The cGAS-STING pathway and mitochondrial metabolism: from mechanistic insights to therapeutic potential in tumor

Abstract

Background

The cyclic GMP-AMP synthase (cGAS)-stimulator of interferon genes (STING) pathway serves as the core cytoplasmic DNA-sensing signaling pathway and represents a pivotal node within the immune-metabolic regulatory network through its coordination with mitochondrial metabolism.

Main body

Emerging evidence suggests that the cGAS-STING pathway plays a dual role in the tumor microenvironment (TME), promoting both anti-tumor immunity and tumor progression. These opposing effects are governed by the spatiotemporal dynamics of STING activation and the metabolic state of mitochondria. Recent studies have revealed that the clinical efficacy of STING agonists is constrained not only by the induction of chronic inflammatory responses but also by tumor metabolic heterogeneity. A comprehensive understanding of the mechanistic crosstalk between the cGAS-STING pathway and mitochondrial metabolism could provide a robust theoretical foundation for developing novel STING-targeted therapeutic strategies based on metabolic modulation.

Conclusion

This review provides a systematic overview of the interactive regulatory mechanisms connecting the cGAS-STING pathway with mitochondrial metabolism and assesses the translational potential of targeting this immune-metabolic axis in the context of precision oncology. The insights discussed herein are intended to support the rational optimization and clinical application of STING agonists.

Introduction

With the success of immune checkpoint inhibitors (ICIs), immunotherapy has become a standard treatment modality and fundamentally changed the treatment paradigm for cancer [1, 2]. Although the remarkable efficacy of ICIs is demonstrated in select patient cohorts, the overall response rates remain limited, underscoring the complexity of tumor immune evasion mechanisms [3]. A systematic dissection of the immune regulatory network within the tumor microenvironment (TME)—particularly the crosstalk between innate immunity and metabolic reprogramming—is thus critical for devising more effective combinatorial strategies [4].

The cGAS-STING pathway serves as a central cytoplasmic DNA sensor in the TME, detecting aberrant tumor-derived DNA, including nuclear DNA from genomic instability or mitochondrial DNA (mtDNA) leaked during cellular stress. This recognition triggers a signaling cascade that culminates in type I interferon (IFN-I) and proinflammatory cytokine production. While the pathway contributes to anti-tumor immune surveillance, it paradoxically facilitates immune evasion and TME remodeling, rendering it an attractive yet challenging therapeutic target [5, 6]. Advances in immunometabolism research have further revealed the dynamic interplay between cGAS-STING signaling and mitochondrial metabolism, positioning this axis as a pivotal regulatory node in the immune-metabolic network. This interaction critically influences tumor initiation, progression, and therapeutic susceptibility by modulating energy metabolism, redox homeostasis, and immune cell functionality [4, 7–9].

Notably, the cGAS-STING pathway exhibits spatiotemporal regulation and functional duality in tumorigenesis. Transient STING activation enhances the maturation of dendritic cells (DCs), promotes CD8⁺ T cell infiltration, and drives tumor regression. Conversely, chronic or excessive activation may induce immune tolerance and foster immunosuppressive microenvironments [10]. This dichotomy stems from persistent STING-driven inflammation and mitochondrial metabolic adaptations, including mtDNA release, tricarboxylic acid cycle (TCA) intermediate accumulation, and reactive oxygen species (ROS) fluctuations—collectively termed metabolic reprogramming [11, 12]. Such complexity underlies the limited clinical efficacy of STING agonists (e.g., ADU-S100, MK-1454), as tumor metabolic heterogeneity differentially modulates STING activation thresholds and therapeutic outcomes [13, 14].

Therefore, systematic elucidation of the mechanistic crosstalk between the cGAS-STING pathway and mitochondrial metabolism could advance our understanding of the metabolic underpinnings of tumor immune evasion and offer novel therapeutic opportunities to enhance STING-targeted therapies. These may include combinatorial metabolic modulation or spatiotemporally controlled pathway activation to overcome current translational challenges [10]. This review comprehensively examines the reciprocal regulatory network linking cGAS-STING signaling with mitochondrial metabolic reprogramming, while evaluating its clinical implications for precision oncology to accelerate the development of STING agonists.

The cGAS-STING pathway

The cGAS-STING pathway represents a highly conserved innate immune surveillance mechanism. Central to this pathway is cGAS, a nucleotidyltransferase family member that contains an unstructured N‑terminal domain and a C‑terminal catalytic domain. cGAS localizes to both the cytoplasm and the nucleus under steady‑state conditions, where it is maintained in an autoinhibited state through binding to nucleosomes [15–18]. Under pathological conditions such as genomic instability, mitochondrial stress, or microbial infection, double-stranded DNA (dsDNA) accumulates aberrantly in the cytosol [19–21]. cGAS undergoes conformational activation upon dsDNA binding, catalyzing the synthesis of the secondary messenger 2′3′-cyclic GMP-AMP (cGAMP) from ATP and GTP substrates. STING is an endoplasmic reticulum (ER)‑localized, dimeric, membrane‑resident protein comprising a short N‑terminal cytosolic region, four transmembrane helices, and a cytosolic ligand‑binding domain with an unstructured C‑terminal tail [22]. The secondary messenger cGAMP engages the ER-localized adaptor protein STING, triggering its oligomerization and ER-to-Golgi apparatus (GA) trafficking [23]. The activated STING complex recruits and activates TANK-binding kinase 1 (TBK1), which phosphorylates the C‑terminal domain of STING and facilitates the recruitment of interferon regulatory factor 3 (IRF3). Phosphorylated IRF3 forms homodimers that translocate to the nucleus to initiate transcription of IFN-I and various inflammatory cytokines, including tumor necrosis factor-alpha (TNF-α), interleukin-1 beta (IL-1β), and interleukin-6 (IL-6) [24, 25] (Fig. 1).

Fig. 1.

cGAS-STING signaling pathway and its anti-tumor immune response. As a cytoplasmic DNA sensor, cGAS can recognize abnormal dsDNA from bacteria, tumor cells, and other sources. Upon recognition, cGAS catalyzes the synthesis of cGAMP from ATP and GTP. Subsequently, cGAMP activates the adaptor protein STING, inducing conformational changes, oligomerization, and trafficking from the endoplasmic reticulum to the Golgi apparatus. In the Golgi apparatus, STING further recruits and activates downstream signaling components, including TBK1 and IRF3. The activated IRF3 dimer then moves to the nucleus, where it induces the transcription of IFN-I and other inflammatory cytokines. The cGAS-STING signaling pathway promotes the maturation and migration of DCs, induces repolarization of TAMs, activates NK cells, and enhances infiltration of CTLs, thereby inducing anti-tumor immune responses. Furthermore, under conditions such as DNA damage, STING can also trigger inflammatory responses and anti-tumor immunity via non-canonical pathways. (Abbreviations: ATM: Ataxia telangiectasia mutated, cGAMP: 2’3’-cyclic GMP-AMP, cGAS: cyclic GMP-AMP Synthase, COP I/II: Coat Protein Complex I/II, CTLs: cytotoxic T lymphocytes, DCs: dendritic cells, dsDNA: Double-Stranded DNA, ER: Endoplasmic Reticulum, IFI16: Interferon gamma-inducible protein 16, IFN: Interferon, IRF3: Interferon Regulatory Factor 3, mtDNA: Mitochondrial DNA, NF-κB: nuclear factor kappa-B, NK cells: natural killer cells, PARP-1: Poly ADP-ribose polymerase 1, STING: Stimulator of Interferon Genes, TAM: tumor-associated macrophage, TBK1: TANK-Binding Kinase 1, TRAF6: TNF Receptor-Associated Factor 6, Ub: Ubiquitin)

In the TME, cGAS-STING pathway activation induces IFN-I production, which promotes the maturation and antigen‑presenting capacity of DCs and enhances CD8⁺ T cell‑mediated anti-tumor responses against tumors expressing MHC class I [26, 27]. Several studies have demonstrated that the cGAS-STING pathway triggers the robust secretion of cytokines and chemokines, such as IL-2 and CXCL9, thereby enhancing natural killer (NK)-cell chemotaxis to tumor sites and augmenting their cytotoxic activity [28]. Moreover, cGAS-STING activation can drive the repolarization of TAMs from a pro-tumorigenic M2 phenotype to an anti-tumor M1 phenotype, thereby reinforcing immune-mediated tumor suppression [29]. Through these mechanisms, the cGAS-STING pathway bridges innate and adaptive immunity, eliciting a more potent anti-tumor immune response. However, monotherapy with STING agonists often leads to adaptive resistance due to excessive signaling activation (usually termed “signal overload”). Mechanistic studies indicate that aberrant STING activation drives the pathological expansion of regulatory B cells (Bregs), which secrete IL‑35 to significantly suppress NK‑cell cytotoxicity, thereby contributing to the immunosuppressive tumor microenvironment [30]. Concurrently, abnormally activated STING signaling engages the IκB kinase (IKK) complex, leading to phosphorylation of the NF‑κB inhibitor IκBα. This event triggers the poly‑ubiquitination and subsequent proteasomal degradation of IκBα, leading to the release, nuclear translocation, and transcriptional activation of nuclear factor kappa-B (NF-κB), a cascade that further promotes the formation of an immunosuppressive microenvironment [31]. These coordinated responses critically shape the immunologic landscape: controlled STING activation establishes an immunogenic microenvironment that suppresses tumor growth, whereas chronic activation paradoxically promotes pro-tumorigenic inflammation and metastatic progression [24, 32–34].

In addition to the well-characterized canonical activation pathways, emerging research has uncovered multiple non-canonical mechanisms of STING activation. Upon cGAMP binding, STING directly engages with the protein kinase R-like endoplasmic reticulum kinase (PERK), initiating signaling via the PERK-eukaryotic translation initiation factor 2 alpha (eIF2α) axis. This cascade promotes a translational program that enhances inflammatory responses and supports cell survival [35, 36]. Notably, in the context of DNA damage, the interferon-inducible protein 16 (IFI16) assembles a signaling complex with DNA damage response factors ataxia telangiectasia mutated (ATM) and poly ADP-ribose polymerase-1 (PARP-1). This complex facilitates STING ubiquitination through TNF receptor-associated factor 6 (TRAF6), leading to NF-κB pathway activation independent of the canonical pathway, thereby contributing to pre-metastatic niche formation during tumor progression [15, 37]. Clinically relevant findings from lung cancer and melanoma models demonstrate that the anti-tumor agent Clofarabine activates the non-canonical STING-NF-κB pathway by inducing direct p53-STING interactions. This activation triggers the secretion of chemokines such as CCL5 and CXCL10, culminating in tumor cell pyroptosis and immunogenic cell death (ICD) [38]. Together, these discoveries significantly broaden the scope of the STING signaling network and establish a molecular framework for developing novel anti-tumor therapeutics.

In summary, the cGAS-STING pathway plays a dual role in regulating tumor immunity, operating through both canonical and non-canonical activation mechanisms. Advances in immunometabolism highlight the bidirectional regulation of this pathway, emphasizing the need to explore its interplay with metabolic reprogramming. Such investigations may pave the way for innovative anti-tumor strategies rooted in immune-metabolic modulation, offering promising solutions to current challenges in translating STING-targeted therapies into clinical practice.

The cGAS-STING pathway and crosstalk with mitochondrial metabolism

Mitochondria serve as central orchestrators of cellular energy metabolism and signal transduction, engaging in a dynamic bidirectional relationship with the cGAS-STING pathway. Within the TME, stressors such as hypoxia, nutrient deprivation, and therapeutic interventions drive mitochondrial metabolic reprogramming in tumor cells. This adaptive response not only meets the bioenergetic demands of tumor cells but also regulates innate immune responses through diverse mechanisms [39, 40]. Mounting evidence indicates that mitochondrial stress induces ROS production and mtDNA release. These released mtDNA molecules function as damage-associated molecular patterns (DAMPs), which are detected by cGAS, thereby initiating STING-dependent inflammatory signaling [9, 40, 41]. Conversely, cGAS-STING pathway activation further reshapes mitochondrial metabolism, perpetuating metabolic reprogramming and fostering a self-reinforcing cycle that promotes tumorigenesis. This metabolic-inflammatory feedback loop extends beyond the traditional model of chronic inflammation-driven tumor progression, unveiling a novel mechanism by which the STING pathway exacerbates tumor malignancy through direct modulation of organelle function. This section comprehensively examines the intricate regulatory interplay between core mitochondrial metabolic pathways—such as oxidative phosphorylation (OXPHOS), the TCA cycle, fatty acid oxidation (FAO), and one-carbon metabolism—and the cGAS-STING signaling axis (Fig. 2). Furthermore, it explores potential therapeutic strategies targeting these metabolic nodes for cancer treatment.

Fig. 2.

cGAS-STING signaling pathway and the crosstalk with mitochondrial metabolism. (Abbreviations: 4-OI: 4-octyl itaconate, ACSL1: acyl-CoA synthetase long-chain family member 1, cGAS: cyclic GMP-AMP synthase, ETC: electron transport chain, FAD: flavin adenine dinucleotide, FADH: flavin adenine dinucleotide (reduced), FAO: fatty acid β-oxidation, G-6-P: glucose-6-phosphate, F-6-P: fructose-6-phosphate, GLUT1: glucose transporter type 1, GSH: glutathione, HIF-1α: hypoxia-inducible factor 1-alpha, HK: hexokinase, KEAP1: Kelch-like ECH-associated protein 1, mtDNA: mitochondrial DNA, NAD⁺: nicotinamide adenine dinucleotide (oxidized), NADH: nicotinamide adenine dinucleotide (reduced), NRF2: nuclear factor erythroid 2-related factor 2, OXPHOS: oxidative phosphorylation, PFK-1: phosphofructokinase-1, PFKFB3: 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase 3, PKM2: pyruvate kinase M2, PPAR-γ: peroxisome proliferator-activated receptor gamma, ROS: reactive oxygen species, SENP3: SUMO-specific protease 3, STING: stimulator of interferon genes, TCA: tricarboxylic acid, TBK1: TANK-binding kinase 1)

Warburg effect and oxidative phosphorylation (OXPHOS)

The “Warburg effect” was first proposed in 1956 and has long been regarded as a hallmark of tumor metabolism, characterized by the preferential reliance of cancer cells on glycolysis over OXPHOS even in the presence of oxygen (Fig. 3) [43, 44]. However, emerging evidence indicates that many cancer cells retain fully functional mitochondria and maintain the capacity for efficient energy production via OXPHOS [45, 46]. These observations underscore the metabolic plasticity and heterogeneity of tumor cells, necessitating a reassessment of the classical Warburg hypothesis. A deeper exploration of the dynamic interplay between glycolysis and OXPHOS provides critical insights into the mechanisms driving tumorigenesis and progression [47, 48].

Fig. 3.

Differences between oxidative phosphorylation, glycolysis, and aerobic glycolysis (Warburg effect)

Notably, the cGAS-STING pathway has been implicated in modulating this metabolic equilibrium, albeit with cell type-specific effects. In immune cells within the TME—such as DCs and macrophages—STING activation stabilizes hypoxia-inducible factor 1-alpha (HIF-1α) and upregulates key glycolytic enzymes, including Glucose Transporter Type 1 (GLUT1), hexokinase 2 (HK2), and pyruvate kinase M2 (PKM2). Concurrently, it enhances IRF3-mediated induction of 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase 3 (PFKFB3) and phosphofructokinase-1 (PFK-1), establishing a self-reinforcing feedback loop [12, 49, 50]. This metabolic reprogramming suppresses OXPHOS while augmenting glycolytic flux, which enables immune cells to thrive under hypoxic conditions. Crucially, the ATP generated through glycolysis sustains the TBK1-IRF3 signaling axis, ensuring robust interferon production and amplifying anti-tumor immune responses [12, 50–52].

In contrast, the functional role of STING in tumor cells diverges significantly. Studies reveal that STING can impede the mitochondrial localization of HK2, thereby attenuating glycolytic activity. This phenomenon correlates with the observed inverse relationship between high STING expression and diminished glycolytic flux in human colorectal cancer [50]. Thus, the spatiotemporal regulation of the glycolysis-OXPHOS balance by STING offers novel perspectives on the crosstalk between tumor immunity and metabolism, while also laying a theoretical foundation for the development of targeted therapeutic interventions.

Tricarboxylic acid cycle (TCA)

The interplay between TCA cycle intermediates and the STING pathway constitutes a pivotal regulatory nexus in immune metabolism within the tumor microenvironment. As the central hub of cellular energy production, the TCA cycle reprogramming not only meets the bioenergetic demands of tumor cells but also actively modulates innate immune signaling through its key metabolites [44]. Notably, succinate, fumarate, and itaconate have emerged as critical immunomodulatory molecules with distinct functional roles [53].

Aberrant succinate accumulation is a hallmark of metabolic dysregulation in the TME. Interestingly, recent studies have revealed that STING pathway activation in infected macrophages elevates succinate levels [12], underscoring the multifaceted role of STING in immune-metabolic crosstalk. Fumarate exhibits a more complex regulatory mechanism: while mitochondrial fumarate accumulation induces membrane hyperpolarization and subsequent mtDNA release [53, 54]—thereby activating innate immunity—fumarate derived from adenylosuccinate lyase (ADSL) competitively inhibits cGAMP-STING binding, paradoxically promoting tumor progression [55]. This dichotomy suggests that fumarate’s immunomodulatory effects are compartment-specific.

Itaconate, a pivotal immunometabolite derived from the decarboxylation of cis-aconitate by aconitate decarboxylase 1 (ACOD1), exhibits precise bidirectional modulation of the cGAS-STING pathway through multifaceted mechanisms. Itaconate inhibits succinate dehydrogenase (SDH), disrupting the TCA cycle and inducing mitochondrial dysfunction, which promotes mtDNA leakage and indirectly activates the cGAS-STING pathway [56]. Additionally, itaconate covalently modifies Kelch-like ECH-associated protein 1 (KEAP1), leading to the activation of the nuclear factor erythroid 2-related factor 2 (NRF2) pathway and subsequent suppression of STING-dependent inflammatory responses [57, 58]. Notably, the STING pathway itself can induce itaconate production, forming a tightly controlled feedback loop that fine-tunes immune activation [59]. Preclinical studies highlight the therapeutic potential of the itaconate derivative 4-octyl itaconate (4-OI), which targets the palmitoylation site (Cys91) of STING [60]. This mechanism provides a novel strategy for metabolic immunotherapy, particularly in modulating the chronic inflammatory tumor microenvironment, with promising implications for cancer treatment.

Furthermore, Nicotinamide Adenine Dinucleotide (NAD⁺), a key coenzyme involved in the TCA cycle, has been shown to regulate the activation of the cGAS-STING pathway. In the Alzheimer’s disease mouse model, decreased NAD⁺ levels lead to the activation of the cGAS-STING pathway, which may be attributed to the critical role of NAD⁺ in DNA repair mechanisms [61, 62]. In the chronic doxorubicin-induced cardiotoxicity mouse model, the cGAS-STING pathway in cardiac endothelial cells enhances the NADase activity of cluster of differentiation 38 (CD38), thereby reducing NAD⁺ levels, inducing mitochondrial dysfunction, and ultimately modulating NAD⁺ homeostasis and mitochondrial bioenergetic metabolism [63].

Collectively, the reciprocal regulation between STING signaling and TCA cycle metabolites forms an intricate network central to tumor immune metabolism, providing a rationale for the development of metabolism-targeted immunotherapies.

Fatty acid oxidation (FAO)

Fatty acid oxidation (FAO) is a critical driver of tumor progression, acting in concert with lipogenesis to sustain tumor proliferation and resistance to therapy. Therapeutically, inhibiting FAO synergizes with chemotherapy to overcome drug resistance, highlighting its clinical relevance [64–67].

At the molecular level, acetyl coenzyme A (acetyl-CoA) homeostasis—governed by carnitine acetyltransferase (CRAT)—links FAO to innate immunity. CRAT deficiency shifts metabolism toward bile acid synthesis, accumulating intermediates that induce mtDNA leakage and cGAS-STING activation [68]. Similarly, the inhibition of ATP-citrate lyase (ACLY) triggers the peroxidation of polyunsaturated fatty acids (PUFAs), compromising mitochondrial integrity and activating cGAS-STING-mediated inflammation. These findings establish a direct mechanistic connection between metabolic perturbation and inflammatory signaling [69].

TBK1 emerges as a linchpin in this network: unphosphorylated TBK1 scaffolds acyl-CoA synthetase long-chain family member 1 (ACSL1) to mitochondrial membranes to promote FAO [70], whereas phosphorylation redirects ACSL1 to the endoplasmic reticulum, favoring fatty acid re-esterification and immune evasion [70–72]. Notably, STING–TBK1 activation in macrophages can induce a metabolic shift from glycolysis to FAO, which in turn dictates functional polarization states [73].

Therapeutically, the regulation of peroxisome proliferator-activated receptor gamma (PPARγ) is particularly salient. STING deficiency stabilizes PPARγ via impaired ubiquitination, driving persistent FAO gene expression [74]. Conversely, STING-activated IRF3 directly suppresses PPARγ transcription, inhibiting adipogenesis [73]. This positions PPARγ as a potential biomarker for STING activity, offering novel diagnostic and therapeutic opportunities in tumor immunometabolism.

One-carbon metabolism

Tumors frequently exhibit a marked dependency on serine to sustain one‑carbon unit production [75]. Acute serine deprivation triggers mitochondrial dysfunction, mtDNA release, and cGAS-STING-mediated anti‑tumor responses [76–80]. However, chronic ER stress impairs serine-derived glutathione synthesis, elevating ROS to attenuate STING signaling [81, 82]. A key regulatory mechanism involves UNC-51-like kinase 1 (ULK1), which phosphorylates STING at Ser366 to promote its degradation, thereby suppressing IRF3 activation and interferon responses [83]. These findings reveal a stress-responsive checkpoint linking amino acid metabolism to innate immunity, suggesting that intermittent serine restriction combined with ULK1 inhibition may offer a viable therapeutic strategy.

The cGAS-STING pathway and crosstalk with mitochondrial quality control

Mitochondrial quality control is a central regulatory mechanism for maintaining cellular metabolic homeostasis. The system orchestrates mitochondrial metabolic functions through the integration of three key biological processes: mitochondrial dynamics, autophagy, and oxidative stress responses [84, 85]. Emerging evidence highlights that dynamic alterations in mitochondrial structure and function regulate OXPHOS efficiency, TCA cycle activity, and ROS homeostasis, thereby significantly influencing the activation of the cGAS-STING pathway [86]. These discoveries establish a molecular link between mitochondrial quality control and innate immune signaling, providing a theoretical framework for understanding tumor metabolic heterogeneity and developing targeted therapeutic strategies.

Mitochondrial dynamics

The equilibrium of mitochondrial dynamics is maintained by a balance between fusion—mediated by mitofusins (MFN1/2) and optic atrophy 1 (OPA1)—and fission, driven by dynamin-related protein 1 (DRP1) [87–89]. These processes are closely tied to cellular energy metabolism. Studies demonstrate that under various stress conditions, mitochondria undergo distinct remodeling patterns: nutrient deprivation promotes fusion and enhances OXPHOS, whereas hypoxia activates DRP1-mediated fission via the HIF-1α-dependent pathway, facilitating a metabolic shift toward glycolysis [90–95]. In multiple solid tumor types, disruption of this delicate balance is closely associated with cancer progression [96, 97].

In senescent glial cells, the downregulation of MFN2 leads to mitochondrial fragmentation and mtDNA leakage, triggering chronic inflammation through the activation of the cGAS-STING pathway [98, 99]. In bone tumor models, MFN2 loss impairs mitochondrial transfer and attenuates STING signaling, promoting metastatic niche formation [100]. Thus, MFN2-mediated mitochondrial fusion may remodel the tumor microenvironment through dual mechanisms: mitigating inflammation caused by mitochondrial dysfunction and enhancing STING-dependent immune responses [101].

Conversely, DRP1-mediated fission contributes to tumor progression via multiple pathways. First, DRP1 overexpression induces mtDNA stress, activating the cGAS-STING pathway and fostering a pro-inflammatory tumor microenvironment [11]. Subsequent cGAS upregulation further promotes DRP1 oligomerization, forming a positive feedback loop that amplifies this effect [102]. Additionally, DRP1 can directly interact with cGAS to suppress mitochondrial ROS accumulation and ferroptosis, promoting tumor growth independently of STING signaling [88].

These findings highlight the importance of maintaining a precise balance between mitochondrial fusion and fission in regulating the cGAS-STING pathway. While mitochondrial fusion exhibits anti-tumor potential, excessive fusion may paradoxically restore tumor cell metabolic capacity by enhancing mitochondrial efficiency, thereby facilitating tumor progression. Consequently, precise modulation of mitochondrial dynamics equilibrium represents a promising anti-tumor strategy. Future research should focus on elucidating tumor-specific regulatory mechanisms of mitochondrial dynamics to guide the development of targeted therapies.

Mitophagy

As a critical component of mitochondrial quality control, mitophagy operates in coordination with mitochondrial dynamics to maintain metabolic homeostasis [103, 104]. When mitochondrial dynamics are disrupted, and excessive fission occurs, fragmented mitochondria that evade clearance can continuously release mtDNA, activating the cGAS-STING pathway [105]. In this process, activated STING upregulates DRP1 expression while suppressing MFN1/2 activity, promoting mitochondrial fragmentation and initiating mitophagy [106–109]. Simultaneously, cytosolic mtDNA can be sensed by cGAS. Upon activation, cGAS directly interacts with Beclin-1 independently of STING, alleviating Rubicon-mediated inhibition and accelerating autophagosome formation to eliminate cytosolic mtDNA [110].

TBK1 serves as a central hub linking mitochondrial dynamics to mitophagy, orchestrating this process through multi-layered regulation. Following ubiquitination of outer mitochondrial membrane proteins, TBK1 is recruited to damaged mitochondria, where it undergoes autophosphorylation and subsequently recruits autophagy proteins such as Focal Adhesion Kinase family interacting protein of 200 kDa (FIP200) and Autophagy-related protein 13 (ATG13) [111]. Concurrently, TBK1 phosphorylates autophagy receptors, including Optineurin (OPTN), Nuclear domain 10 protein 52 (NDP52), and p62, thereby enhancing their binding affinity for ubiquitin chains and Microtubule-associated protein 1 light chain 3 (LC3) [112–114], and facilitating the execution of mitophagy.

Under energy stress, the AMP-activated protein kinase (AMPK)-ULK1 signaling axis further activates TBK1, promoting mitophagosome formation and autophagosome maturation via DRP1 phosphorylation, ultimately clearing damaged mitochondria [115, 116]. Notably, TBK1-dependent phosphorylation of autophagy adaptors, combined with a self-amplifying feedback loop involving mitochondrial ubiquitin chains, significantly enhances mitophagic efficiency. Figure 4 illustrates the crosstalk between the cGAS-STING pathway and mitophagy [117–121].

Fig. 4.

cGAS-STING signaling pathway and the crosstalk with mitophagy. Damaged mitochondria are typically characterized by depolarization and a reduction in mitochondrial membrane potential (ΔΨm), which further impedes the translocation of PINK1 into the IMM. Consequently, PINK1 accumulates and stabilizes on the OMM. The accumulated PINK1 phosphorylates Parkin at serine 65 (Ser65) and its substrate ubiquitin (pSer65-Ub), thereby activating the E3 ubiquitin ligase activity of Parkin. Activated Parkin subsequently catalyzes the formation of polyubiquitin chains on outer mitochondrial membrane proteins, which facilitates the recruitment of autophagy receptors such as OPTN and NDP52. Through their LIR motifs, OPTN and NDP52 bind to LC3, thereby tethering ubiquitinated mitochondria to autophagosomes and enabling their selective degradation via mitophagy. Furthermore, this autophagic process interacts with the cGAS-STING signaling pathway, jointly contributing to the fine-tuned regulation of intracellular homeostasis. (Abbreviations: AMPK: AMP-Activated Protein Kinase, Atg13/14/101: Autophagy-Related 13/14/101, cGAMP: 2′3′-cyclic GMP-AMP, cGAS: cyclic GMP-AMP Synthase, DNA: Deoxyribonucleic Acid, DRP1: Dynamin-Related Protein 1, dsDNA: Double-Stranded DNA, IκB: Inhibitor of κB, IMM: inner mitochondrial membrane, IRF3: Interferon Regulatory Factor 3, LC3: Microtubule-associated protein 1 light chain 3, LIR: LC3-interacting region, MFN1/2: Mitofusin 1/2, mtDNA: Mitochondrial DNA, NDP52: Nuclear dot protein 52, NF-κB: Nuclear Factor κ-Light-Chain-Enhancer of Activated B Cells, OMM: outer mitochondrial membrane, OPTN: Optineurin, Parkin: Parkin E3 Ubiquitin Ligase, PIK3C3: Phosphatidylinositol 3-Kinase Catalytic Subunit Type 3, PINK1: PTEN-induced kinase 1, STING: Stimulator of Interferon Genes, TBK1: TANK-Binding Kinase 1, Ub: Ubiquitin, ULK1: Unc-51 Like Autophagy Activating Kinase 1, VPS15/34: Vacuolar Protein Sorting 15/34)

The elucidation of mitophagy’s dynamic regulatory mechanism offers novel insights into tumor therapeutics. Particularly under excessive fission conditions, the cGAS-STING pathway promotes mitophagy, mitigating chronic inflammation caused by mtDNA leakage. This reflects an intrinsic self-regulatory mechanism of the cGAS-STING pathway, with potential implications for modulating the tumor inflammatory microenvironment. Recent studies demonstrate that impaired mitophagy can enhance the Warburg effect and drive tumor progression [122–124], suggesting that maintaining mitophagy function may represent a key mechanism by which the cGAS-STING pathway suppresses tumor metabolic reprogramming. Future research should identify the critical regulatory determinants of STING signaling in tumor metabolic reprogramming, including its activation timing, duration, cell-type specificity, and crosstalk with other pathways, to enable the development of precise therapeutic interventions.

Oxidative stress

Oxidative stress is pivotal in maintaining mitochondrial function and determining cell fate by modulating the balance between ROS production and elimination. Under physiological conditions, ROS act as essential signaling molecules within the cellular metabolic network. In the TME, ROS exhibit spatiotemporal specificity, with compartment-specific ROS exerting unique regulatory roles [90, 125–129]. The functional outcomes of ROS signaling depend critically on their intracellular concentration, as pathological accumulation can lead to oxidative damage and genomic instability [130].

Accumulating evidence indicates that mitochondrial-derived ROS regulate cGAS-STING pathway activity through multiple mechanisms. Mitochondrial ROS can trigger mtDNA leakage, while nuclear ROS compromise nuclear membrane integrity, releasing genomic DNA fragments [131, 132]. These cytosolic DNA species activate the cGAS-STING cascade, a process that is negatively regulated by NRF2, a key transcription factor involved in antioxidant defense. NRF2 modulates Pirin expression to establish a feedback loop [133]. Under chronic inflammation or hypoxia, sustained ROS-STING activation promotes the degradation of glutathione peroxidase 4 (GPX4) and exacerbates chromatin fragmentation via the ROS-JNK pathway, thereby amplifying cGAS-STING signaling [134, 135]. This cascade not only intensifies oxidative damage but may also promote a pro-inflammatory microenvironment conducive to tumor progression.

However, some studies have found that ROS is essential for STING-mediated anti-tumor immunity. In DCs, ROS accumulation upregulates SUMO-specific protease 3 (SENP3) and enhances its interaction with IFI204, facilitating IFI204 desumoylation. This molecular event significantly strengthens STING-dependent IFN-I signaling, augmenting DC anti-tumor immune functions [136]. Leveraging this mechanism, researchers have developed nanodelivery systems, such as Mn/Fe-based nanomaterials, which synergistically induce ROS production and activate the STING pathway. These systems promote ICD and activate DCs, NK cells, and cytotoxic T lymphocytes (CTLs), eliciting potent anti-tumor immune responses [137–141]. Furthermore, the combination of glutathione (GSH) depletion and STING activation has demonstrated promising anti-tumor efficacy [142–144].

These nanocomplexes provide a novel platform for investigating ROS-cGAS-STING interplay in the tumor microenvironment [139, 145–148]. However, the dual role of the ROS-cGAS-STING axis complicates the design of therapeutic approaches. Future research should prioritize spatiotemporal modulation of ROS-STING signaling to clarify the relationship between tumor microenvironment profiles and therapeutic outcomes, enabling a shift from nonspecific oxidative damage to controlled immune activation for more targeted cancer immunotherapy.

Potential of combined targeting of the cGAS-STING pathway and mitochondrial metabolism in cancer therapy

Recent advances in immunometabolism research have highlighted the cGAS-STING pathway and mitochondrial metabolism as pivotal interconnected nodes in tumor biology. This growing understanding underscores the therapeutic potential of simultaneously targeting these pathways, offering a novel and promising strategy for cancer treatment. While interventions targeting key components of the cGAS-STING pathway or mitochondrial metabolic enzymes have demonstrated significant efficacy in preclinical and clinical models, systematic exploration of their coordinated regulation and the development of combined therapeutic approaches remain in their infancy. This section reviews current clinical trials targeting the cGAS-STING pathway and mitochondrial metabolic enzymes and discusses the rationale and potential impact of combination therapies in oncology.

Anti-tumor strategies targeting the cGAS-STING pathway

The development of STING agonists represents a primary therapeutic strategy for modulating the cGAS-STING pathway. Clinically relevant agonists fall into two categories: cyclic dinucleotides (CDNs) and non-CDN small molecules [139–142]. The two agonist classes exhibit significant differences in their pharmacological properties, which profoundly influence their applicability in combination therapy strategies. CDN-based agonists such as ADU-S100 and MK-1454 have shown synergistic anti-tumor activity when combined with immune checkpoint inhibitors [143, 144]. However, their clinical application faces limitations, including short plasma half-life, poor bioavailability, the typical need for intratumoral administration, and the potential to induce systemic inflammatory toxicity [155]. In contrast, non-CDN agonists like diABZI and SNX2811 exhibit superior pharmacokinetic profiles, oral bioavailability, metabolic stability, and chemical tractability, making them more favorable for clinical translation [146–149]. Moreover, diABZI has been found to cross the cell membrane efficiently, enter the cytoplasm, and bind STING, thereby overcoming the cellular uptake bottleneck associated with CDN-based drugs [160]. Consequently, non-CDN STING agonists, due to their excellent cell permeability, metabolic stability, and suitability for systemic administration, demonstrate significant advantages for the treatment of systemic tumors, the design of oral dosing regimens, and combination therapies with immune- and metabolism-targeted agents. Particularly in tumors with complex microenvironments where CDNs penetrate poorly, non-CDN agonists offer greater possibilities for developing combination immunotherapies based on metabolic reprogramming [161]. Table 1 provides an overview of ongoing Phase I/II clinical trials evaluating CDN and non-CDN STING agonists.

Table 1.

STING agonists for clinical application in tumors

Types of Agonist	Drug Name	Subject Tumors	Combination	Phase	Clinical trial registration number
CDNs	ADU-S100	HNSCC	Pembrolizumab	II	NCT03907141
	BMS-986,301	Solid tumors	± Nivolumab and Ipilimumab	I	NCT03956680
	BI 1,387,446	Solid tumors	± Ezabenlimab	I	NCT04147234
	BI 1,703,880	Solid tumors	Ezabenlimab	I	NCT05471856
	E7766	Advanced solid tumor and Lymphoma	Alone	I	NCT04144140
		Urinary bladder neoplasm	Alone	I	NCT04109092
	IMSA101	Oligoprogressive solid tumor malignancies	PULSAR, ICI	II	NCT05846659
	MK- 1454	HNSCC	± Pembrolizumab	II	NCT04220866
		Solid tumors and Lymphomas	± Pembrolizumab	I	NCT03010176
	MK-2118	Solid tumors	± Pembrolizumab	I	NCT03249792
	SB 11285	Advanced solid tumors	± Atezolizumab	I	NCT04096638
	TAK-676	NSCLC, TNBC, HNSCC	Pembrolizumab/Radiation-therapy	I	NCT04879849
		Solid tumors	Pembrolizumab/Radiation-therapy	I	NCT04420884
Non-CDN STING agonists	DMXAA	Solid Tumors	Alone	I	NCT00856336
		Solid Tumors	Alone	I	NCT00832494
		Non-small Cell Lung Cancer	Carboplatin and Caclitaxel	I/II	NCT00832494
	GSK3745417	Solid tumors	± Dostarlimab	I	NCT03843359
	HG381	Solid Tumors	Alone	I	NCT04998422
	KL340399	Solid Tumors	Alone	I	NCT05549804
	ONO-7914	Solid tumors	± ONO-4538	I	NCT06535009
	SNX281	Solid tumors	± Pembrolizumab	I	NCT04609579

The clinical translation of STING agonists as a novel immunotherapeutic strategy faces multiple challenges that reflect broader difficulties in developing immune modulators. A central challenge is the need for precise regulation of STING activation—moderate activation induces robust anti-tumor immune responses, while excessive activation may trigger cytokine-mediated toxicity [150]. This dilemma is further complicated in immunosuppressive tumor microenvironments, particularly in tumors with limited T cell infiltration, where monotherapy often yields suboptimal responses [151]. Moreover, although intratumoral injection remains the mainstay of administration to minimize systemic toxicity, its clinical utility is constrained by practical limitations, especially in patients with multiple metastatic lesions [146].

Contemporary research is progressing in three key strategic directions to address the clinical translation challenges associated with the development of STING agonists. The first direction centers on the design of next-generation agonist molecules with enhanced selectivity, achieved through structural optimization to enable more precise dosing and an expanded therapeutic window. The second direction involves the exploration of combination therapy regimens that integrate STING agonists with immunotherapeutic agents and complementary pharmacological interventions, aiming to overcome treatment resistance driven by tumor microenvironment heterogeneity. The third direction focuses on developing intelligent drug delivery systems based on stimuli-responsive nanomaterials, which are specifically engineered to facilitate tumor-targeted drug release and enable accurate spatiotemporal regulation of STING pathway activation [152, 153].

Potential of targeting mitochondrial metabolism to enhance STING immune activation

In response to the translational challenges of STING agonists, researchers have increasingly explored the synergistic potential of targeting mitochondrial metabolism. This approach is grounded in the observation that mitochondrial metabolic reprogramming directly modulates cGAS-STING pathway activity, while STING activation reciprocally regulates cellular metabolic states, establishing a unique immunometabolic network [166]. Therefore, rational combination strategies should aim to co-regulate these two axes, to establish a positive feedback loop for immunogenic activation while effectively circumventing resistance mechanisms.

Targeting key enzymes in the TCA cycle can regulate the accumulation of TCA metabolites, thereby influencing the cGAS-STING pathway and innate immunity. The TCA cycle serves as a central metabolic hub in tumor cells, providing essential precursors for the biosynthesis of macromolecules. CPI-613, an inhibitor of this pathway, demonstrated significant therapeutic efficacy in a phase III clinical trial for metastatic pancreatic cancer (NCT03504423) through dual inhibition of pyruvate dehydrogenase (PDH) and α-ketoglutarate dehydrogenase (α-KGDH) [167, 168]. Its mechanism may involve blocking TCA metabolite accumulation and directly or indirectly modulating the cGAS-STING pathway. Mutations in isocitrate dehydrogenase (IDH) generate the oncometabolite (R)-2-hydroxyglutarate (R-2-HG), which can suppress T cell function by enhancing the transcriptional activity of T cell nuclear factors and interfering with polyamine biosynthesis [169]. Therefore, combining mutant IDH1/2 inhibitors with STING agonists holds promise for reversing this immunosuppressive state and reactivating T cell anti-tumor activity. Targeting the electron transport chain (ETC) complex I is another important strategy for modulating tumor metabolism. The classic ETC complex I inhibitor metformin exerts a dose-dependent biphasic effect on tumor metabolism: therapeutic concentrations substantially reduce TCA cycle flux, whereas low-dose exposure may promote metastasis through activation of the FAO pathway [98]. A preclinical study found that the combination of metformin and the STING agonist 2’3’-cGAMP exhibited a synergistic effect in activating the STING/IRF3/IFN-β pathway in pancreatic cancer cells, demonstrating its potential for combination therapy [170]. Beyond direct enzyme inhibition, emerging strategies targeting mitochondrial protein translation offer alternative approaches for ETC modulation [171, 172].

As a critical energy source for tumor cells under stress, the rate-limiting enzyme for FAO, carnitine palmitoyltransferase 1 A (CPT1A), is a highly promising target. Preclinical studies indicate that CPT1A inhibitors can effectively activate the cGAS-STING pathway and enhance neutrophil-mediated tumor elimination. Although the clinical development of specific CPT1A inhibitors is still in its early stages, their mechanism is well-defined, providing a clear rationale for combination therapy [173]. Therapeutic targeting of one-carbon metabolism has a well-established clinical history. Methotrexate, which inhibits dihydrofolate reductase (DHFR) and thymidylate synthase (TYMS), revolutionized the treatment of childhood acute lymphoblastic leukemia (ALL) in the mid-20th century [174, 175]. Recent research has found that methotrexate reduces extracellular adenosine and enhances host STING-mediated anti-tumor immunity [176]. Since then, small-molecule inhibitors targeting metabolic enzymes have continued to expand, as summarized in Table 2, providing valuable opportunities for combination strategies with STING agonists.

Table 2.

Mitochondrial enzymes for clinical application in tumors

Pathway	Enzyme	Inhibitors	Subject Tumors	Combination	Phase	Clinical trial registration number
TCA cycle	IDH (mutant)	AG-881	Acute Myeloid Leukemia	Alone	I	NCT02492737
			Glioma	AG-120	I	NCT03343197
			Glioma	Alone	I	NCT02481154
	KGDHC	CPI-613	Advanced Malignancies	Alone	I/II	NCT00741403
			Recurrent Small Cell Lung Cancer	Alone	I	NCT01931787
			Pancreatic Cancer	Chemotherapy	I	NCT01835041
ETC	Complex I	Tamoxifen	Bladder Cancer	Alone	II	NCT02197897
			Neuroendocrine Tumors	Alone	II	NCT03870399
			Breast Neoplasms	LY353381	III	NCT00034125
		Metformin	Breast Cancer	Alone	III	NCT01101438
			Advanced Malignant Solid Neoplasm	Sapanisertib	I	NCT03017833
			Colorectal Cancer	Alone	III	NCT05921942
			Prostate Cancer	Enzalutamide	I	NCT02339168
		IACS-010759	Advanced Malignant Solid Neoplasm	Alone	I	NCT03291938
	Complex III	Resveratrol	Colon Cancer	Alone	I	NCT00256334
			Lymphangioleiomyomatosis	Sirolimus	II	NCT00433[70–72]6
	Complex IV	Doxorubicin	Advanced Solid Tumours	Pantoprazole	I	NCT01163903
			Solid Tumors	Gemcitabine, and Velcade	I	NCT00500422
			Malignant Gynecologic Tumours	Carboplatin	II	NCT00189410
		Photofrin	Brain Tumor	Alone	I	NCT01682746
			Bile Duct, Gallbladder, or Pancreas Cancer	Alone	II	NCT00003923
			Esophageal Cancer	Alone	II	NCT00002935
		Fenretinide	Cervical Cancer	Alone	III	NCT00003075
			Solid Tumors	Alone	I	NCT00003250
			Brain and Central Nervous System Tumors	Alone	II	NCT00006080
			Small Cell Lung Cancer	Alone	II	NCT00009971
One carbon metabolism	SHMT2	Methotrexate	Non-melanoma Skin Cancers	Alone	II	NCT05315128
			Gestational Trophoblastic Neoplasia	Alone	III	NCT00003702
			Brain and Central Nervous System Tumors	Leucovorin	II	NCT00082797
		Pemetrexed	Non-small Cell Lung Cancer	AZD2171	II	NCT00410904
			Solid Tumors	LY573636	I	NCT01215916
			Neuroendocrine Tumors	Alone	II	NCT00424723
			Advanced Breast Cancer	Cisplatin	Ⅳ	NCT01143974
		Raltitrexed	Solid Tumor	ICI D1694	I	NCT00002902
			Gastric Cancer	Paclitaxel	II	NCT03083613
			Recurrent Childhood Acute Myeloid Leukemia	Alone	I	NCT00003528

Beyond the aforementioned strategies of directly targeting metabolic enzymes, modulating mitochondrial membrane channels quality control mechanisms (including dynamics, autophagy, and redox homeostasis) also shows synergistic potential. The mitochondrial outer membrane channel protein voltage-dependent anion channel 2 (VDAC2) plays a critical role in maintaining metabolic homeostasis by regulating the transport of key metabolites, including ATP/ADP and NAD⁺/NADH. VDAC2 dysfunction disrupts metabolic balance and induces mtDNA leakage, which in turn robustly activates the cGAS-STING pathway and IFN-I response [177, 178]. The regulation of these mitochondrial quality control systems can establish a positive feedback loop with STING pathway activation [172, 179–181]. The positive feedback between STING activation and mitochondrial ROS production is particularly noteworthy, providing a theoretical basis for novel dual-targeted “metabolic-immune” nanosystems. These intelligent delivery platforms enable the spatiotemporally precise coregulation of metabolic reprogramming and immune activation, offering a promising solution to clinical challenges such as tumor heterogeneity and therapeutic resistance [182].

Future research should prioritize elucidating the specific interaction mechanisms between key metabolic pathways, including glycolysis and lipid metabolism, and the STING signaling axis. Such insights will be critical for optimizing personalized combination therapy regimens.

Conclusion and prospect

The cGAS-STING pathway and mitochondrial metabolism form a crucial regulatory axis in tumor immunity, with their dynamic interaction influencing the balance between anti-tumor immune responses and cancer progression. Recent evidence shows that mitochondrial metabolites, such as succinate and itaconate, act as endogenous signaling molecules that can modulate the activation threshold and duration of the STING pathway through allosteric regulation and epigenetic mechanisms. In turn, prolonged STING signaling can create a bidirectional feedback loop by inducing mitochondrial dysfunction, which is characterized by disturbances in mitochondrial dynamics, mitophagy, oxidative stress, and metabolic reprogramming. Understanding these regulatory interactions more deeply may open new therapeutic avenues for enhancing cancer immunotherapy.

Future research should focus on three key translational directions. First, developing spatiotemporally controlled STING agonist delivery systems is essential. These systems would enable precise modulation of drug release kinetics, thereby optimizing anti-tumor efficacy while minimizing inflammatory toxicity. Second, integrating metabolomic profiling with immune microenvironment characterization into a comprehensive biomarker framework will refine patient stratification and therapeutic monitoring. Third, considering the significant variability in the cGAS-STING-mitochondrial metabolism axis across different tumor types, tailored therapeutic strategies must account for the unique metabolic and immunological features of each tumor microenvironment.

In summary, the cGAS-STING pathway and mitochondrial metabolism form a dynamic and bidirectional regulatory axis, and their balance determines the relationship between immune homeostasis and pathological inflammation. Therefore, understanding the spatiotemporal coordination between mitochondrial metabolism and STING signaling is crucial for designing next-generation immunotherapies that combine metabolic reprogramming with innate immune regulation. This approach provides a rational treatment framework for advanced precision treatments covering oncology, infectious diseases, and autoimmune disorders.

Abbreviations

4-OI

4-Octyl itaconate

ACLY

ATP-citrate lyase

ACOD1

Aconitate decarboxylase 1

ACSL1

Acyl-CoA synthetase long-chain family member 1

ADSL

Adenylosuccinate lyase

ALL

Acute lymphoblastic leukemia

AMPK

AMP-activated protein kinase

ATM

Ataxia telangiectasia mutated

ATG13

Autophagy-related protein 13

cGAMP

Cyclic GMP-AMP

cGAS

Cyclic GMP-AMP synthase

CDN

Cyclic dinucleotide

CD38

Cluster of differentiation 38

COPⅠ/Ⅱ

Coat Protein Complex Ⅰ/Ⅱ

CRAT

Carnitine acetyltransferase

CPT1A

Carnitine palmitoyltransferase 1 A

CTLs

Cytotoxic T lymphocytes

DAMPs

Damage-associated molecular patterns

DCs

Dendritic cells

DHFR

Dihydrofolate reductase

DRP1

Dynamin-related protein 1

dsDNA

Double-stranded DNA

eIF2α

Eukaryotic initiation factor 2 alpha

ER

Endoplasmic reticulum

ETC

Electron transport chain

FAO

Fatty acid β-oxidation

FAD

Flavin adenine dinucleotide

FADH

Flavin adenine dinucleotide hydrate

FIP200

Family interacting protein of 200 kDa

GA

Golgi apparatus

GLUT1

Glucose transporter type 1

GPX4

Glutathione peroxidase 4

GSH

Glutathione

HIF-1α

Hypoxia-inducible factor 1-alpha

HK2

Hexokinase 2

ICD

Immunogenic cell death

ICIs

Immune checkpoint inhibitors

IFI16

Interferon gamma-inducible protein 16

IDH

Isocitrate dehydrogenase

IFN-I

Type I interferon

IL-1β

Interleukin-1 beta

IL-6

Interleukin-6

IRF3

Interferon regulatory factor 3

KEAP1

Kelch-like ECH-associated protein 1

LC3

Microtubule-associated protein 1 light chain 3 (LC3)

LIR

LC3-interacting region

MFN1

Mitofusin 1

MFN2

Mitofusin 2

mTOR

Mechanistic target of rapamycin

mtDNA

Mitochondrial DNA

NAD⁺

Nicotinamide adenine dinucleotide (Oxidized)

NADH

Nicotinamide adenine dinucleotide (Hydride)

NDP52

Nuclear dot protein 52

NF-κB

Nuclear factor kappa-B

NK

Natural killer

NRF2

Nuclear factor erythroid 2-related factor 2

OMM

Outer mitochondrial membrane

OPTN

Optineurin

OXPHOS

Oxidative phosphorylation

PARP-1

Poly ADP-ribose polymerase 1

PDH

Pyruvate dehydrogenase

PERK

Protein kinase R-like endoplasmic reticulum kinase

PFK-1

Phosphofructokinase-1

PFKFB3

6-Phosphofructo-2-kinase/fructose-2,6-bisphosphatase 3

PINK1

PTEN-induced kinase 1

PKM2

Pyruvate kinase M2

PPARγ

Peroxisome proliferator-activated receptor gamma

pSer65-Ub

Substrate ubiquitin (Ser65-Phosphorylated)

PUFA

Polyunsaturated fatty acids

ROS

Reactive oxygen species

SDH

Succinate dehydrogenase

SENP3

SUMO-specific protease 3

STING

Stimulator of interferon genes

TAMs

Tumor-associated macrophages

TBK1

TANK-binding kinase 1

TCA

Tricarboxylic acid cycle

TME

Tumor microenvironment

TNF-α

Tumor necrosis factor-alpha

TRAF6

TNF receptor-associated factor 6

Ub

Ubiquitin

ULK1

Unc-51-like kinase 1

VDAC2

Voltage-dependent anion channel 2

α-KGDH

α-Ketoglutarate dehydrogenase

ΔΨm

Mitochondrial membrane potential

Author contributions

KZ: Writing – review & editing, Writing – original draft, Conceptualization. SYC: Validation, Writing – original draft, Conceptualization. NW: Validation, Writing – original draft, Conceptualization. QW: Validation, Writing – original draft. XRW: Validation, Writing – original draft. KH: Validation, Writing – original draft. JLZ: Funding acquisition, Writing – review & editing. FMK: Writing – review & editing, Supervision, Funding acquisition, Conceptualization. All authors read and approved the final manuscript.

Funding

This work was supported by grants from the Noncommunicable Chronic Diseases-National Science and Technology Major Project (No.2024ZD0521103), Tianjin Public Health Science and Technology Major Youth Project (No.24ZXGQSY00090), Science and Technology Project of Haihe Laboratory of Modern Chinese Medicine (No.GZY-KJS-2025-056), National Clinical Research Center for Chinese Medicine Acupuncture and Moxibustion Open Funding Project (No.NCRCOP2023007), Tianjin Key Research Projects in Traditional Chinese Medicine (No.2025011), Hebei Provincial Administration of Traditional Chinese Medicine Research Project (No.T2025083 & No.T2025059), Pilot Demonstration Project for the Inheritance and Innovative Development of Traditional Chinese Medicine in Nankai District, Tianjin (No.20240204019), and Tianjin Municipal Key Disciplines and Key Specialties Construction Program in Medicine (TJYXZDXK-010 A).

Data availability

Data sharing is not applicable to this article as no datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.