antibodies, has demonstrated the ability to activate CD8⁺ T cells in both the blood and tumor, leading to favorable responses [118].

Improving hypoxic conditions

Improving hypoxic and hypoglycemic conditions in the TME can attenuate T cell exhaustion, as evidenced by a decrease in the expression of exhaustion markers, an increase in the expression of effector or memory-like genes, and an increase in the response rate to immunotherapy [22]. Studies have shown that inhibiting VEGF-dependent angiogenesis in the TME with axitinib can relieve T cell hypoxia and limit the exposure of tumor-infiltrating T cells to metabolic stress, significantly improving the efficacy of immunotherapy. Axitinib in combination with a PD-1 antibody has been FDA-approved for first-line treatment of advanced renal cell carcinoma [21]. Additionally, research indicates that hypoxic conditions in tumors lead to histone H3K27 methylation, which drives the terminal exhaustion program in T cells. Overexpression of the H3K27 histone demethylase Kdm6b effectively overcomes hypoxia, restores the effector functions of exhausted T cells, and promotes anti-tumor immunity [119]. This suggests that immunotherapies targeting transcription factor activity, in conjunction with strategies to alleviate hypoxia, may have a synergistic effect, making the improvement of hypoxic conditions a viable strategy to enhance immunotherapy.

Metabolic modulators

Metabolic modulators such as 2-deoxy-D-glucose (2-DG), which acts as an inhibitor of glycolysis, have been shown to reduce the expression of terminal exhaustion markers such as Tim-3 and LAG-3 in T cells while increasing T cell persistence, suggesting that the strategy of short-term inhibition of glycolysis preserves T cell exhaustion without compromising their effector functions and shows promise as a potential metabolic intervention strategy to maintain the antitumor activity of T cells [25].

In summary, regulating mitochondrial metabolism through the above-mentioned pathways to achieve metabolic reprogramming of T_ex cells may theoretically and empirically repair their impaired effector functions, reverse the exhausted state, and thus enhance their responsiveness to antitumor immunotherapy.

Enhancement of mitochondrial Biogenesis to optimize T cell function

As mentioned above, in the TME, tumor-infiltrating T cells show a gradual decrease in the expression of the transcriptional co-activator PGC-1α, which plays a central role in mitochondrial biogenesis. PGC-1α is able to co-operate with a number of transcription factors, such as PPAR and nuclear respiratory factor 1 (NRF-1), and co-regulate the expression of genes involved in mitochondrial biogenesis, metabolism, and function. These genes encode proteins, enzymes, and other regulatory factors that are key components of mitochondrial biogenesis. Overexpression of PGC-1α has been shown to promote mitochondrial biogenesis, which in turn optimizes T cell function and improves the efficacy of PD-1 blockade therapy [15].

Similarly, metabolic reprogramming of CAR-T cells by modulation of PGC-1α, when used for immunotherapy, is effective in inducing mitochondrial biogenesis, which not only activates gene expression programs associated with effector functions but also promotes a more durable memory T cell state. Treatment of an immunodeficient animal model inoculated with A549 human lung adenocarcinoma cells with CAR-T cells demonstrated significantly improved therapeutic efficacy [120]. This finding further highlights the importance of modulating mitochondrial biogenesis to enhance the antitumor function of T cells.

Regulation of antioxidants and ROS homeostasis to restore exhausted T cell function

Use of antioxidants

ROS is a by-product of cell metabolism. Moderate amount of ROS can promote the activation and proliferation of T cells, but excessive ROS can lead to cell damage and functional failure. Regulating the homeostasis of antioxidants and ROS may be an effective strategy to restore exhausted T cell function.

Exogenous antioxidants

Maintenance of normal T cell effector and memory functions requires a good balance between ROS generation and antioxidant scavenging. As mentioned above, prolonged antigenic stimulation impairs ATP production and triggers aberrant mROS generation, which in turn inhibits the self-renewal capacity of T cells and accelerates their terminal differentiation [21, 22]. However, exogenous antioxidants, such as N-acetylcysteine (NAC), effectively blocked mROS accumulation driven by persistent antigens and restored the proliferative capacity, effector functions, and expression of memory cell-related genes in T_ex cells, thereby enhancing anti-tumor immune responses in vitro and in vivo [22]. Similarly, other mitochondria-targeted antioxidants, such as Trolox (a water-soluble vitamin E analog), have been shown to restore the cytokine production capacity of chronically stimulated T cells [22]. Studies have also shown that the use of mitochondria-targeted antioxidants, such as mitoquinone (MitoQ) and piperidine nitroxide (MitoTempo), can correct MMP depolarization defects and reduce intracellular levels of ROS in HBV-specific CD8⁺ T cells, restoring the expansion capacity and cytokine secretion of these cells, and significantly increasing cell viability and antiviral activity [121].

Regulation of the antioxidant enzyme system in vivo

In addition, scavenging excessive ROS in T cells by regulating the activity of the antioxidant enzyme system in vivo may also inhibit T cell exhaustion, thereby restoring their antitumor function. CAT is a major intracellular antioxidant enzyme that maintains the intracellular ROS balance by scavenging H₂O₂. CAT supplementation can reduce mROS and overall ROS production within CD8⁺ T cells, improve T cell activation, reduce the proportion of cells in an exhausted state, and increase central memory-like phenotypes, thereby enhancing self-renewal capacity and antitumor functions in vivo, demonstrating potential long-term benefits [122].

The above evidence suggests that either the use of exogenous antioxidants or activation of the in vivo antioxidant enzyme system to eliminate excess ROS in the TME and maintain the balance between ROS and antioxidants can inhibit the T cell exhaustion program, promote the differentiation of memory-like phenotypes, and restore proliferative and effector functions, thereby enhancing anti-tumor immune effects.

Application of antioxidants in targeting ROS and clinical limitations

Reducing ROS levels in the TME can have varied effects on T cells and tumor cells [123]. Oxidative stress caused by ROS can damage mtDNA and proteins, leading to impaired OXPHOS and reduced ATP production. Antioxidants, by neutralizing ROS, alleviate oxidative stress and protect cells from damage, thereby presenting a theoretical therapeutic value [124]. Preclinical studies have shown that antioxidants restore mitochondrial function and enhance T cell energy metabolism by reducing ROS levels. This improves T cell effector function, reduces T cell exhaustion, and inhibits the expression of immune checkpoint molecules such as PD-1 and CTLA-4 on T-cells, thereby reinvigorating T-cells and enhancing their anti-tumor activity [122, 125].

Tumor cells often rely on high ROS levels to maintain their survival and proliferation. Reducing ROS levels in the TME not only restores T cell function but also adversely affects tumor cells. Specifically, lowering ROS levels can induce apoptosis in tumor cells by disrupting redox balance and activating pro-apoptotic pathways [126]. Additionally, lowering ROS levels can sensitize tumor cells to immune attack by diminishing their ability to induce immunosuppressive mechanisms. This includes the downregulation of immune checkpoint ligands and the suppression of T_regs activity, thereby creating a more favorable environment for T-cell-mediated anti-tumor responses [127].

This differential effect, where antioxidants both enhance T cell function and impair tumor cell survival, highlights the potential of antioxidant therapy to improve anti-tumor immunity and enhance the efficacy of cancer treatments. However, most research on antioxidants is based on preclinical models, and although these results are encouraging, the application of antioxidants in a clinical setting still faces numerous challenges. These preclinical models may not fully replicate the complexity and heterogeneity of human tumors. In clinical trials, the efficacy of antioxidants is often less than expected [128,129,130]. Additionally, the dose and mode of administration of antioxidants significantly impact their effectiveness. Preventive doses (low doses) can protect both normal and tumor cells, while therapeutic doses (high doses) can inhibit cancer cell growth without affecting the growth of normal cells [131]. Oral antioxidants, with lower bioavailability, might affect the efficacy of chemotherapeutic drugs, whereas intravenous administration can act quickly, better control the dosage, and potentially enhance synergistic effects with chemotherapy [132, 133]. Improper use can lead to reduced efficacy [131]. More importantly, the use of antioxidants may bring about potential off-target effects, leading to decreased efficacy or even severe adverse reactions. This is primarily because they not only act on the intended sites but may also interact with other biomolecules, causing non-specific effects or side effects, such as diminishing the effectiveness of other therapeutic drugs or directly harming healthy tissues [124].

In the treatment of NSCLC, ICIs and platinum-based chemotherapies (such as cisplatin and carboplatin) are commonly used as first-line treatments, with their efficacy confirmed in multiple clinical trials [134,135,136,137]. However, when combining antioxidants, potential interactions must be carefully considered. Antioxidants protect normal cells by reducing ROS levels, but they may also neutralize the ROS generated by platinum drugs, thereby weakening their anti-cancer effects [123, 138,139,140]. Additionally, excessive use of antioxidants may interfere with T cell activation and function, which is critical for the efficacy of ICIs [141, 142].

In summary, although antioxidants have potential advantages in targeting ROS, their clinical limitations cannot be ignored. Therefore, when considering the combination of antioxidants with other standard therapies, it is crucial to thoroughly evaluate their interactions and potential side effects to ensure optimal therapeutic outcomes. Future research should focus on optimizing combined treatment regimens and developing personalized treatment strategies to enhance efficacy and reduce side effects.

Regulation of the ROS generation pathway

By intervening in the key links of ROS generation, reducing ROS accumulation, and maintaining intracellular ROS balance, it is possible to inhibit T cell exhaustion, restore their activity, and enhance their antitumor ability.

Nicotinamide adenine dinucleotide (NAD⁺) plays an important role in redox reactions. As a mediator of redox reactions, NAD⁺ participates in glycolysis, the TCA cycle, and FAO to form NADH, which is a key hydrogen donor for the synthesis of ATP during OXPHOS in the mitochondria, accompanied by the generation of ROS. Supplementation with Nicotinamide riboside (NR) improves mitochondrial health, reduces ROS production, promotes cytokine secretion, and aids in the control of melanoma in mice [20]. Similarly, the addition of Niacinamide (NAM) blocked the development of inhibitory receptors, alleviated the suppression of IL-2 and TNFα expression in T_ex cells, and promoted T cell differentiation towards effector states, which is closely related to ROS regulation [143].

Intracellular potassium ions

Recent studies have also elucidated the relationship between intracellular potassium (K⁺) ions and ROS. It has been found that deletion of the sodium-potassium pump (Na⁺/K⁺ ATPase) in T cells and a decrease in intracellular potassium ion concentration promotes ROS accumulation, exacerbates tonic signaling, and leads to T cell exhaustion. Both antioxidant therapy and exogenous potassium ion supplementation prevented ROS accumulation and inhibited exhaustion of Na⁺/K⁺ ATPase α1 subunit (ATP1A1)-deficient T cells in vitro [144].

In conclusion, maintaining intracellular ROS at appropriate levels and preventing excessive accumulation by regulating the balance between antioxidants and ROS is essential for inhibiting T cell exhaustion in the TME. At the same time, it can also revive T_ex cells, restore their proliferative capacity and antitumor activity, and effectively enhance the immune efficacy of T_ex cells.

Remodeling T cell fate through mitochondrial quality control

Activation of mitochondrial autophagy to restore mitochondrial health

Accumulating evidence suggests that mitochondrial dynamics play a critical role in driving T cell phenotypic and functional transitions and can influence the fate trajectory of CD8⁺ T cells [14]. Therefore, mitochondrial quality control may be an emerging strategy for restoring T_ex cells. As previously described, TILs in the TME exhibit attenuated mitochondrial autophagic activity, leading to the accumulation of depolarized mitochondria, which in turn contributes to terminal T cell exhaustion [20]. Mitochondrial autophagy is potentially beneficial in counteracting T cell exhaustion by removing damaged or inefficient mitochondria and promoting the generation of new mitochondria. Recent studies have shown that urushiol A, a mitochondrial autophagy enhancer, improves the adaptation of CD8⁺ T cells in the TME by activating Pink1-dependent mitochondrial autophagy in T cells and promoting the expansion of T_scm cells, thereby significantly enhancing the persistence and anti-tumor activity of CD8⁺ cytotoxic T lymphocytes. This effect is closely associated with a decrease in MMP, an increase in CD95 expression, and activation of the ERK1/2-ULK1 axis, highlighting the potential of inducing mitochondrial autophagy and metabolic adaptation for intervention in the field of tumor immunotherapy [145, 146].

Regulation of mitochondrial dynamics

An imbalance in mitochondrial dynamics can lead to structural and functional dysfunction, affecting the number and distribution of mitochondria in the cell and exacerbating T cell exhaustion. Therefore, regulating mitochondrial fission and fusion may help restore T cell function. Key proteins involved in these processes, such as Drp1 and optic atrophy 1 (Opa1), are important targets for regulation.

Drp1 is the primary regulatory protein involved in mitochondrial fission, responsible for dividing the mitochondrial network into smaller fragments. In T cells, Drp1 activation can be achieved through various pathways, including phosphorylation and dephosphorylation [57]. In the TME, high levels of ROS lead to cellular damage, disrupting the balance of mitochondrial dynamics and resulting in mitochondrial fragmentation and dysfunction. Hyperactivation of Drp1 further impairs T cell energy metabolism by inhibiting MMP [57].

Opa1 is a key protein involved in the fusion of the inner mitochondrial membrane, essential for maintaining mitochondrial morphology and function. In T cells, the expression and function of Opa1 are crucial for mitochondrial integrity and energy metabolism [147]. Opa1 expression is regulated by multiple factors, including transcription factors, post-translational modifications, and MMP [148]. During T cell exhaustion, Opa1 expression is typically downregulated, leading to impaired mitochondrial fusion and decreased mitochondrial function. The loss of Opa1 also results in increased mtDNA damage and ROS production, further exacerbating T cell exhaustion [149].

Antioxidants, as an effective therapeutic strategy, can neutralize free radicals and reduce intracellular oxidative stress, thereby protecting mitochondria from damage and preserving their dynamic balance [150]. Antioxidants such as NAC and glutathione (GSH) can reduce ROS levels, inhibit the hyperactivation of Drp1, and thus decrease mitochondrial fragmentation. This helps to restore mitochondrial function, improve T cell energy metabolism, and enhance effector functions [150]. Huang et al. demonstrated that in CAR-T cells, antioxidant treatment can mitigate ΔΨm depolarization defects in exhausted CD8⁺ T cells, possibly through the regulation of Drp1, and exhibit reduced mitochondrial fragmentation, thereby enhancing the persistence and anti-tumor efficacy of CAR-T cells [151]. Additionally, antioxidants can protect Opa1 from oxidative damage by reducing ROS levels, maintaining its normal expression and function. This helps to preserve mitochondrial fusion, enhance mitochondrial function, and improve T cell survival and effector functions [148, 152, 153].

Beyond Drp1 and Opa1, other regulators play roles in T cell mitochondrial dynamics and function, providing a more comprehensive understanding of the complexity of mitochondrial dynamics in T cell exhaustion.

CD137 (4-1BB), expressed in CD8⁺ TILs, acts as a TCR-independent co-stimulatory signal that promotes the exhaustion program as well as the proliferation and terminal differentiation of CD8⁺ T cells [154]. Williams et al. successfully restored exhausted T cells co-expressing 4-1BB and LAG-3 by using anti-LAG-3 in combination with 4-1BB monoclonal antibody, enhancing their anti-tumor immune response [155]. Teijeira et al. observed that stimulation of 4-1BB activation resulted in restoration of mitochondrial mass and function in human and mouse CD8⁺ T cells, a process dependent on OPA-1-mediated mitochondrial fusion and biogenesis, thereby increasing mitochondrial mass and MMP in CD8⁺ T cells, demonstrating that mitochondrial morphology and function are critically linked to the enhanced antitumor effector activity of CD137 co-stimulated T cells [156].

Serine/threonine kinase liver-associated kinase B1 (LKB1) has been implicated as a key regulator of the mitochondrial membrane disruption caused by OPA-1 deficiency. By regulating TCA cycle metabolism and transcriptional remodeling, LKB1 affects the effector functions of Th17 cells, and deletion of LKB1 restores IL-17 expression in mitochondrial membrane-damaged Th17 cells, suggesting that the LKB1-mitochondrial axis helps T cells adapt to different immune and organizational environments and may be a potential target for future mitochondrial therapy [157].

Complement C1q binding protein (C1QBP) is a mitochondrial protein that affects mitochondrial biogenesis through the AMPK/PGC-1α signaling pathway and mitochondrial morphology by increasing Drp1 expression and phosphorylation at Ser616. Mice with knockdown of the C1QBP gene had more dysfunctional T_ex cells and higher levels of inhibitory receptor expression. This demonstrates that improving T cell mitochondrial plasticity and metabolic adaptation through C1QBP enhances the antitumor function of CAR-T cells, providing a novel way to improve their antitumor immunotherapeutic efficacy [158].

Novel insights: potential drug development and clinical implications

Development of potential drugs targeting mitochondria

Although existing research has highlighted the critical role of mitochondrial dysfunction in T cell exhaustion, several unresolved questions and potential new hypotheses remain to be explored. Here are some potential directions for drug development:

Restoring T cell glycolysis and OXPHOS function by targeting mitochondrial metabolic enzymes is a promising direction. Preclinical studies have shown that targeting enzymes such as hexokinase, pyruvate kinase, and complexes I, II, III, and IV can restore T cell metabolism, thereby enhancing their effector function and persistence [22, 159, 160]. Developing small-molecule drugs that specifically target these metabolic enzymes could benefit patients who do not respond well to current treatments.

Factors such as DNA methyltransferase 3 A (DNMT3A) and ROS accumulation play significant roles in T cell exhaustion [21, 159, 161]. Targeting these metabolic and epigenetic regulators to modulate T cell function is a promising strategy to enhance T cell immune function. By regulating T cell metabolism, such as enhancing FAO, promoting mitochondrial biogenesis [120], reducing oxidative stress, modulating miRNA expression [162], and regulating MMP and calcium ion balance [163], it is possible to improve T cell function and, consequently, the efficacy of immunotherapy. These strategies may be particularly relevant for patients with T cell exhaustion due to impaired mitochondrial function.

Combination therapy strategies show great potential in overcoming the limitations of current treatments and enhancing the effectiveness of immunotherapy. New mitochondrial-targeted drugs, such as mtDNA repair agents and mitochondrial membrane stabilizers, along with metabolic modulators (e.g., PDK inhibitors, FASN inhibitors), epigenetic modifiers (e.g., DNA methyltransferase inhibitors (DNMTi), lysine methyltransferase inhibitors (KMTi)), and signaling pathway modulators (e.g., AKT inhibitors, STAT3 inhibitors), are key areas for future development. Combining these drugs with PD-1/PD-L1 inhibitors or other novel ICIs (e.g., LAG-3 and TIGIT inhibitors) and CAR-T cell therapies can target multiple pathways, including mitochondrial structure, function, metabolism, and signaling. This multifaceted approach can improve and restore T cell function, thereby enhancing the anti-tumor immune response.

Future research should further investigate the specific mechanisms of these new targets and combination therapy strategies, develop novel drugs, and evaluate their safety and efficacy in preclinical and clinical trials. This will provide more effective tools for cancer immunotherapy. For more details on ongoing and upcoming clinical trials, as well as those completed but awaiting publication, see Supplementary Table 1.

Application and emerging therapeutic strategies in clinical trials

Progress in the clinical trials of mitochondrial-targeted CAR-T therapy

CAR-T cell therapy involves the genetic engineering of T cells to express chimeric antigen receptors (CARs), enabling them to specifically recognize and kill tumor cells expressing specific antigens. This approach not only enhances the effector function and persistence of T cells but also addresses the limitations of traditional chemotherapy and radiotherapy in relapsed or refractory cases. Current clinical trials in CAR-T cell therapy are focused on several key areas:

Enhancing T cell effector function and persistence: Second-generation CAR-T cell therapies incorporate co-stimulatory molecules (such as CD28 or 4-1BB) into the CAR structure, significantly enhancing T cell activation and persistence. This enhancement is closely related to improvements in mitochondrial quality and function, allowing T cells to remain active for longer periods in the body, thereby more effectively recognizing and killing tumor cells. Multiple clinical trials have shown that second-generation CAR-T cell therapies demonstrate significant efficacy and acceptable safety in B-cell acute lymphoblastic leukemia (B-ALL) and diffuse large B-cell lymphoma (DLBCL). For example, in the ZUMA-1 trial (NCT02348216), KTE-C19 (axicabtagene ciloleucel) achieved a complete remission rate of 39% in patients with DLBCL. Compared to first-generation CAR-T cells, the addition of co-stimulatory domains in CARs has significantly improved treatment outcomes [164].

Third-generation CAR-T cell therapies further optimize the CAR structure by including a single-chain variable fragment (scFv), a CD3ζ domain, and two tandem co-stimulatory domains. Preclinical studies suggest that third-generation CAR-T cells further enhance T cell persistence and effector function. However, in small-scale clinical trials, third-generation CAR-T cells did not show significantly superior anti-tumor activity compared to second-generation CAR-T cells [165, 166].

Exploring new immune cell types: Chimeric Antigen Receptor Natural Killer (CAR-NK) cell therapy is an emerging immunotherapy that genetically engineers NK cells to specifically recognize and kill tumor cells. Unlike CAR-T therapy, NK cells do not require prior activation and can rapidly recognize and kill tumor cells without specific antigen stimulation, with better tolerability and fewer treatment-related side effects. This therapy has shown good safety and preliminary efficacy in patients with relapsed or refractory hematologic malignancies and solid tumors [167,168,169]. Studies indicate that tumor-infiltrating NK cells have small and fragmented mitochondria, limiting their tumor immune surveillance [170]. Umbilical cord blood-derived CAR-NK cells, when expanded with “feeder cells” and/or cytokines, exhibit enhanced metabolic adaptability and increased cytotoxicity against CD19-positive tumors [171]. Future efforts to improve the effector function and persistence of CAR-NK cells may focus on enhancing mitochondrial metabolic adaptability, addressing the limitations of traditional immunotherapies in certain tumor types and providing new treatment options for patients.

For more detailed information on the current status of ongoing and planned CAR-T therapy clinical trials, including those completed but not yet published, see Supplementary Table 2.

Other emerging strategies targeting mitochondria

Several emerging therapeutic strategies targeting mitochondria, such as metabolic modulators, mitochondria-targeted drugs, antioxidants, epigenetic modifiers, and combination therapies with ICIs or CAR-T, are being explored. These strategies aim to alter T cell metabolism, protect mitochondrial function, optimize CAR-T cell mitochondrial function, alleviate immune suppression, reduce the proportion of exhausted T cells, and restore and enhance T cell effector function, thereby improving the efficacy of immunotherapy.

Metabolic Modulators: Multiple Phase I/II clinical trials have shown that the pyruvate dehydrogenase kinase (PDK) inhibitor dichloroacetate (DCA), used alone or in combination with chemoradiotherapy, demonstrates good safety and preliminary efficacy in some patients with advanced solid tumors [172, 173]. In addition to PDK inhibitors, other metabolic modulators have shown promise. The PI3Kδ inhibitor idelalisib demonstrated significant activity in the Phase II DELTA trial and has been approved by the FDA for the treatment of relapsed chronic lymphocytic leukemia. Furthermore, in vitro studies have shown that PI3Kδ inhibitors can improve T cell expansion and CAR-T cell function, suggesting their potential as a complementary therapy to enhance the effectiveness of CAR-T cells [174, 175].

Epigenetic Modifiers: Several clinical studies have confirmed that DNMTi combined with PD-1 inhibitors show good safety and preliminary efficacy in some patients with advanced solid tumors [176,177,178,179,180]. Another type of epigenetic modifier, KMTi, such as EZH2 inhibitors, can restore the expression of suppressed genes by inhibiting the methylation of histone, thereby showing potential anti-tumor effects. A pilot study by Hussain et al. (ETCTN 10183) evaluated the combination of tazemetostat (an EZH2 inhibitor) and pembrolizumab in advanced urothelial carcinoma. Preliminary results showed that the combination was well tolerated and induced durable responses in chemotherapy-resistant patients with poor prognosis, providing valuable insights into the benefits and risks of this combined therapy [181].

Combination with ICIs: Multiple Phase I/II clinical trials in advanced solid tumors, such as melanoma, have shown that LAG-3 inhibitors in combination with PD-1 inhibitors can reduce oxidative stress damage in T cells, protect mitochondrial function, and thereby enhance T cell effector function and persistence. This combination therapy, particularly in patients with high LAG-3 expression, has demonstrated higher objective response rates and longer PFS [182].

Several Phase I/II clinical studies have evaluated the safety and efficacy of TIGIT and PD-1/PD-L1 co-inhibition in various cancers [96,97,98], showing that this multi-target combination therapy exhibits good anti-tumor activity in some patients, particularly in hepatocellular carcinoma and melanoma [183, 184]. The combination activates optimal T cell anti-tumor function mediated by CD226, relieving TIGIT and PD-1-mediated T cell inhibition, and further enhancing T cell anti-tumor activity [185].

Future directions and perspectives

Advancements in genetic and cellular engineering, particularly CRISPR-Cas9 technology, offer new opportunities to address the challenges of immune resistance and T cell exhaustion. CRISPR-Cas9 can precisely modify specific sequences in the genome, enabling the repair of mtDNA mutations and restoration of mitochondrial function. It can also regulate genes associated with mitochondrial biogenesis (e.g., PGC-1α, mitochondrial transcription factor A (TFAM), NRF1) and dynamics (e.g., Drp1, MFN1/2, Opa1), thereby enhancing mitochondrial number and function and restoring homeostasis [20, 157, 186, 187].

CAR-T cell therapy has shown remarkable success in treating hematological malignancies such as lymphoma, multiple myeloma, and leukemia [188,189,190,191]. However, its effectiveness against solid tumors is limited by rapid exhaustion of CAR-T cells within the TME. This exhaustion is characterized by reduced proliferation and effector function, and is further exacerbated by the harsh conditions of the TME, including hypoxia and high levels of ROS. These factors significantly impact the persistence and functionality of CAR-T cells.

To enhance the effectiveness of CAR-T cells in solid tumors, optimizing their manufacturing from metabolic and epigenetic perspectives is crucial. Such optimization can inhibit exhaustion and promote the generation of more effective or long-lived memory cells, thereby improving anti-tumor efficacy and longevity [187, 192]. Recent studies have shown that enhancing the mitochondrial function of CAR-T cells using CRISPR-Cas9 technology can significantly improve their survival and killing capacity in the TME [192]. Additionally, epigenetic interventions during the manufacturing process can prevent exhaustion or induce a memory phenotype, making CAR-T cells more effective at targeting and eliminating tumor cells [187].

Moreover, some tumor cells express mitochondria-related metabolic enzymes, and CAR-T cells can be engineered to specifically target these enzymes [193]. Combining CRISPR-Cas9 with CAR-T cell technology can further enhance CAR-T cell function by repairing mitochondrial dysfunction and improving metabolic adaptability and persistence, thus enhancing therapeutic efficacy [194].

Continued research is needed to discover more mitochondria-targeted drugs and develop innovative technologies and strategies. For example, antioxidants, NAD⁺ supplements, glycolysis inhibitors (e.g., 2-deoxyglucose), Akt inhibitors, PGC-1α agonists, and mitochondrial protectors like coenzyme Q10 (ubiquinone) [195], PQQ [196, 197], and resveratrol [198] must be empirically validated. These studies are anticipated to open new therapeutic avenues and advance our understanding of the immune system’s complexity, potentially leading to revolutionary breakthroughs in cancer therapy.

Conclusion

Although immune checkpoint blockade therapy has shown significant clinical benefits in treating advanced tumors, challenges such as immune resistance persist. Emerging evidence indicates that mitochondrial dysfunction plays a critical role in T cell exhaustion. This review highlights the dysregulation of energy metabolism, mitochondrial dynamics, and ROS levels, and their impact on T cell function. Strategies to restore mitochondrial function, such as using CRISPR-Cas9 technology and optimizing CAR-T cell therapy, show promise in enhancing antitumor immune efficacy. Further research is essential to identify more mitochondria-targeted drugs and to develop innovative technologies. These efforts are expected to open new therapeutic avenues and advance our understanding of the immune system’s complexity, leading to breakthroughs in cancer therapy.