Vascular NADPH Oxidases and Atherothrombotic Stroke

Abstract

Oxidative stress constitutes a main molecular mechanism underlying cardiovascular diseases (CVDs). This pathological mechanism can be triggered by NADPH oxidases (NOXs), which produce reactive oxygen species (ROS). In fact, the different NOXs have been associated with myocardial infarction, atherothrombosis, and stroke. More specifically, we will focus on the implications of NOXs in atherothrombotic stroke. Each NOX member participates in a different way in the several stages of this disease: endothelial dysfunction, immune cell infiltration, foam cell genesis, vascular smooth muscle cells (VSMC) proliferation, and atherosclerotic plaque formation. Additionally, some NOXs are involved in plaque instability, thrombosis, ischemic stroke, and ischemia-reperfusion injury (IRI). Interestingly, the effects of NOXs in this pathology depend on the specific homolog, the cell type in which they are activated, and the stage of the disease. In this review we summarize the most up-to-date information about the implications of vascular NOXs in each of these processes. Finally, we highlight some limitations and future perspectives on the study of NOXs in CVDs.

1. Introduction

Atherothrombotic stroke is a cardiovascular disease (CVD) with a high global incidence. It occurs when a thrombus forms after the rupture of an atherosclerotic plaque, and it travels across the bloodstream, finally obstructing blood flow in the brain parenchyma [1]. This disease progresses through several phases, beginning with the development of an atherosclerotic plaque. Initially, endothelial cells are damaged by humoral stimuli, leading to inflammation and the transfer of low-density lipoproteins (LDL) into the vessel intima [2]. Reactive oxygen species (ROSs) produced from different cell types oxidize these LDLs (oxLDLs). Immune cells attracted by the damaged endothelium infiltrate and accumulate oxLDLs, becoming foam cells [3]. These cells accumulate and secrete growth factors, causing the invasion of the subendothelial space by vascular smooth muscle cells (VSMC). VSMC proliferate and switch their phenotype (sometimes becoming foam cells), contributing to plaque development. This plaque can either occlude the artery or rupture, leading to thrombus formation, which could potentially cause an ischemic stroke [4].

One of the most relevant molecular mechanisms underlying atherothrombotic stroke is oxidative stress, characterized by an imbalance between ROS levels and antioxidant defenses in the cells. In this context, ROSs overreact, modifying lipids, proteins, and DNA and causing cell damage, which may trigger different pathologies [5]. ROS are mainly produced by the mitochondrial electron chain, the NADPH oxidase family (NOX), nitric oxide synthases, xanthine oxidase, and cytochrome P450 [6]. Interestingly, the NOX family is the primary ROS source in the vasculature, and its overactivation is a key factor in several CVDs, such as atherothrombotic stroke [7].

2. NADPH Oxidases in the Vessel Wall

NOXs are a family of enzymes that produce ROS as their main catalytic product. This family is composed of seven different members: five homologs (NOX1–NOX5) and two dual oxidases (DUOX1, DUOX2). These members differ in their regulatory subunits, their expression across different cell types, and their intracellular localization [8]. Specifically, NOX1, NOX2, NOX4, and NOX5 are expressed in the vascular wall and monocytes/macrophages, while DUOXs present lower levels [9]. NOX3 is mainly restricted to the inner ear [10]. In this section, we summarize the distribution of the main NOX members in endothelial cells, VSMC, and monocytes/macrophages (Figure 1).

NOX1 was described in 2000 in various mammalian cell lines [11]. Its activation requires regulatory subunits (p47phox, NOXA1), and its main product is O₂^•−. NOX1 is expressed by endothelial cells [12] and VSMC [13] in the vascular wall and by monocytes [14]. Intracellularly, NOX1 is detected in the nucleus, plasma membrane, endosomes, and peroxisomes [15]. In VSMC, NOX1 is expressed in caveolin-coated areas of the plasma membrane [16]. Little is known about its localization in endothelial cells and monocytes. NOX1 upregulation has been associated with CVDs such as atherosclerosis, hypertension, and diabetes [17,18,19].

NOX2 (gp91phox or phagocytic NOX) was the first described member, and it produces O₂^•−. It is a highly glycosylated protein, and its activation depends on phosphorylation and the recruitment of p22phox, p47phox, and p67phox subunits. NOX2 is expressed in endothelial cells, VSMC, and monocytes in different subcellular localizations [15]. For instance, in macrophages, NOX2 is located at intracellular membranes at baseline, translocating to the plasma membrane upon activation [20]. In endothelial cells, NOX2 is associated with the actin cytoskeleton and cellular protrusions [21]. By contrast, NOX2 localization in VSMC remains unknown. NOX2 has been associated with diabetes, myocardial infarction, oxLDL accumulation, and thrombosis [22,23].

NOX4 (initially named “Renox”) was first described in the kidney [24]. NOX4 only requires the p22phox subunit to produce ROS and appears to be active under physiological conditions [25]. The primary effect of this oxidase is mediated by H₂O₂. NOX4 is expressed in endothelial cells [26], VSMC [16], and monocytes [27]. In endothelial cells, it is located in the endoplasmic reticulum [21]; in VSMCs, in focal adhesions, mitochondria, and the nucleus [16,28,29]; and in macrophages, in the endoplasmic reticulum [27]. There is some debate about the role of NOX4 in CVDs, as it has been described to have both protective and damaging roles [30].

NOX5 was the last discovered member of the family. It is highly expressed in the testis, spleen, and lymph nodes [31]. It is absent in the rodent genome, which complicates its study in the experimental rodent models. Interestingly, NOX5 is the only member regulated by Ca⁺⁺ levels. In the human vasculature, NOX5 is present at all vascular segments [32]: endothelial cells [33], VSMC [34], and monocytes [35]. It is located in the nucleus, endoplasmic reticulum, and plasma membrane [36]. In VSMC, it is abundant in cholesterol-rich areas of the plasma membrane and translocates to rafts upon activation [37]. In dendritic cell-derived monocytes, NOX5 is expressed on the outer membrane of the mitochondria [38]. Recently, NOX5 has been found to interact with the actin cytoskeleton [39]. NOX5 is implicated in several CVDs, such as diabetes, atherosclerosis, and myocardial infarction [40].

3. NADPH Oxidases in Atherosclerosis

3.1. NADPH Oxidases, Endothelial Dysfunction and Inflammation

Interestingly, decades ago, endothelial dysfunction was established as an early predictor of adverse outcomes in heart failure [41]. Moreover, endothelial dysfunction and oxidative stress levels have been related to cardiovascular events in patients suffering from coronary artery diseases [42]. Exercising and the Mediterranean diet are two different interventional changes that reduce endothelial dysfunction and, as widely known, the risk of CVDs [43,44].

NOX members are activated by humoral factors associated with CVDs, leading to endothelial dysfunction. For example, Ang II stimulates NOX2 [45,46], NOX4 [47], and NOX5 [48,49] activity/expression in human endothelial cells. ET-1 activates NOX2 in pigs [50] and NOX5-derived ROS in human endothelial cells [48], contributing to endothelial dysfunction. Moreover, oxLDLs induce NOX2 activity and endothelial dysfunction in human primary coronary artery endothelial cells [51]. The role of NOX4 in endothelial dysfunction in response to LDL/oxLDL stimulation is debated; some studies describe a pathological role in human umbilical vein endothelial cells (HUVEC) [52,53], while others suggest a protective role in LDL-receptor knock-out mice [54]. Finally, lysophosphatidylcholine activates NOX5 through increased Ca⁺⁺ levels in human aortic endothelial cells (HAEC) [55].

In addition, NOXs are involved in endothelial cell apoptosis. For instance, NOX1 regulates apoptosis in human sinusoidal endothelial cells [12]. The localization of NOX2 and NOX4 appears closely linked to apoptosis in HUVEC. Specifically, caspase 3 activity associates with NOX2- and NOX4-derived ROS production in response to homocysteine [56]. Finally, our group previously showed that NOX5 induces apoptosis in HAECs through the unfolded protein response [57,58].

These proapoptotic effects of NOXs create a proinflammatory environment by releasing cytokines from endothelial cells. For instance, NOX2 in endothelial cells mediates in vivo production of granulocyte-monocyte colony-stimulating factor (GM-CSF), likely through toll-like receptor (TLR) stimulation [59]. Likewise, flagellin activates NOX4 by TLR5 receptor, increasing ICAM-1 expression and IL-8 secretion in HAEC. In fact, NOX4 knock-down reduced the ICAM-1 expression, IL-8 secretion, and transendothelial migration of immune cells across the endothelium [60]. Similar effects have been observed in LPS-induced HAEC, as NOX4 increased ICAM-1 and VCAM-1 expression in the absence of the Cyb5r3 cytochrome [61]. Finally, NOX4 is involved in ferroptosis-induced inflammation in HUVECs by modulating the Nrf-2 transcription factor [62]. Moreover, endothelial NOX5 in diabetic Akita mice increased inflammation via MCP-1 and TLR4 [63]. Tobacco smoke extract induced CXCL16 production via NOX5 activation in human umbilical arterial endothelial cells (HUAEC), which regulates endothelial dysfunction and immune cell adhesion [64]. Prostaglandins (PGs) are also implicated in this process. For instance, in liver sinusoidal endothelial cells, COX-2-derived PGE2 acts over NOX1 and NOX4, regulating the balance between ROS and NO production, which may lead to endothelial dysfunction [65]. Finally, in immortalized HAEC (teloHAEC), NOX5 increases PGE2 production by COX-2 activation [49]. All the studies mentioned are summarized in Figure 2.

As discussed above, endothelial-derived inflammation plays a key role during atherosclerosis development. However, the inflammatory mediators that aggravate the development of atherosclerosis plaques present different origins. Among them, chronic inflammatory diseases could be highlighted. For instance, patients with rheumatoid arthritis present an increased risk of subclinical atherosclerosis related to the activation of different proinflammatory pathways [66]. The oxidative systems seem to directly participate in this mechanism. ROS production in response to galectin-3 promotes inflammation and apoptosis of human chondrocytes [67]. In addition, particulate matter activates NOX2 and enhances inflammation in synovial fibroblasts of rheumatoid arthritis [68]. Interestingly, microbiota dysfunction leads to an increase of pathological bacteria, which in turn increases ROS production and is directly related to diabetes-induced increased risk of atherosclerosis. The LPS/TLR/NOX axis is the main molecular mechanism involved in this process [69]. This pathway can also lead to amyloidosis, which can turn into systemic inflammation through NOX2 activity [70] or mediate pancreatic β-cells apoptosis through NOX1 activity [71], in both cases increasing the risk of atherosclerosis. Therefore, NOXs promote different sources of inflammation that can accelerate or increase the risk of this pathology.

3.2. NADPH Oxidases, Immune Cell Infiltration and Foam Cells

In this proinflammatory context, endothelial cells upregulate adhesion molecules to facilitate immune cell adhesion and infiltration. NOX activity and expression closely correlate with adhesion molecule expression. NOX2 overexpression in murine endothelial cells increases leukocyte adhesion and VCAM-1 expression. In mice with cardiac hypertrophy produced by Ang II infusion, NOX2 enhances immune cell infiltration into the heart [72]. Similarly, salusin-β (a prohypertensive peptide) increases monocytic adhesion via NF-κB in VSMC through NOX2 [73]. Again, NOX4 overexpression in endothelial cells exhibits a protective role, reducing the recruitment of immune cells caused by Ang II in vivo [74]. However, NOX4 also promotes the expression of ICAM-1 and VCAM-1 in response to TNF-α in HUVEC [75]. Finally, NOX5 increases VCAM-1 and ICAM-1 expression and mononuclear cell infiltration in HUVEC [76]. This pathway could be induced by oxLDLs [55]. In a myocardial infarction model, mice overexpressing NOX5 in endothelial cells upregulate cardiac VCAM-1 levels [77].

Changes in the extracellular matrix (ECM) can facilitate immune cell infiltration into the subendothelial space of the vessel wall. Interestingly, p22phox regulates vascular wall elastic fiber composition in vivo, correlating with the MMP-12/TIMP-1 ratio [78]. Besides, these alterations in the ECM allow VSMCs to proliferate and migrate into the atherosclerotic lesion. In this case, NOX1 and NOX2 seem to be involved in this process. These NOX homologs respond to asprosin, a newly discovered human hormone, increasing VSMC proliferation and migration [79]. Ang II-induced NOX1 activation promotes VSMC migration and proliferation via MMP-9 [80]. NOX2 activation by Ang II induces vascular remodeling in VSMC and adventitial fibroblasts [81,82]. Additionally, NOX2 is implicated in the development of atherosclerotic plaque in vivo, which seems to be related to a reduction in the MMP-9 activity [83]. Finally, NOX4 influences the remodeling of vascular adventitial fibroblasts in a diabetic model of rats [84]. The relationship between NOX5 and MMP expression/activity in the vascular context remains unclear.

NOXs also play a crucial role in foam cell formation by mediating lipid oxidation. NOX1 may play a dual role in this process. On the one hand, NOX1 activation by TLRs increases ROS production and foam cell generation [85]. On the other hand, NOX1 participates in atherosclerotic plaque propagation by mediating LDL pinocytosis [86]. NOX2 has been implicated in macrophage apoptosis during efferocytosis, contributing to increased inflammation and immune cell infiltration into the plaque [87]. Furthermore, NOX2 promotes lipid accumulation in VSMC, another source of foam cells [73]. There is no literature suggesting a direct relationship between NOX4 or NOX5 and foam cells, although NOX5 has been associated with VSMC migration in hypertensive patients [88]. All these studies are summarized in Figure 3.

3.3. NADPH Oxidases, Plaque Development, Plaque Rupture, and Thrombosis

Foam cells stimulate neighboring VSMC, promoting their migration and proliferation within the atherosclerotic plaque, increasing its size. NOX1A mediates VSMC migration, proliferation, and differentiation into macrophage-like cells in ApoE^−/− mice [89]. These effects of NOX1 on VSMC have been widely described in vitro and in vivo [90,91,92]. Interestingly, NOX2 activity follows NOX1 in VSMC, promoting cell migration through a second peak of ROS [81]. Although the effect of NOX2 on VSMC migration was already described [73], this study highlights a temporal relationship between two NOXs in a specific cell type.

NOX2 expressed in endothelial cells contributes to vascular remodeling by increasing VSMC proliferation, indicating a paracrine effect derived from NOX activity [93]. The precise role of NOX4 in VSMC migration and differentiation remains unclear, although some studies suggest an anti-synthetic and anti-proliferative role [94]. Some authors indicate that this effect is the result of a coordinate balance between NOX1 and NOX4 activities in VSMCs [94,95]. Although limited information exists on NOX5 and VSMC, great progress has been made recently. VSMC extracted from hypertensive patients presents greater migratory capacity mediated by NOX5 activation [96]. Regarding vascular calcification, NOX5 promotes the switch of VSMC to a synthetic phenotype that secretes Ca⁺⁺-loaded vesicles [97]. This effect of NOX5 in VSMC-mediated vascular calcification may appear in humans in response to smoking [34].

Several in vivo models of atherosclerosis served to evaluate the effects of NOXs on plaque development. NOX1 produces pro-atherosclerotic effects in diabetic mice, as its deletion decreases chemokines secretion, immune infiltration, and profibrotic markers expression [98]. Interestingly, ET-1 production from endothelial cells is linked to NOX1 pro-atherosclerotic effects in vivo [99]. Dietary supplementation with blackberry prevented NOX1 effects in vivo [100]. NOX2 plays a pro-atherosclerotic role. It was demonstrated that air pollutants aggravated atherosclerosis in vivo by a NOX2-derived mechanism, which could be inhibited by the antioxidant capacity of melatonin [101]. NOX2 inhibition by gp91ds-tat causes the regression of the atherosclerotic plaques in ApoE^−/− mice [83]. NOX4 deletion in ApoE^−/− mice accelerates atherosclerosis in response to partial carotid artery ligation under a high-fat diet, suggesting a protective role [102]. Conversely, in LDLr^−/− mice, NOX4 exhibits a pathological profile in response to a high-fat diet, increasing plaque burden [54]. NOX4 deletion in VSMC prevents plaque development in response to Ang II [103]. In coronary artery disease patients, NOX5 appears overexpressed in atherosclerotic plaques, localizing in endothelial cells in early lesions and in VSMC in advanced lesions [32]. Furthermore, NOX5 colocalizes with immune cells in plaques from patients with atherosclerosis [104]. Regarding experimental atherosclerosis, endothelial overexpression of human NOX5 in knock-in mice does not affect plaque development [105]. However, NOX5 deletion in rabbits aggravates atherosclerosis in response to a high-fat diet [106]. Therefore, it is unclear if NOX5 promotes pathology or acts as a protective response to vascular damage.

NOXs can also modulate the stability of atherosclerotic plaques. NOX1 activation in VSMC promotes the development of unstable plaques [107]. In a model combining carotid branch ligation, renal artery constriction, and a high-fat diet in ApoE^−/− mice, NOX2 inhibition favors the production of more stable plaques [108]. NOX4 expression in patients with carotid artery stenosis directly correlates with plaque stability and inversely correlates with caspase-3 activity, suggesting a protective role [109]. The role of NOX5 in plaque stability remains unknown.

The instability of atherosclerotic plaques leads to rupture and thrombosis. In this stage of the disease, NOXs play a crucial role. Pharmacological inhibition or genetic depletion of NOX1 reduces collagen-dependent thrombosis in a FeCl₃ carotid occlusion model [110]. However, studies where NOX1 is specifically depleted from platelets suggest that NOX1 is involved in thrombin-induced thrombosis. By contrast, NOX2 seems to be involved in collagen-induced thrombosis [111]. On the contrary, another group describes that NOX2 is dispensable for arterial thrombosis in large vessels [112]. Finally, in a carotid occlusion model comparing a triple knock-out mouse (NOX1^−/−, NOX2^−/− and NOX4^−/−) to NOX-specific knock-out mice, it was concluded that NOX4 does not play a relevant role in pulmonary thromboembolism and ex vivo platelet aggregation, while NOX1 and NOX2 act as prothrombotic enzymes [113]. The role that NOX5 plays in thrombosis remains unclear. As can be observed, there is still controversy regarding the role of NOXs in thrombosis. All the described studies are summarized in Figure 4.

4. NADPH Oxidases in Thrombosis and Stroke

4.1. NADPH Oxidases, BBB Disruption and Stroke

Once the thrombus occludes a cerebral artery, an ischemic stroke occurs. In this context, NOX1 seems to alter BBB permeability. Oxygen/glucose deprivation increases ROS production by NOX1 in human brain endothelial cells (hBMEC), reducing the expression of adherent proteins and promoting permeability [114]. GKKT136901, a NOX1/4 inhibitor, prevents the increased permeability of hBMEC caused by methamphetamine, restoring ZO-1 and VE-cadherin expression [115]. In a model of traumatic brain injury, NOX1 is upregulated in endothelial cells from the neurovascular unit, overproducing O₂^•– and increasing permeability by the dysregulation of the TJs [116]. However, the relevance of NOX1 in MCAO models seems to be limited [117,118].

NOX2 presents similar effects on BBB permeability and stroke. At oxygen and glucose deprivation conditions, NOX2-derived ROS promote hBMEC permeability [119]. Likewise, methamphetamine induces NOX2 activation, dysregulating ZO-1, occludin, and claudin 5 and increasing BBB permeability in rat microvascular endothelial cells in vivo [120]. In an ischemic stroke model, NOX2 expression increased in the brain twelve hours, one day, and two days after the reperfusion. Interestingly, NOX2 knock-out mice subjected to ischemic stroke present lower infarct volume and edema and exhibit improved neurological outcomes compared to control littermates [121]. Nevertheless, other studies suggest a limited role for NOX2 in these models [117].

Little information is available about the role of endothelial NOX4 in BBB permeability. NOX4 expression increases after stroke in human patients and experimental models. NOX4 knock-out mice subjected to MCAO are prevented from BBB leakage and neuronal apoptosis. Similar effects are observed when using VAS2870, a specific NOX4 inhibitor [117]. Finally, endothelial NOX4 knock-out mice exhibit smaller infarct sizes after MCAO [122].

Finally, the role of NOX5 in BBB disruption and stroke remains poorly understood. Our group showed that in a model of aging performed in endothelial NOX5 knock-in mice, ZO-1 and occludin expression decreased, which was accompanied by memory loss [123].

4.2. NADPH Oxidases and Immune Infiltration in the Brain

After the onset of ischemia, immune cells infiltrate the brain in coordinated temporal waves. Initially, infiltrative immune cells present a proinflammatory phenotype, turning into anti-inflammatory resolving cells at later stages of the disease [124]. Scarce information is available about NOX’s effects at this point. NOX2 and NOX4 have been described as BBB-integrity disruptors during autoimmune encephalomyelitis, regulating the production of anti-inflammatory and proinflammatory mediators [125]. Additionally, NOX2 is related to the developmental process of microglia in the cerebral cortex of mice. More specifically, NOX2 promotes the infiltration of macrophages in the developing tissue, indicating a role in peripheral immune cell infiltration into the brain [126]. In a model of chronic restrain stress, several NOX subunits and proinflammatory markers appear upregulated in the cerebrovascular endothelium, indicating a relationship between NOXs and cerebral inflammation during adult life [127].

In CVDs, the findings are similar. In a subarachnoid hemorrhage model, splenectomy reduces the infiltration of inflammatory cells and NOX2 expression in cardiac and cerebral tissues [128]. NOX activity seems to mediate ATP-derived NETosis after MCAO, an activity produced by infiltrating neutrophils [129]. In addition, NOX2 negatively regulates TIPE-2, a protein that inhibits the infiltration of peripheral immune cells [130]. Although scarce, the works carried out to date suggest that NOXs play a relevant role in the infiltration of immune cells after ischemic stroke.

4.3. NADPH Oxidases and Ischemia-Reperfusion Injury

NOXs have been deeply studied in the ischemia-reperfusion injury (IRI), damage associated with an excessive production/accumulation of ROS after the recanalization of the infarcted area. Some authors suggest that NOX1, NOX2, and NOX4 play a relevant role in this process since their expression increases after reperfusion in a nylon-induced MCAO model. In this study, PI3Kγ^−/− mice reduce the upregulation of NOXs, which resulted in reduced neutrophil infiltration and a reduction in MMP-9 expression and brain damage [131]. Nonetheless, in a similar model of IRI in rats, NOX2 and NOX4 expression, but not NOX1, were upregulated at protein levels in the ischemic tissue [132].

In any case, ROS produced during IRI plays a dual role. In a mouse model of MCAO and reperfusion, NOX inhibition by apocynin after damage displayed different effects depending on the timing. After the resolution of stroke, apocynin reduces inflammation and promotes angiogenesis. However, one week and two weeks after the event, it increases inflammation and reduces angiogenesis [133].

NOX1 mediates ROS production after oxygen/glucose deprivation and reoxygenation in murine brain endothelial cells [114]. In male rats subjected to MCAO followed by reperfusion, NOX1 expression increases in neurons of the peri-infarcted region. Interestingly, NOX1 inhibition increases newborn cell survival in this region, which produces an improvement in functional recovery [134].

In a similar model, NOX2 knock-out mice present increased revascularization of the infarcted area after three days [135]. Then, miR-652 reduced NOX2 expression, ROS production, and tissue injury one day after reperfusion in the brains of rats subjected to MCAO [136]. NOD2 seems to activate NOX2 during IRI in vivo, increasing the secretion of proinflammatory mediators in the initial days after reperfusion [137]. Moreover, NOX4 targeting by miR-454 reduced ROS production in neuron-like SH-SY5Y cells, exhibiting a protective effect. Likewise, this miRNA reduced infarct size, edema, and cell death in the brains of rats one day after reperfusion [138]. In addition, NOX4 has been associated with TLR4 activation one day after IRI. TLR4 inhibition reduced NOX4 expression, oxidative stress, and neuronal apoptosis in vivo [139]. Interestingly, several drugs that exert neuroprotective effects in IRI in rodents seem to act by downregulating NOX4 [140,141,142]. Additionally, apocynin and NADPH treatment protected the brain tissue from inflammation and injury in an IRI model, a protection mediated by NOX2 and NOX4 downregulation [143]. NOX2 and NOX4 overexpression seem to be acute responders to IRI, as their levels increase three hours after reperfusion, returning to baseline after one day [144].

There is little information available about NOX5 implications in IRI. Nonetheless, Casas et al. recently described a key role for this oxidase. NOX5 produces ROSs in the first hour after reoxygenation, while NOX4 acts later, with a peak of activity after four hours. Interestingly, hypoxia increased NOX5 expression in hBMEC. In addition, in an immune and endothelial cell knock-in model for NOX5, BBB leakage and infarct volume were increased, and the neurological outcomes worsened after ischemia-reperfusion [145].

5. Conclusions

The atherothrombotic stroke is strongly influenced by NOX activity in each stage. NOXs promote endothelial dysfunction by two main mechanisms: apoptosis of endothelial cells and the secretion of proinflammatory mediators such as cytokines and PGs. These mechanisms seem to be triggered by LPS/TLR signaling, which activates the pathological role of NOX1, NOX2, NOX4, and NOX5. These homologs also increase the expression of adhesion molecules such as ICAM-1 and VCAM-1 in endothelial cells, easing immune cell infiltration. In fact, TNF-α and NF-κB are the main signaling mediators that induce this process. NOXs are also involved in VSMC migration and proliferation, which may promote atherosclerotic plaque progression. The changes in the composition of the ECM produced by NOXs may also ease the invasion of the subintima space by immune cells and VSMCs. In fact, NOX1, NOX2, and NOX4 have been demonstrated to promote plaque progression in vivo, while NOX5 seems to have the opposite effect.

In the ischemic stroke stage of the disease, NOX1, NOX2, and NOX5 might promote BBB impairment by endothelial cell apoptosis and TJ dysregulation. On the contrary, NOX4 seems to play a protective effect in this stage. Finally, in the IRI produced after stroke resolution, NOX2, NOX4, and NOX5 increase the infarct size, potentially by a burst in ROS production.

6. Future Perspectives

Although NOXs have been widely studied, there is still a great lack of information that can be focused from different approaches.

(i)

More cell type-specific knock-out/knock-in in vivo models would improve the current knowledge. Most of the reviewed studies demonstrate that each individual NOX homolog plays a specific role in each cell type. Therefore, understanding which specific NOX is playing the pathological effect in each stage of the disease would help to address the situation;

(ii)

More integrative studies that delve deep into the interconnection between different NOXs or their paracrine effect in other cell types should be performed. As atherothrombosis involves different cell types in its pathogenesis, it should be interesting to use cocultures, organoids, or organ-on-chip devices mimicking the vessel wall to contextualize the effect of NOXs in a more complex biological system;

(iii)

There is an urgent need to develop isoform-specific NOX inhibitors and study these enzymes as potential therapeutical targets in CVDs [146]. These molecules have demonstrated efficacy in different diseases in vivo. For instance, NCATS-SM7270, a NOX2 inhibitor, protects from traumatic brain injury [147]. In fact, GKT137831, a NOX1/NOX4 inhibitor has reached phase 2 clinical trials, used to slow down diabetic kidney disease [148]. The combination of NOX-specific inhibitors with a cell-specific drug delivery system applied in target stages of the disease could be a promising therapeutic strategy not only in atherothrombosis but also in other CVDs.

To sum up, although great efforts have been made to study NOXs in atherothrombotic stroke, there is still much to discover.