Knock, knock: Who’s there? Nox1

Abstract

It has become apparent that major sources of reactive oxygen species (ROS) in mammalian cells are the NADPH oxidases. These multimeric enzymes are composed of membrane bound catalytic subunits, or Nox enzymes and a small docking subunit termed p22phox. Depending on the Nox isoform, cytosolic regulatory subunits also play a role in their activation. There are 7 Nox isoforms. The first identified was Nox2, also termed gp91^phox, which is present in phagocytic cells and is responsible for the oxidative burst. Unlike Nox2, the other Nox enzymes exhibit sustained production of ROS at somewhat lower levels. Their modes of activation differ, as do their cellular distribution. Nox 1-4 are expressed in vascular cells in rodents. Nox 5 has been identified in atherosclerotic lesions of humans but is not expressed in rodents. Stimulation of cells with angiotensin II, cytokines, catecholamines, high glucose, or mechanical stretch promote NADPH oxidase activation, in large part due to recruitment of cytosolic subunits to the membrane subunit, leading to formation of the functional enzyme complex that can transfer electrons molecular oxygen and formation of superoxide (O₂ ^·–).¹ Superoxide serves as a progenitor for other ROS, including hydrogen peroxide (H₂O₂), peroxynitrite (OONO^-), and hypochlorous acid (HOCl^-). Enhanced activity the NADPH oxidases and increased expression of its subunits occur in many pathological conditions including hypertension, diabetes, atherosclerosis, cardiac hypertrophy and heart failure. Studies of genetically altered mice have shown that deletion of various NADPH oxidase subunits protect against these diseases, while their overexpression promotes pathology.^2-4

A consequence of excessive ROS formation is loss of endothelial nitric oxide (NO), which can occur by rapid radical-radical reactions with O₂ ^·– and other radicals, or can occur because of dysfunction of the endothelial NO synthase. NO has protective, anti-inflammatory and anti-atherosclerotic effects. Loss of NO therefore promotes expression of endothelial adhesion molecules, enhances platelet adhesion and promotes vascular smooth muscle proliferation, all critical aspects of atherosclerotic lesion formation.⁵

Given the enormous amount of basic knowledge supporting a role of ROS in cardiovascular diseases, it initially seemed logical that antioxidant vitamins might be beneficial in preventing or reversing disorders such as atherosclerosis and hypertension. Unfortunately, this has turned out not to be the case. Numerous large clinical trials have failed to show any benefit of treatment with vitamin E, vitamin C, beta-carotene and related antioxidants.^6-8 Surprisingly, some studies have actually shown that high doses of vitamins paradoxically worsen outcomes. As an example, the HOPE-TOO study showed that high dose vitamin E actually increased the incidence of heart failure.⁹ The explanations for failure of these antioxidant trials remains unclear but might reflect a lack of efficacy of the antioxidants used. For example the rates of reaction of vitamin E and vitamin C are about one million times slower than that of O₂ ^·– with NO, such that one would not expect that oral treatment with these antioxidants would prevent this reaction. ROS also have beneficial effects, such as bacterial and viral killing, modulation of cell growth and proliferation and cell signaling. Thus, non-specific scavenging of all ROS might lead to untoward effects. Related to this, the Nox enzymes have unique subcellular locations, and non-targeted drugs might either not reach these sites or reach the wrong sites and scavenge the wrong radicals. The timing of treatment, often started well after the onset of disease, might not be able to improve clinical outcomes. For the reasons above and others, there has been substantial interest in first identifying which Nox enzymes are involved in various pathological processes, and second developing specific inhibitors that target these pathological Nox isoforms and leave others alone.

In the issue of Circulation, Grey et al have taken an important step in this direction.¹⁰ These investigators used three methods, siRNA, genetically altered mice and a novel pharmacological agent to show that Nox1 is a key player in the vascular pathology caused by hyperglycemia and diabetes. They found that high glucose stimulates mRNA expression of Nox1 but not Nox 4 in cultured human endothelial cells. Moreover, high glucose increased ROS production in these cells and this was prevented by small interfering RNA to Nox1 or by a novel Nox1 inhibitor GKT13783. To translate these cell culture findings to a clinical scenario, the authors bred mice lacking either Nox1 or Nox4 with ApoE^-/- mice that are prone to develop atherosclerosis. Induction of diabetes in ApoE^-/- mice led to severe aortic atherosclerosis, and this was markedly reduced in mice lacking Nox1 but not in mice lacking Nox4. An amazing finding was that treatment with the Nox1 inhibitor GKT137831 also reduced development of aortic atherosclerosis in the diabetic ApoE^-/- mice. Of interest, Nox1 deletion and GKT137831 had no effect on body weight, serum cholesterol, systolic blood pressure, and triglyceride and LDL levels in diabetic versus control mice. Thus, these interventions seemed to reduce atherosclerosis by acting downstream of classical risk factors. Some of these downstream events likely include expression of pro-inflammatory and pro-fibrotic genes including the monocyte chemoattractant peptide 1 (MCP1), the vascular cell adhesion molecule 1 (VCAM1), fibronectin, CTGF and collagen IV, which were all reduced in Nox1^-/- mice and by GKT137831 therapy. These molecular signals were accompanied by reduced inflammation, evidenced by reduced vascular macrophage adhesion and infiltration in Nox1^-/- mice and GKT137831 treated animals (Figure).

Figure 1.

The role of Nox1 in diabetes associated atherosclerosis. High glucose activates Nox1, which in turn produces reactive oxygen species (ROS). These promote expression of pro-inflammatory and pro-fibrotic genes that lead to vascular inflammation and atherosclerotic lesion development. The Nox1 inhibitor GKT137831 prevents ROS production in response to high glucose and decreases vascular inflammation, fibrosis and plaque formation.

Several purported NADPH oxidase inhibitors have been used experimentally, but most of these are either non-specific or ineffective. Diphenylene iodonium, commonly used as such an agent, actually inhibits all flavin containing enzymes including the nitric oxide synthases, complex 1 of the mitochondria¹¹ and xanthine oxidase¹² and thus cannot be used as a specific NADPH oxidase inhibitor. Apocynin has extensively been used as an NADPH oxidase inhibitor, but in most cases seems to work principally as a non-specific antioxidant.¹³ The chimeric peptide gp91ds-Tat is composed of nine amino acids from the coat protein of HIV and nine amino acids of gp91^phox (Nox2). This peptide binds to p47^phox and interferes with assembly of p47^phox and gp91^phox. It effectively inhibits the Nox2 based NADPH oxidases, but cannot be given orally because it is a peptide¹⁴. Other agents, including the antibiotic peptide PR39, aminoethyl benzenesulfono fluoride (AEBSF) and the benzo(b)pyran-4-one derivative, S17834 have been used experimentally as NADPH oxidase inhibitors, but have off-target effects that have limited their application. A compound recently developed by Vasopharm, known as VAS2870 inhibits all NADPH oxidases, and is under development for treatment of peripheral artery disease.

An obvious problem with inhibiting all the NADPH oxidases is that blockade of Nox2 or gp91^phox would lead to a condition analogous to chronic granulomatous disease, in which lack of the neutrophil NADPH oxidase leads to repeated bacterial infections and early death. For this reason, it would be highly desirable to develop selective inhibitors of Nox1, Nox4 and perhaps Nox5 that do not affect Nox2. In an effort to accomplish this, Laleu et al at Genkyotex used high throughput screening to identify a series of compounds that inhibited Nox1 and Nox4.¹⁵ Structure activity relationship studies identified a pyrazolopyridine dione which was subsequently named GKT137831. Studies of membranes of cells expressing either Nox1, Nox2, Nox4 or xanthine oxidase showed that GKT137831 is 10-fold more specific for Nox 1 and Nox4 than for Nox2 and that it has essentially no ability to inhibit xanthine oxidase. Subsequent studies have shown that GKT137831 attenuates hypoxia induced pulmonary vascular cell proliferation.¹⁶ This agent also protects hepatocytes against apoptosis in culture in response to TNFα and Fas Ligand stimulation and reduces hepatic fibrosis in response to bile duct ligation in vivo.¹⁷ Phase I human studies have demonstrated good safety and tolerability following single and multiple oral doses of GKT137831 and a favorable pharmacokinetic profile.

The fact that GKT137831 inhibits both Nox1 and Nox4 can be beneficial because both isoforms have been implicated in conditions such as hypertension, heart failure, liver fibrosis and vascular hypertrophy. A remaining issue is the role of Nox5, which accumulates in human atherosclerotic lesions and can contribute substantially to ROS production at these sites.¹⁸ It is unclear how best to inhibit Nox5 or if GKT137831 has any activity against this isoform. It is possible that calcium channel blockers, by limiting intracellular calcium, would prevent Nox5 activation.

Atherosclerosis in diabetes is an enormous clinical problem, causing myocardial infarction, limb loss, ischemic bowel disease, renal disease and stroke. It is often diffuse and relentlessly progresses despite therapy with commonly used drugs. An inhibitor of Nox1, which could reduce oxidative stress and prevent accelerated atherosclerosis in diabetes, would provide an additional therapeutic option in these difficult to treat patients. Additional studies of agents such as GKT137831 are clearly warranted.