Oxidative stress, extracellular matrix targets, and idiopathic pulmonary fibrosis

Abstract

Idiopathic pulmonary fibrosis (IPF) is an interstitial lung disease characterized by progressive fibrosis of the alveolar interstitium. The pathogenesis is thought to involve abnormal reepithelialization and dysregulated remodeling of the extracellular matrix after alveolar injury. There is growing evidence through human and animal studies that oxidative stress plays a role in this dysregulation. Markers of oxidative stress have been identified in the lungs of IPF patients and aberrant antioxidant activity exacerbates pulmonary fibrosis in animal models. In addition, the extracellular matrix is a critical component in regulating cellular homeostasis and appropriate wound healing. Recent investigations support that the matrix is a target of oxidative stress in the lung and IPF. Extracellular matrix degradation products, produced by reactive oxygen species, may promote fibrogenesis by influencing epithelial, mesenchymal, and inflammatory cell activity. The impact of the interactions of oxidative stress and the matrix of the lung remains unclear and may prove to be an important target for new therapies in IPF. Utilizing oxidative enzymes, antioxidants, or the matrix as therapeutic targets to control oxidative stress in IPF will continue be an area of active research and innovative discoveries in the coming years.

Introduction

The lung is continually exposed to higher oxygen levels than other tissues. Furthermore, exogenous oxidants, as well as pollutants, can augment oxidant production and activate inflammatory cells to generate additional reactive oxygen and nitrogen species. The lung protects itself against these oxidants with protective antioxidants and antioxidant enzymes. Various disease states in the lung involve dysregulation of this balance through excessive oxidant production or decreases in antioxidants. Idiopathic pulmonary fibrosis (IPF) is a lung disease characterized by progressive fibrosis of the alveolar interstitium [1,2]. Reactive oxygen species (ROS) and markers of oxidative stress are evident in human IPF [3,4] and levels of ROS negatively correlate with pulmonary function in IPF and may predict disease severity [5].

Idiopathic pulmonary fibrosis

Diagnosis and pathological findings

Idiopathic Pulmonary Fibrosis is an interstitial lung disease characterized by severe and progressive fibrosis of the alveolar interstitium. In the United States, the prevalence of IPF is estimated to be 42.7 per 100,000 and the disease incidence to be 16.3 per 100,000 [6]. Patients develop symptoms of dyspnea (shortness of breath) and nonproductive cough with presentation between 50 and 70 years of age. IPF is slightly more common in males than females [2] and has a dismal prognosis with a 5-year mortality rate between 50 and 70% [1,7].

From the time of IPF diagnosis, there is a mean survival of 3–5 years [1,7]. A diagnosis of IPF is made from a thorough history and physical, chest radiography, pulmonary function tests, high-resolution computed tomography (CT), and lung biopsy [1,7]. Patients typically present with a history of greater than 3 months of dyspnea and a nonproductive cough. On physical exam, bilateral dry inspiratory crackles may be appreciated at the lung bases. Chest radiography shows ground glass opacities and CT analysis shows irregular thickening of the alveolar septa. As fibrosis of the lung progresses, the normal lung architecture becomes distorted under the tension of the fibrosis. This change is often described as a “honeycomb” appearance of the lung [8].

The gold standard of diagnosis of IPF is still pathologic examination of lung tissues. Histologically, IPF has a pattern of usual interstitial pneumonia (UIP), which is characterized by areas of immature and mature fibrosis (temporal heterogeneity) and alveolar inflammation with intervening areas of normal tissue architecture [8]. On H&E staining, myofibroblastic foci are present, which are light-staining areas of spindle-shaped mesenchymal cell expansion among collagen and matrix deposition. These foci are randomly dispersed throughout the lung and are a marker of active disease [8]. Inflammation is also present and is assessed through bronchoalveolar lavage and interstitial microscopy, which shows the presence of macrophages, neutrophils, eosinophils, mast cells, and lymphocytes [2,8].

Pathophysiology

Pulmonary fibrosis can occur in various situations: due to an unknown stimuli (idiopathic); environmental/occupational exposure, i.e., asbestos, silica; induced by pharmacological agents, i.e., bleomycin; radiation exposure; and associated with other primary diseases such as collagen vascular diseases or familial forms [2,7]. These variations of IPF differ by their proposed pathogenic factors Table 1.

Table 1.

The various types of pulmonary fibrosis and proposed pathogenesis

Type of pulmonary fibrosis	Stimuli	Proposed pathogenic factors
Idiopathic	Unknown	Aberrant wound healing [9,10]; Profibrotic proteins, i.e., TGF-β [11–14]; oxidative stress [15–17]; initial inflammation [18–20];
Environmental/occupational	Asbestos; silica; paraquat	Particle transition metals, i.e., free iron, and oxidative stress [21,22]; inflammation [23–25]; profibrotic proteins, i.e., TGF-β [22,26];
Pharmacologic agents	Bleomycin amiodarone	Agent-induced alterations in oxidants/antioxidants [27–29];
Radiation-induced	Radiation therapy to the chest	oxidative stress [30]; Loss of antioxidants [31]; profibrotic proteins, i.e., TGF-β [32,33]; inflammation [34];
Hereditary familial IPF	Genetic mutations	Surfactant protein C and A1 [35], TERT and TERC genes [36]
Collagen vascular diseases	Collagen abnormality	Autoimmune tissue injury and aberrant matrix deposition [37,38].

The underlying processes of pulmonary fibrosis are currently thought to involve the presence of persistent stimuli or injury, aberrant wound healing, and dysregulated repair/remodeling of the lung that results in fibrosis. Epithelial injury is thought to be one of the initial steps in the pathogenesis of pulmonary fibrosis. Furthermore, analysis of UIP lung biopsies revealed a significant loss in type I epithelial cells in fibroblastic foci and areas of lung deterioration, along with increases in epithelial apoptosis markers [39]. After epithelial cell death, the basement membrane of the alveolar surface is left denuded and exposed. Epithelial wound healing or reepithelialization is a concerted effort by various cells types to restore the lung after an injury or cell death. Mesenchymal cells, such as fibroblasts and myofibroblasts, promote new matrix synthesis to form a suitable scaffold on which epithelial cells can repopulate lost cells. This process is thought to be disrupted and dysregulated in the development of IPF. Transforming growth factor-β (TGF-β) is a profibrotic protein released primarily by alveolar epithelial cells and macrophages within the lung. TGF-β has multiple functions including inducing the expression of collagens, proteoglycans, and matrix components by fibroblast/myofibroblasts, is chemotactic to macrophages and fibroblasts, and can lead to the development of fibrotic lesions [11,13,40,41].

While the pathogenesis of IPF remains unclear, inflammation and oxidant/antioxidant imbalances within the lung are believed to be involved [5,42]. The role of inflammation has not been delineated in human disease and is a controversial issue, which is likely one reason why many investigators consider pulmonary fibrosis to be primarily a disease of abnormal wound repair. Therapeutic studies that target inflammation, such as corticosteroids, have failed to show clinical benefits [43–45]. However, several studies have highlighted associations between the presence of inflammatory cells and the disease prognosis. Neutrophilia and eosinophilia are seen in the bronchoalveolar lavage fluid (BALF) of 70–90 and 40–60% of IPF patients, respectively [7]. This increase in inflammatory cells has been associated with a worse prognosis and mortality in some clinical studies [18,46,47]. Immune activation and inflammation have been shown to play an important role in fibrosis models [25,48].

Inflammatory cells can damage the lung through the release of oxidative species, proteases (i.e., matrix metalloproteinases, elastase), peroxidases (i.e., myeloperoxidase), cytokines, and growth factors [49,50]. These findings collectively suggests that inflammation, while not necessarily the primary mechanism of IPF pathogenesis, may contribute to a profibrotic environment by affecting the wound-repair process, level of oxidative stress, and the extent of remodeling. Studies support that chemotactic factors and neutrophils are present in IPF [20,51] and increase the likelihood of disease progression and lack of response to immunosuppressive agents [1,18,52]. In IPF, inflammatory cells may also release exaggerated amounts of reactive oxygen species [17].

Inflammatory cells may impact IPF pathogenesis by enhancing the oxidative imbalance in the lung. Studies show increased reactive oxygen species production in leukocytes from the serum and evidence of enhanced oxidative stress in the plasma and bronchoalveolar lavage fluid of IPF patients [3,4]. Levels of oxidative stress have been shown to negatively correlate with aspects of pulmonary function in IPF patients and may provide information about disease severity [5]. Given this evidence, free radicals are thought to play an important role in IPF pathogenesis, potentially through both direct and indirect mechanisms.

Oxidative Stress in the Lung

Oxidative stress is frequently defined as the imbalance of oxidant production and antioxidant defenses, where oxidants dominate and lead to cellular dysfunction and tissue damage. When considering oxidative stress, the lung is somewhat unique due to its exposure to relatively higher oxygen tensions than other tissues. The oxygen pressure of inhaled air is 20 kPa (150 mm Hg). Pressures in venous blood flow are around 6 kPa (45 mm Hg) and may be as low as 0.13 kPa in some tissues, while the oxygen at the alveoli of the lung is ~13.3 kPa (100 mm Hg) [53]. Thus, the lung is constantly facing relatively high oxygen tensions, which may augment oxidative insults. Under normal conditions, the ability of the lung to maintain an oxidative balance and a nontoxic pulmonary environment is likely due to a combination of mechanisms including protective antioxidants, low metabolic demands, and low levels of transition metals. Unregulated production of reactive oxygen species (ROS) and reactive nitrogen species (RNS) in the lung and other tissues can lead to an imbalance in oxidants relative to antioxidants leading to oxidative and nitrosative stress, respectively.

Reactive species

Reactive oxygen species are formed from one-electron reduction of diatomic oxygen and subsequent one-electron reductions to achieve more reactive oxygen by-products, such as superoxide radical, hydrogen peroxide (H₂O₂), and hydroxyl radical (•OH) either spontaneously or through enzyme catalysis [54–56]. Hydroxyl radicals can be formed from the reaction of H₂O₂ with transition metal ions or the breakdown of peroxynitrite and is one of the most reactive radicals produced in biological systems. It acts on local substrates at diffusion-limited rates of between 10⁹ and 10¹¹ M⁻¹ s⁻¹ [54], with a half-life of only milliseconds. The reactivity of superoxide and hydrogen peroxide is approximately 10^1–3 M⁻¹ s⁻¹ [57].

In vivo, other mechanisms for the generation of free radical and reduction–oxidation products include electron leak from mitochondrial metabolism, enzymatic reactions (oxidases and peroxidases) and release from activated leukocytes through the NADPH oxidase-catalyzed oxidative burst [55]. The mitochondrial respiration chain is extremely active and generates superoxide when electrons leak from the energy-producing system. Enzymatic production of reduction–oxidation products can occur through the reaction of xanthine oxidase with hypoxanthine in the presence of O₂, producing superoxide, xanthine, urate, and H₂O₂. Xanthine oxidase has been shown to be upregulated with lung injury and is dependent on iron stores for its increased activity [58–60].

Activated leukocytes (neutrophils, macrophages, eosinophils) are another primary source of reduction–oxidation products through enzymes such as NADPH oxidase, myeloperoxidase, and eosinophil peroxidase. These oxidases catalyze one- or two-electron reductions to form superoxide and H₂O₂, respectively, that contribute to the killing of microbes, intracellular and extracellular signaling [61,62], and potential damage to host tissues when released from these cells [55].

NADPH oxidases are important sources of reduction–oxidation products in many other noninflammatory cells in the lung. Indeed, several recent studies indicate that ROS production by NADPH oxidases plays a central role in the pathogenesis of pulmonary fibrosis and inflammation [25,63–65]. The Nalp3 inflammasome has been shown to have a central role in animal models of pulmonary fibrosis and is activated by reduction–oxidation by-products of NADPH oxidase in the lung [25]. A recent article has found that NADPH oxidase-4 plays a central role in myofibroblast activation [63], a cell known to be important in matrix remodeling and the progression of pulmonary fibrosis [63–65], further highlighting the active role ROS play in the pathogenesis of pulmonary fibrosis.

Myeloperoxidase (MPO) is a second enzyme found in neutrophil azurophillic granules that aids in bacterial killing, by utilizing H₂O₂ and chloride ions to produce toxic hypochlorous acid (HOCl) or bleach. MPO is a highly cationic enzyme allowing it to localize to cell surfaces, such as endothelial and epithelial surfaces, through interactions with glycosaminoglycan chains. HOCl can participate with hydroxyl radical in the fragmentation of extracellular matrix components such as hyaluronan and glycosaminoglycan side chains like heparan sulfate [66–68] (see below). MPO has also been shown to catalyze the metabolism of nitric oxide (·NO) making it unavailable for modulating vascular tone [69]. Eosinophil peroxidase is the eosinophil equivalent of MPO and has similar biological activity.

Nitrogen

An excess of various nitrogen-containing species within a system leads to nitrosative stress. Nitric oxide (·NO) is an important nitrogen species produced by nitric oxide synthase enzymes through metabolism of L-arginine. There are three NOS enzymes, two of which are constitutively expressed: endothelial eNOS, neuronal nNOS, and one inducible NOS (iNOS/NOS2). Twenty times more ·NO can be produced by iNOS than the other enzymes [55]. iNOS activity is induced by external stimuli such as bacterial lipopolysaccharide [70]. Nitric oxide mediates the relaxation of smooth muscle cells in the cardiovascular and pulmonary systems. However, in addition to these beneficial effects of ·NO, it can contribute to pathologic processes, especially when produced in large quantities. Notably, nitric oxide has been shown to be important in the pathogenesis of pulmonary fibrosis, especially when produced by iNOS [71,72]. Covalent reactions can occur between ·NO and NO-derived species with biological molecules such as proteins, DNA, lipids and amino acids, which can modify the function of these molecules [73].

The radicals produced by reduction of oxygen can react with freely diffusible nitric oxide to form additional radical species, in effect inactivating nitric oxide, which is a potent signaling molecule (i.e., the inactivation of NF-κB [74]) and vaso-relaxant [70]. Peroxynitrite anion (ONOO⁻) can be formed through the diffusion-limited reaction of superoxide with nitric oxide [75] or by reactions between hydrogen peroxide and nitrite [76]. The antioxidant enzyme superoxide dismutase, which is highly expressed in the lung, acts to keep superoxide levels low, thus preserving nitric oxide function [77]. Peroxynitrite is a powerful oxidant that can modify tyrosine residues producing nitrotyrosine and is also a potent oxidizer of thiols. At a physiologic pH, the protonated form of peroxynitrite (peroxynitrous acid) will decompose into hydroxyl radical and nitrogen dioxide. One-electron reduction reactions of ·NO will form nitrite, nitrogen dioxide, and nitrate. Notably, myeloperoxidase can catalyze nitrite-dependent nitration of tyrosine residues which can further promote nitrosative tissue damage [78].

A study by Saleh et al. provides support for the involvement of nitrosative stress in IPF lungs. They found that the lungs of IPF patients had increased NOS expression and nitrotyrosine modifications of proteins, suggesting that the IPF lungs were exposed to elevated nitrosative stress compared to healthy controls [79]. Animal studies further suggest that ·NO signaling stimulates increased production of remodeling proteins, such as TGF-β and MMPs in pulmonary fibroblasts [71]. iNOS-null animals that were exposed to silica inhalation developed significantly less pulmonary fibrosis than wild-type mice, supporting a profibrotic role for ·NO produced by iNOS in the lung [72].

Oxidative Stress and IPF—Evidence from Patient Populations

The underlying pathogenesis of pulmonary fibrosis is currently thought to involve the presence of a persistent stimuli or injury, such as oxidative stress, and dysregulated repair of the lung that results in fibrosis. Several studies have found evidence of increased oxidative stress in idiopathic pulmonary fibrosis. Reduction–oxidation products and free radicals of oxygen metabolism are difficult to measure directly in tissues; thus many clinical studies have utilized biological markers of oxidative reactions for assessment, such as modified carbonyls, proteins, DNA, and lipids.

Bronchoalveolar lavage is a technique used to sample the epithelial lining fluid of the lung. 8-Isoprostane, a product of free radical-mediated lipid peroxidation, is increased in the bronchoalveolar lavage fluid (BALF) of IPF patients [80], as well as, in exhaled breath condensate along with hydrogen peroxide [81]. Exhaled ethane, a second marker of lipid peroxidation, is increased in patients with interstitial lung disease [82] and mirrored the patients’ P_aO₂ levels and clinical course, as patients with significantly elevated levels of ethane died or rapidly deteriorated. Oxidized proteins with carbonyl modifications are increased in the BALF of nonsmoking IPF and sarcoidosis patients [83]. Increased carbonyl-modified proteins have also been shown in systemic sclerosis, IPF, eosinophilic pneumonia, and allergic alveolitis [84]; however, proteomic studies reveal that more low-molecular-weight proteins are altered in IPF BALF [85]. These oxidative markers negatively correlated with pulmonary function in all of these studies.

Antioxidant enzyme status is also altered in patients with IPF. Glutathione levels in alveolar epithelial lining fluid are decreased in IPF lungs [16,42]. Markart et al. show that IPF patients appear to compensate for oxidative stress with increased expression of Nrf-2, a redox-sensitive antioxidant transcription regulator, and significant increases in low-molecular-weight antioxidants during fibrotic phases [86]. Notably, these antioxidants are insufficient to counter-balance the oxidative stress. Kinnula et al. report significant decreases in extracellular superoxide dismutase in fibrotic regions of UIP lungs [87], which suggests that oxidative stress would be increased in these areas. The loss of antioxidants in the lung and abnormal cellular signaling for antioxidant expression may have a role in IPF pathogenesis.

Oxidative stress and animal models of IPF

Animal models of pulmonary fibrosis have offered opportunities to further evaluate the role of oxidative stress in alveolar injury, inflammation, and fibrosis development. The stimuli commonly used to initiate pulmonary fibrosis in these animal models are bleomycin (intratracheal, subcutaneous, or intraperitoneal administration), asbestos, or silica, which are administered intratracheally or via an inhalation chamber. Bleomycin forms a complex with redoxactive iron, molecular oxygen, and DNA, resulting in DNA strand breaks [88]. Bleomycin also produces superoxide and hydroxyl radicals that can damage cell membranes, lipids, and proteins [89]. Asbestos-induced pulmonary fibrosis studies show increased superoxide production directly through transition metal reactions and indirectly through oxidative bursts from recruited neutrophils and macrophages exposed to asbestos [24,90].

It has also been demonstrated that antioxidants can prevent bleomycin- and asbestos-induced pulmonary fibrosis. Examples include studies showing protection against pulmonary fibrosis with N-acetylcysteine and desferoxomine administration [29], treatment with lecithinized superoxide dismutase, and extracellular superoxide dismutase overexpression [91], as well as protection when there is decreased ROS production in NADPH oxidase knockouts or knock-downs. [64,92]. In contrast a lack of superoxide dismutase exacerbates bleomycin-induced fibrosis [93]. In bleomycin- and asbestos-induced pulmonary fibrosis, extracellular superoxide dismutase (EC-SOD) protects by limiting both inflammation and fibrosis development [93–95]. Importantly, reactive oxygen species produced by asbestos fibers have been shown to directly activate profibrotic TGF-β in the lung [22,26], which is important in fibrosis development. Furthermore, antioxidant treatment with superoxide dismutase was able to inhibit asbestos fiber-induced activation of latent TGF-β. These studies indicate that ROS can directly contribute to profibrotic activation of TGF-β, and that antioxidants may useful in preventing this effect.

Similar to the asbestos model, intratracheal instillation of silica results in an acute accumulation of inflammatory cells (neutrophils, macrophages, lymphocytes, and occasional eosinophils) in the alveolar spaces and interstitium [96–98], damage to epithelial cells [99,100], and subsequent collagen deposition [101–103] and fibrosis development [103,104]. Models of radiation-induced pulmonary fibrosis have also shown a role for ROS as treatment with MnSOD has been shown to inhibit fibrosis [105,106]. The absence of EC-SOD promotes fibrosis in various inhalation injury models and treatment with SOD-mimetic agents, such as TBAP, are also protective [93,95,107–109].

Many studies have identified several potential mechanisms through which the presence of oxidative stress in the lungs can lead to increased inflammation and fibrosis. Reactive oxygen species can alter inflammation through the activation of nuclear factor-kappa B (NF-κB) and activator protein-1 (AP-1) [110,111]. These redox-sensitive transcription factors can bind to promoter regions in DNA and control the gene expression of a host of genes including those controlling proinflammatory cytokines, growth factors, and apoptotic signals [110]. NF-κB activation occurs in alveolar epithelial cells after asbestos exposure [23,112]. Oxidative stress is also evident in the fibro-proliferative response. In vitro, H₂O₂ can stimulate the proliferation of cultured human fibroblasts [113]. Furthermore, fibroblasts isolated from IPF lungs are capable of inducing apoptosis in epithelial cells in vitro [114], further promoting the cycle of alveolar damage and abnormal repair. Hydrogen peroxide, as a diffusible factor, can lead to increased epithelial cell death, as well [50]. Finally, oxidative species can lead to direct degradation of the extracellular matrix (ECM) and can control proteolytic degradation through the activity of matrix metalloproteinases and tissue inhibitors of MMPs [115,116].

Molecular targets of oxidants in IPF

Extracellular matrix

The extracellular matrix is critical for maintaining a strong structure that can withstand mechanical stretch and recoil of the lung as well as providing the architectural support for normal epithelial growth. The ECM of the lung is composed of several major components including collagens, elastin, fibronectin, proteoglycans, hyaluronan, and laminin [117,118]. Pulmonary fibrosis is characterized by often drastic changes in the ECM, which can be the result of excessive matrix deposition (an increase in collagen deposition [2,119,120]), impairment in ECM degradation, and resolution or a combination of these two. Thus, ECM changes become very complex over the pathogenic course of IPF. Fig. 1 depicts the potential changes that may occur in a normal lung and after epithelial injury to an alveoli.

Fig. 1.

Within a normal uninjured alveoli (A), several processes create a homeostatic environment. The epithelium remains intact (1) along with an intact extracellular matrix (2) that binds and localizes many proteins such as growth factors and cytokines. Because there is no alveolar injury, inflammatory cell recruitment is not necessary (3) and oxidants/antioxidants are in balance. When lung injury occurs (B), the epithelium becomes denuded (1) and the ECM is degraded by enzymatic and oxidative mechanisms (2). The cellular injury, ECM degradation products, and cytokines cause inflammatory cell influx (3) and oxidative stress (4). This complex environment results in impaired epithelialization and dysregulated matrix deposition (5).

Heparan sulfate proteoglycans comprise a membrane-bound core protein with attached sulfated polysaccharide side chains [121]. Polyanionic proteoglycans, like heparan sulfates, have the ability to bind highly cationic proteins and transition metals. The binding of transition metals makes them potential sites for metal-catalyzed redox chemistry within the body. Furthermore, cationic proteins such as myeloperoxidase and eosinophil peroxidase bind to the ECM and are sites of additional radical production such as hypochlorous acid from Cl⁻ and hydrogen peroxide [67,68].

Hyaluronic acid, a glycosaminoglycan, and syndecan proteoglycans are prominent in the lung and are highlighted here, representing potential ECM targets for ROS. These components can function in various ways within the ECM: (1) bind and localize soluble and insoluble ligands, i.e., growth factors, TGF-β, FGF, cytokines; (2) act as a soluble paracrine or autocrine factors when the ectodomain is shed; (3) maintain receptor abilities for internalization of ligands; (4) facilitate leukocyte migration and trafficking [122]. Syndecans are also linked to the actin cytoskeleton and have key roles in controlling cellular migration, proliferation, and homeostasis. Known ligands for syndecans that are of importance with regard to inflammation and fibrosis include TGF-β1 and 2, HGF, VEGF, PDGF-AA, FGF, cytokines, and chemokines such as IL-8, MCP-1, and TNF-α [122–125]. Syndecans can also bind other ECM components such as fibronectin and laminin and can bind enzymes such as neutrophil elastase, tissue plasminogen activator, and extracellular superoxide dismutase. Syndecan-1 and −4 can bind elastase in dermal wound fluids protecting them from their inhibitors and modulating the proteolytic potential of the microenvironment [126]. They can also bind MPO, which may promote increased oxidative stress, as described above.

The ECM can be degraded in two primary ways: enzymatic cleavage (MMPs, hyaluronidases, heparanase) or oxidative cleavage, which can transform the ECM into soluble effector molecules. Heparanase is an endoglycosidic enzyme that cleaves HS side chains [127] and is capable of cleaving syndecan-1 [128]. Syndecan core protein ectodomains can be shed from the cell surface through proteolytic cleavage of the juxtamembrane region. Matrilysin or MMP 7 is a protease that binds to heparan sulfate [129] and induces shedding of the syndecan-1 ectodomain in a model of acute lung injury induced by bleomycin [130]. Indeed, MMP 7 is believed to play a central role in the pathogenesis of pulmonary fibrosis [131,132]. Consistent with this, MMP 2 and MMP 9 can also shed syndecan ectodomains in vitro [133]. This shedding can be regulated by tissue inhibitors of matrix metalloproteinases, such as TIMP 3 [134].

The role of oxidative stress in extracellular matrix dysregulation can be highlighted by recent studies of hyaluronic acid and the heparan sulfate proteoglycan syndecan-1 [135–137]. Hyaluronic acid shedding has been shown in the lungs of IPF patients [138]. Shedding of syndecan-1 is also significantly elevated in the BALF of IPF patients. In both the asbestos and the bleomycin mouse models of pulmonary fibrosis, syndecan-1 is shed into the BALF during the inflammatory and fibrotic phases of injury.

Notably, the absence of EC-SOD in the lung results in exaggerated syndecan-1, heparan sulfate, and hyaluronic acid fragmentation or shedding [135–137,139]. Combined with data that demonstrate oxidants can directly lead to shedding and fragmentation of these ECM components in vitro, these findings suggest that oxidants are directly contributing to fragmentation and shedding of the ECM components in models of pulmonary fibrosis. The importance of oxidative fragmentation/shedding of high-molecular-weight hyaluronan to low-molecular-weight species is highlighted by studies demonstrating that high-molecular-weight hyaluronan has antifibrotic and anti-inflammatory activity, but low-molecular-weight hyaluronan has proinflammatory and profibrotic activity in the lung [140,141]. Shed syndecan-1 can also promote fibrosis in several unique ways. Shed syndecan-1 induces neutrophil chemotaxis (which can be inhibited by EC-SOD in vivo and in vitro), inhibits alveolar reepithelialization, and stimulates fibrogenic TGF-β release. These studies show the important role that oxidative stress has in modulating the ECM and how the oxidative by-products can promote fibrosis in the lung.

Studies also indicate that oxidative fragmentation of HS side chains can occur through hypochlorite species generated by MPO [66–68,142] and through hydroxyl radicals generated by xanthine oxidase [143,144]. This is particularly important in sites of inflammation and neutrophil influx. Potential oxidative reactions can occur to the core protein itself or the polysaccharide side chains. Protein backbone oxidation is more complex and significant cleavage occurs only with very reactive oxygen species such as hydroxyl radicals [145]. Oxidation and cleavage of the core protein can occur through hydrogen abstraction from a central α-carbon and subsequent reaction with oxygen to form a peroxyl radical [146]. This radical can undergo conversion to an α-C alcohol and the peptide bond can be cleaved by hydrolysis or to an alkoxyl species resulting in cleavage of the peptide bond. Polysaccharide side chain fragmentation can occur through hydrogen abstraction from any of the C–H bonds on the sugar residue creating a C-centered radical, called an α-hydroxyalkyl radical (·C (OH)RR′) [145]. This radical can then be converted to a peroxyl radical in the presence of oxygen and undergo chain hydrolysis or can undergo β-scission of the glycosidic bond which would fragment the chain [145,146].

Matrix metalloproteinases

ECM degradation and turnover is also regulated by the activity of matrix metalloproteinase enzymes (MMPs) and their tissue inhibitor counterparts (TIMPS). MMPs are matrix-degrading proteinases (currently a total of 22) shown to be upregulated in models of pulmonary fibrosis [116,147]. The majority of MMPs are synthesized as proenzymes and activated by proteolysis of a cysteine–zinc prodomain, called a “cysteine switch” [115,148]. ROS are also capable of activating MMPs, increasing their transcription, and deactivating proteases [148–150]. Thus, oxidants may play a significant role in unregulated activity of MMPs in pulmonary fibrosis. The substrates of MMPs are extracellular matrix components and soluble factors and include, but are not limited to, the following: (1) MMP 1, 8, and 13 are collagenases targeting collagens I, II, III, VII, X, gelatin, and pro-TNF-α; (2) MMP 2 and 9 are gelatinases targeting type IV and V collagen, gelatin, elastin, fibronectin, pro-TGF-β, and pro-TNF-α; (3) MMP 3, 10, and 11 are stomelysins that target proteoglycans, laminin, fibronectin, gelatin, and pro-TNF-α; and (4) MMP 7 (matrilysin) targets proteoglycans, collagens, laminin, decorin, gelatin, and fibronectin [115]. Tissue inhibitors of metalloproteinases (TIMPs 1–4) are extracellular or membrane-bound enzymes that bind tightly to MMPs to inhibit their degradative activity [115].

In IPF patients, MMP 2 and 9 and TIMPs 1 and 2 are elevated in areas of alveolar damage and at disrupted basement membranes [151]. McKeown et al. report increases in MMP 3, 7, 8, and 9 in BALF from IPF patients with levels higher in patients with earlier mortality [152]. Rosas et al. report increases in MMP 1 and 7 in serum, BALF, and lung tissue in IPF patients, suggesting they may be blood biomarkers for IPF [132]. Increases in MMP 7 expression have also been reported [131]. MMP 7-null mice are also protected from bleomycin-induced pulmonary fibrosis [131]. Animal studies show similar results with increases in MMP 2 and 9 in the fibrotic phase of bleomycin-induced fibrosis [116]. Cabrera et al. report that an overexpression of MMP 9 diminishes bleomycin-induced fibrosis [153]. In asbestos-induced fibrosis, MMP 9 and MMP 2 are important during the inflammatory and fibrotic phases of disease pathogenesis, respectively [116]. Inhibitors of MMPs have also been successful in protecting against asbestos-induced pulmonary fibrosis [116]. While there are differences in findings related to the role of MMPs in pulmonary fibrosis pathogenesis, the balance of MMPs and TIMPs is likely to play important roles during the course of fibrogenesis in the lung.

Antioxidants in the Lung

The lung expresses a variety of antioxidant resources to protect against oxidative stress within the tissue. These antioxidant defenses include low-molecular-weight antioxidants (glutathione, vitamin E, uric acid, etc.); metal-binding proteins (transferrin, lactoferrin, etc.); thiol-containing proteins with redox regulatory activity (thioredoxin, peroxiredoxin, and glutaredoxins); enzymes that degrade H₂O₂ (catalase and glutathione peroxidases); mucins; detoxifying enzymes (glutathione-S-transferases); and superoxide dismutases. These antioxidants create a homeostatic system that functions to scavenge oxidative species and radicals that can damage cellular and matrix components. Several of these antioxidants have been shown to be important in IPF and are highlighted below.

Catalase is an important scavenger of H₂O₂ expressed within the alveolar epithelium and inflammatory cells of the lung. H₂O₂ has been shown to be an activator of pulmonary fibroblasts from IPF lungs and catalase can inhibit this activation [50]. A recent study reports that catalase administration to asbestos-treated wild-type mice protects against pulmonary fibrosis development, by inhibiting the production of H₂O₂ through Rac1 GTPase-stimulated NADPH oxidase [154,155]. Similarly, in a rat model of asbestosis, extended administration of polyethylene glycol (PEG)-linked catalase for 20 days decreases fibrosis and collagen deposition in the lung [155]. It should be noted that patients with acatalesemia do not have pulmonary fibrosis, thus the loss of catalase itself is not causative in pulmonary fibrosis.

Glutathione and alpha-tocopherol are important low-molecular-weight antioxidants found in the lung. Glutathione has been shown to be decreased in the epithelial-lining fluid [16] and in fibrotic lesions of IPF lungs. Furthermore, one study reports that the administration of oral N-acetylcysteine can increase glutathione levels in BAL fluid [156] and sputum [15], and within alveolar epithelial cells to reduce oxidant production [157], suggesting that antioxidants can be exogenously modified in the lung. Administration of aerosolized glutathione to a small number of IPF patients resulted in a shift in the lung’s oxidant–antioxidant balance toward the later [158]. In a recent study, alpha-tocopherol, commonly known as vitamin E, was reported to be elevated in the BAL fluid of IPF patients and was emulated during fibrosis in the bleomycin-animal model [86].

Superoxide dismutase (SOD) was first described by McCord and Fridovich in 1969 [159,160]. There are three SOD enzyme isoforms including intracellular CuZn SOD (SOD1), mitochondrial manganese MnSOD (SOD2), and extracellular CuZn SOD (SOD3). Extracellular superoxide dismutase (EC-SOD) was identified by Marklund et al. in 1982 and is highly expressed in the lung [161,162]. When the epithelium of the lung is exposed to oxygen or a noxious stimuli, extracellular antioxidants have a critical role in preventing oxidative stress.

EC-SOD has been implicated in the pathogenesis of pulmonary diseases involving oxidative stress [94,163,164]. EC-SOD is an active extracellular scavenger of superoxide free radicals by catalyzing the dismutation of superoxide into hydrogen peroxide and oxygen. This occurs at a rate constant of N10⁹ M⁻¹ s-¹ [164]. EC-SOD is highly expressed in the vasculature [165] and functions to preserve nitric oxide bioavailability within various organ system [77,166–168] by removing superoxide that can deplete NO. Fig. 2 depicts the structure and functional domains of EC-SOD. This enzyme is highly expressed in the lung and localizes to cell surfaces by binding to heparan sulfate species [164,169–172] and type I collagen [139,163] through its matrix-binding domain (MBD).

Fig. 2.

Schematic of EC-SOD structure and heparin affinity. (A) The EC-SOD monomer contains an enzymatic functional domain (gray), a unique matrix-binding domain (MBD) at the carboxyl terminus (black) which is composed of arginine (R) and lysine (K) residues, variable free cysteine residues that can participate in disulfide bonding (Cys), and an N-linked glycosylation site. The positively charged MBD makes the site suitable for binding to highly negatively charged heparin species in the ECM. (B) EC-SOD tetramer affinity for the matrix can be regulated by proteolytic removal of the matrix-binding domain (MBD).

Enzymatic cleavage of any of the four MBDs will decrease or abolish the affinity of EC-SOD for the matrix and cell surfaces (Fig. 2B). EC-SOD has three heparin affinity types: no affinity (type A), moderate affinity (type B), and high affinity (type C). Trypsin or endoproteinase treatment of EC-SOD, which targets lysine residues, can abolish or weaken the matrix-binding affinity of EC-SOD [169]. This supports the important role of the cluster of basic amino acids in the C-terminus of EC-SOD.

EC-SOD has been shown to play an important role in several models of pulmonary fibrosis including bleomycin-, asbestos-, and radiation-induced fibrosis. Knockout mice lacking EC-SOD throughout their tissues have significantly more lung fibrosis, acute lung injury, and inflammation dominated by a neutrophil influx due to bleomycin and asbestos intratracheal administration [93,95,109]. EC-SOD distribution in the lungs of wild-type mice also changes, as it is lost from the parenchyma where it normally resides and increases in air spaces in fibrosis models [109,173,174], and after hyperoxia [107]. EC-SOD appears to be exerting its anti-inflammatory and antifibrotic effects by inhibiting oxidative degradation of matrix components, as discussed above.

Therapeutic Approaches—Controlling Oxidative Stress and Treatment

Despite recent studies and advances in the understanding of the pathogenesis and clinical course, there are currently no effective therapies for IPF, aside from lung transplantation, which comes with its own complications. Several options are available, such as anti-inflammatory agents; however, there are very few to no clinical studies that show improvements in progression-free survival, functional capacity, or quality of life [175]. While corticosteroids have been used over the last 50 years, Flaherty et al. report that fewer than 20% of patients have improvement with steroid therapy [45,176]. Chronic, low-dose prednisone may be a maintenance therapy in responsive patients, but is not recommended for all IPF cases [7,175].

Immunosuppressive/cytotoxic agents, such as azathioprine, which impairs leukocyte proliferation [175] and cyclophosphamide, an alkylating agent, are used in patients who are nonresponsive to steroids and have shown favorable results in 15–20% of IPF cases [1]. Raghu et al. completed a study of high-dose prednisone therapy versus high-dose prednisone plus azathioprine and found no significant differences in clinical measures, such as forced vital capacity (FVC) and diffusion capacity of carbon monoxide (DLCO), with either therapy [177]. Cyclophosphamide has shown no survival benefit in studies [43,178], and has a profound side-effect profile [175,178] which limits its utility.

Finally, antifibrotic agents such as colchicine [179] have been tried, but have been unsuccessful in humans. Colchicine functions by decreasing collagen formation through fibroblasts and macrophages [175] and its metabolites have the ability to scavenge free radicals; however, this also leads to the production of secondary radicals [180,181]. While colchicine showed promise in vitro and in animal models [182,183], it has shown no survival or lung function benefit in clinical IPF [44,184], thus having limited clinical utility in IPF therapy.

Amidst the grim outlook of current therapies, the IPF research community continues to find new molecular targets and therapeutic options, some of which have antioxidant activity. Pirfenidone is a pyridone molecule that has anti-inflammatory and antifibrotic effects in both in vitro and in vivo pulmonary studies. It can also scavenge hydroxyl and superoxide free radicals [185–187]. It was successful in abrogating bleomycin-induced fibrosis in animal models [188,189], by decreasing TGF-β expression and subsequent collagen deposition. In a Phase II open-label trial, Pirfenidone was effective in improving both 1-year survival to 78% (compared to 70% reported in other studies) and stabilizing or improving lung function (diffusing capacity (DLCO) and forced vital capacity) in patients with advanced IPF [190]. Pirfenidone is not currently approved in the United States.

Many studies have focused on carnosine, N-acetylcysteine, and SODs. Administration of carnosine, a free radical scavenging peptide, in mice decreases inflammatory and fibrotic markers in bleomycin-induced fibrosis [191] and is available for human administration. N-Acetylcysteine, one of the most highly studied thiol-containing agents, scavenges H₂O₂ (hydrogen peroxide), OH (hydroxyl radical), and HOCl (hypochlorous acid) and promotes glutathione synthesis [156]. The IFIGENIA trial (a randomized, double-blinded, placebo-controlled trial) reported that the addition of high-dose NAC to the standard therapy of prednisone and azathioprine can significantly slow IPF progression compared to standard therapy alone when evaluated on forced vital capacity, DLCO, and a composite physiologic index [192,193]. One study reported that aerosolized NAC resulted in improved oxygen saturation and CT image changes in IPF patients; however, it had no effect on pulmonary function or quality of life [194]. It remains unclear if there is a survival benefit from NAC therapy.

Antioxidant mimetics may be another potential therapeutic strategy. A recent review discusses these and additional antioxidant mimetics in detail [195]. Small-molecular-weight SOD mimetics, such as metalloporphyrins, have been effective in limiting radiation-induced lung injury, oxidative stress, inflammation, and bleomycin-induced pulmonary fibrosis in animal models [108,196,197]. The metalloporphyrins have several distinct antioxidant actions including scavenging superoxide (SOD-like activity), hydrogen peroxide (catalase activity), peroxynitrite, and lipid peroxidases [195] and are not readily metabolized in vivo [195,196]. While it is hoped that novel antioxidant therapies will provide therapeutic benefit to patients, it is clear that oxidative stress is just one component of pulmonary fibrosis. Thus, combination therapies of antioxidants with other antifibrotic agents may be a more rationale approach to future therapeutic investigations.

Lung transplantation is the only current option that prolongs survival in IPF patients. Considerations for lung transplantations should be made early on in the disease course, as the wait list time is around 46 months, during which time many patients with advanced disease die prior to transplant [7,175]. The 5-year survival posttransplant is approximately 40% [198]. Lung transplantation is also associated with increased pulmonary and systemic oxidative stress during the posttransplant period [199].

Final Discussion

The current belief in IPF pathogenesis is that cellular injury, which is often repetitive, acts as the inciting event for fibrosis development and oxidative imbalance within the lung. Causes of the oxidative stress include, but are not limited to, the cellular injury, transition metal exposure, inflammation, or drugs that participate in reduction–oxidation reactions. The importance of oxidative modifications to the extracellular matrix and how this alters cellular responses in human IPF remain open and underinvestigated areas. Recent studies show that ECM degradation products do have biological functions and may add to the progression of pulmonary fibrosis (Fig. 3). While there is not sufficient evidence that the ECM products are initiating or causative factors in IPF, they appear to promote a profibrotic environment.

Fig. 3.

Oxidative stress can accompany or be caused by various stimuli in the lung, i.e., repetitive cellular injury, transition metal or particulate exposure, noxious drugs, inflammation, and enzymatic activity such as NADPH oxidases. This oxidative imbalance can result in degradation of ECM components and a loss of protective antioxidants. This leaves the tissue susceptible to increased inflammation, profibrotic signals, and aberrant wound healing—all of which may contribute to the progression of pulmonary fibrosis.

The current literature on the pathogenesis of tissue fibrosis focuses primarily on the roles of epithelial, mesenchymal, and inflammatory cells. Specifically, the role for oxidative shedding of matrix components and the effects of the shed species during tissue injury remain unclear. Extracellular superoxide dismutase (EC-SOD) is the most abundant antioxidant enzyme in the extracellular space of many tissues where it is localized through binding to matrix components such as heparan sulfates. Novel evidence is available that supports the importance of oxidative stress in the ECM, such as with syndecan-1 or hyaluronic acid, and that antioxidants, like EC-SOD, have primary roles in protecting the matrix and preventing detrimental downstream consequences (Fig. 3).

In addition to scavenging oxidants once they are produced, it may also be possible to directly inhibit their production from their source. Novel strategies to inhibit oxidant production by any of these complex enzymes or transition metal systems may provide ideal targets for therapeutic intervention for IPF patients.

This review highlights the important role that oxidative stress has in the pathogenesis of idiopathic pulmonary fibrosis and emphasizes the importance of the extracellular matrix in the pathogenesis of pulmonary fibrosis. Oxidative degradation of the extracellular matrix may prove to be a good therapeutic target given that: (1) intact ECM is critical for appropriate wound healing; (2) ECM degradation by-products are biologically active, and (3) the ECM localizes many cytokines, growth factors, and enzymes shown to potentiate fibrosis. Additional investigations into antioxidant therapeutics are necessary to elucidate their full potential, especially with regard to extracellular matrix degradation. The clinical arena of IPF needs more effective therapies and while antioxidants alone may not be the complete answer, combination therapies that include antioxidants, or inhibitors of oxidant generation, may contribute to future effective therapies for this disease.