Role of NADPH oxidase 2-derived reactive oxygen species in cardiac electrophysiological disorders

ABSTRACT

Oxidative stress mediated by NADPH oxidase 2 (NOX2), a major source of reactive oxygen species in the myocardium, is an important mechanism that underlies cardiac arrhythmias. NOX2 is activated by the precise assembly of multiple protein subunits. It plays a critical role in initiating and sustaining atrial fibrillation and ventricular arrhythmias by disrupting calcium homeostasis, differentially regulating ion channel function, and inducing structural remodeling. Notably, a close association between NOX2 and arrhythmias is well-established, and interventions targeting NOX2 have demonstrated therapeutic potential. However, the signaling pathways through which NOX2 modulates myocardial electrophysiology under different pathological states remain unclear. Moreover, the dynamic coupling of NOX2 with upstream risk factors, such as abnormal cardiac load and metabolic diseases, remains poorly characterized. In this review, we summarize the structure, activation mechanisms, arrhythmogenic mechanisms, and targeted strategies of NOX2, providing a comprehensive reference for understanding its role in cardiac electrical remodeling and developing antiarrhythmic therapies.

Introduction

Arrhythmias, particularly atrial fibrillation (AF) and ventricular arrhythmias (VAs), are highly prevalent cardiovascular complications that severely affect clinical prognosis. Their underlying mechanisms remain incompletely understood, and precise, safe targeted therapies remain lacking [1,2].

NADPH oxidase 2 (NOX2) is a major myocardial source of reactive oxygen species (ROS), and its aberrant activation drives excessive ROS production, leading to cardiac electrical remodeling and structural remodeling [3,4]. Notably, elevated atrial oxidative stress is closely associated with AF; excess superoxide anions in AF primarily stem from abnormally upregulated NOX2 activity. Elevated atrial NOX2 activity also correlates with an increased risk of postoperative AF in patients undergoing cardiac surgery [5–7]. The recently identified X-ROS mechanoredox signaling pathway indicates that NOX2 regulates cardiac excitation – contraction coupling under physiological conditions. However, under pathological states, dysregulation of this pathway causes calcium (Ca²⁺) homeostasis disturbance and directly promotes arrhythmogenesis [8,9]. Importantly, NOX2 activation is well-established as a common pathway linking various risk factors, including pressure overload, hypoxia, metabolic diseases, rapid pacing, and structural heart disease, to the development of arrhythmias [10,11]. However, the precise mechanisms linking NOX2 activity to alterations in myocardial electrophysiology remain unclear across different pathological contexts, and the relationship between NOX2 and established risk factors (e.g. abnormal cardiac load and metabolic disease) warrants further investigation.

To our knowledge, comprehensive reviews focused specifically on the arrhythmogenic mechanisms and targeted therapeutic strategies of NOX2 are currently lacking. Therefore, we focus on NOX2 as a major target, summarizing its structural characteristics, activation mechanisms, differential regulatory roles between physiology and pathology, atria and ventricles, different disease stages and etiological backgrounds, its upstream and downstream signaling networks, and clinical translational potential.

Discovery, structure, and activation of NOX2

Discovery and structural characteristics of NOX2

In 1978, Segal et al. first identified cytochrome b558 (later designated NOX2/gp91phox) in neutrophil membranes and demonstrated its specific absence in patients with chronic granulomatous disease [12,13]. In the late 1980s, the gene encoding its catalytic subunit was successfully cloned [14,15]. Subsequent studies further reported that NOX2 functional activation relies on a multi-subunit assembly system, successively identifying major components, including the membrane subunit p22phox [16] and cytosolic regulatory subunits p47phox, p67phox, p40phox, and the small GTPase Rac [17–20].

Structurally, NOX2 possesses a conserved catalytic core comprising an N-terminal transmembrane domain and a C-terminal dehydrogenase domain. Its N-terminus contains six transmembrane helices responsible for chelating two heme groups, whereas the C-terminal dehydrogenase domain specifically binds flavin adenine dinucleotide and the substrate NADPH. In terms of assembly, NOX2 forms the membrane-bound heterodimer cytochrome b558 as the core complex, composed of the catalytic subunit gp91phox (NOX2) and the regulatory subunit p22phox. The coordinated assembly of these cytosolic subunits is required for electron transfer, ultimately mediating the generation of superoxide anion [3].

Activation mechanism of NOX2

NOX2 activation requires ordered protein – protein interactions and is tightly regulated to prevent excessive superoxide production. In resting cells, the flavocytochrome b558 heterodimer (NOX2 and p22phox) resides in the membrane, whereas p47phox, p67phox, p40phox, and Rac remain in the cytosol.

Cellular stimulation induces phosphorylation-dependent conformational changes that expose interaction motifs and promote cytosolic subunit translocation to the membrane. Activating stimuli vary by tissue and include phorbol myristate acetate, N-formylmethionyl-leucyl-phenylalanine, opsonized zymosan, and bacteria in neutrophils [21], as well as G protein-coupled receptor agonists (e.g. angiotensin II (Ang II) [22,23] and endothelin-1), growth factors, cytokines, and mechanical stress in cardiac tissue [24].

NOX2 activation begins with phosphorylation of the p47phox subunit at multiple serine residues (ranging from Ser303 to Ser379) [25–27] by upstream kinases, including protein kinase C (PKC), mitogen-activated protein kinase (MAPK) [25], and protein kinase B [28]. Phosphorylation induces a conformational change in p47phox, relieving the autoinhibitory state of tandem SH3 domains, which exposes the SH3 domains to bind with the proline-rich region on the p22phox subunit [29]. Subsequent phosphorylation of p47phox and p40phox also exposes their respective PX domains [30]. Binding to phosphatidylinositols (e.g. phosphatidylinositol 3-phosphate [PI3P]) anchors the cytosolic p47phox–p67phox–p40phox complex to the cell membrane [30]. Membrane-associated p47phox then recruits the activator subunit p67phox to directly bind to NOX2 [31]. It also facilitates incorporation of p40phox, which further stabilizes the complex – membrane association by binding to PI3P. Rac activation depends on 3-phosphorylated phosphoinositides catalyzed by members of the phosphoinositide-3-kinase family. Among these, phosphatidylinositol-3,4,5-trisphosphate specifically recruits various Rac activators to the plasma membrane [32]. Finally, GTP-bound Rac is recruited to the membrane and sequentially binds to NOX2 [33] and p67phox [34] to complete the complex. Once assembled, NOX2 catalyzes vectorial electron transfer across the membrane (or phagosomal membrane formed by internalization) to reduce O₂, generating O₂•⁻ (Figure 1) [35,36].

Figure 1.

Mechanism of NOX2 activation and ROS generation in the heart under pathological conditions. Under physiological conditions, ROS production is tightly balanced by antioxidant defense systems. Pathological stimuli, including pressure/volume overload, hypoxia/ischemia, diabetes/obesity, and tachypacing, disrupt this balance. These signals, including Ang II, growth factors, cytokines, and mechanical forces, activate receptor-mediated signaling pathways, including PI3K/AKT, PKC, and MAPK. These kinases phosphorylate the cytosolic subunits of NOX2 (p47phox, p67phox, p40phox) and promote the translocation of the Rac1/2 GTPase, leading to the assembly of the functional NOX2 complex at the membrane. The assembled complex catalyzes the conversion of NADPH to NADP⁺ +H⁺, driving ROS production. ROS, reactive oxygen species; Ang II, angiotensin II; PI3K, phosphoinositide-3-kinase; AKT, protein kinase B; PKC, protein kinase C; MAPK, mitogen-activated protein kinases.

The relationship between NOX2 activation and AF

Commonalities and differences in NOX2 across different causes of AF

Various AF risk factors, including pressure overload, hypoxia, metabolic diseases, rapid pacing [37], cardiac structural lesions, acute inflammation, and cardiac surgery promote atrial remodeling and AF susceptibility by activating the atrial NOX2 pathway [38]. Although the initial activating signals differ among various etiologies, NOX2-mediated oxidative stress serves as a common key node linking multiple risk factors to AF.

Pressure overload selectively activates NOX2 in the left atrium, triggering oxidative stress that modifies and impairs ryanodine receptor 2 (RyR2), thereby increasing sarcoplasmic reticulum (SR) Ca²⁺ leak [39]. Intermittent hypoxia can induce mitochondrial ROS production via a NOX2-dependent pathway [40], leading to downregulation and structural remodeling of atrial connexins Cx40 and Cx43 [41]. Regarding metabolic factors, obesity triggers AF by activating the pro-arrhythmic transcription factor PITX2 through NOX2 activation [42]. Conversely, diabetes increases AF risk by activating the calcium/calmodulin-dependent protein kinase II (CaMKII) signaling pathway via oxidative stress [43]. Rapid pacing also causes oxidative damage by inducing NOX2 expression and mitochondrial ROS production; such damage can activate PKCε in a frequency-dependent manner and upregulate the I_KACh current, driving electrical remodeling of AF [44]. Among structural cardiac lesions, severe mitral regurgitation is associated with elevated NOX2 activity and expression in the atrial myocardium [45]. Similarly, the pro-arrhythmic mechanism of hyperuricemia is closely associated with oxidative stress [46]. Additionally, acute inflammatory conditions, such as community-acquired pneumonia, can activate NOX2 via endotoxins and induce acute AF [47]. Atrial NADPH oxidase activity in patients post-cardiac surgery is an independent risk factor for postoperative AF [6,48]. Its activity levels are significantly higher in patients who develop postoperative AF than in those who maintain sinus rhythm and represent the strongest independent predictor of postoperative AF. This finding is strongly supported by humoral biomarkers; elevated urinary oxidative stress biomarkers (e.g. F2-isoprostane) during the perioperative period are associated with an increased risk of postoperative AF, providing direct clinical evidence for the core role of oxidative stress [49].

The renin – angiotensin system is a critical factor regulating AF onset and progression. Clinical meta-analyses have confirmed that angiotensin-converting enzyme inhibitors/angiotensin receptor blockers (ACEI/ARB) drugs can significantly reduce the overall risk of AF, with the most pronounced preventive effects observed in patients with heart failure and left ventricular hypertrophy [50,51]. Mechanistically, excessive activation of local cardiac Ang II can induce atrial dilation, electrophysiological disturbances, and the onset of AF, independent of systemic blood pressure and ventricular structural changes [52]. Moreover, Ang II can trigger oxidative stress by activating NOX2, subsequently modulating downstream kinase pathways and participating in atrial electrical and structural remodeling [22]. Within this framework, endothelial cells play a significant role in AF structural remodeling. Ang II activates NOX2 in endothelial cells, promoting endothelial – mesenchymal transition (EMT) and local inflammation that significantly drive myocardial fibrosis, which may be a critical mechanism in the formation of the atrial substrate [53]. This NOX2 activation is a critical step in Ang II -induced AF development [54,55], whereas Pak1 may be involved in the negative regulation of this oxidative stress pathway, influencing atrial remodeling progression [56]. Interventions targeting the renin – angiotensin system hold clinical promise in patients with cardiac structural remodeling, likely through multilevel modulation of oxidative stress and electrical remodeling [57]. The pro-arrhythmic effect of the Ang II – NOX2 axis extends beyond atrial remodeling to mediate ventricular-level injury via the AT₁ receptor. In a rapid atrial pacing (RAP) model, AT₁ receptor activation significantly upregulates ventricular NOX2 expression, triggers oxidative stress (increased F₂-isoprostanes and LOX-1), and induces coronary microcirculatory dysfunction and myocardial injury. Notably, this process can be completely blocked by irbesartan, providing a mechanistic explanation for non-coronary angina in patients with paroxysmal AF [58].

In summary, although different etiologies activate NOX2 through distinct initial signals, they all converge on the NOX2-ROS signaling hub, driving AF through three major mechanisms: Ca²⁺ homeostasis imbalance, ion channel remodeling, and structural remodeling. This common mechanism provides a theoretical basis for AF prevention and treatment strategies targeting NOX2.

Dynamic changes in NOX2 across different stages of AF

The role of NOX2 varies significantly across different stages of AF progression, exhibiting a dynamic pattern from being a core driver in the early phase to playing a diminished role in the advanced stage.

During the initiation and early maintenance of AF, NOX2-mediated oxidative stress is a core driving factor. Mechanistic studies in patients with human AF reported that atrial oxidative stress (increased superoxide) stems from enhanced activity of NAD(P)H oxidase, which contains the gp91phox subunit, rather than changes in its mRNA expression levels [59]. This conclusion was further supported at the transcriptomic level, with microarray studies reporting that atrial tissues from patients with AF exhibit upregulated pro-oxidant gene expression and downregulated antioxidant gene expression [60]. These data establish oxidative stress as a major driver of AF pathogenesis at the gene level.

Clinical biomarker studies further reveal the critical role of NOX2 in early AF. Serum levels of soluble NOX2-derived peptide (sNOX2-dp) and urinary isoprostane are significantly elevated in patients with paroxysmal and persistent AF [5]. In goat AF models and postoperative patients with AF, Rac1-NADPH oxidase activity and NOX2/p22phox expression in the left atrium are significantly upregulated in the early stage of the disease (within 2 weeks), independent of inflammatory cell infiltration [61].

As AF progresses, the dominant role of NOX2 gradually diminishes. In long-term AF, ROS are primarily derived from uncoupled nitric oxide synthase and mitochondrial oxidases. sNOX2-dp and urinary isoprostane levels show no significant changes in patients with permanent AF [5]. This mechanistic distinction explains the differential therapeutic efficacy of statins, which are effective in early-stage AF (by inhibiting the Rac1–NOX2 axis) but ineffective in persistent AF [61].

These findings suggest that NOX2-mediated oxidative stress may be a major driver of the onset and early maintenance of AF, rather than a passive consequence of persistent disease. This dynamic pattern provides an important rationale for developing stage-specific antioxidant therapeutic strategies for AF.

Multilevel mechanisms of NOX2-derived ROS-mediated AF: From triggered activity, reentrant substrate, to structural remodeling

Triggered activity

NOX2 activation induces calcium (Ca²⁺) handling dysfunction, promotes NOX-derived ROS generation, and impairs Ca²⁺ homeostasis in atrial myocytes, thus representing a critical arrhythmogenic mechanism for inducing AF. This mechanism involves Ca²⁺ leak from the SR and activation of Ca²⁺ signaling pathways [2,62,63]. Structurally, NOX2 is expressed on the sarcolemma and T-tubule membrane of cardiomyocytes, positioned in close proximity to RyR2 [8], thus facilitating functional coupling.

NOX2-derived ROS induce abnormal Ca²⁺ handling by acting on RyR2, significantly increasing AF susceptibility, as observed across multiple experimental models, including rapid pacing, drugs (e.g. ibrutinib), pressure overload, and metabolic diseases (e.g. obesity). Specifically, the anticancer drug ibrutinib activates NOX2, leading to extensive ROS production. ROS oxidize and activate CaMKII, resulting in the phosphorylation of RyR2 (at the Ser2814 site), SR Ca²⁺ leak, atrial structural remodeling, and increased AF susceptibility. NOX inhibitors can reverse this process [64], highlighting the therapeutic potential of targeting NOX2-derived oxidative stress. In a pressure overload model, mechanical stretch from transverse aortic constriction activates NOX2, directly causing oxidative modification of RyR2 and triggering SR Ca²⁺ leak. Using an RyR2 stabilizer addresses this leak and effectively prevents AF, confirming a causal relationship [39]. In an obesity model, diet-induced obesity activates NOX2 to generate ROS, which upregulates the expression of the transcription factor PITX2. This upregulation triggers Ca²⁺ channel remodeling, significantly increasing AF burden. Intervention with mitochondrial antioxidants or NOX2 inhibitors can reverse this electrical remodeling and mitigate AF susceptibility [42,65]. However, despite different triggers, the NOX/ROS signaling pathway converges with the RyR2 Ca²⁺ release channel; altering RyR2 function induces abnormal Ca²⁺ handling, representing a core convergent mechanism underlying AF across multiple etiologies (Table 1).

Table 1.

Comparison of NOX2 expression, activation and downstream effects between atrium and ventricle.

Common and Specific Mechanisms	Atrium	Ventricle	Commonalities
NOX2 Expression and Activity Level	Paroxysmal/postoperative/early AF: Significant upregulation of NOX2 (gp91phox) protein expression and activity [5,49,61,74] Controversial findings in persistent AF [5,61] Permanent AF: NOX2 expression returns to baseline with no significant elevation [5,61]	Increased NOX2 activity with unchanged protein expression [94,200] Upregulated NOX2 protein expression [201,202]	NOX2 activity is upregulated under pathological stimulation, accompanied by synchronous elevation of oxidative stress biomarkers
Upstream Activating Triggers	RAP [58], surgical trauma [6,64], infection [47], and drugs [64]	Ethanol [93], Duchenne muscular dystrophy [99,102,114] Indirect activation of ventricular NOX2 secondary to RAP [58]	Canonical upstream: Ang II/AT1R axis as the main activating pathway of NOX2 [56,58,83,110] Pressure overload [39,203] and metabolic diseases can also activate NOX2 [42,98,116]
Downstream Effects	Ca2+ homeostasis: NOX2/ROS induces RyR2 oxidation/phosphorylation and NCX dysfunction, leading to Ca2+ overload and SR Ca2+ leak [39,56,64] Ion channels: Increased INa,L [72], decreased ICa,L [70], and increased K+ current [44] Structural remodeling: Myofibril degradation [37], atrial fibrosis, left atrial enlargement [75,83], and Cx40/Cx43 remodeling [41]	Ca2+ homeostasis: Oxidative modification of SERCA and RyR2 leading to Ca2+ overload [88,93,94,99,102,110] Ion channels: Increased INa,L [111,112], increased ICa,L [110,111] and decreased K+ current [97,136,204] Structural remodeling: Cx43 remodeling [114], interstitial fibrosis, and ventricular systolic dysfunction [116,203]	Jointly trigger Ca2+ homeostasis dysregulation, ion channel remodeling, and myocardial fibrosis

Abbreviations: AF, atrial fibrillation; RAP, rapid atrial pacing; Ang II, Angiotensin II; AT₁R, Angiotensin II type 1 receptor; ROS, reactive oxygen species; RyR2, ryanodine receptor 2; SR, sarcoplasmic reticulum; NCX, sodium-calcium exchanger; I_Na,L, the late sodium current; I_Ca,L, L-type calcium current; APD, action potential duration; ERP, effective refractory period; SERCA, sarcoplasmic/endoplasmic reticulum calcium ATPase.

NOX2-derived ROS-mediated ion channel remodeling

Oxidative stress is a critical mechanism driving ion channel remodeling in atrial myocytes during AF, with NOX2-derived ROS playing a crucial role. This remodeling differentially regulates multiple ionic currents, ultimately shortening the action potential duration (APD), inducing triggered electrical activity, and slowing conduction. This establishes the electrophysiological substrate for AF initiation and maintenance via reentry [66,67]. Notably, the development of atrial AF is closely associated with dysfunction of various K⁺ channels [68,69]. For example. evidence from a diet-induced obesity model demonstrates that NOX2 activation triggers atrial electrical remodeling [42,65]. The primary mechanism involves NOX2-derived ROS activating PITX2 and PKC, which specifically upregulates ultrarapid delayed rectifier potassium current (I_Kur) and slow delayed rectifier potassium current (I_Ks), inhibits the expression and function of the Nav1.5 channel, significantly reduces peak sodium current (I_Na) density, which leads to conduction disturbances, and mediates a decrease in L-type calcium current density (I_CaL). This cascade collectively results in a marked shortening of APD. Beyond the kinase pathway, ROS can directly regulate K⁺ channels through post-translational modifications. Hydrogen peroxide significantly enhances I_Kur through the high redox sensitivity of K⁺ channels [70]. For example, sulfonation of C581 at the COOH-terminus of the K_V1.5 channel, which encodes I_Kur, is elevated in diseased human hearts. This modification redirects the channel from recycling to degradation, dynamically regulating membrane expression and current density [71]. Additionally, the acetylcholine-sensitive K⁺ current is another critical factor in AF electrical remodeling. This current is specifically upregulated by NOX2-derived oxidative damage via frequency-dependent activation of PKCε; NOX2 inhibition can block this pathway, preventing AF [44]. Hydrogen peroxide itself can also inhibit the sodium-calcium exchanger (NCX) current and I_CaL. Collectively, this suppression of inward currents, combined with the enhancement of outward K⁺ currents, shortens APD [70]. In aged and fibrotic atria, acute oxidative stress rapidly oxidizes and activates CaMKII, enhancing the late sodium current (I_Na,L), which further induces phase-3 early afterdepolarizations (EADs) and mediates focal triggered activity that initiates AF. Notably, this process can be effectively inhibited by CaMKII inhibitors or specific I_Na,L blockers [72].

Structural and gap junction remodeling and atrial substrate perpetuation

Activation of the NOX2–ROS axis directly induces structural remodeling. In a rapid pacing model, rapid atrial electrical activity generates ROS by activating the transforming growth factor-β and NOX2 signaling pathways; these two pathways are interconnected and lead to cardiomyocyte myofibril degradation [37]. In an obstructive sleep apnea model, intermittent hypoxia induces oxidative stress by activating NOX2, resulting in the downregulation and decoupling of atrial connexins (Cx40/Cx43), thereby promoting AF [41]. In an obesity model, NOX2 activation induces ion channel remodeling, directly increases AF, and slows down conduction velocity. Moreover, NOX2 knockout significantly alleviates fibrosis and improves conduction, confirming its core role [42,65]. In terms of upstream signals, Rac1 GTPase can enhance mineralocorticoid receptor activity by upregulating 11β-HSD2 expression, thereby driving fibrosis. This process is strongly associated with increased NADPH oxidase activity, which establishes the Rac1–NOX2–fibrosis axis [73]. RAP significantly enhances superoxide anion (O₂•⁻) generation in the left atrium and left atrial appendage, and this mechanism is associated with Rac1-mediated upregulation of NOX and xanthine oxidase activities. The exacerbation of such oxidative stress results in reduced nitric oxide bioavailability, which in turn promotes endothelial dysfunction, inflammatory responses, and tissue remodeling, further reinforcing the structural abnormalities of the atrial substrate [74].

NOX2-mediated ROS triggers AF by activating the CaMKII, MAPK, and PKC pathways

Multiple pathological stimuli, including Ang II, ibrutinib, diabetes mellitus, and heart failure, converge on the NOX2–ROS – CaMKII axis to drive AF [55]. Central to this pathway is the ability of NOX2-derived ROS to oxidize and activate CaMKII [43]. Activated CaMKII then phosphorylates major targets, including RyR2, which result in SR Ca²⁺ leak and triggered activity [64,75], and Nav1.5, which causes heterogeneous spatial distribution of I_Na and conduction heterogeneity [76]. Inhibition of NOX2 or CaMKII can reverse this electrical and structural remodeling, confirming this axis as a common core mechanism underlying AF initiation and progression.

Various pathological stimuli, including heart failure, aging, and binge alcohol consumption, induce ROS production by activating NOX2, which in turn triggers the MAPK pathway, especially the JNK2 isoform, thereby driving AF onset and progression [77,78]. Activated JNK2 inhibits the transcription of Cx43 through c-Jun phosphorylation, leading to impaired intercellular electrical coupling [79]. JNK2 can also directly phosphorylate or transcriptionally upregulate CaMKII, promote RyR2-mediated SR Ca²⁺ leak, and induce diastolic Ca²⁺ waves and triggered activity [80–82]. Ang II upregulates cathepsin K via the NOX2–ROS – p38 MAPK axis, facilitating collagen degradation and AD [83]. Plasma microvesicles collected from AF patients activate the MKK4/JNK2 pathway by binding to platelet CD36, further contributing to a prothrombotic state [84]. In a chronic alcohol exposure model, sustained JNK2 activation exerts a compensatory effect in maintaining ventricular systolic function, although at the expense of enhanced atrial arrhythmogenicity [85]. Intervention studies report that ACEI/ARB drugs alleviate fibrosis and AF substrate formation by inhibiting the Ang II – NOX2–ROS – MAPK pathway,53 whereas specific inhibition of JNK2 (e.g. Alda-1) can directly block JNK2 activity independently of ALDH2, reversing abnormal Ca²⁺ handling and AF susceptibility [86].

NOX2–ROS drives atrial ion channel remodeling and promotes AF by activating distinct PKC isoforms. In obesity models, the NOX2–PKCα/δ axis suppresses I_Na and enhances I_Ks, leading to shortened APD and slowed conduction [42,65]. In rapid pacing models, the NOX2–PKCε axis upregulates I_KACh in a frequency‑dependent manner (Figure 2) [44].

Figure 2.

Mechanisms by which NOX2-derived ROS promote atrial fibrillation. Following activation by upstream pathological signals, NOX2 generates ROS, driving arrhythmogenic remodeling through multiple pathways. Structurally, NOX2-mediated ROS activate JNK2, downregulate connexins Cx40 and Cx43, disrupt intercellular electrical coupling, and provide a structural substrate for reentry. At the electrophysiological level, ROS differentially regulate ion channels either directly or via activating downstream kinases, such as PKC and CaMKII: they suppress I_Na and I_Ca,L, while enhancing I_Kur, I_KACh, and I_Ks. These changes collectively shorten atrial APD and ERP, promoting the maintenance of reentry. Furthermore, NOX2–ROS signaling directly oxidizes the RyR2, or activates CaMKII via oxidation, which can also be phosphorylated or transcriptionally upregulated by JNK2. CaMKII phosphorylates RyR2 and triggers diastolic SR Ca²⁺ leak. Elevated cytosolic Ca²⁺ activates NCX, generating I_Ti that induces DADs. Activated CaMKII also enhances I_Na,L, evoking EADs. These abnormalities, including aberrant Ca²⁺ handling, impaired ion channel function, and structural remodeling, collectively create a pro-arrhythmic substrate that promotes the initiation and maintenance of atrial fibrillation. ROS, reactive oxygen species; JNK2, c-Jun N-terminal kinase 2; PKC, protein kinase C; CaMKII, calcium/calmodulin-dependent protein kinase II; APD, action potential duration; ERP, effective refractory period; RyR2, ryanodine receptor 2; SR, sarcoplasmic reticulum; NCX, sodium-calcium exchanger; I_Ti, transient inward current; DADs, delayed afterdepolarizations; I_Na,L, the late sodium current; EADs, early afterdepolarizations.

Multilevel regulatory mechanisms of the NOX2–ROS axis in VAs

NOX2/ROS induces Ca²⁺ handling abnormalities in VAs

Disruption of Ca²⁺ homeostasis in cardiomyocytes is a key factor in the induction of VA by NOX2/ROS. Its targets are the key Ca²⁺ -handling proteins in the SR. Oxidative stress (e.g. H₂O₂) can exert inverse regulation on the SR calcium ATPase (SERCA) and NCX through thiol oxidative modification – specifically inhibiting SERCA activity and enhancing NCX activity. These two effects synergistically deplete SR Ca²⁺ content rapidly, leading to reduced transient Ca²⁺ amplitude and acute systolic dysfunction [87,88]. Sympathetic activation-induced SR Ca²⁺ overload and diastolic Ca²⁺ leak constitute the classical pathogenesis of catecholaminergic polymorphic ventricular tachycardia (CPVT) [89,90]. Ca²⁺ cycling disorders mediated by NOX2/ROS are closely linked to CPVT. ROS derived from NOX2 can directly oxidize and modify RyR2, thereby increasing its probability of opening and inducing diastolic SR Ca²⁺ leak. The formation of inter-subunit disulfide bonds (Cys1078 and Cys2991) in RyR2 is the key structural basis for this pathological leak [91]. However, SERCA itself is also a critical target of oxidative modification; oxidation of its cysteine 674 (C674) inhibits Ca²⁺ reuptake and causes mitochondrial Ca²⁺ overload, ultimately initiating cardiomyocyte apoptosis and progressing from reversible electrical disorders to irreversible structural remodeling and heart failure [92]. In addition, NOX2/ROS mediates the oxidative activation of CaMKII. The activated CaMKII further exacerbates RyR2-mediated Ca²⁺ leak through phosphorylation modification [93,94]. Disruption of Ca²⁺ homeostasis induces characteristic ventricular arrhythmias in CPVT by triggering delayed afterdepolarizations (DADs) [95,96]. These pathological mechanisms have been verified in multiple models, including metabolic (high-fat diet) [94], toxic (ethanol) [93], and diabetic models [97,98], and can be reversed by NOX2 inhibition, highlighting the universality of this pathway.

During physiological myocardial stretching, the microtubule-dependent X-ROS signaling pathway is rapidly activated, prompting NOX2 to generate low levels of ROS. ROS act on RyR2 near the SR, sensitizing them, and triggering Ca²⁺ sparks [8]. This process finely regulates excitation-contraction coupling, accelerates the increase of Ca²⁺ transients, and enhances contractility – a mechanism that constitutes an important component of the heart’s Frank – Starling law [9]. However, this balance is disrupted under pathological conditions (e.g. Duchenne muscular dystrophy [DMD]), where the expression and activity of NOX2 are abnormally elevated (by up to 5-fold). This leads to the occurrence of spontaneous diastolic Ca²⁺ release events in cardiomyocytes and increased SR Ca²⁺ leak [99]. Additionally, the density and properties of the microtubule network undergo changes, resulting in excessive activation of X-ROS signaling [100]. Excessive ROS directly modifies RyR, inducing SR Ca²⁺ leak and spontaneous Ca²⁺ waves [101]. Conversely, ROS activate CaMKII [102], disrupting intracellular Ca²⁺ homeostasis and action potentials. These effects contribute to Ca²⁺ -dependent VAs [8] (Figure 3).

Figure 3.

Schematic illustration of NOX2-mediated oxidative stress in ventricular arrhythmogenesis. Mechanical forces activate NADPH oxidase 2 (NOX2) to generate stretch-induced reactive oxygen species (ROS), termed X-ROS, which modulates cardiomyocyte excitation-contraction coupling under physiological conditions. In pathological states, ROS oxidizes Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) to enhance ryanodine receptor 2 (RyR2) activity, triggering sarcoplasmic reticulum (SR) Ca²⁺ leak and elevated diastolic Ca²⁺ levels, which in turn induce delayed afterdepolarizations (DADs) via the sodium-calcium exchanger (NCX)-mediated transient inward current (I_Ti). Additionally, risk factors such as high-fat diet, hyperglycemia, ethanol, epinephrine, tachycardia, and angiotensin II (ang II) activate NOX2 to generate ROS, remodeling cardiac ion channels by suppressing transient outward potassium current (I_to)and upregulating late sodium current (I_Na,L)and L-type calcium current (I_Ca,L), thereby inducing early afterdepolarizations (EADs). Furthermore, ROS modulate sarcoplasmic/endoplasmic reticulum calcium ATPase (SERCA)-mediated Ca²⁺ uptake and mitochondrial redox status, exacerbating mitochondrial Ca²⁺ overload to finally induce ventricular arrhythmias. Downregulation of connexins 40/43 also contributes to impaired intercellular coupling. ROS, reactive oxygen species; CaMKII, Ca²⁺/calmodulin-dependent protein kinase II; RyR2, ryanodine receptor 2; SR, sarcoplasmic reticulum; DADs, delayed afterdepolarizations; NCX, sodium-calcium exchanger; I_Ti, transient inward current; Ang II, angiotensin II; I_to, transient outward potassium current; I_Na,L, late sodium current; I_Ca,L, L-type calcium current; EADs, early afterdepolarizations; SERCA, sarcoplasmic/endoplasmic reticulum calcium ATPase.

Inhibiting NOX2 [99] or genetically blocking the oxidative activation of CaMKII [102] can restore abnormal Ca²⁺ handling, improve contractile function, and effectively prevent arrhythmia. Computational model studies have also reported that normalization intervention on X-ROS signaling can significantly reduce the risk of Ca²⁺ -dependent arrhythmias [100].

In summary, the X-ROS signaling pathway serves as a key regulator of mechanochemical coupling under physiological conditions, whereas its excessive activation under pathological conditions becomes a core link driving electrical remodeling and contractile dysfunction. In diabetic and obese models, palmitate-induced mitochondrial ROS is amplified by NOX2 (ROS-induced ROS release, RIRR), leading to mitochondrial ROS bursts and impaired function. Abnormal Ca²⁺ leak in the SR caused by palmitate can promote arrhythmias and heart failure [103].

Moreover, sustained or intense oxidative stress can steer reversible electrophysiological disturbances toward irreversible structural damage by activating downstream kinases. Pathological stimuli such as Ang II activate NADPH oxidase (dependent on the p47phox subunit), generating ROS that directly oxidize methionine residues (Met281/282) in the regulatory domain of CaMKII, leading to its sustained activation in the absence of Ca²⁺ signaling. This “oxidative activation” of CaMKII exacerbates Ca²⁺ -handling abnormalities and directly drives cardiomyocyte apoptosis [104].

NOX2/ROS induces electrophysiological remodeling

In addition to Ca²⁺ homeostasis, NOX/ROS directly regulates the electrophysiological properties of cardiomyocytes. At the ion channel level, these effects are complex and disease specific. In models of Takotsubo cardiomyopathy and acute hyperglycemia, high concentrations of epinephrine or hyperglycemia activate the NOX – ROS – PKC pathway, leading to Na⁺ channels dysfunction and I_Ca,L or to a reduction in transient outward potassium current (I_to). These changes cause abnormalities in APD and induce arrhythmias [97,105,106]. Clinical studies have confirmed a close correlation between long QT syndrome (LQTS) and stress cardiomyopathy. The Behr group proposed that reduced myocardial repolarization reserve is the intrinsic cause of QT prolongation under stress [107]. Abnormal activation of the NOX2/ROS pathway impairs myocardial repolarizing K⁺ channel function at multiple levels [108]. It suppresses I_to and I_Kur currents by downregulating Kv4.2/Kv4.3 and Kv1.5/Kv2.1 channel expression, respectively, and also reduces the rapidly activating delayed rectifier potassium current (I_Kr) current via phosphorylation modification mediated by multiple kinases [109]. Defects in distinct K⁺ channels correspond to specific LQTS subtypes: dysfunction of Kv7.1, Kv11.1 and K_ir2.1 leads to LQT1, LQT2 and LQT7 respectively [96].

In an Ang II model, the regulatory mechanism is more sophisticated. It enhances I_Na and I_Ca,L to establish an excitability foundation [110]. Conversely, it also directly induces triggered activities such as EADs by increasing the I_Na,L involving the NOX2-PKA axis and the NOX2-CaMKII axis [111]. Additionally, Ang II can directly inhibit the transcription of the cardiac sodium channel (SCN5A) via the NOX-ROS-NF-κB pathway, leading to a long-term reduction in Na⁺ [112]. The NOX2/ROS pathway serves as a core mechanism regulating Nav1.5 channel function and mediating the pathogenesis of LQT3 [90,113]. NOX2/ROS amplifies pathological I_Na,L, prolongs APD and induces EADs by regulating transcription, directly oxidizing cysteine residues of Nav1.5 or activating multiple kinase pathways [108]. In addition, NOX2/ROS participates in the development of LQT8 via oxidative modification of Ca²⁺ channels and calmodulin. LQT8 caused by gain-of-function mutations in CACNA1C is characterized by incomplete inactivation of I_Ca,L and increased Ca²⁺ influx [89,95]. Activation of NOX2 further elevates Cav1.2-mediated Ca²⁺ current through the CaMKII pathway and exacerbates repolarization abnormalities [111] (Table 1).

The role of NOX/ROS is more consistent in terms of gap junction regulation. Oxidative stress disrupts Cx43 function and causes abnormal electrical conduction by modifying its post-translational state. In the DMD model, NOX2-derived ROS oxidatively activate CaMKII, leading to detrimental hypophosphorylation of Cx43 at the S325/S328/S330 sites. This induces lateralization of Cx43 distribution and excessive hemichannel activity, increasing conduction heterogeneity and arrhythmia susceptibility; this process can be reversed by NOX2 genetic intervention [102,114]. In a type 1 diabetes model, cardiac conduction abnormalities were shown to be strongly associated with changes in Cx43 expression, distribution, and tyrosine phosphorylation/nitration modifications [115]. Therefore, despite the differences in upstream etiologies and specific modification sites (e.g. serine vs. tyrosine), NOX/ROS-driven Cx43 dysfunction serves as a key link between structural remodeling and electrical conduction disorders.

At the structural remodeling level, NOX/ROS indirectly and persistently impairs the anatomical basis of electrical conduction by inducing myocardial fibrosis and hypertrophy. For instance, in a diabetes model, hyperglycemia activates Rac1 and the downstream NADPH oxidase, triggering endoplasmic reticulum stress and inflammatory responses that lead to myocardial fibrosis, hypertrophy, and dysfunction. Targeted inhibition of Rac1 or NADPH oxidase effectively blocks this process and improves cardiac function [116]. This structural remodeling further exacerbates the heterogeneity and instability of myocardial electrical conduction.

NOX2/ROS regulate cardiac ion channel function and induce channelopathies and arrhythmias

Long QT syndrome

Downregulation of the expression and function of cardiac K⁺ channels induced by abnormal activation of the NOX2/ROS signaling pathway is a key pathological mechanism underlying LQTS. The repolarization of cardiac action potentials and the stability of the QT interval rely on outward repolarizing currents carried by K⁺ channels including Ito, I_Kur, I_Kr, I_Ks and inward rectifier potassium current (I_K1). Elevated endogenous ROS predominantly exerts an inhibitory effect on various cardiac K⁺ channels and broadly suppresses repolarizing K⁺ currents at the transcriptional, protein expression and functional levels [108] (Table 2). Mechanistically, ROS downregulates the expression of Kv4.2/Kv4.3 and Kv1.5/Kv2.1 channels via the thioredoxin and glutaredoxin systems, thereby directly inhibiting I_to [97] and I_Kur [117,118]. It also indirectly suppresses K⁺ channel activity through phosphorylation modification by activating downstream signaling cascades such as RTK-SRC, PKA and PKC [109], and directly modulates Kv11.1 function to reduce I_Kr [119]. Dysfunction of distinct K⁺ channels corresponds to specific LQTS subtypes: attenuation of Kv7.1-mediated I_Ks causes LQT1, reduction of Kv11.1-related I_Kr leads to LQT2, and loss-of-function of K_ir2.1 accompanied by decreased I_K1 results in LQT7 (Andersen-Tawil syndrome) [96,120–122]. Defects in these K⁺ channels ultimately lead to insufficient outward repolarizing currents in cardiomyocytes, prolonging ventricular APD and QT interval, increasing repolarization dispersion, and triggering EADs, which markedly raise the risk of life-threatening arrhythmias such as torsade de pointes.

Table 2.

Effect of NOX2/ROS on cardiac ion channel currents associated with channelopathies.

Protein/Channels	Changes by NOX2/ROS and Electrophysiological Disorders	Channelopathies	References
Nav1.5	NOX2/ROS directly oxidizes cysteine residues of the channel, delaying channel inactivation and causing pathological elevation of INa,L, prolonged APD, elevated plateau potential and phase 2 EADs; NOX2/ROS inhibits SCN5A transcription, reduces INa, slowed phase 0 depolarization, ventricular conduction block and increased transmural repolarization dispersion	LQT3 BrS	[111–113,125,126,205]
Cav1.2	NOX2/ROS oxidizes cysteine residues of Cav1.2 and activates CaMKII, leading to increased ICa,L (LQTS); Congenital loss-of-function mutations in CACNA1C/CACNB2/CACNA2D1 underlie SQT4–6; NOX2/ROS further suppresses ICa,L and reduces inward Ca2+ current (SQTS)	LQT8 SQT4/SQT5/SQT6 BrS3/BrS4/BrS11	[96,111,132,206–208]
RyR2	NOX2/ROS induces S-glutathionylation and S-nitrosylation of RyR2; Promotes disulfide bond formation at C1078 and C2991 of RyR2 subunits, raising channel open probability; ROS activates CaMKII, and CaMKII-mediated phosphorylation of RyR2 further exacerbates channel leak	CPVT	[88,90,91,134]
SERCA2a	NOX2/ROS induces thiol oxidation and inhibits SERCA2a activity, resulting in impaired SR Ca2+ reuptake	CPVT	[87,92]
NCX	NOX2/ROS modification upregulates NCX transport activity and enhances its forward mode (Ca2+ efflux and Na+ influx), generates ITi and directly triggers DADs	CPVT	[87,122,123]
Kv4.2/Kv4.3 (Ito), Kv1.5/Kv2.1 (IKur)	NOX2/ROS downregulates channel transcription and protein expression; PKC mediated phosphorylation inhibits channel function, decreased outward repolarizing Ito and IKur, prolonged APD and QT interval (LQTS); KCNE3 mutations, NOX2/ROS further upregulates Ito, increased right ventricular transmural repolarization dispersion (BrS)	LQTS BrS	[117,118,120,209,210]
Kv11.1 (IKr), Kv7.1 (IKs), Kir2.1 (IK1)	NOX2/ROS suppresses IKr, IKs and IK1 prolonged APD and QT interval (LQT1/LQT2/LQT7); NOX2/ROS upregulates IKr in a concentration-dependent manner, leading to abnormally enhanced outward currents, accelerated repolarization, shortened APD and QT interval (SQT1)	LQT1/LQT2/LQT7 SQT1	[96,119,120,211–213]

Abbreviations: AF, atrial fibrillation; RAP, rapid atrial pacing; Ang II, Angiotensin II; AT₁R, Angiotensin II type 1 receptor; ROS, reactive oxygen species; I_Na,L, the late sodium current; APD, action potential duration; EADs, early afterdepolarizations; LQT, Long QT syndrome; CaMKII, calcium/calmodulin-dependent protein kinase II; I_Ca,L, L-type calcium current; SQT, Short QT syndrome; BrS, Brugada syndrome; RyR2, ryanodine receptor 2; CPVT, catecholaminergic polymorphic ventricular tachycardia; SERCA, sarcoplasmic/endoplasmic reticulum calcium ATPase; SR, sarcoplasmic reticulum; NCX, sodium-calcium exchanger; I_Ti, transient inward current; DADs, delayed afterdepolarizations; PKC, protein kinase C; I_to, transient outward potassium current; I_Kur, ultrarapid delayed rectifier potassium current; I_Kr, rapidly activating delayed rectifier potassium current; I_Ks, slow delayed rectifier potassium current; I_K1, inward rectifier potassium current.

The NOX2/ROS pathway modulates the expression and function of the cardiac Na⁺ channel Nav1.5 (encoded by SCN5A), resulting in abnormally increased I_Na,L and impaired I_Na. This serves as a core molecular mechanism for LQT3 [90,113]. Ang II activates the NOX2-ROS-NF-κB signaling axis, which promotes nuclear translocation of NF-κB and its binding to the SCN5A promoter [112]. This process inhibits SCN5A transcription, reduces I_Na, delays Na⁺ channel inactivation, and elevates I_Na,L. Meanwhile, NOX2-derived ROS oxidatively activates the PKA and CaMKIIδ pathways, and PKA activated by ROS upregulates I_Na and I_Ca,L, and elevates the action potential plateau [110]. In addition, ROS directly oxidizes cysteine residues on Nav1.5 or activates multiple regulatory pathways including PKA, PKC, and CaMKII, to inhibit I_Na and enhance I_Na,L [108]. Pathologically elevated I_Na,L significantly prolongs APD and induces EADs.

By oxidatively modifying Ca²⁺ channels and calmodulin and disrupting cardiac Ca²⁺ homeostasis, the NOX2/ROS pathway increases I_Ca,L and prolongs APD, representing an important mechanism for Ca²⁺ -related LQTS subtypes including LQT8 and LQT14–17. LQT8 (Timothy syndrome) caused by gain-of-function mutations in CACNA1C is characterized by incomplete inactivation of I_Ca,L, enlarged window current, and augmented Ca²⁺ influx [89,95]. NOX2 activation induces massive ROS accumulation, which further upregulates Cav1.2-mediated I_Ca,L via oxidative activation of the CaMKII pathway and exacerbates abnormal Ca²⁺ influx in cardiomyocytes [111]. LQT14–17, caused by mutations in CALM1/2/3 and TRDN, are inherently accompanied by impaired inhibitory function of RyR2 and disrupted Ca²⁺ homeostasis. Oxidative stress mediated by NOX2/ROS further aggravates repolarization disorders and QT prolongation [89,95].

Arrhythmias in the three major clinical LQTS subtypes (LQT1, LQT2 and LQT3) are all initiated by EADs [90]. Along with inhibiting outward K⁺ currents such as I_Ks and I_Kr, impairing myocardial repolarization reserve and prolonging APD, ROS also increases I_Na,L by activating CaMKII [111,123]. The combined effects cause repeated opening of calcium channels and eventually trigger phase 2 EADs.

Short QT syndrome

NOX2-derived ROS differentially modify cardiac ion channel proteins, alter the function of distinct channel subtypes, and remodel cardiac repolarizing currents. These changes synergistically shorten APD and the QT interval. From the perspective of acquired oxidative stress, ROS amplify the proarrhythmic effects of congenital gene mutations and are involved in the electrophysiological disorders across all subtypes of short QT syndrome (SQT1–SQT8).

The regulation of voltage-gated potassium channels (Kv) by ROS exhibits subtype dependence. Under most pathological conditions, ROS downregulate outward repolarizing K⁺ currents including I_to and I_Kur via direct thiol oxidation or activation of PKC and tyrosine kinase/SRC pathways. However, the modulation of Kv11.1 (I_Kr) by ROS depends on concentration and the microenvironment (Table 2). Exogenous ROS shifts the voltage dependence of Kv11.1 inactivation and accelerates channel activation, thereby increasing I_Kr. SQT1 (KCNH2) is caused by gain-of-function mutations in K⁺ channels, which lead to pathological elevation of I_Kr. The resultant excess outward repolarizing currents cause accelerated myocardial repolarization and QT shortening [120].

In terms of Na⁺ channel regulation, SQT7 results from loss-of-function mutations in SCN5A, which reduce the basal Na⁺ current. NOX2/ROS-mediated transcriptional repression of SCN5A further attenuates Na⁺ influx, impairs phase 0 depolarization of action potentials and indirectly shortens the plateau phase, ultimately contributing to QT interval shortening [96,124,125].

Regarding calcium channel regulation, SQT4–SQT6 are attributed to loss-of-function mutations in CACNA1C, CACNB2 and CACNA2D1. These mutations reduce I_Ca,L, serving as the congenital genetic basis for QT shortening [89]. ROS generated by activated NOX2 under physiological and pathological conditions directly oxidizes cysteine residues on Cav1.2 or inhibits I_Ca,L through PKC and CaMKII signaling pathways, thus accelerating repolarization [108].

Brugada syndrome

Approximately 20% of patients with Brugada syndrome (BrS) carry loss-of-function mutations in SCN5A. The reduced channel expression and decreased I_Na constitute the primary mechanism underlying abnormal depolarization in this disorder [126–128]. The core pathological features of BrS include impaired myocardial conduction, increased transmural action potential dispersion, and myocardial fibrosis in the right ventricle [129,130]. NOX2/ROS further suppresses I_Na at both transcriptional and post-translational levels. Activated NOX2 produces hydrogen peroxide, which oxidatively activates NF-κB. The activated NF-κB binds to the SCN5A promoter, thereby inhibiting gene transcription, reducing Nav1.5 mRNA and protein expression, decreasing the number of functional Na⁺ channels, and lowering the total Na⁺ current [112]. Meanwhile, ROS directly oxidizes cysteine residues on the α-subunit of Nav1.5, alters channel gating properties, and reduces I_Na [131], which exacerbates local conduction delay in the right ventricular outflow tract (RVOT).

Combined dysfunction of K⁺ and Ca²⁺ channels contributes to repolarization disorders in BrS. The R99H mutation in KCNE3 (MiRP2) relieves its inhibitory effect on Kv4.3, and additional oxidative modification by NOX2/ROS leads to abnormal elevation of I_to [120]. Loss-of-function mutations in CACNA1C, CACNB2, and CACNA2D1 intrinsically reduce I_Ca,L [89,132]. ROS oxidizes cysteine residues on the α1C subunit of Cav1.2 to further inhibit I_Ca,L, resulting in insufficient inward current during the action potential plateau phase. The elevated I_to together with reduced I_Ca,L amplifies transmural repolarization dispersion in the right ventricle [133].

Catecholaminergic polymorphic ventricular tachycardia

Multiple pathological stimuli and metabolic disorders can activate the myocardial NOX2 signaling axis and trigger oxidative stress-induced Ca²⁺ dyshomeostasis, which lays a redox foundation for the development of catecholaminergic polymorphic ventricular tachycardia (CPVT). NOX2-derived ROS primarily disrupt SR Ca²⁺ cycling via oxidative modification of key cardiac Ca²⁺ regulatory proteins, leading to diastolic Ca²⁺ leak, the core pathological hallmark of CPVT. ROS directly mediate redox modification of RyR2, increasing its S-glutathionylation and S-nitrosylation levels. In addition, ROS promotes disulfide bond formation between C1078 and C2991 residues on RyR2 subunits [91], which converts RyR2 into a pathologically leaky Ca²⁺ channel [88,90,108]. Meanwhile, ROS bidirectionally modulate Ca²⁺ transporters through thiol oxidation: they inhibit the Ca²⁺ reuptake activity of SERCA2a and enhance the transport function of NCX. These effects rapidly deplete SR Ca²⁺ stores, elevate cytoplasmic Ca²⁺ concentrations and further exacerbate Ca²⁺ dyshomeostasis [87,92]. Furthermore, NOX2-derived ROS induce oxidative activation of CaMKII. Activated CaMKII phosphorylates RyR2 to aggravate SR Ca²⁺ leak, and also modulates Na⁺ channels to increase I_Na,L and prolong APD, thereby raising arrhythmogenic susceptibility [93,110].

Ca²⁺ dyshomeostasis mediated by NOX2/ROS ultimately triggers DADs and contributes to the characteristic ventricular arrhythmias in CPVT [95,96]. The classical pathogenesis of CPVT refers to SR Ca²⁺ overload and diastolic Ca²⁺ leak induced by sympathetic activation [89,90]. Sustained SR Ca²⁺ leak caused by excessive NOX2 activation leads to abnormally elevated local cytoplasmic Ca²⁺ levels, which switches NCX to its forward transport mode and generates I_Ti, consequently inducing DADs. With increased sympathetic tone and catecholamine release, phosphorylation mediated by PKA [134] and CaMKII further augments Ca²⁺ release via RyR2 and worsens Ca²⁺ overload. Once the amplitude of DADs reaches the depolarization threshold of cardiomyocytes, ectopic triggered activities occur [122,123]. (Table 2)

The pathology-dependent dual role of NOX2

Although the preceding sections have primarily focused on the pathogenic role of NOX2 in cardiac electrophysiological disorders, NOX2-derived ROS can also exert protective or beneficial effects under specific physiological and pathological conditions.

It has been demonstrated that Ang II pretreatment reduces myocardial infarct size following ischemia/reperfusion injury through NOX2 activation. This cardioprotective effect is mediated by NOX2-enhanced ROS production, which in turn triggers mitochondrial ROS release, activates JNK and p38 MAPK pathways, and inhibits lipid peroxidation [135]. Similarly, rapid cardiac pacing produces a preconditioning effect and increases NADPH oxidase activity in SR-enriched microsomal fractions. This NOX2 activation enhances S-glutathionylation of RyR2—a redox modification that maintains faster Ca²⁺ release rates during periods of increased cardiac activity. This mechanism represents an important component of tachycardia-induced myocardial preconditioning [88].

Under physiological conditions, myocardial stretch rapidly activates NOX2 at the sarcolemmal and t-tubular membranes via microtubule-dependent mechanotransduction [8,101]. Locally produced ROS sensitize adjacent RyR2 channels, triggering a burst of Ca²⁺ sparks that fine-tune excitation-contraction coupling. Recent studies using NOX2 knockout mice have demonstrated that stretch-induced ROS accelerate the rise phase of Ca²⁺ transients and enhance myocardial contractility, thereby contributing to the Frank-Starling mechanism – the heart’s intrinsic adaptive response to increased preload [9].

The most striking example of the context-dependent role of NOX2 comes from studies of Chagas disease. In a mouse model of acute Chagasic cardiomyopathy, genetic deletion of NOX2 (gp91phox knockout) unexpectedly promoted a proarrhythmic phenotype, characterized by an increased incidence of EADs, APD alternans, and elevated arrhythmia susceptibility [136].

It is worth noting that NOX2 also performs important protective functions in other physiological systems. For example, in phagocytes, the NOX2-mediated “respiratory burst” is a key innate immune mechanism for defending against microbial invasion [10,137]; loss of NOX2 function leads to chronic granulomatous disease, characterized by recurrent and severe infections [138]. However, these non-cardiac functions of NOX2 are beyond the scope of this review and will not be discussed in detail here.

Furthermore, the connection between NOX2/ROS and ferroptosis warrants attention. Ferroptosis is a novel form of iron-dependent, lipid peroxidation-driven programmed cell death, whose core mechanism is intimately linked to intracellular redox imbalance. Studies have demonstrated that in cancer stem cells, redox-sensitive transcription factors such as nuclear factor erythroid 2-related factor 2 (NRF2), BTB-and-CNC homologue 1 (BACH1), and hypoxia-inducible factor-1α (HIF-1α) maintain redox homeostasis by regulating antioxidant pathways and metabolic reprogramming, thereby influencing stemness maintenance, EMT, and therapeutic resistance [139]. These findings suggest that NOX2-derived ROS may not only contribute to arrhythmogenesis through classical pathways (Ca²⁺ homeostasis imbalance, ion channel remodeling, and structural remodeling) but also exert broader regulatory effects on cardiomyocyte survival and death decisions by influencing ferroptosis-related pathways. Future studies should explore the cross-regulation between NOX2/ROS and ferroptosis in cardiac electrophysiological disorders, which may provide novel insights for the development of antiarrhythmic therapeutic strategies.

Therapeutic potential and limitations of NOX2-targeted intervention

Inhibitors targeting the NOX2 cytosolic regulatory complex

Targeting p47phox-p22phox interaction

p47phox is a key cytosolic regulatory subunit of NOX2. Its phosphorylation and subsequent binding to the membrane subunit p22phox represent the rate-limiting steps in the NOX2 assembly process, making it the mainstream target for inhibitor design. Inhibitors targeting the interaction between p47phox and p22phox primarily fall into two categories: peptide-based and small-molecule inhibitors.

Gp91ds-tat is a peptide-based inhibitor that was designed in 2001 [140]. It is an 18-amino-acid chimeric peptide that specifically inhibits NOX2 [141] and does not affect the activity of NOX1, NOX4, or xanthine oxidase [142]. This synthetic chimeric peptide comprises nine amino acids derived from the HIV-tat sequence (for internalization) and nine amino acids from the intracellular loop B sequence of NOX2. It competitively inhibits the binding of p47phox to p22phox and is effective in reducing oxidative stress and tissue damage in animal models of Ang II-induced hypertension and traumatic brain injury [8,143]. However, limitations inherent to its peptide nature, such as poor cell penetration, susceptibility to degradation, and rapid clearance, severely hinder its clinical translation [144].

Among small-molecule inhibitors, ebselen (SPI-1005) is the most extensively studied compound, with the fastest progress in clinical translation. It has entered Phase II clinical trials for multiple diseases [145,146]. Ebselen targets the bis-SH3 domain of p47phox, interfering with its binding to p22phox, inhibiting the membrane translocation of p47phox [147], and downregulating the expression of NOX2 subunits [146]. Clinical trials have been conducted for conditions including stroke, hearing loss, and COVID-19, demonstrating substantial application potential [145,148–152].

Although LMH001 exhibits a high inhibitory activity [153], its stability and target mechanism have been questioned in independent studies [154]. Additionally, CPP11G/CPP11H (obtained through rational drug design), bivalent inhibitors (e.g. compound 33) [27,155], and the natural product celastrol can efficiently inhibit p47phox-p22phox interaction [10,156–162]. Novel small-molecule SH3 domain binders, such as the indole derivatives C6/C14 [163] and 2-aminoquinoline compounds [164], also show potential for specific inhibition of this target (Table 3).

Table 3.

Targets and cardiovascular effects of NOX2 inhibitors.

Inhibitor	Target/Mechanism	Cardiovascular Effect	Study Type	References
gp91ds-tat	Targets p47phox-p22phox interaction	Inhibits Ang II-induced aortic O2•− generation and alleviates hypertension; potently suppresses vascular O2•− production	In vivo	[140]
LMH001	Targets p47phox-p22phox interaction	Attenuates Ang II-induced endothelial NOX2 activation and O2•− production, ameliorates Ang II-associated hypertension, vascular inflammation, and reduces aortic aneurysm incidence	In vivo	[153]
Celastrol	Targets p47phox-p22phox interaction	Activates Nrf2 signaling pathway, inhibits NOX2 and AT1R, and protects against Ang II-induced injury in human umbilical vein endothelial cells	In vitro (human cells)	[214]
Celastrol	Targets p47phox-p22phox interaction	Suppresses NOX2-mediated ROS production and calcification in aortic valve interstitial cells, alleviates calcific aortic valve disease	In vitro + In vivo	[215]
AEBSF	Interferes with p47phox binding	Attenuates α1-adrenergic receptor-induced cardiac hypertrophy	In vitro	[216]
Phox-I1 analogues	Targets p67phox-Rac GTPase interaction	Inhibits platelet NOX2 activation, ROS generation and platelet aggregation, exerts anti-thrombotic effects without prolonging bleeding time	In vitro + In vivo	[167]
NSC23766	Targets Rac GTPase	Inhibits acute Ang II-induced NOX-derived O2•− production in endothelial cells	In vitro	[168]
Statins	Targets Rac GTPase	Reduces atrial NOX activity, myocardial oxidative stress and perioperative cardiac complications	Human studies + In vitro	[48]
Statins	Targets Rac GTPase	Rapidly suppresses platelet NOX2 activation, oxidative stress and thrombogenesis	Human + In vitro	[170,171]
Statins	Targets Rac GTPase	Alleviates pulmonary hypertension and pulmonary vascular remodeling	In vivo	[172]
Statins	Targets Rac GTPase	Attenuates myocardial injury and apoptosis induced by coronary microembolization	In vivo	[175]
Statins	Targets Rac GTPase	Improves left ventricular remodeling, myocardial fibrosis and cardiac function post-myocardial infarction	In vivo	[217,218]
GSK2795039	Specific NOX2 catalytic inhibitor	The first validated selective NOX2 inhibitor with consistent efficacy both in vitro and in vivo	In vitro + In vivo	[176]
GSK2795039	Specific NOX2 catalytic inhibitor	Ameliorates doxorubicin-induced cardiomyopathy and cardiomyocyte necroptosis	In vitro + In vivo	[181]
GSK2795039	Specific NOX2 catalytic inhibitor	Stabilizes vulnerable atherosclerotic plaques	In vitro + In vivo	[180]
GSK2795039	Specific NOX2 catalytic inhibitor	Improves post-infarct cardiac remodeling and mitochondrial function	In vitro + In vivo	[178]
GSK2795039	Specific NOX2 catalytic inhibitor	Reduces myocardial oxidative stress and cardiac hypertrophy	In vitro + In vivo	[177]
Apocynin	Non-selective NOX inhibitor	Mediates cardioprotection against Ang II-induced myocardial ischemia-reperfusion injury	In vitro + In vivo	[135]
Apocynin	Non-selective NOX inhibitor	Attenuates Ang II-induced endothelial mitochondrial ROS and eNOS uncoupling	In vitro	[219]
Apocynin	Non-selective NOX inhibitor	Inhibits Ang II-triggered arrhythmogenic signaling pathways	In vitro	[111]
Apocynin	Non-selective NOX inhibitor	Regulates tachycardia-induced myocardial preconditioning	In vitro	[220]
Apocynin	Non-selective NOX inhibitor	Reduces oxidative stress, cardiomyocyte apoptosis and improves cardiac function in post-infarct heart failure	In vivo	[221]
Apocynin	Non-selective NOX inhibitor	Improves Ca2+ handling and myocardial contractility in dystrophic cardiomyopathy	In vitro	[99]
Apocynin	Non-selective NOX inhibitor	Attenuates diet-induced ventricular arrhythmias	In vivo	[94]
Apocynin	Non-selective NOX inhibitor	Ameliorates ethanol-induced hypertension and endothelial diastolic dysfunction	In vivo	[222]
DPI	Non-selective NOX inhibitor	Blocks Ang II-induced cardiac hypertrophy	In vitro + In vivo	[223]
DPI	Non-selective NOX inhibitor	Alleviates obesity-associated myocardial hypertrophy	In vivo	[201]
DPI	Non-selective NOX inhibitor	Inhibits PA-induced ROS generation and improves endothelial vasodilation	In vitro	[224]
VAS2870	Non-selective NOX inhibitor	Reduces vascular ROS in hypertensive animals and restores endothelial diastolic function	In vivo	[225]
VAS2870	Non-selective NOX inhibitor	Reverses myocardial NOS uncoupling and improves Ca2+ handling in dystrophic myocardium	In vitro	[99]

Abbreviations: Ang II, angiotensin II; O₂•⁻, Superoxide; Nrf2, nuclear factor erythroid 2-related factor 2; AT₁R, Angiotensin II type 1 receptor; eNOS, endothelial nitric oxide synthase; PA, palmitic acid; NOS, nitric oxide synthase.

Targeting p67phox

Peptides derived from p67phox, corresponding to overlapping segments of the 259–279 region of p67phox, self-assemble into aggregates. These aggregates specifically bind and sequester p67phox, thereby inhibiting NOX2 activity. This mechanism mimics the potential intramolecular auto-inhibition of p67phox and provides a novel strategy for targeting the interaction interface [165].

Phox-I1 class inhibitors target the interaction between p67phox and Rac GTPases, thereby effectively inhibiting NOX2 activity. However, these inhibitors do not affect bleeding time in animal models, demonstrating their potential as antithrombotic agents [166,167].

Targeting Rac GTPase

Rac GTPases are key upstream regulators of NOX2 activation. In 2004, Gao et al. [168] identified NSC23766 as a small-molecule that specifically binds to the guanine nucleotide exchange factor (GEF)-binding pocket of Rac1. It selectively inhibits the activation of Rac by GEFs such as Trio or Tiam1, making it a crucial tool for studying the Rac-NOX2 pathway and related diseases.

Statins (e.g. atorvastatin and rosuvastatin) inhibit the isoprenylation of Rac1, blocking its recruitment to the NOX2 complex, thereby exerting rapid antioxidant effects. Preoperative statin therapy significantly reduces the atrial NADPH oxidase activity and the oxidative stress in patients. Additionally, through a Rac1-mediated mechanism, it inhibits the production of myocardial O2•⁻ and ONOO⁻, effectively lowering the risk of complications such as post-cardiac surgery AF [48]. These results indicate that statins exert cardioprotective effects partly by inhibiting Rac1-NOX2, providing a mechanistic explanation for their clinical pleiotropic effects [169–175].

Direct catalytic inhibition and pan-NOX inhibitors

Pan-NOX inhibitors act directly on the catalytic active site of the enzyme or simultaneously affect multiple NOX isoforms. Regarding specific NOX2 catalytic inhibitors, GSK2795039 is the first small‑molecule NOX2‑specific inhibitor validated to be effective in vivo, which acts by competitively binding to NADPH. It significantly inhibits NOX2 activity at both semi‑recombinant and cellular levels, reducing ROS production [176–181]. This inhibitor exerts cardioprotective effects with underlying mechanisms in various cardiovascular disease models. In myocardial infarction models, GSK2795039 attenuates oxidative stress and mitochondrial dysfunction, reduces cardiomyocyte apoptosis and infarct size, and improves cardiac remodeling and function by inhibiting NOX2 [178]. In chemotherapy‑induced cardiotoxicity models, GSK2795039 alleviates myocardial injury by suppressing NOX2‑mediated oxidative stress [179,181]. In vulnerable atherosclerotic plaque models, GSK2795039 stabilizes plaque structure by inhibiting the NOX2‑ROS‑MerTK axis [180]. Furthermore, GSK2795039 can mimic the genetic knockout of Nox2 in vivo, exhibiting high selectivity for other NOX isoforms and related oxidases [177]. Regarding nonspecific or pan-NOX inhibitors, apocynin is a nonspecific NOX inhibitor [182]. However, as a prodrug, it requires activation by myeloperoxidase (MPO) to inhibit the membrane translocation of p47phox. It also has limitations such as low inhibitory potency (IC₅₀ ~10 μM) and lack of specificity, rendering it an unsuitable tool compound or drug candidate [183,184]. However, Apocynin possesses inherent antioxidant activity as an ROS scavenger [185,186]. In an ibrutinib-induced AF model, apocynin can reduce AF susceptibility by inhibiting NOX2/4 expression, suggesting its potential for AF intervention [64].

Diphenyleneiodonium (DPI) is an often-used laboratory compound, and acts at low micromolar or nanomolar doses. It is a broad-spectrum flavoprotein blocker that can extensively inhibit various flavin-containing enzymes (e.g. XO and eNOS) with extremely poor specificity [187,188]. VAS2870/VAS3947 belong to the triazolopyrimidine class of pan-NOX inhibitors. Both inhibit NOX through covalent alkylation of the conserved active-site cysteine residues, thereby affecting the first step of the reaction (NADPH binding and FAD reduction) and blocking NOX at the start of the catalytic cycle. They also lack isoform specificity [10,35,189].

Indirect regulation: Inhibitors targeting auxiliary proteins

The phospholipase A2 (PLA2) activity of Prdx6 is crucial for NOX2 activation. Its inhibitor MJ33 and the PLA2 inhibitory peptide (PIP) that mimics pulmonary surfactant protein A can effectively inhibit NOX2 activation, providing a new target for the treatment of diseases such as acute lung injury [190–192].

Other inhibitors

4-(2-Aminoethyl) benzenesulfonyl fluoride (AEBSF) is a serine protease inhibitor that inhibits NOX enzyme activity by interfering with the binding of the cytosolic subunit p47phox [193]. Whether AEBSF directly inhibits NOX or exerts its effect by acting on signaling pathways related to p47phox activation remains unclear.

Myricetin reduces ROS production by inhibiting gp91phox/p47phox assembly, thereby blocking the activation of the JAK-STAT1 signaling pathway and significantly alleviating lipopolysaccharide (LPS)-induced acute lung injury [194].

Natural polyphenols such as curcumin and resveratrol have also shown potential for NOX2 inhibition. Curcumin and its derivatives (DMC and BDMC) effectively inhibit the matrix invasion ability of monocytes in a monocyte – macrophage differentiation model [195]. This effect is achieved by inhibiting protein kinase C delta (PKCδ) activity, downregulating NOX2 expression and p47phox membrane translocation, and blocking the NADPH oxidase/ROS signaling pathway, thereby demonstrating potential for anti-atherosclerotic effects.

Resveratrol can significantly reduce ROS production [196]. It also increases expression of inflammatory factors (iNOS, COX-2, NO, PGE2) induced by carbon black nanoparticles (CBNPs) in lung epithelial cells by inhibiting PKC-α and Nox2-related signaling pathways. This demonstrates its ability to combat oxidative stress and inflammation through a multitarget mechanism.

The novel small-molecule compound TG15-132 [85] significantly inhibits NOX2 activity and the expression of downstream inflammatory factors in cell models. It also exhibits good brain permeability (brain/plasma concentration ratio > 5) and is safe in animals, indicating its potential as a neuroprotective agent.

Conclusion

In summary, NOX2 serves as a nodal point linking multiple pro-arrhythmic risk factors, and contributes to AF and VAs by causing cardiac electrophysiological disorders. Targeted NOX2 therapy faces certain limitations. Its efficacy is stage-dependent; statins prevent postoperative AF by inhibiting the Rac1–NOX2 axis [48], but are ineffective in late-stage AF [61]. Although NOX2 overexpression increases AF vulnerability, studies have reported that it only acts as a synergistic pathogenic factor [197]. Complete NOX2 inhibition carries the risk of inducing immunodeficiency, whereas its moderate inhibition may balance efficacy and safety to a certain extent in chronic granulomatous disease [198]. Its function is also pathologically context-dependent, exerting cardioprotective effects in Chagas disease [136]. Nevertheless, clinical evidence that antioxidant strategies can reduce the risk of postoperative AF and potentially decrease recurrence after cardioversion provides a valuable basis for antioxidant-targeted therapeutic approaches [199]. Therefore, future efforts should focus on developing highly selective NOX2 inhibitors and exploring therapeutic strategies based on disease stages and pathological contexts.

Funding Statement

This work was supported by the National Natural Science Foundation of China under Grant [82370327]; and the National Key Research and Development Program of China under Grant [2022YFA1104300].

Disclosure statement

No potential conflict of interest was reported by the author(s).

During the preparation of this work the authors used Doubao to assist with language polishing, English expression optimization, and manuscript revision. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.

Data availability statement

Data sharing is not applicable to this article as no data were created or analyzed in this study.