Mitochondria reactive oxygen species signaling-dependent immune responses in macrophages and T cells

Summary:

Mitochondria are key regulators of immune cell function, going beyond their traditional role in ATP and metabolite production to support anabolic processes and act as hubs for intracellular signaling. A key aspect of this signaling function is the production of mitochondrial reactive oxygen species (mtROS), which act as critical second messengers in both adaptive and innate immune regulation. Immune cells maintain an optimal concentration of mtROS to maintain physiological responses and, excessive or lack of mtROS production contributes to chronic inflammation, autoimmunity, and cancer. Here, we review the molecular mechanisms controlling mtROS production and detoxification, their role in shaping macrophage and T cell fate and function, and their implications for disease pathogenesis.

Introduction

Mitochondria play a pivotal role in immune cell function, extending beyond their classical role in ATP production to serve as key regulators of metabolism, signaling, and immune responses^1,2. They generate essential metabolites that support anabolic processes and influence immune cell differentiation and activation³. Additionally, mitochondria act as hubs for intracellular signaling by producing reactive oxygen species (ROS). For decades, mitochondrial ROS (mtROS) were predominantly considered damaging metabolic byproducts, contributing to oxidative stress, cellular dysfunction, and disease pathogenesis⁴. This perspective framed mtROS as harmful agents that disrupt immune homeostasis, drive immune exhaustion and fuel chronic inflammation and autoimmunity. However, emerging research has reframed this narrative, revealing mtROS as indispensable signaling molecules that orchestrate immune cell function, metabolic adaptation, and pathogen defense. Rather than being merely cytotoxic, mtROS serve as crucial regulators of immune activation, cytokine signaling, and cell fate decisions⁵. However, excessive mtROS production contributes to chronic inflammation and autoimmune disorders, underscoring the need for precise regulatory mechanisms to maintain immune balance^6,7. This review examines the evolving roles of mtROS in regulating innate and adaptive immunity, with a particular focus on acute activation of macrophages and T cells, emphasizing newly emerging concepts supported by both genetic and pharmacological evidence.

Primer on mitochondrial ROS

The mitochondrial electron transport chain (ETC) complexes I and III are the primary sites where electrons lead to the incomplete reduction of oxygen and the formation of superoxide (O₂•⁻) in both the mitochondrial matrix and intermembrane space^8–10. Superoxide dismutases, with SOD1 in the intermembrane space and SOD2 in the matrix, catalyze the conversion of superoxide (O₂•⁻) into hydrogen peroxide (H₂O₂). mtROS generation is not constant but highly dynamic, fluctuating based on metabolic demands, substrate availability, and mitochondrial efficiency¹¹. When ATP demand is high, mitochondria utilize the proton gradient efficiently, leading to a lower mitochondrial membrane potential (ΔΨm) as protons flow through ATP synthase (complex V) to generate ATP. This reduces electron leakage and consequently lowers mtROS production. Conversely, when ATP demand is low, the proton gradient builds up, causing hyperpolarization (higher ΔΨm). This increases the probability of electron leakage, especially at mitochondrial complex I and complex III, leading to elevated mtROS production. Thus, mild mitochondrial uncoupling or controlled dissipation of ΔΨm can mitigate excessive ROS generation, reduce oxidative stress while maintain efficient energy production.

Mitochondrial complex I primarily generates superoxide within the mitochondrial matrix, especially under conditions of elevated NADH/NAD⁺ ratios and high ΔΨm. Moreover, mitochondrial complex I can produce superoxide through reverse electron transport (RET), a process in which electrons move counter to the conventional electron transport chain flow¹². RET occurs when a highly reduced coenzyme Q (CoQ) pool and a strong proton-motive force (Δp) drive electrons backward from the CoQ pool onto the flavin mononucleotide (FMN) of complex I, reducing the FMN which can donate a pair of electrons to NAD+ to form NADH, or pass one electron to oxygen to generate superoxide¹³. This mechanism is closely linked to succinate oxidation, as succinate dehydrogenase (SDH, mitochondrial complex II) transfers electrons to the CoQ pool, increasing its reduction state and facilitating RET at complex I^12,14. When ΔΨm reaches a critical threshold, electrons from the over-reduced CoQ pool are pushed back into complex I, ultimately accumulating at the flavin mononucleotide (FMN) site where excessive O₂•⁻ generation occurs as a key consequence of RET. Mitochondrial complex III produces superoxide on both sides of the mitochondrial inner membrane, the matrix and the intermembrane space through the ubiquinone (Q) cycle^15,16. Superoxide generated in the intermembrane space can traverse to the cytosol via the voltage dependent anion channel (VDAC)¹⁷. During the Q cycle, ubiquinol (QH₂) donates electrons, with one transferred to cytochrome c via the Rieske [2Fe-2S] cluster and the other moving toward the Qi site via cytochrome b. Under conditions such as a high proton-motive force, an elevated ubiquinol/ubiquinone (CoQH₂/CoQ) ratio, or inhibition of electron flow (e.g. <1 %O₂), semiquinone at the Qo site can directly transfer electrons to molecular oxygen, generating superoxide, which is released into both the intermembrane space and the matrix¹⁸. This dual localization allows superoxide produced at mitochondrial complex III to impact both the mitochondrial matrix and cytosolic signaling pathways. It is important to note in addition to mitochondrial complex I and III, there are nine other potential sites of ROS generation in the mitochondrial ETC¹⁹. However, whether these other sites are necessary for immune functions during physiology or pathology is not well understood.

H₂O₂ serves as a critical signaling molecule and is tightly regulated within cells to balance its roles in redox signaling and oxidative damage^4,20. An optimal level of H₂O₂ sustains immunity (Figure 1A). As a signaling agent, H₂O₂ primarily modifies thiol groups in proteins, particularly cysteine residues, leading to changes in protein function, localization, or interactions (Figure 1B)^21,22. It reacts with transition metal centers, selenoproteins, and selective thiol proteins, with peroxiredoxins (PRDXs) and glutathione peroxidases (GPXs) being key targets²¹. PRDXs and GPXs act as both antioxidant defenses and redox sensors, ensuring that H₂O₂ concentrations remain controlled. Due to their rapid reaction rates with H₂O₂, PRDXs and GPXs limit its diffusion and protect less reactive thiol proteins from oxidation. However, for H₂O₂ to participate in signaling, specialized mechanisms facilitate its reaction with target proteins, such as direct transfer from PRDXs to signaling proteins in a redox relay. This allows selective oxidation of regulatory proteins involved in cellular responses. To prevent excess accumulation, antioxidant enzymes, including catalase, GPXs, and PRDXs, rapidly detoxify H₂O₂, converting it into water and oxygen. Additionally, the thioredoxin and glutathione systems help recycle oxidized proteins back to their reduced forms, maintaining redox homeostasis. By integrating both signaling and detoxification mechanisms, cells ensure that H₂O₂ functions as a controlled messenger without causing oxidative damage.

Figure 1: An optimal mitochondrial H2O2 signaling is crucial for maintaining immunity, through cysteine-mediated redox regulation.

(A) We propose that immune cells require an optimal concentration of H₂O₂ to maintain their function. (B) H₂O₂ and electrophiles such as fumarate and its derivatives (dimethyl fumarate (DMF), monomethyl fumarate (MMF)) or itaconate and its derivatives (dimethyl itaconate (DI), 4-ocytl itaconate (4-OI)) react with cysteines to confer their biological effects in immune cells. Initial H₂O₂ exposure creates reversible cysteine modifications (disulfide bonds, sulfenic acid) that can be reduced back to free thiols through the glutathione/thioredoxin antioxidant systems, maintaining redox homeostasis. However, continued H₂O₂ exposure leads to irreversible cysteine oxidation to sulfinic and sulfonic acid derivatives, potentially resulting in protein dysfunction.

While controlled H₂O₂ signaling is essential for cellular function, dysregulation of redox homeostasis can lead to oxidative stress. A prime example of this ROS-driven cellular damage is ferroptosis, a regulated form of cell death initiated by iron-dependent lipid peroxidation, primarily driven by the Fenton reaction and polyunsaturated fatty acid (PUFA) peroxidation²³. The Fenton reaction generates highly reactive hydroxyl radicals (•OH) from H₂O₂ in the presence of Fe². These hydroxyl radicals attack PUFA-containing phospholipids in cellular membranes, initiating a lipid peroxidation chain reaction. This process begins with lipid hydrogen abstraction, forming lipid radicals (L•), which react with molecular oxygen to generate lipid peroxyl radicals (LOO•), ultimately accumulating toxic lipid peroxides (L-OOH) that disrupt membrane integrity and trigger ferroptosis. In contrast, monounsaturated fatty acids (MUFAs) help protect against ferroptosis by competing with PUFAs for incorporation into phospholipids.

To counteract oxidative damage, immune cells utilize multiple antioxidant systems, including the cyst(e)ine-glutathione(GSH)-GPX4, CoQ10-FSP1, squalene, and BH4-DHFR pathways, which reduce L-OOH to lipid alcohols (L-OH) and prevent ferroptosis. These protective mechanisms remain effective unless GPX4 is inhibited or GSH levels are depleted, which can occur through direct inhibition of GPX4 or reduced cystine uptake through system Xc–, a cystine/glutamate antiporter^24,25. Lipid remodeling enzymes such as acyl-CoA synthetase long-chain family member 4 (ACSL4) and lysophosphatidylcholine acyltransferase 3 (LPCAT3) further influence ferroptosis susceptibility by incorporating PUFAs into membrane phospholipids, making cells more vulnerable to oxidative damage^26,27. Additionally, immune cells also employ the transcription factor NRF2 to upregulate antioxidant proteins and maintain NADPH and GSH production, enhancing protection from ROS-induced damage²⁸ (Figure 2). We suggest that immune cells establish distinct redox microdomains within the cell, allowing for compartmentalized H₂O₂ signaling in specific locations while maintaining robust antioxidant capacity elsewhere to neutralize harmful ROS species like superoxide (O₂⁻), hydroxyl radicals (-OH), and lipid peroxides (L-OOH)²⁰. This spatial organization of redox activity may be crucial for balancing essential immune signaling functions with protection against oxidative damage.

Figure 2: Mitochondrial ROS Regulation: From Redox Signaling to Ferroptotic Cell Death.

The mitochondrial electron transport chain (ETC) generates superoxide (O₂•⁻) that is converted to hydrogen peroxide (H₂O₂) by superoxide dismutases (SOD1 in the mitochondrial intermembrane space and SOD2 in the mitochondrial matrix). H₂O₂ serves as a key signaling molecule through redox-dependent cysteine modifications of target proteins. Simultaneously, cellular antioxidant systems regulate ROS levels and prevent oxidative damage. A major antioxidant defense system involves NRF2 activation, which drives NADPH and glutathione (GSH) production. Glutathione peroxidases (GPXs), particularly GPX4, and peroxiredoxins (PRDXs) neutralize H₂O₂ and lipid peroxides. When these protective mechanisms are compromised, excessive ROS can trigger ferroptosis, an iron-dependent cell death pathway. In this process, ferrous iron (Fe²⁺) catalyzes the Fenton reaction, converting H₂O₂ to highly reactive hydroxyl radicals (OH·) that initiate lipid peroxidation of polyunsaturated fatty acids (PUFAs). Multiple protective mechanisms prevent ferroptosis, including: the cystine/glutamate antiporter (Xc⁻) that imports cysteine for GSH synthesis; GPX4 that neutralizes lipid peroxides; ferroptosis suppressor protein 1 (FSP1) that utilizes the Coenzyme Q-NADPH antioxidant system to reduce lipid peroxides independently of GSH; incorporation of monounsaturated fatty acids (MUFAs) that resist peroxidation; and tetrahydrobiopterin (BH4) that acts as a radical-trapping antioxidant. This balanced regulation of ROS determines whether their effects manifest as controlled redox signaling that supports immune function or destructive oxidative damage leading to cell death.

Mitochondrial ROS in Innate Immunity

An early link connecting mtROS to the innate immune response has been the discovery that tumor necrosis factor receptor (TNFR)-associated factors (TRAFs) regulate intracellular ROS as downstream signal transduction mediators^29,30. TRAF recruitment to TNFR-related proteins induces mitochondrial ROS production, which in turn influences nuclear factor kappa B (NF-κB) activation and apoptosis²⁹. TLR signaling also stimulates glutathione synthesis which is necessary for optimal IL-1β production³¹. Furthermore, lipopolysaccharide (LPS) activation of Toll-like receptor 4 (TLR4)-induced inflammation by neutrophils is dependent on mitochondrial respiratory complex I. Notably, inhibition of mitochondrial complex I by metformin or rotenone has been shown to reduce ROS production, suppress NF-κB activation, and lower proinflammatory cytokine expression, ultimately mitigating acute lung injury³². Engagement of Toll-like receptors (TLRs) leads to mitochondrial recruitment to phagosomes and a subsequent increase in mtROS production³³. Importantly, this work has revealed mtROS as an additional effector mechanism for bacterial killing. Interestingly, mtROS appears to play a critical role specifically in TLR-mediated classical macrophage responses but not in interleukin-4 (IL-4)-driven alternative activation, highlighting a pathway-specific dependency on mitochondrial redox signaling in macrophage biology³⁴.

TLR-dependent signaling occurs within minutes to hours activating a cascade of signaling pathways to generate distinct pro-inflammatory cytokines including IL-1β, TNF-α and IL-6 and the anti-inflammatory cytokine IL-10. Recent work has demonstrated that the mitochondrial ETC is dispensable for TLR3 or TLR4 induced early production of pro-inflammatory cytokines³⁴. However, the mitochondrial ETC is necessary for acute (<4 hrs) IL-10 release and production of IL-10 at later (>24hrs) time points. Macrophages with genetic loss of mitochondrial complex III function release less IL-10 following TLR3 or TLR4 stimulation and exhibit increased susceptibility to inflammatory challenges such as influenza A virus infection and LPS-induced endotoxic shock³⁴. Importantly, restoring mitochondrial complex III function without generating superoxide using alternative oxidase (AOX) failed to restore IL-10 release, highlighting the specific requirement for complex III-derived mtROS³⁴. Supporting these findings, another recent study has found that pharmacological inhibition of mitochondrial complex III superoxide generation with S3QEL 1.2, a specific suppressor of mtROS production from the mitochondrial complex III Qo site³⁵, decreases IL-10 in LPS-activated bone marrow derived macrophages (BMDMs) in vitro. Interestingly, S3QEL 1.2 administration in vivo enhances immunotherapy efficacy likely through diminishing IL-10-mediated immunosuppression³⁶.

Mechanistically, these studies collectively reveal a multi-tiered regulatory model where mtROS influences IL-10 at both transcriptional and post-translational levels in a temporally coordinated manner. At the transcriptional level, superoxide production from mitochondrial complex III at the 24-hour timepoint supports sustained expression of c-Fos, a critical subunit of the activator protein 1 (AP1) transcription factor complex required for TLR4-induced IL-10 gene expression³⁶. This transcriptional control represents a delayed regulatory mechanism that maintains IL-10 production during prolonged activation. Consistent with this model, genetic loss of mitochondrial complex II can also decrease IL-10 production at 24 hours, presumably by reducing electron flux into mitochondrial complex III and consequently diminishing ROS-dependent transcriptional activation³⁷. Complex III ROS also appears to influence the post-translational regulation of IL-10, particularly during early activation phases. This is evidenced by experiments showing that activation of protein kinase A (PKA) rescues IL-10 release in mitochondria complex III-deficient BMDMs following LPS stimulation³⁴. These findings establish a direct link between TLR signaling, mitochondrial complex III ROS generation, and IL-10 regulation through both rapid post-translational mechanisms and sustained transcriptional control (Figure 3A), though the specific protein targets of ROS-mediated oxidation in these pathways remain to be identified.

Figure 3: Mitochondrial Electron Transport Directionality Dictates Distinct ROS-Mediated Immune Responses.

(A) During forward electron transport, electrons flow from NADH to complex I, then to coenzyme Q (Q), complex III, cytochrome c, and finally to complex IV where oxygen is reduced to water. Complex III generates superoxide (O₂⁻•) that is released into the intermembrane space (IMS) and the matrix. Complex III-derived ROS is essential for both transcriptional regulation of IL-10 through activating transcription factors and post-translational control of IL-10 secretion. (B) Reverse electron transport (RET) occurs under conditions of high protonmotive force and a highly reduced coenzyme Q pool. When succinate accumulates and is oxidized by complex II (succinate dehydrogenase), electrons flow to coenzyme Q. Due to the highly reduced state of the Q pool and elevated membrane potential, electrons can flow backwards into complex I, leading to superoxide production into the matrix. RET-derived ROS promotes gasdermin D (GSDMD) activation, enhancing its pore-forming activity and facilitating the release of pro-inflammatory cytokines such as IL-1β.

In contrast to the role of mitochondrial complex III ROS in regulating the anti-inflammatory cytokine IL-10, mitochondrial complex I ROS production has been linked to the control of pro-inflammatory cytokine production and release upon chronic TLR signaling in vitro and in vivo. Initial metabolomic studies have identified an accumulation of succinate in LPS-activated macrophages at 24 hours, resulting from both enhanced glutamine-dependent anaplerosis and reduced oxidation through succinate dehydrogenase (SDH)^38,39. Importantly, subsequent research has suggested that SDH-mediated succinate oxidation, actually results in an excessive reduction of the coenzyme Q pool in LPS activated BMDMs⁴⁰. Further, in response to LPS-induced metabolic reprogramming mitochondrial membrane potential is increased. These conditions—a reduced coenzyme Q pool and elevated membrane potential—create a favorable environment for reverse electron transport at mitochondrial complex I, where electrons flow backwards through the complex, generating superoxide⁴⁰ (Figure 3B).

Multiple recent studies have provided evidence for this mechanism, using chemical biology approaches and genetic tools to demonstrate that LPS-stimulated macrophages can produce superoxide through RET at mitochondrial complex I. However, to date, unequivocal causal evidence linking RET to macrophage inflammatory cytokine production is still lacking. siRNA silencing of SDHA or inhibition of complex II using multiple inhibitors in macrophages reduces succinate oxidation and decreases IL-1β production, which correlates with reduced RET-mediated ROS production^14,37,40. These studies highlight a critical role for succinate oxidation in inflammatory cytokine production, but do not specifically isolate RET as electrons donated at mitochondrial complex II could also contribute to mitochondrial complex III derived-ROS. Recently, the ND6^P25L mouse model, which harbors a mitochondrial DNA (mtDNA) mutation altering mitochondrial complex I function has been used to isolate the role of RET in inflammatory cytokine production. Multiple studies have suggested that ND6^P25L cells have reduced RET-mediated mtROS production without altering NADH turnover, TCA cycle activity, or ATP production^40,41. BMDMs from ND6^P25L mice show reduced IL-1β release in vitro⁴⁰. A limitation of this study is that RET-induced superoxide was measured 24 hours after LPS stimulation, whereas activation of the NLRP3 inflammasome by LPS combined with ATP or nigericin was assessed within 6 hours. Interestingly, microglia in ND6^P25L mice display attenuated inflammatory responses in vivo, however, inflammatory cytokine expression is not impaired in these cells in vitro⁴². Importantly, other studies investigating ND6^P25L mutant cells have found impairments in mitochondrial complex I function including reduced NADH oxidation, potentially limiting the broad value of this tool to isolate RET^43–45. This is especially relevant in the context of mitochondrial complex I impairment as multiple studies have demonstrated that macrophages lacking NDUFS4, an accessory subunit of mitochondrial complex I, display enhanced inflammatory cytokine production in response to TLR activation^46,47. NDUFS4 deficiency specifically increases mtROS generation from mitochondrial complex I in macrophages while reducing overall ETC activity^46,47. Crucially, treatment of NDUFS4-deficient macrophages with antioxidants reduces inflammatory cytokine production^46,47, suggesting mitochondrial complex I derived-ROS serves as the likely pro-inflammatory signal. Critically, it remains unclear whether increase in mtROS in NDUFS4-deficient cells is due to RET or some alternative ROS generating mechanism.

Combined, these studies using a variety of pharmacologic and genetic approaches reveal a compelling correlation between increased mitochondrial complex I ROS production and enhanced inflammatory output. However, the relationship is complex and sometimes contradictory. Macrophages lacking the evolutionarily conserved signaling intermediate in Toll pathway (Ecsit) protein, which is known to interact with mitochondrial complex I³³, also display elevated complex I ROS generation and reduced ETC activity⁴⁸. Yet, unlike NDUFS4-deficient cells, ECSIT-deficient macrophages do not show enhanced inflammatory cytokine production in response to LPS stimulation⁴⁸. ECSIT likely ablates complex I function while NDUFS4 generates a hypomorph and allows NADH oxidation to generate superoxide⁴⁹. Further, it is important to note that acute pharmacologic inhibition of complex I does not increase inflammatory cytokines upon various PRR stimulation at 24 hours⁵⁰. The loss of NDUFS4 preserves the NADH-dehydrogenase module within complex I, likely allowing superoxide generation⁴⁹, whereas acute pharmacological inhibition disrupts this process and prevents superoxide production. We suggest that future studies aiming to clarify the role of mitochondrial complex I in inflammation under acute or chronic conditions should utilize genetic models, such as NDUFS2 ablation⁵¹ or other strategies that fully disrupt mitochondrial complex I activity. Simultaneously, the expression of Saccharomyces cerevisiae NDI1 should be employed to restore mitochondrial complex I dependent electron transport without superoxide generation⁵², enabling a precise distinction between the metabolic and redox-dependent functions of mitochondrial complex I in immune signaling.

The mechanisms underlying how mtROS drives pro-inflammatory cytokine production downstream of TLR activation remained largely undefined. In Listeria monocytogenes-infected macrophages, one identified pathway involves mtROS-induced intermolecular disulfide linkage of NEMO (NF-κB essential modulator)⁵³. This redox-dependent modification activates the IκB Kinase complex, triggering downstream ERK1/2 and NF-κB signaling pathways that culminate in pro-inflammatory cytokine secretion. While this direct oxidative modification of NEMO represents an elegant example of how mtROS can translate metabolic changes into altered gene expression, it likely represents just one of several parallel mechanisms. Other redox-sensitive transcription factors and signaling intermediates, including additional components of the NF-κB pathway, MAP kinases, and various transcriptional co-activators, may also be regulated by mtROS through distinct molecular interactions. This multilayered redox regulation enables fine-tuning of pro-inflammatory responses across diverse pathogenic challenges and cellular contexts. While there is growing evidence for how mtROS influences classical macrophage activation, our understanding of ROS involvement in alternative (IL-4) macrophage activation remains limited. Although complex III-derived ROS appears dispensable for initial alternative activation³⁴, this doesn’t preclude potential roles for other mitochondrial ROS sources or more nuanced temporal regulation of redox signaling during IL-4 polarization. The selective nature of ROS-mediated transcriptional regulation likely involves a combination of factors: localized ROS production in specific cellular compartments, differential thiol reactivity among transcription factors and co-activators, and the temporal dynamics of antioxidant responses that shape the duration and intensity of redox signaling. This compartmentalized approach may allow cells to use similar reactive molecules to exert distinct transcriptional outcomes depending on the activation context.

NLRP3 inflammasome activation occurs in response to stress signals, leading to the recruitment of ASC (apoptosis-associated speck-like protein) and the assembly of the inflammasome. ASC facilitates caspase-1 activation, which cleaves pro-IL-1β and pro-IL-18 into their active forms. Caspase-1 also cleaves gasdermin D (GSDMD), releasing its N-terminal fragment, which forms membrane pores, driving IL-1β and IL-18 release and pyroptotic cell death⁵⁴. A commonly used in vitro assay involves stimulating cells with LPS (signal 1) for a few hours to induce the expression of NLRP3 inflammasome components, followed by treatment with nigericin or extracellular ATP (signal 2) for minutes to activate the NLRP3 inflammasome. Several studies propose links between mtROS and different aspects of inflammasome activation^55–58. While genetic evidence using cells deficient in mitochondrial complex I or III superoxide production does not support a role for mtROS in initial NLRP3 inflammasome activation⁵⁹. The genetic and biochemical findings support a critical role for mtROS in controlling IL-1β release through the regulation of GSDMD oxidation^60–63, the key effector of pyroptotic cell death. Disabling RET induced ROS at mitochondrial complex I in the ND6^P25L model diminishes release of IL-1β likely through controlling GSDM⁴⁰. Furthermore, mtROS promotes the oligomerization and pore-forming activity of GSDMD^60,61, which enhances pyroptotic cell death and IL-1β release. The molecular basis for this redox sensitivity centers on GSDMD cysteine 192, with mutation of this residue impairing pore formation and pyroptosis induction⁶². Thus, the major input of mtROS into NLRP3 inflammasome likely resides not in the initial activation but through the regulation of pyroptosis.

The role of mitochondria and ROS in NLRP3 inflammasome activation appears contradictory in some studies, however, these discrepancies can largely be attributed to differences in experimental approaches and the distinction between acute and chronic activation conditions. For instance, genetic deletion of mitochondrial complex I, as seen in the NDUFS4 deficient model, increase pro-inflammatory cytokines like TNF-α and IL-6^46,47, but reduces initial NLRP3 inflammasome activation⁶⁴. This effect is likely due to phosphocreatine depletion that decreases ATP resulting in diminished NLRP3 activation^58,59,65. Additionally, conditions associated with metabolic syndrome or chronic inflammation e.g. Alzheimer’s Disease may result in persistently elevated ROS production from multiple sources, creating a pro-inflammatory environment that facilitates pyroptosis⁶⁶. These findings suggest that mitochondria contribute to NLRP3 inflammasome activation and pyroptosis through multiple interconnected mechanisms.

The tricarboxylic acid (TCA) cycle acts not only as an anabolic hub but also generates signaling metabolites that modulate immune responses by influencing mtROS and directly modifying cysteine residues⁶⁷. Metabolic profiling studies have identified TCA cycle alterations upon TLR activation in macrophages with succinate accumulation correlating with inflammatory cytokine production^14,38–40. Notably, the TCA metabolites itaconate and fumarate can also accumulate following macrophage activation and function as electrophiles controlling cystine modification of distinct proteins to control inflammation^68–75 (Figure 4). This cysteine-targeted signaling represents a mechanistic convergence with mtROS, establishing TCA metabolites as key integrators of metabolic and redox-dependent immune regulation. When studying the immunomodulatory effects of these metabolites, researchers often employ synthetic mimetics to elucidate the mechanisms of their endogenous counterparts^69,72,74. However, it is critical to note that endogenous itaconate and fumarate share some, but not all, molecular targets with their synthetic mimetics, such as dimethyl itaconate (DI), 4-octylitaconate (4-OI) or dimethyl fumarate (DMF), presenting a significant challenge for precise mechanistic interpretation⁷⁶.

Figure 4: Endogenous itaconate and fumarate function as electrophiles that directly modify immune signaling through cysteine alkylation of target proteins.

Left) Following TLR4 or IFNAR activation, ACOD1 (IRG1) expression increases, catalyzing the conversion of cis-aconitate to itaconate. Itaconate modifies cysteines in GAPDH to reduce glycolytic flux and also inhibits peroxiredoxin-5 (PRDX5) through a non-covalent mechanism, elevating mitochondrial H₂O₂ levels that enhance cGAS-STING signaling and type I interferon production. Itaconate can also inhibit succinate dehydrogenase (SDH) activity preventing succinate oxidation to fumarate and potentially altering mitochondrial ROS production. Right) Like itaconate, endogenous fumarate acts as an electrophile capable of cysteine modification. Fumarate is known to inhibit GAPDH similarly to itaconate and can also inactive KEAP1, which induces NRF2 activity. Fumarate hydratase (FH) activity is reduced by TLR and IFNAR signaling, while fumarate accumulation promotes mitochondrial DNA and RNA release that activates type I interferon responses. Importantly, synthetic derivatives of these metabolites—including dimethyl itaconate (DI), 4-octyl itaconate (4-OI), dimethyl fumarate (DMF), and monomethyl fumarate (MMF)—act as significantly stronger electrophiles than their endogenous counterparts. Synthetic itaconate derivatives exerts anti-inflammatory effects through multiple pathways including directly alkylating cysteine residues on gasdermin D (GSDMD) to reduce late inflammasome activation, KEAP1 to activate NRF2-dependent antioxidant pathways and signaling, JAK1 to inhibit downstream signaling, and ATF3 to inhibit IκBζ activity. Chemical fumarate mimetics show similar effects on KEAP1 and NRF2 activity. However, their enhanced electrophilic activity and expanded protein targets distinguish them functionally from the more selective effects of endogenous itaconate and fumarate, highlighting the importance of distinguishing between physiological metabolite signaling and pharmacological electrophile effects in immune regulation.

Itaconate production rapidly increases during the first 24 hours following macrophage activation as Immune-responsive gene 1/Aconitate Decarboxylase 1 (IRG1/ACOD1) becomes highly upregulated, catalyzing the decarboxylation of cis-aconitate to generate substantial quantities of itaconate^68,77,78. Initially characterized for its antimicrobial properties⁷⁹, IRG-1 deficient macrophages exhibit enhanced inflammatory cytokine production in vivo, suggesting a potential role for itaconate in curtailing inflammatory responses^80,81. Further work has continued to place itaconate as a central immunomodulatory metabolite with several distinct mechanisms of action.

Initial studies found that itaconate inhibited SDH, preventing succinate oxidation^68,82. These data suggest itaconate functions as an endogenous brake on mitochondrial electron transport chain activity during inflammation. Beyond its effects on SDH, itaconate’s electrophilic properties enable direct modification of proteins through cysteine alkylation similar to the mechanism of H₂O₂-mediated signaling. Synthetic derivatives of itaconate directly reduces NLRP3 inflammasome assembly, thereby suppressing IL-1β production⁷⁰. Similarly, itaconate derivatives can also modify multiple cysteine residues on JAK1, inhibiting its kinase activity and impairing IL-4 dependent macrophage polarization⁷². The physiological importance of these findings remains unclear as these studies have used chemical mimetics and not endogenous itaconate. Further, the observed effects occur at early time points following macrophage activation prior to significant endogenous itaconate production in these cells. Critically though, endogenous itaconate does confer tolerance to late inflammasome activation potentially by modifying gasdermin D, and limiting its processing and pore-forming activity that would otherwise drive pyroptosis⁷¹. Further, a recent study has revealed that endogenous itaconate plays a key role in regulating inflammation in human macrophages. The absence of endogenous itaconate results in elevated cytokine production, including TNFα, IL-6, IL-1β, and CXCL10, underscoring its essential function as an immunomodulatory feedback mechanism in innate immune responses. In this system, itaconate has been proposed to suppress inflammatory cytokine production by modifying cysteine residues on IRAK4, thereby hindering its activation and facilitating NF-κB degradation⁸³.

Itaconate accumulation can also activate NRF2 activity in activated macrophages^69,76. Nrf2 activation drives the expression of enzymes involved in glutathione synthesis and regeneration, potentially limiting mtROS and creating a negative feedback loop to resolve inflammation. Interestingly besides buffering ROS, NRF2 has been proposed to act as a transcriptional repressor directly reducing pro-inflammatory cytokine expression in macrophages⁸⁴. Although the exact mechanism by which endogenous itaconate drives NRF2 induction is unknown, synthetic derivatives can alkylate and inhibit Keap1, the upstream inhibitor of Nrf2⁷³. In addition to altering ROS, itaconate and it’s chemical derivatives also regulate inflammation by modifying glycolytic enzymes in macrophages, including GAPDH, and LDHA, thereby inhibiting glycolysis and promoting an anti-inflammatory state^85,86. These findings highlight the diverse immunometabolic roles of itaconate and license it as a key immunometabolic regulator with implications for treating inflammatory diseases. Importantly, Itaconate derivatives like 4-OI and DI show stronger electrophilic activity then endogenous itaconate, suggesting potential therapeutic applications⁷⁶. Dimethyl itaconate regulates inflammation through the induction of electrophilic stress⁶⁹. Specifically, DI initiates an integrated stress response that activates ATF3, a transcription factor that subsequently suppresses IκBζ expression, thereby inhibiting IL-17-mediated inflammatory pathways. This represents a distinct mechanism from endogenous itaconate’s effects and operates independently of the Nrf2 pathway commonly associated with electrophilic compounds. Importantly, using DI in a mouse psoriasis model reduces inflammation, suggesting therapeutic potential for autoimmune diseases mediated by the IL-17–IκBζ axis⁶⁹. These findings indicate that diethyl itaconate (DI) is a stronger electrophile than endogenous itaconate, resulting in distinct molecular targets and effects. The cellular sensors linking electrophilic stress to transcriptional reprogramming remain poorly defined, as does the full spectrum of proteins modified by itaconate and its derivatives.

While synthetic itaconate exerts anti-inflammatory effects, recent studies have uncovered context-dependent pro-inflammatory functions including the activation of type I interferon production^76,87. Importantly, recent work has suggested that itaconate may enhance type I interferon production independent of its electrophilic activity. Specifically, endogenous itaconate can non-covalently inhibit peroxiredoxin 5 (PRDX5), which increases mitochondrial derived H₂O₂ in activated macrophages, contributing to the amplification of type I interferon responses⁸⁸. This non-covalent mechanism reveals a functional divergence from itaconate’s electrophilic activities. Although much of the mechanistic work exploring itaconate’s regulation of interferon signaling has been performed in vitro, in vivo data supports these findings^88,89. Interestingly, besides its role in promoting initial antiviral responses, a recent study has proposed that itaconate also enhances inflammatory cytokine production in alveolar macrophages in response to LPS in contrast to its previously described anti-inflammatory effects⁹⁰. These studies demonstrate the growing complexity of endogenous itaconate’s immunologic effects across different cell types and inflammatory contexts and highlight the need for additional studies to further define the scenario specific effects of itaconate and its derivatives in diverse disease states.

Building on the electrophilic and immunomodulatory properties of itaconate, recent investigations have revealed fumarate, another TCA cycle metabolite, can also impact inflammatory responses. While both metabolites share electrophilic properties enabling cysteine modifications, their generation and downstream effects reveal distinct regulatory patterns in immune signaling networks. Recent studies have suggested that fumarate hydratase (FH), the enzyme that converts fumarate to malate, serves as a critical restraint on inflammatory cytokine production in epithelial cells and type I interferon production in macrophages respectively^74,75. Genetic deletion of FH leads to fumarate accumulation, which has been proposed to promote the release of mitochondrial RNA or DNA into the cytosol, driving a robust type I interferon response. These findings suggest that supraphysiological fumarate accumulation may have pro-inflammatory effects through mitochondrial stress and nucleic acid release. Importantly, a recent study has found that early during viral infections metabolic rewiring of the aspartate-arginoinosuccinate shunt (AAS) generates fumarate that promotes interferon-β production. Inhibition of the AAS, which reduces fumarate, suppresses interferon responses in vivo; further licensing fumarate as antiviral metabolite in physiologic contexts⁹¹. Interestingly, in different physiologic contexts, endogenous fumarate can also function as an electrophile capable of activating the NRF2 pathway to sustain antioxidant capacity⁹². This is likely a response to the electrophilic nature of fumarate, which can conjugate with GSH and potentially deplete cellular GSH levels^93,94.

The translational significance of these electrophilic metabolites is exemplified by DMF, an FDA-approved treatment for multiple sclerosis and psoriasis. DMF possesses stronger electrophilic properties than endogenous fumarate, leading to more robust cysteine modifications and anti-inflammatory effects primarily attributed to Nrf2 activation through KEAP1 alkylation⁹⁵. DMF covalently modifies cysteine residues in KEAP1 to activate the NRF2 pathway, enhancing antioxidant responses and reducing inflammation^96,97. Additionally, DMF suppresses NF-κB signaling, decreasing the production of pro-inflammatory cytokines independent of NRF2 activation^98–100. A chemical proteomic study has identified specific cysteine residues in proteins like PKCθ that are sensitive to DMF modification, providing insights into its immunomodulatory mechanisms¹⁰¹.

Collectively, these studies in macrophages reveal an integrated network where mitochondrial ROS and electrophilic TCA cycle metabolites converge on cysteine-targeted signaling to regulate immune function. While complex I-derived ROS enhances pro-inflammatory IL-1β release and complex III-generated ROS drives IL-10 secretion, metabolites like itaconate and fumarate function as endogenous electrophiles that modify specific target proteins to control range of immune responses. This convergence establishes TCA cycle intermediates as critical coordinators of metabolic and redox-sensitive immune regulation. Synthetic electrophiles like DMF and DI/4OI demonstrate promising anti-inflammatory effects, though future research must clarify the hierarchical relationships between these pathways and identify the physiological contexts in which these metabolites dictate specific immunological outcomes.

Mitochondrial ROS in T Cell Biology

Just as mtROS orchestrate critical aspects of innate immune cell function, they also serve as essential signaling molecules in T cell biology. T cell receptor (TCR) engagement triggers a complex cascade of signaling events and metabolic adaptations where mtROS play crucial regulatory roles^102–104. This transient increase in mtROS activates downstream TCR signaling pathways, influencing T cell activation and antigen specific proliferation. While the fundamental importance of mtROS in initial TCR signaling has been extensively reviewed elsewhere¹⁰⁵, recent advances have deepened our understanding of how these reactive molecules shape T cell fate decisions and functional outcomes in both physiological and pathological contexts.

Several studies have demonstrated that excessive ROS can impair Th17 cell differentiation and function with key metabolic enzymes, which control TCA cycle flux and ROS generation, implicated in this regulation. Both pyruvate dehydrogenase kinase 1 (Pdhk1) and glutaminase (Gls) deficiency suppress Th17 cell differentiation by altering mitochondrial metabolism and increasing ROS^106,107. Beyond these enzymes, antioxidant systems that neutralize ROS are equally important for optimal T cell metabolism and Th17 cell function. T cell-specific deletion of glutamate cysteine ligase (Gclc), the rate-limiting enzyme for glutathione synthesis, severely disrupts ROS balance and impairs MYC dependent metabolic reprogramming to sustain T cell proliferation as well as Th17 cell IL-22 production during infection^108,109. Building on this work, a recent study has demonstrated sex-specific differences in ROS regulation, with androgen receptor signaling restricting glutaminolysis and mitochondrial respiration in male T_h17 cells, resulting in lower Th17 cell-intrinsic ROS production compared to female Th17 cells; which potentially links ROS to the higher prevalence of Th17 cell-mediated diseases observed in women¹¹⁰. Collectively, these studies reveal that oxidative stress inhibits Th17 cell differentiation and activity in a variety of contexts.

In contrast, a growing body of evidence demonstrates that low concentrations of ROS signaling, specifically from mitochondria, are positive regulators of Th17 cell differentiation and function. IEX-1 deficient mice exhibit enhanced mtROS production following T cell activation which promote Th17 cell differentiation, an effect that could be reversed by treatment with mitochondria-targeted antioxidants, confirming the specific role of mtROS¹¹¹. Another, mechanistic link between mtROS and Th17 cell differentiation has been recently proposed, demonstrating that high glucose intake promotes Th17 cell differentiation by increasing mtROS, which activate TGF-β from its latent form¹¹². This finding aligns with earlier work showing that mtROS is required for SMAD signaling downstream of TGF-β receptor activation in other cell types, suggesting a conserved role for mtROS signaling in TGF-β-dependent processes¹¹³. Beyond growth factor signaling, calcium homeostasis also influences mtROS production in Th17 cells. Calcium signaling through Stromal interaction molecule 1 (STIM1) has been shown to regulate mitochondrial membrane potential and ETC function in Th17 cells, with STIM1 deletion leading to increased mtROS production and a shift toward a non-pathogenic Th17 cell phenotype¹¹⁴. Additionally, age-associated increases in inflammatory Th17 cell cytokine production are linked to defects in mitochondrial bioenergetics and elevated mtROS, which could be ameliorated by metformin treatment that normalized ROS balance¹¹⁵. Taken together, these studies reveal a critical dichotomy in how ROS influences Th17 cell biology: excessive oxidative stress inhibits Th17 cell differentiation and function while, basal mtROS serves as an essential signaling intermediate that dictates Th17 cell development and the degree of effector activity. This biphasic relationship positions ROS signaling as a potential therapeutic target in Th17 cell-driven inflammatory diseases, where specific cellular ROS pools could be used to modulate Th17 responses depending on the disease context.

The balance between pro-inflammatory Th17 cells and immunosuppressive regulatory T cells (Tregs) represent a critical axis that determines outcomes in autoimmunity, infection, and tissue homeostasis. Given their opposing functions and reciprocal developmental pathways, it is particularly noteworthy that these cell populations display opposing responses to ROS signaling. In contrast to Th17 cells, reduced antioxidant defenses resulting in elevated ROS appear to enhance Treg cell functionality. Deletion of peroxiredoxin II (PRDXII), an intracellular antioxidant that eliminates H₂O₂, significantly attenuated experimental colitis in mice while increasing Treg cell number and function¹¹⁶. Similarly, mice deficient in both glutathione peroxidase 1 (GPX1) and catalase display attenuated colitis, hyper-functional Treg cells, and impaired Th17 cell differentiation¹¹⁷. These findings suggest that elevated cellular ROS may enhance immunoregulatory mechanisms through Treg cell potentiation, opposite to effects observed in Th17 cells¹⁰⁶. Interestingly, while elevated cellular ROS appear to favor Treg cell differentiation and function, Treg cell-specific impairment of glutathione synthesis, compromises Treg cell function¹¹⁸. These effects operate independent from elevated ROS, as limiting serine availability restored function of Gclc-deficient Treg cells, suggesting GSH in Treg cells regulates cellular activity primarily through metabolic reprogramming rather than ROS neutralization¹¹⁸.

Although genetic impairment in antioxidant defenses resulting in elevated ROS appear to enhance Treg cell development and activity, high concentrations of mtROS appear detrimental to Treg cell function. Treg cells from individuals with autoimmunity and from experimental autoimmune encephalomyelitis mouse models showed elevated mitochondrial oxidative stress, DNA damage, and increased cell death¹¹⁹. Scavenging of mtROS in these Treg cells reverses these effects, preventing cell death, and attenuating autoimmune responses. Environmental factors also influence Treg cell function through mitochondrial mechanisms, as demonstrated by high salt exposure, which promotes autoimmunity by disrupting mitochondrial respiration, increasing mtROS, and downregulating FOXP3 transcription factor expression¹²⁰. Collectively, these findings reveal an intriguing paradox in T cell redox regulation: while elevated total cellular ROS appears to promote Treg cell development and function while inhibiting Th17 cells, mitochondrial-specific ROS exhibits the opposite effect—enhancing Th17 function while impairing Treg cell stability. This apparent contradiction likely reflects the diverse sources and compartmentalization of ROS within cells, where mtROS pools may activate distinct signaling pathways compared to cytosolic or peroxisomal ROS. Future studies should aim to delineate how specific ROS sources and subcellular localization dictate the differential development and function of Th17 versus Treg cells, potentially revealing new therapeutic targets for modulating this critical immune balance in inflammatory and autoimmune diseases.

CD8⁺ T cells undergo marked metabolic adaptations upon activation to support proliferation, effector function, and persistence. Recent studies have highlighted how mitochondrial fitness and redox balance critically influence CD8⁺ T cell fate decisions, particularly during chronic antigen stimulation (Figure 5). Specifically, CD8⁺ T cells with lower mitochondrial membrane potential (ΔΨm) and reduced mtROS production demonstrate improved metabolic fitness, superior persistence, and enhanced tumor-killing capacity¹²¹. Supporting this concept, chronic antigen exposure impairs mitochondrial oxidative phosphorylation and increases mtROS production, limiting CD8⁺ T cell self-renewal and stemness - effects that could be reversed by specifically targeting mtROS¹²². These findings suggest that while basal concentrations of mtROS may be necessary for TCR signaling and T cell function, excessive mtROS accumulation can drive T cell dysfunction and impair long-term persistence.

Figure 5: ROS-Dependent Regulation of CD8+ T Cell Fate and Function.

Mitochondrial ROS serves as a critical determinant of CD8+ T cell biology, with distinct concentration thresholds governing cellular outcomes across different activation states. Under physiological conditions (top panel), low basal concentrations of H₂O₂ are essential for maintaining both naïve and memory T cell populations. However, chronic antigen exposure (middle panel) disrupts redox balance, leading to progressive H₂O₂ accumulation that drives CD8+ T cell exhaustion and impaired effector function. When H₂O₂ generates lipid radicals concentrations at sufficient levels in T cells to trigger ferroptosis (bottom panel). In this pathway, excessive H₂O₂ generation leads to hydroxyl radical (•OH) formation via the Fenton reaction, which attacks polyunsaturated fatty acids (PUFAs) in cellular membranes to produce lipid peroxides (•LOOH). This cascade of oxidative damage overwhelms cellular antioxidant defenses, particularly the glutathione/GPX4 system, ultimately resulting in membrane destruction and cell death. The progressive increase in mtROS from physiological signaling to pathological oxidative stress illustrates how ROS homeostasis critically governs T cell biology, with implications for immunotherapy, chronic infections, and cancer immunity.

Consistent with this model, multiple recent studies have found that chronic antigen stimulation is associated with elevated quantities of mtROS that appear to drive CD8⁺ T cell exhaustion. Studies of HBV-specific CD8⁺ T cells have demonstrated that mitochondrion-targeted antioxidants could partially restore the function of exhausted cells by neutralizing excess mtROS, directly implicating ROS signaling in CD8⁺ T cell exhaustion¹²³. While PD-1 signaling contributes to early metabolic defects and elevated mtROS in exhausted T cells¹²⁴, subsequent work reveals that these redox abnormalities become fixed over time, with chronically stimulated T cells generating increased mitochondrial superoxide that compromises their function even when PD-1 is blocked as in the case of immune checkpoint blockade therapies¹²⁵. This suggests that while PD-1 signaling may initiate metabolic dysfunction, prolonged exhaustion establishes mtROS-generating pathways that operate independently of immune checkpoints. Independent of PD-1 pathways, mitochondrial quality control also influences mtROS production, as tumor-infiltrating CD8⁺ T cells accumulate damaged mitochondria due to impaired mitophagy, resulting in elevated mtROS that appears to drive epigenetic reprogramming toward exhaustion¹²⁶. Environmental factors further amplify these effects, with the combination of continuous antigen stimulation and hypoxia synergistically increasing mtROS generation to rapidly accelerate exhaustion¹²⁷. Recently, a potential mechanistic link has suggested that mtROS accumulation promotes redox-dependent stabilization of HIF-1α, potentially facilitating the transition from precursor to terminally exhausted CD8⁺ T cell states¹²⁸.

Importantly, while multiple studies associate increased mtROS with CD8⁺ T cell exhaustion, much of this evidence remains correlative. Defining the causality in these relationships requires genetic tools that can specifically isolate mtROS production from other mitochondrial functions. Recent work from our group using a genetic approach that couples mitochondrial complex III deficiency with AOX expression, a system that restores electron transport and metabolic function without generating complex III-derived superoxide³⁴ has revealed surprising and context-specific requirements for mtROS in CD8+ T cell biology. CD8⁺ T cells lacking mitochondrial complex III show three major impairments during viral infection: severely diminished proliferation, failure to generate memory precursors, and development of a phenotype resembling exhaustion early during acute infection. Restoring electron transport with AOX largely rescued proliferation and reversed the exhausted phenotype, as well as rescued initial memory precursor formation suggesting this process to do not require mitochondrial complex III-derived mtROS. However, these memory cells still fail to complete their differentiation or persist long-term (Steinert at al. Nature Immunology. Online July 16^th In Press). These findings demonstrate that mitochondrial complex III-derived mtROS is essential for terminal memory CD8⁺ T cell development and/or maintenance but is not causal for the development of an exhausted phenotype. This work highlights how genetic approaches that can isolate specific mitochondrial signals in physiologically relevant contexts. Moving forward, applying these genetic tools and others that directly dampen mROS across diverse immune challenges will be essential to delineate the precise targets of specific mtROS sources in shaping T cell responses.

T cell exhaustion represents a major mechanism by which tumors evade immune control, with CD8⁺ T cells often rendered dysfunctional within the tumor microenvironment. Beyond driving exhaustion, altered redox balance impacts multiple aspects of T cell antitumor function. Tumor-infiltrating CD8⁺ T cells often exhibit fragmented mitochondria that generate excessive mtROS, with ROS scavengers restoring their activity¹²⁹. Genetic manipulation of mitochondrial function further demonstrates that increased oxidative stress impairs CD8⁺ T cell antitumor responses¹³⁰, while CD8⁺ T cells responding to radiation therapy after initial immunotherapy failure, demonstrate lower ROS accumulation¹³¹. mtROS also influences T cell migration within tumors, affecting efficient tumor infiltration¹³². Importantly, redox balance extends beyond CD8⁺ T cells, as intratumoral Treg cells undergo apoptosis due to oxidative stress yet maintain suppressive function through adenosine signaling, contributing to immunotherapy resistance¹³³. These findings have therapeutic implications, as evidenced by a recent clinical study showing that asparagine deprivation, which temporally modulates ROS in CD8⁺ T cells, enhances responses to immune checkpoint blockade in patients with recurrent metastatic nasopharyngeal carcinoma¹³⁴. Collectively, these studies highlight how mitochondrial ROS balance influences multiple facets of T cell antitumor immunity and suggest that modulating mtROS could enhance immunotherapy efficacy. However, it is essential to recognize that most existing therapeutic paradigms are built upon correlative observations rather than causal mechanisms. As the field moves forward, we must critically reevaluate the existing literature on mtROS in T cell function, acknowledging that correlative associations no matter how consistent across multiple studies may not reflect true causal relationships when tested with definitive genetic approaches. This recognition should guide future research efforts and the development of more precisely targeted therapeutic strategies.

An emerging role for ROS in regulating antitumor immune responses involves ferroptosis, an iron-dependent form of regulated cell death driven by the accumulation of lipid peroxides. Initial studies identified CD8⁺ T cells as inducers of ferroptosis in cancer cells, revealing an additional mechanism of immune-mediated tumor control. IFNγ released from CD8⁺ T cells can downregulate the expression of cystine transport systems in tumors, impairing cystine uptake and promoting lipid peroxidation and ferroptosis¹³⁵. Subsequent work has shown that specific fatty acids can synergize with this mechanism, as T cell-derived IFNγ combined with certain lipids enhances tumor ferroptosis and anti-tumor immunity¹³⁶. These findings suggest promising therapeutic approaches combining immunotherapy with ferroptosis activators to enhance tumor killing.

However, recent evidence has identified that T cells themselves may be more vulnerable to ferroptosis than certain cancer cells, complicating therapeutic approaches. Early studies have documented that T cell lipid peroxidation induces ferroptosis and prevents immunity to infection, with GPX4-deficient T cells rapidly accumulating membrane lipid peroxides and undergoing ferroptotic cell death¹³⁷. This vulnerability has been confirmed in the tumor context by pharmacological screening, which identifies CD8⁺ T cells as more sensitive to GPX4 inhibitors than tumor cells¹³⁸. Further supporting this concept, recent work has demonstrated that maintaining the glutathione (GSH)/GPX4 axis is critical for CD8⁺ T cell metabolic fitness and antitumor function, with a distinct gene signature of GSH metabolism correlating with improved clinical responses and survival across several human cancers¹³⁹. This fits with the observation that CD36-mediated uptake of fatty acids by tumor-infiltrating CD8⁺ T cells functions as an inducer of lipid peroxidation and ferroptosis, reducing cytokine production and impairing antitumor function¹⁴⁰. Importantly, blocking CD36 or inhibiting ferroptosis in CD8⁺ T cells effectively restores their antitumor activity and enhances efficacy in combination with checkpoint blockade therapy.

Ferroptotic vulnerability extends beyond CD8⁺ T cells to other immune populations. Regulatory T cells depend on GPX4 to prevent lipid peroxidation and ferroptosis during activation, with Treg cell-specific deletion of GPX4 reducing tumor growth, but enhancing colitis by impairing Treg cell function^141,142. Treg cells show a preference for PUFA incorporation compared to other effector T cells potentially driving autoimmunity in some contexts¹⁴³. Similarly, follicular helper T (Tfh) cells show heightened susceptibility to ferroptosis during the germinal center reaction, with the selenium-GPX4 axis serving as a critical survival mechanism¹⁴⁴. Interestingly, supplementation of selenium increased Tfh cell numbers and enhances the antibody response to influenza vaccination. Beyond T cells, ferroptosis of tumor-associated neutrophils also influences antitumor immunity¹⁴⁵, as their death releases oxygenated lipids that limit T cell activity, yet their immunosuppressive function is abrogated when ferroptosis is inhibited. The varying sensitivities to ferroptosis and mixed immunologic effects across cell types present both challenges and opportunities for cancer therapy. Moving forward, it will be crucial to better understand the specific metabolic inputs and cell-specific antioxidant defense mechanisms that determine ferroptotic vulnerability in particular immune compartments. Such insights could inform more targeted approaches that selectively induce ferroptosis in tumor cells while preserving or enhancing immune cell function, potentially improving outcomes of cancer immunotherapy.

Challenges and Opportunities

Despite advancements in ROS detection techniques, major technical challenges persist, including the lack of specificity in current methods, the difficulty in distinguishing different ROS species, and the complexities of measuring ROS within specific cellular compartments¹⁴⁶. Many conventional detection assays, such as commercial fluorescent probes, suffer from cross-reactivity and experimental artifacts, while traditional methods like electron paramagnetic resonance and chemiluminescence are often hindered by sensitivity limitations and technical complexity¹⁴⁷. The transient nature of ROS and their rapid interactions further complicate accurate quantification, necessitating the use of multiple well-validated approaches to ensure reliable measurements. Differentiating between various ROS species (i.e., O₂•⁻, •OH, H₂O₂) is particularly difficult, as commonly used probes frequently react with multiple oxidants, leading to potential misinterpretations¹⁴⁶. To address these issues, researchers incorporate complementary detection strategies, including high-performance liquid chromatography, mass spectrometry¹⁴⁸, and genetically encoded fluorescent sensors such as HyPer7¹¹, to enhance specificity and enable real-time tracking of ROS fluctuations.

Another major challenge in ROS research is the ability to measure ROS production within specific cellular compartments. ROS are generated in multiple organelles, including mitochondria, peroxisomes, the endoplasmic reticulum, and the plasma membrane, each contributing to distinct cellular processes. Targeted fluorescent probes and genetically encoded sensors have been developed for compartment-specific ROS measurement, but their application is not without limitations. For example, mitochondria-targeted boronate probes (MitoB) allow for selective detection of H₂O₂ within mitochondria¹⁴⁹, and roGFP-based sensors can monitor redox changes in different organelles¹⁵⁰. Furthermore, MitoNeoD, a mitochondria-targeted probe, is preferentially converted to MitoNeoOH by superoxide without intercalating with DNA, producing a fluorescent signal that more accurately reflects superoxide quantities than MitoSOX^40,151. To investigate ROS-dependent immune responses in vivo, we propose generating mice that conditionally express a genetically encoded sensor for detecting H₂O₂ or lipid peroxides in specific cell types. As a complementary approach, an emerging method, Matrix-Assisted Laser Desorption-Ionization Imaging Mass Spectrometry (MALDI-IMS), offers spatial mapping of biomolecules directly from tissue sections¹⁵². MALDI-IMS can detect oxidative stress biomarkers and ROS-reactive probes, identifying oxidative modifications in lipids and proteins, such as 4-hydroxynonenal (4-HNE) and malondialdehyde (MDA). Additionally, exogenous ROS-sensitive probes (e.g., hydroethidine derivatives or BODIPY-based tags) can form stable oxidation products for high-resolution ROS activity mapping^153,154. By correlating these molecular signatures with histological features, MALDI-IMS enables precise visualization of ROS distribution in disease models.

While measuring ROS in immune cells is important, we advocate for tools that specifically perturb mitochondria-dependent ROS production to assess causal links to immune responses. Site-specific mitochondrial electron transport chain (ETC) inhibitors, such as S1QELs (complex I) and S3QELs (complex III), can suppress superoxide formation without disrupting normal electron transport, though they lack cell-type specificity^35,155,156. Alternatively, mice engineered to conditionally express alternative oxidase (AOX) coupled to loss of specific mitochondrial complex III subunits can eliminate superoxide production while maintaining electron transport^34,59,157. Similarly, conditional mitochondrial complex I subunit loss coupled with NDI1 expression disables superoxide production without affecting electron transport function^52,158,159. One caveat is that the lack of proton pumping by NDI1 and AOX diminishes mitochondrial ATP generation, though glycolysis can compensate to maintain ATP levels. Genetic models targeting antioxidant proteins like SODs, PRDXs, and GPX4 can further dissect the role of elevated superoxide, H₂O₂, and lipid peroxides in immune responses, although compensatory antioxidant defenses may confound results^160,161.

While current genetic models offer valuable insights into ROS signaling, they often lack the temporal and spatial control needed to establish causality in physiological contexts. To overcome these limitations, emerging genetic tools offer promising alternatives. D-tag mice represent one such approach, engineered to express a modified degron tag fused to proteins of interest, enabling targeted protein degradation via the ubiquitin-proteasome system¹⁶². When specific small molecule degraders are administered, they bind to the D-tag, recruiting E3 ubiquitin ligases that mark the tagged protein for proteasomal degradation, allowing precise, reversible control of protein concentrations in defined cell populations. Complementing this approach, D-Alanine oxidase (DAO) expressing mice provide a system for controlled, localized H₂O₂ generation¹⁶³. These mice can be engineered to conditionally express DAO in specific tissues or cell types, which catalyzes the oxidative deamination of administered D-alanine, producing pyruvate, ammonia, and importantly, H₂O₂¹⁶⁴. By controlling the timing and dose of D-alanine delivery (via diet, drinking water, or injection), researchers can generate defined concentrations of H₂O₂ in targeted tissues, directly testing causal relationships between H₂O₂ signaling and immune responses. Together, these innovative genetic systems provide powerful tools for dissecting the causal roles of specific proteins and ROS-mediated signaling in immune regulation under physiologically relevant conditions.

Finally, cysteine proteomics has become a key approach for mapping H₂O₂-mediated protein oxidation and redox signaling¹⁶⁵. Techniques like biotin-switch labeling, isotope-coded affinity tagging (ICAT), and chemoproteomic probes (e.g., dimedone-based DAz-2) selectively capture oxidized cysteines for mass spectrometry analysis^166–168. Click chemistry-based probes with azide or alkyne tags enhance labeling efficiency and specificity, while quantitative proteomics tools like oxidation-dependent tandem mass tags (TMTs) enable dynamic profiling of redox-sensitive cysteines^169–171. Once specific cysteines are identified as critical for H₂O₂ signaling, mutagenesis is essential to confirm functional relevance. Cysteine-to-serine (Cys-to-Ser) or cysteine-to-alanine (Cys-to-Ala) mutations prevent oxidation without altering protein function, distinguishing regulatory modifications from oxidative stress byproducts. Advanced mutations, such as cysteine-to-aspartate (Cys-to-Asp) or cysteine-to-glutamate (Cys-to-Glu), mimic irreversible oxidation states (-SO₂H or -SO₃H) to model chronic oxidative stress. By integrating cysteine mutagenesis with functional assays, structural biology, and proteomics, researchers can dissect ROS-driven pathways in immune activation, cytokine production, and pathogen defense. Knock-in mutations at endogenous loci in mice will help establish causality in redox biology, differentiating oxidative regulation from secondary damage in specific immune cell contexts.

Conclusion

In summary, mtROS have emerged as crucial regulators of immune function, influencing both innate and adaptive immunity through complex metabolic and signaling pathways. While excessive mtROS production drives inflammation, cytokine expression, and immune cell exhaustion, controlled mtROS generation is essential for immune activation, differentiation, and damage resolution. The intricate balance between mitochondrial metabolism, ROS production, and immune regulation is further shaped by TCA cycle intermediates such as succinate, itaconate, and fumarate, which modulate inflammatory responses via distinct yet overlapping molecular mechanisms. The therapeutic potential of targeting mtROS, either by limiting excessive oxidative stress or leveraging electrophilic metabolites like DMF and itaconate derivatives, presents exciting opportunities for treating inflammatory diseases, autoimmunity, and cancer. Future research should focus on defining the precise redox-dependent mechanisms governing immune cell fate, identifying key cysteine targets of mtROS signaling, and developing novel strategies to modulate mitochondrial metabolism for therapeutic benefit. By integrating metabolic and redox biology, new avenues for fine-tuning immune responses in health and disease can be deciphered.

Mitochondrial reactive oxygen species (mtROS) serve as critical signaling molecules that regulate immunity. This review examines the spectrum of mtROS effects, ranging from essential roles in innate immune activation and T cell differentiation to pathological inflammation and T cell dysfunction.