Pre‐clinical evidence of a dual NADPH oxidase 1/4 inhibitor (setanaxib) in liver, kidney and lung fibrosis

Abstract Fibrosis describes a dysregulated tissue remodelling response to persistent cellular injury and is the final pathological consequence of many chronic diseases that affect the liver, kidney and lung. Nicotinamide adenine dinucleotide phosphate (NADPH)‐oxidase (NOX) enzymes produce reactive oxygen species (ROS) as their primary function. ROS derived from NOX1 and NOX4 are key mediators of liver, kidney and lung fibrosis. Setanaxib (GKT137831) is a first‐in‐class, dual inhibitor of NOX1/4 and is the first NOX inhibitor to progress to clinical trial investigation. The anti‐fibrotic effects of setanaxib in liver, kidney and lung fibrosis are supported by multiple lines of pre‐clinical evidence. However, despite advances in our understanding, the precise roles of NOX1/4 in fibrosis require further investigation. Additionally, there is a translational gap between the pre‐clinical observations of setanaxib to date and the applicability of these to human patients within a clinical setting. This narrative review critically examines the role of NOX1/4 in liver, kidney and lung fibrosis, alongside the available evidence investigating setanaxib as a therapeutic agent in pre‐clinical models of disease. We discuss the potential clinical translatability of this pre‐clinical evidence, which provides rationale to explore NOX1/4 inhibition by setanaxib across various fibrotic pathologies in clinical trials involving human patients.

to restore normal tissue architecture and preserve tissue function. 3 However, severe or persistent tissue damage can lead to fibrosis due to chronic inflammation and dysregulated repair. [1][2][3][4] Here, tissue remodelling and repair processes occur simultaneously, characterized by excessive ECM deposition and the formation of permanent scar tissue. [1][2][3][4] As such, fibrosis can lead to distorted tissue architecture and impaired organ function. [1][2][3] In some diseases, such as liver cirrhosis, diabetic nephropathy and idiopathic pulmonary fibrosis (IPF), extensive fibrosis-induced tissue remodelling can lead to organ failure and death. [1][2][3][4] The onset and progression of fibrosis is orchestrated by many profibrotic metabolites, including reactive oxygen species (ROS). 5 ROS are highly reactive oxygen-derived molecules encompassing free radicals (e.g., superoxide anion [O 2 •-]), and non-radical species (e.g., hydrogen peroxide [H 2 O 2 ]). 6,7 Oxidative stress denotes the state in which ROS levels exceed the protective capacity of cellular antioxidant defence systems. 7 Whilst controlled generation of ROS is important for normal physiological processes, such as cellular signalling and antimicrobial immunity, excessive ROS production can contribute to pathophysiological consequences, including fibrosis development and persistence. 5,7,8 Specifically, ROS activate and mediate the effects of profibrotic cytokines, namely transforming growth factor-beta (TGFβ). 5,9 In turn, TGFβ triggers ROS production and suppresses cellular antioxidant levels, which induces oxidative stress and contributes to fibrosis progression. 5,9 ROS can be generated as by-products of the mitochondrial electron transport chain. 5 Specific enzymes can also produce ROS, including xanthine oxidase, cytochrome P450 oxidases, cyclooxygenases, lipoxygenases and nicotinamide adenine dinucleotide phosphate (NADPH)-oxidases (NOX). 5,8 The majority of these enzymes generate ROS as by-products of their enzymatic activities, whereas NOX enzymes produce ROS as their primary catalytic endproduct, either as O 2 •or H 2 O 2 via a NADPH-dependent reduction of molecular oxygen. 8,10,11 The NOX enzyme family constitutes seven members: NOX1-5 and dual oxidases (DUOX) 1 and 2. 10,11 The NOX isoforms differ in how they are regulated, their subcellular localization and the type of ROS produced (Table 1). 11,12 Recent evidence has demonstrated the importance of NOX enzymes in the development of tissue inflammation and fibrosis. 10 NOX1 and NOX4 have been shown to drive fibrotic pathologies in various organs, including the liver, [13][14][15][16][17][18] kidney [19][20][21][22][23][24][25] and lung. [26][27][28][29] Thus, the pharmacological inhibition of NOX1/4 offers a potentially promising therapeutic intervention for a range of fibrotic pathologies, particularly those relating to the liver, kidney and lung.
Setanaxib, formerly GKT137831, is a first-in-class NOX1/4 dual inhibitor that blocks the activity of NOX1/4, therefore reducing ROS production and concurrent harmful fibrotic effects ( Figure 1). 11,30 Setanaxib is the first NOX inhibitor to progress through pre-clinical testing to clinical development. To date, setanaxib has been investigated in patients with type 2 diabetes and albuminuria (phase 2; NCT02010242), 31  The dysregulated liver repair process is thought to be initiated by epithelial injury, which is often associated with the release of dangerassociated molecular patterns (DAMPs) and pathogen-associated molecular patterns (PAMPs). 10 DAMPs and PAMPs are recognized by hepatic-resident macrophages or recruited monocytic cells, triggering liver inflammation. 10 Liver inflammation encompasses the recruitment and activation of immune cells by proinflammatory mediators secreted from injured hepatocytes, namely chemokines, tumour necrosis factor-alpha (TNFα) and other mediators and their subsequent infiltration into liver tissue. 10 Injured hepatocytes and activated immune cells also secrete profibrotic mediators, such as TGFβ, which alongside proinflammatory mediators, activate resident hepatic stellate cells (HSCs) and trigger their transdifferentiation into myofibroblast-like cells that have proinflammatory and profibrotic properties. 10,39 Oxidative stress is a key aetiological factor that initiates liver fibrosis. NOX1/4-derived ROS, generated by hepatocellular injury, enhance the progression of liver fibrosis by stimulating type I collagen production and mediating the profibrotic effects of TGFβ, including HSC activation and the continued production of proinflammatory cytokines and ECM components by myofibroblasts ( Figure 2

| Setanaxibinliverfibrosis: pre-clinicalevidence
Setanaxib has been shown to replicate the hepatoprotective effects seen in multiple NOX1/4 KO models of liver fibrosis. In mice with the superoxide dismutase 1 G37R mutation (SOD1mu), which enhances NOX-mediated ROS production, setanaxib suppressed liver fibrosis in CCl 4 -and BDL-treated SOD1mu mice. 13 This was reflected by reduced hepatic collagen deposition and alpha-smooth muscle actin (α-SMA) expression to a similar extent to those in WT mice. 13 Macrophage infiltration and activation, and TNFα mRNA expression, were also lowered in CCl 4 -treated SOD1mu and WT mice following setanaxib treatment, indicating the suppression of liver inflammation in these mice. 13 Consistent with these data, pretreatment of primary culture-activated HSCs from SOD1mu mice with setanaxib reduced the expression of profibrotic genes, including collagen α1(I), and suppressed ROS production to similar levels as setanaxib-treated WT HSCs. 13 In line with these observations, ROS production and HSC activation were reduced in primary HSCs isolated from BDL-treated mice following pre-treatment with setanaxib, as shown by the suppression of procollagen α1(I), α-SMA and TGFβ expression. 16 Observations of reduced oxidative stress, hepatocyte apoptosis and liver fibrosis were seen in setanaxib-treated mice following BDL compared to those treated with solvent. 16 Similar results were TA B L E 1 Tissue distribution and regulation of NOX/DUOX isoforms.  vehicle-treated rats. 15 PHT is caused by increased intrahepatic resistance to portal blood flow in chronic liver diseases, most commonly cirrhosis, which in turn is caused by ROS-induced liver fibrosis and the concurrent formation of hyperdynamic circulation. 45,46 Hyperdynamic circulation can also be induced by various vascular mediators, including NO, whose synthesis and release is driven by endothelial nitric oxide synthase (eNOS), and vascular endothelial growth factor (VEGF). 15 Importantly, treatment with setanaxib has been shown to decrease H 2 O 2 -induced mesenteric VEGF expression in PHT rats induced with PPVL, alongside reduced eNOS phosphorylation and NO production in mesenteric arteries compared with vehicle-treated rats. 15

| Clinicaltranslatability
Pre-clinical studies demonstrate that the effects of genetic NOX1/4 deficiency are consistent with those of setanaxib in human liver cells and animal models of liver fibrosis.
In liver biopsy samples from patients with stage 2/3 autoimmune hepatitis, NOX4 expression levels were higher in both myofibroblasts and hepatocytes compared with these cells from control patients. 16

| RoleofNOX1/4inkidneyfibrosis
Kidney fibrosis is a common pathological consequence of progressive kidney disease that is most commonly caused by diabetes mellitus and hypertension and may ultimately lead to end-stage kidney disease, regardless of aetiological cause. 10,47 As with liver fibrosis, cells. 19,22,23 Thus, given that the effects of setanaxib in human cell lines replicate those seen in NOX4 KO models, these data suggest that benefits elicited by setanaxib in pre-clinical models could potentially be translated to human patients, warranting further exploration in a clinical trial setting.

| RoleofNOX1/4inlungfibrosis
Lung fibrosis is characterized by chronic injury repair and ECM deposition in the interstitial-alveolar spaces and is associated with many of the interstitial lung diseases. 52,53 Excessive accumulation of fibrotic tissue can lead to reduced lung compliance and increased respiratory effort, as well as hypoxaemia and pulmonary hypertension caused by the collapse of alveolar structural and functional integrity. Together, these factors manifest clinically as the progressive deterioration of respiratory mechanics and gas exchange, which may eventually progress to respiratory failure and death. 52,54 Lung fibrosis has diverse aetiologies including drugs, chemical insults, radiation, occupational exposures and connective tissue diseases; alternatively, it may be 'idiopathic' in nature. 55 Damage-induced alveolar epithelial cell death causes the release of profibrotic mediators, including TGFβ, which activate resident lung fibroblasts and trigger their differentiation into myofibroblasts, resulting in heightened ECM synthetic capacity and resistance to apoptosis. 10,53,54 Importantly, TGFβ induces NOX4-mediated ROS production in lung fibroblasts, which sustains myofibroblast differentiation and drives fibrosis progression in lung disease ( Figure 4). 10,56 Injured alveolar epithelial cells also release proinflammatory mediators, which trigger the infiltration of activated immune cells to the site of injury. 10,54 Immune cells in turn amplify inflammatory signalling and contribute to fibrosis development by exacerbating myofibroblast differentiation and ECM deposition. 10,54 Cellular senescence is also thought to contribute to the pathology of lung fibrosis, in which senescent cells secrete senescence-associated secretory phenotype factors to promote inflammation, tissue remodelling and cell growth. Additionally, setanaxib improved the susceptibility of fibroblasts to apoptosis, reflected by the increased caspase-3 activity. 28

| Clinicaltranslatability
Several lines of pre-clinical evidence have demonstrated that effects of genetic NOX1/4 deficiency are consistent with those of setanaxib in human lung cells and animal models of lung fibrosis. 63 In hypoxia-exposed human pulmonary artery endothelial and Given that NOX4 expression is upregulated in lung fibroblasts isolated from IPF patients, and that the protective effects of setanaxib in human cell lines replicate those seen in NOX4 KO models, these data strengthen evidence suggesting that the benefits of setanaxib treatment seen in pre-clinical models could potentially be translated to human patients, warranting further exploration in a clinical trial setting.

| PER S PEC TIVE SANDFUTURE DIREC TIONS
While there is a wealth of pre-clinical evidence that supports the Although several studies have demonstrated that the genetic deletion or pharmacological inhibition of NOX1/4 confers protective effects against fibrosis, conflicting evidence exists suggesting that NOX1/4 deficiency may promote fibrosis. 64,65 This was shown in mice subjected to unilateral ureteral obstruction (UUO), a model F I G U R E 4 NOX1/4-mediated ROS production in lung fibrosis. ECM, extracellular matrix; MSC, mesenchymal stem cell; NADPH, nicotinamide adenine dinucleotide phosphate; NOX, NADPH oxidase; ROS, reactive oxygen species; TGFβ, transforming growth factorbeta.
of tubular stress in the kidneys leading to kidney fibrosis in chronic kidney disease (CKD), where tubulointerstitial fibrosis and tubular epithelial cell apoptosis were significantly increased, and peritubular capillary density was significantly decreased, in NOX4 KO mice compared with WT mice. 65 Importantly, oxidative stress was not reduced but increased in NOX4 KO mice subjected to UUO compared with WT mice, suggesting that NOX4 plays an antioxidant role. 65 Collectively, these data therefore suggest that NOX4 may protect against kidney fibrosis in CKD by counteracting oxidative stress, lowering apoptosis and maintaining microvascularization in kidney tubular cells. 65 The selective inhibition of NOX1/4-driven ROS production by setanaxib is supported by multiple lines of evidence. However, it has been proposed that effects induced by setanaxib are independent of NOX1/4 activity and the compound may instead modulate ROS metabolism through other mechanisms, for example peroxidase inhibition. [66][67][68] Thus, the pharmacological characterization of setanaxib in terms of its selectivity and mode of action requires further exploration.
To clearly distinguish between a true NOX inhibitor and molecules with ROS-scavenging and/or assay-interfering properties, a series of biochemical assays have previously been used. [66][67][68] Here, setanaxib demonstrated substantial interference with peroxidase-dependent assays and potently inhibited the H 2 O 2 -producing activity of xanthine oxidase in the absence of NOX enzymes; these effects potentially originated from direct xanthine oxidase inhibition or H 2 O 2 scavenging, or the non-specific inhibition of horseradish peroxidase used in the assay. 66,68 Thus, while setanaxib is the most widely recognized NOX1/4 dual inhibitor, its inhibitory action on ROS production may not fully stem from the inhibition of NOX1/4, but rather from an unspecified redox mechanism that merits further investigation. 66,68 Therefore, while there have been significant advances in the understanding of the pathophysiology underlying liver, kidney and lung fibrosis, the precise roles played by NOX1/4 in their pathophysiology and the selective inhibitory action of setanaxib require further exploration.

ACK N O WLE D G E M ENTS
The authors acknowledge Krassimir Mitchev, MD, PhD, Calliditas Therapeutics AB, London, UK, for publication coordination and Ryan Haines, BSc, Costello Medical, London, UK and Olivia Wakeman, BSc, and Sarah Jayne Clements, PhD, Costello Medical, Cambridge, UK, for medical writing and editorial assistance based on the authors' input and direction.

FU N D I N GI N FO R M ATI O N
This study was sponsored by Calliditas Therapeutics AB. Support for third-party writing assistance for this article was funded by

CO N FLI C TO FI NTE R E S T
VJT, KJD, NJT: The authors confirm that there are no conflicts of interest. CS: Former employee and shareholder of Calliditas Therapeutics Suisse SA.

DATAAVA I L A B I L I T YS TAT E M E N T
Data sharing is not applicable to this article, as no new data have been generated.