Balancing Reactive Oxygen Species in the Epigenome: NADPH Oxidases as Target and Perpetrator

Abstract

Significance: NADPH oxidases are important sources for regulated generation of reactive oxygen species (ROS). The main ROS produced are superoxide and hydrogen peroxide, both of which are redox signaling molecules in the context of various cellular functions. Redox imbalance due to excessive or insufficient ROS is a hallmark of pathophysiological aspects, including cancer development and progression. Recent Advances: Epigenetic silencing of NADPH oxidases by hypermethylation of their promoter region or of the genes required for their assembly and activity occurs in diseases, such as lung cancer, and may represent an early stage of neoplastic transformation. Critical Issues: Loss of ROS-mediated signaling by epigenetic silencing may promote tumorigenesis. Conversely, increased oxidative stress caused by oncogene-induced overexpression of NADPH oxidases may also drive epigenetic instability. Thus, the cellular redox balance is likely vital in carcinogenesis. Future Directions: NADPH oxidases may serve as prognostic tumor biomarker, especially when their individual expression is confined to accessible tissues, such as mucosal epithelia or blood. Further validation of NADPH oxidase/dual oxidase enzymes as candidate markers will require well controlled, large-scale clinical data sets. This review is focused on NADPH oxidases as targets of epigenetic changes in cancer and on the emerging role of ROS as inducers of epigenetic changes. Antioxid. Redox Signal. 18, 1937–1945.

Introduction

T he NADPH oxidase (Nox) family of proteins is unusual in that their sole function is the production of reactive oxygen species (ROS), although the species of ROS produced varies depending on isoform of the oxidase. ROS are typically regarded as a by-product of functioning or damaged mitochondria and of other enzyme systems, such as lipoxygenases or peroxisomal oxidases. In contrast, Nox generate a defined burst of ROS as a controlled response to particular stimuli, and thus should be considered as significant contributors toward redox signaling. There are seven members of this family, Nox1–5 and dual oxidase (Duox) 1–2 (Fig. 1). The catalytic Nox domain is anchored by six to seven transmembrane passages in plasma or endomembranes, and is usually assembled as a heterodimeric complex. Duox enzymes are distinguished from their Nox counterparts by the presence of an additional extracellular peroxidase homology domain, whose biological function is still debated. All Nox/Duox enzymes require a unique subset of partner proteins for maturation and catalytic activity. Heterodimers are formed by association with smaller transmembrane proteins. Nox1–4 assembles with p22^phox, whereas Duox proteins partner with DuoxA1–2. These partner proteins serve several functions, including stabilizing the Nox/Duox enzyme, directing the dimer to particular cellular locations and facilitating protein–protein interactions that are necessary for complete enzyme assembly and activity. An active Nox complex contains usually several proteins with the Nox2-dependent phagocyte oxidase being composed of at least six components.

FIG. 1.

The basic structure of the Nox/Duox enzymes. All members of the Nox/Duox family share the basic Nox structure with six transmembrane domains. Duox enzymes possess also a peroxidase homology domain connected via an additional transmembrane domain to a region containing EF hand domains, a feature they share with Nox5. Known activating partner proteins are presented alongside the general oxidase structure (for Nox1–3 represented in circles). Duox, dual oxidase; Nox, NADPH oxidase. (To see this illustration in color, the reader is referred to the web version of this article at www.liebertpub.com/ars.)

The Nox/Duox family has wide tissue distribution, ranging from the innate immune system, kidney, brain, thyroid, and skeletal muscle to testis and inner ear (32). Almost all cell types and tissues express at least one Nox, either constitutively or inducible, with many cells harboring several oxidases. The subcellular localization of these enzymes, when active, is usually the plasma membrane or membranes forming specialized vesicles (redox signalosomes). Unusual locations, such as mitochondria and nuclear compartments have been reported (33, 55). Interestingly, not all of the Nox/Duox enzymes release the same ROS. For instance, Nox4 and the Duox enzymes seem to generate only hydrogen peroxide in comparison to the superoxide that the other oxidases produce. This is unusual considering the shared catalytic domain structure of these enzymes. The core structure of Nox is composed of six transmembrane domains, two of which contain histidine residues required for heme coordination. The loops between these membrane-embedded segments contain regions that are required for assembly and activity of the oxidase. On the intracellular side, the carboxyl terminus harbors the NADPH and FAD binding domains. The electrons they accept are passed on via the hemes to the outer membrane, where they will reduce molecular oxygen to superoxide (5) (Fig. 2). It is unresolved how hydrogen peroxide can be produced directly by Nox4 and Duox enzymes. Some Nox/Duox family members display distinct features, such as calcium sensitivity, due to the presence of calcium-binding EF hands, a motif commonly found in calcium-binding proteins, in their amino termini (Nox5, Duox1–2) or by being apparently constitutively active (Nox4).

FIG. 2.

Generation of superoxide by the phagocyte oxidase. The phagocyte oxidase Nox2 is currently the best characterized member of the Nox family. This scheme illustrates the manner in which electrons from the NADPH donor are passed along the structure of the oxidase, passing through the centrally coordinated hemes to reduce molecular oxygen to superoxide at the outer membrane face.

The Contribution of Nox Enzymes to Cellular Signaling

Although the best characterized oxidase, the phagocyte oxidase (Nox2), is known for its critical role in host defense, the other members of the Nox/Duox family seem to be more involved in cell signaling. Several publications have placed Nox4 within the TGF-β signaling cascade and Nox4-deficient mice are resistant to TGF-β-driven, bleomycin-induced lung fibrosis (9, 27, 65). In vascular smooth muscle cells, Nox1 and Nox4 are the predominant Nox. They exhibit unique responses to stimulation with the platelet-derived growth factor or with angiotensin II. These stimuli upregulate Nox1, but not Nox4, resulting in a brief burst of superoxide (36). Reducing Nox1 expression downregulated the redox-sensitive p38 mitogen-activated protein kinase (MAPK) and Akt signaling pathways (36). These signaling pathways have also been associated with Nox1 in human and rat hepatoma cells (61). Herein, the increased signaling through the epidermal growth factor receptor correlated with increased Nox1 expression, again acting upstream of p38 MAPK and Akt. Nox4, on the other hand, can regulate the expression of antioxidant enzymes through activation of the Nrf2 transcription factor (8). Duox1-derived ROS modulated the expression of the mucin protein, MUC5A, in airway epithelia through a TGF-α signaling axis (64). Additionally, cross talk between the different Nox enzymes has emerged, with increased expression and activity of one oxidase being responsible for upregulation of another oxidase (14, 44). Considering the well-documented correlation between increased ROS generation and cancer, it is not surprising that dysregulation of Nox has been linked to tumorigenesis (54, 67). Although most of our current knowledge focuses on upregulation of activated NADPH oxidases (Nox1, Nox4) in cancer, examples exist where epigenetic silencing of oxidases (Duox1, Duox2) and thus diminished ROS generation appears to be an early event during tumor development and progression.

The Involvement of Nox Enzymes in Carcinogenesis

In a physiological cellular setting, evidence suggests that Nox1 and Nox4 may be necessary for replicative senescence in tissues, where they are highly expressed such as the lining of the blood vessels, the stomach, and the colon (4, 35). In these tissues, a state of senescence may be triggered by ROS as part of an antiangiogenic program. This enforced growth arrest may act as protection from transformation. This observation has been reported for both Nox1 and Nox4 in endothelial cells and in thyrocytes transduced with oncogenic H-Ras (62, 72). However, attempting to attach a candidate tumor suppressor designation to these proteins seems too facile. Clearly, these proteins exert cell signaling functions related and necessary to normal cell physiology and in some context these functions coincide with the terminal differentiated cell phenotype. However, it appears that Nox enzymes can find themselves co-opted in pursuit of a tumorigenic phenotype in various stages of transformation.

The contribution of Nox enzymes to carcinogenesis has been described for several types of cancer. For instance, Nox4 was reported to be at least partially responsible for the transformation of melanoma cells, and inhibition of Nox4 prevented the development of endothelial tumors in nude mice and restricted the invasiveness of gliomablastoma in rats (6, 48, 75). The p22^phox subunit, required for Nox1–4 activity, has been implicated in the inactivation of the tumor suppressor tuberin in renal carcinoma cells (7). Fibroblasts, engineered to overexpress Nox1, produced tumors when introduced into nude mice (66), suggesting a mitogenic function for Nox1. A close relationship between increased expression of Nox1 and activating mutations of K-Ras was uncovered in a screen of human colon cancer tissues (37, 62). The same report detected increased levels of Nox1 in a transgenic mouse expressing oncogenic K-Ras^G12V in the intestinal epithelium. While this may appear to contradict the H-Ras-induced senescence associated with Nox1 and Nox4 mentioned earlier (62, 72), H-Ras and K-Ras seem to exert distinct effects with H-Ras inducing growth arrest in lung and colon and K-Ras promoting increased proliferation (46, 57). In general, excessive ROS generation is thought to increase the mitogenic potential of tumors transformed with oncogenic Ras (46). Knockdown of Nox1 in rat kidney cells expressing constitutively active K-Ras resulted in a change from a transformed phenotype toward a more normal cell behavior. As mainly Nox1 and Nox4 have been connected to enhanced ROS production in tumors, these oxidases are likely the source for ROS as a means to produce epigenetic changes in other tumor-associated genes.

Duox enzymes are known for their involvement in thyroid hormone biosynthesis. Genetic mutations, such as substitutions and deletions causing frame shifts often cause loss of Duox2-DuoxA2 activity, leading to hypothyroidism (25, 52). An inverse correlation between Duox expression and thyroid cancer was detected. While the mRNA levels of Duox1 and Duox2 displayed no change between normal tissue and thyroid tumors, over half of the tissues showed decreased Duox protein expression in thyrocytes (34). A report evaluating Duox activity in thyroid nodular lesions found a decrease in Duox activity in these tumor tissues (22). Similarly, Duox1–2 enzymes are highly expressed in healthy airway epithelia (18, 21), but their expression is epigenetically silenced in lung cancer (42). The frequent and early downregulation of Duox in diverse lung tumors suggests an important physiological role for Duox in maintaining a fully functional lung epithelial barrier as discussed in following chapters. In contrast to these observations, colon adenoma microarray analysis demonstrated significantly upregulated expression of Duox2 message in tumor versus adjacent normal tissue (31). Elevated expression of Duox1 and Duox2 mRNA in hepatocellular carcinoma specimen was linked to poor prognosis (41).

This review is focused on Nox as targets of epigenetic changes in cancer and on the emerging role of ROS as inducers of epigenetic changes. As the examples demonstrate, localized ROS within the cellular environment are powerful signaling molecules, and one can assume that increased concentrations of ROS will have effects upon redox-sensitive epigenetic modifications.

A Brief Introduction to Epigenetic Modifications

Epigenetic modifications comprise methylation of cytosines (at sites referred to as CpG islands) and a range of modifications of the amino terminal tails of histones, including phosphorylation, acetylation, methylation, ubiquitination, and sumoylation.

Methylation of DNA does not occur uniformly across the genome and is typically found to favor promoter regions of the genome with approximately half of all promoters containing CpG islands (12). Retrotransposons and repeated sequences account for the majority of the remainder. However, the epigenome is not fixed in place and is responsive to environmental cues. In fact, the communication between these two systems can be as rapid as 4 h after exposure to airborne particulate matter (3), where long interspersed elements (LINE) were found to undergo a decrease in methylation. In fact, the same study found a correlation between methylation patterns and seasonal environmental changes for LINEs, thus indicating a high level of responsiveness in the epigenome.

Another level of modulating gene expression is through the acetylation of histones, which bind to and wrap the DNA into conformations that facilitate or inhibit gene expression. Typically, the acetylation of the N-terminal tails of the histone proteins will cause an increase in gene expression as the DNA is in a looser state called euchromatin, where transcription can proceed with more ease. The enzymes responsible for the addition of the acetyl residues to the histones are called histone acetyl transferases (HATs). Another family of enzymes termed histone deacetylases (HDACs) removes these acetyl residues, causing remodeling of the chromatin to an inactive state.

Epigenetic modifications are reversible and thus the epigenome is not static; rather it is dynamic and this plasticity relays information between the environment and the genome. Epigenetic changes may occur as a result of Nox/Duox-regulated signaling pathways. Alternatively, the Nox/Duox enzymes may themselves be subject to epigenetic regulation. Our work as discussed here focuses on epigenetic changes occurring in Duox enzymes.

Epigenetic Silencing of Duox Enzymes

As mentioned earlier, Duox1–2 enzymes are not only expressed in the thyroid, but also in mucosal barrier tissues and, in particular, on the apical side of the lung epithelium. Each Duox subunit associates with a particular partner protein, DuoxA1 and DuoxA2, to form a functional Duox complex. These complexes mature via the endoplasmic reticulum–Golgi pathway and translocate to the plasma membrane (24, 43, 47). The maturation factor DuoxA1 can be alternatively spliced to yield at least three different isoforms. These isoforms result from alternative splicing at two splice sites. These sites are to be found in the intron preceding exon 3 and within exon 6 (47). Variants lacking exon 3 are incapable of restoring oxidase activity when coexpressed with Duox1 (43, 47). The most active isoform, DuoxA1–2 (also termed DuoxA1α) is also capable of cross-functionality with Duox2 to produce ROS, but this association alters the cellular location of the active Duox2 enzyme (43).

The gene arrangement of the two Duox proteins and their respective maturation factors warrants discussion (Fig. 3). All four genes are found clustered within the same compact locus on chromosome 15 with Duox1 and Duox2 facing away from each other on the genome. This organization is conserved across a wide variety of species. However, the maturation factors are also arranged in a head to head configuration with their cognate Duox. It has been suggested that the intergenic regions between these pairs have a bidirectional promoter organization (11, 74). Linking the Duox1 and DuoxA1 genes is a highly GC-rich region containing many CpG islands. There is no TATA box, but the intergenic region contains binding sites for the transcription factor Sp1. The intergenic region between Duox2 and DuoxA2 is configured differently, displaying normal GC content. Only the DuoxA2 promoter region features a TATA box. As a result, any loss of expression occurring as a result of promoter hypermethylation is likely to affect all Duox-related proteins.

FIG. 3.

The unique chromosomal arrangement of the genes coding for Duox enzymes and their maturation factors. Duox1–2 located on chromosome 15 in a head to head arrangement with each other and their respective maturation factors. The regions between the different genes are smaller than normal, which has implications for the transcription of these genes. The DuoxA1 maturation factor can be spliced to yield different isoforms, whose functionality differs depending on the exons they contain.

Loss of Duox1, DuoxA1, and Duox2 expression in lung cancer cell lines and lung tumor tissues by epigenetic silencing occurs frequently (42). In 9 out of 12 lung cancer lines and in ∼60% of the analyzed lung tumor samples, hypermethylation and silencing of Duox1–2 and DuoxA1 was observed (Fig. 4). Hypermethylation seems to take place throughout the bidirectional promoter regions. The expression and function of Duox1 and Duox2 was recovered in response to treatment with 5′-aza-2′-deoxycytidine. Often several modifications, methylation and histone deacetylation, are found layered upon the genome, acting as reinforcement for inhibition of expression. Interestingly, the sensitivity to DNA methylation appears to be the only epigenetic modification regulating Duox expression in lung tumors as treatment of lung cancer cell lines with trichostatin A, an inhibitor of histone deacetylation, did not restore Duox protein expression.

FIG. 4.

The intergenic promoter regions of Duox1/DuoxA1 and Duox2/DuoxA2 contain multiple CpG islands. The promoter elements of Duox1/DuoxA1 and Duox2/DuoxA2 have been mapped in detail. The CpG-rich regions display differential methylation in normal versus lung tumor tissue. The line graphs beneath the promoter structure show the location of the promoter mapped and the dashes represent CpG sites. The pie charts demonstrate the percentage methylation of this sites in NHBE (normal human bronchial epithelial) cells versus either H661 or A549 (lung cancer) cell lines.

Given the localization of the Duox enzymes to barrier tissues and mucosal epithelial surfaces, it is important to consider their regulation in response to environmental factors. Downregulation of Duox enzymes was observed in smokers and patients with chronic obstructive pulmonary disease, conditions in which hypoxia is common (49). Epigenetic regulation is often a means of communication and interaction between the environment and the genome, and the presence or absence of Duox1–2 may represent an ideal model for further study of the mechanics of this interaction.

Epigenetic Modifications Affecting Nox in General

The current paucity of published information regarding epigenetic modification of Nox/Duox enzymes is not indicative of a lack of such regulation of these genes. On the contrary, vast amounts of CHiP data have been logged in databases through the ENCODE project and are available for perusal through sites, such as the human histone modification database (76). For instance, the histone modification pattern of Duox1 in lymph tissue has been logged into this database, originating from a large-scale screening (69). Close examination of the genes included in these screens will likely yield useful information about the modes of regulation these genes are subject to.

DNA methylation of the Nox5 gene has been linked to downregulation of Nox5 mRNA expression. In lung cancer cell lines, treatment with 5′-aza-2′-deoxycytidine caused at least a fourfold increase in Nox5 mRNA expression in four of seven lung cancer lines tested (63). Immortalized bronchial epithelial cells derived from a healthy donor showed comparable upregulation of Nox5 mRNA. However, the origin of these cells was an aged smoker. Thus, epigenetic silencing of Nox5 in lung cancer is still inconclusive and awaits more extensive studies. It is also worth mentioning that there are multiple isoforms of Nox5, including a short form reminiscent of Nox1–4 (20). The study mentioned here investigated the Nox5 short form, which is often found in cultured cell lines, but this Nox5 enzyme is not characterized well in respect of regulation and function. Considering the proximity of the Nox5 locus (15q23) to the Duox chromosomal arrangement (15q15–15q21) and the observed methylation-mediated transcriptional silencing of Duox, the implications regarding Nox5 silencing are interesting. It has been suggested that the Nox5 short isoform can bind to and inactivate the longer functional variants (53). Recently, hypermethylation of the Nox5 promoter was also observed in fetal ventricular septal defect (77).

The expression of the phagocyte oxidase (Nox2) appears to be subject to epigenetic regulation during the IFN-γ-mediated immune response. Treatment of myelomonocytic cells with this cytokine caused increased expression of both, Nox2 and p67^phox (17). This coordinated response ensures that a functioning oxidase complex will be produced, and arises from the fact that the promoters of these genes share a unique cis element that is necessary for the attachment of CREB binding protein (CBP), a HAT. HAT binding promotes the recruitment of other transcription factors to the promoters of these genes, allowing their increased expression (17).

The apparent effect of the environment on epigenetic changes is also reflected in increased cellular oxidative stress as a result of harmful environmental conditions. Diesel exhaust and cigarette smoke are well-known epigenetic modulators. In fact, cigarette smoke is capable of upregulating the expression of cytosolic phospholipase A₂ via a Nox-dependent pathway that relies on p300/CBP HAT proteins (10). The effect of p300/CBP recruitment on Nox2 gene expression is sensitive to the effects of TNFα (38). Both of these effects were observed in human tracheal smooth muscle cells.

ROS Produced by Nox/Duox as Mediator of Epigenetic Changes

Excessive ROS production can alter gene expression by generating toxic DNA metabolites. Further, dependent on the concentration and species produced, protein function will be altered via post-translational modifications (oxidation, S-glutathionylation), demonstrating the far-reaching effects of these chemical moieties on gene/protein expression and functionality. It is not surprising therefore to discover that ROS exert changes on the epigenome (Fig. 5).

FIG. 5.

A model of ROS effects on the epigenome. ROS, either hydrogen peroxide or superoxide, can be Nox/Duox-derived or mitochondrial, peroxisomal, or the product of enzymes, such as 5-LOX or XO. ROS generated by Nox/Duox activity can be downregulated through epigenetic mechanisms. Alternatively, ROS produced by Nox/Duox may mediate epigenetic changes in other genes. 5-LOX, 5-lipoxygenase; XO, xanthine oxidase; DNMT, DNA methyltransferase; HDAC, histone deacetylase; ROS, reactive oxygen species. (To see this illustration in color, the reader is referred to the web version of this article at www.liebertpub.com/ars.)

Lim and coworkers reported that ROS are capable of inducing changes in the methylation status of the E-cadherin promoter in hepatocellular carcinoma (40). E-cadherin is classified as a tumor suppressor and its silencing is a common carcinogenic event. This silencing is mediated through epigenetic modification of its promoter as opposed to inactivating mutations. Promoter hypermethylation and subsequent downregulation of E-cadherin expression was observed when hepatocellular carcinoma cells were exposed to H₂O₂, leading to the binding of the transcription factor SNAIL to regions within the E-cadherin promoter. The attachment of SNAIL appeared to act as a signal for recruitment of DNA methyltransferase 1 (DNMT1) and HDAC1 resulting in E-cadherin silencing. As the ROS source was either exposure to ROS or treatment with ROS-generating compounds, such as menadione or phenazine methosulfate, it is yet not apparent how this effect might be induced in a pathophysiological context. Although the expression of a range of antioxidant enzymes was measured in this study, including SOD1 and SOD2, only SOD1 protein expression was decreased. Hong and coworkers linked ROS generated by the Nox5 short isoform to the hypermethylation and subsequent suppression of the p16 tumor suppressor gene in oesophageal adenocarcinoma cells exposed to low pH (28). Another recent report examined the relationship between norepinephrine treatment and the repression of PKCɛ gene expression as a result of promoter hypermethylation (73). This link was found to be ROS-dependent, which the authors attributed to Nox1. The contribution of mitochondrial ROS was not sufficiently addressed, and the observed phenotype may, in fact, rely on more than one ROS generator. Indeed, there is evidence of cross talk between Nox and mitochondria (16, 70), where Nox-derived ROS seems to be partially responsible for the development of nitrate tolerance-associated endothelial dysfunction, a condition initiated by mitochondrial ROS

ROS-mediated epigenetic regulation can also be channeled through histone acetylation. Nickel ions are known to inhibit gene expression partly through inhibition of the acetylation of histone H4 through binding to an amino terminal histidine residue. Additionally, exposure of cells to nickel ions generates ROS, leading to inhibition of acetylation and subsequent gene suppression (29). However, the effect of ROS on epigenetic gene regulation is not straightforward. For example, thalidomide treatment increased acetylation of histone H4 via ROS (1). The study investigating the effects of nickel ions used human hepatoma cells, while the thalidomide experiments were performed in erythroid progenitor cells. The Nox expression profiles in these two cell types is likely quite different. This is an important point as under the ROS umbrella, a number of different species can be found. For example, cellular responses may differ depending whether superoxide or hydrogen peroxide is being generated. This question is especially relevant as some members of the Nox/Duox family are incapable of releasing superoxide, while others can generate both species. Recently, the effect of hydrogen peroxide on the expression of RUNX3, a transcription factor and known tumor suppressor, were investigated (30). Hydrogen peroxide induced increased expression and activity of DNMT1 and HDAC, thus causing epigenetic silencing of RUNX3. Analysis of expression and activation profiles of Nox together with probing carefully the type of ROS generated will give more insight in common principles of oxidative epigenome modulation.

The redox status of the cell is a finely tuned mechanism. In regard to ROS produced by an active Nox enzyme, a system of checks and balances exists to prevent or at least counteract nonspecific consequences. For instance, the expression of catalase is influenced by ROS through both direct and indirect means. The promoter regions of catalase and of Oct-1, one of catalase's transcription factors, contain CpG islands. Both genes are subject to hypermethylation upon exposure to H₂O₂, potentially a way to promote transformation by decreasing catalase expression during tumor development (45, 56).

While the examples given here suggest that ROS promote epigenetic changes, they do not address the mechanism(s) by which this occurs. How will oxidants whether Nox- or mitochondria-derived promote these changes? Redox regulation of DNMTs could explain ROS-induced silencing. In fact, O'Hagan and coworkers observed relocalization of a silencing complex that included DNMT1 and DNMT3b from non-GC- to GC-rich areas of the genome upon exposure of cells to exogenous hydrogen peroxide (51). Investigations into the redox-sensitive regulation of the FOXO4 transcription factor yields further clues (15). In response to ROS, the p300 HAT domain binds to FOXO4, thereby inhibiting its activity. This is achieved through specific cysteine residues distributed throughout the FOXO4 protein, which act as redox sensors. Upon oxidation, cysteine–thiol disulfide bridges facilitate the formation of a covalent complex between p300 and FOXO4. Although, p300 in this context is not acting in an epigenetic regulatory capacity, the fact that this protein is capable of association through disulfide bridges could have further, as yet, unknown implications for its role in histone acetylation and for ROS-induced epigenomic modifications in general.

Nox as Biomarkers in Cancer

Specific and sensitive markers for diagnosis, early detection, and risk assessment are acutely needed in cancers where late diagnosis is still common. Current strategies for detecting lung cancer, for example, are either too costly and invasive for screening programs or not specific and sensitive enough. Analysis of remote media, namely, blood and sputum, holds great promise, although specificity and sensitivity of existing approaches still requires improvement (2, 39, 59). To ensure tissue-specific, early detection of the different forms of nonsmall cell lung cancer, the most prevalent lung cancer, the recovery and testing of sputum in high-risk groups, such as smokers, seems the most promising avenue. It is now apparent that a panel of highly predictive biomarkers might be required. Such a panel may consist of a combination of DNA methylation-based and oncogenic markers. Several candidate genes have been identified, but more characterization and validation are necessary. The potential of Nox for inclusion in a biomarker panel is still uncertain. Oncogenic, activating mutations in Nox have not yet been observed. However, many cancers and, in particular, oncogenic Ras-driven tumors display increased oxidative stress due to upregulation of Nox/Duox enzymes (19). It seems rather unlikely that an increase in Nox expression can serve as a specific tumor biomarker as many other diseases, including acute hypoxia or asthma, also lead to upregulation of Nox/Duox (26, 58, 68). In contrast, Duox1 or DuoxA1 promoter hypermethylation might fit the bill in light of the predominantly lung-specific expression of these genes and their early silencing in a high percentage of lung cancers, although large-scale studies with well-defined tumor specimens are missing (Fig. 6). Duox enzymes are unlikely to fall into the category of tumor suppressors as absence of Duox function seems not to lead to a spontaneous tumor development and their effects on growth or transformation in cell culture are subtle (23, 42, 60). However, recent studies indicate an important role of Duox1 in wound healing (42, 50, 71), a process that when deregulated is now regarded as potential basis for oncogenesis (13). Thus, further studies linking Duox silencing in the lung epithelium with tissue injury and tumor development are warranted.

FIG. 6.

Schematic summary of altered Nox activity and subsequent downstream effects in cancer.

Innovation

The contribution of Nox/Duox enzymes to redox signaling is well established, but their dual role as target and source of epigenetic remodeling is only beginning to emerge. As a target of modifications, Duox enzymes are frequently epigenetically silenced in lung cancer via promoter hypermethylation. On the other hand, oxidative stress can initiate chromatin remodeling and aberrant DNA methylation, and Nox enzymes constitute a key ROS generator in cancer cells. As some anticancer therapeutics use selective oxidative stress as a proapoptotic approach, future efforts in deciphering the spectrum of beneficial versus detrimental ROS will be of particular interest.

Footnotes

Acknowledgments

The authors acknowledge support of Science Foundation Ireland (SFI Stokes award and SFI-PI award to U.G.K.).

Abbreviations Used

References

Aerbajinai

, Zhu

, Gao

, Chin

, Rodgers

. Thalidomide induces gamma-globin gene expression through increased reactive oxygen species-mediated p38 MAPK signaling and histone H4 acetylation in adult erythropoiesis. Blood, 110:2864–2871. 2007.

Anglim

, Alonzo

, Laird-Offringa

. DNA methylation-based biomarkers for early detection of non-small cell lung cancer: an update. Mol Cancer, 7:81. 2008.

Baccarelli

, Wright

, Bollati

, Tarantini

, Litonjua

, Suh

, Zanobetti

, Sparrow

, Vokonas

, Schwartz

. Rapid DNA methylation changes after exposure to traffic particles. Am J Respir Crit Care Med, 179:572–578. 2009.

Banfi

, Maturana

, Jaconi

, Arnaudeau

, Laforge

, Sinha

, Ligeti

, Demaurex

, Krause

. A mammalian H+ channel generated through alternative splicing of the NADPH oxidase homolog NOH-1. Science, 287:138–142. 2000.

Bedard

, Krause

. The NOX family of ROS-generating NADPH oxidases: physiology and pathophysiology. Physiol Rev, 87:245–313. 2007.

Bhandarkar

, Jaconi

, Fried

, Bonner

, Lefkove

, Govindarajan

, Perry

, Parhar

, Mackelfresh

, Sohn

, Stouffs

, Knaus

, Yancopoulos

, Reiss

, Benest

, Augustin

, Arbiser

. Fulvene-5 potently inhibits NADPH oxidase 4 and blocks the growth of endothelial tumors in mice. J Clin Invest, 119:2359–2365. 2009.

Block

, Gorin

, New

, Eid

, Chelmicki

, Reed

, Choudhury

, Parekh

, Abboud

. The NADPH oxidase subunit p22phox inhibits the function of the tumor suppressor protein tuberin. Am J Pathol, 176:2447–2455. 2010.

Brewer

, Murray

, Arno

, Zhang

, Anilkumar

, Mann

, Shah

. Nox4 regulates Nrf2 and glutathione redox in cardiomyocytes in vivo. Free Radic Biol Med, 51:205–215. 2011.

Carnesecchi

, Deffert

, Donati

, Basset

, Hinz

, Preynat-Seauve

, Guichard

, Arbiser

, Banfi

, Pache

, Barazzone-Argiroffo

, Krause

. A key role for NOX4 in epithelial cell death during development of lung fibrosis. Antioxid Redox Signal, 15:607–619. 2011.

10.

Cheng

, Lin

, Lee

, Hsu

, Kou

, Yang

. Cigarette smoke extract regulates cytosolic phospholipase A2 expression via NADPH oxidase/MAPKs/AP-1 and p300 in human tracheal smooth muscle cells. J Cell Biochem, 112:589–599. 2011.

11.

Christophe-Hobertus

, Christophe

. Delimitation and functional characterization of the bidirectional THOX-DUOXA promoter regions in thyrocytes. Mol Cell Endocrinol, 317:161–167. 2010.

12.

Costello

, Plass

. Methylation matters. J Med Genet, 38:285–303. 2001.

13.

Coussens

, Werb

. Inflammation and cancer. Nature, 420:860–867. 2002.

14.

Damiano

, Fusco

, Morano

, De Mizio

, Paterno

, De Rosa

, Spinelli

, Amente

, Frunzio

, Mondola

, Miot

, Laccetti

, Santillo

, Avvedimento

. Reactive oxygen species regulate the levels of dual oxidase (Duox1-2) in human neuroblastoma cells. PLoS One, 7:e34405. 2012.

15.

Dansen

, Smits

, van Triest

, de Keizer

, van Leenen

, Koerkamp

, Szypowska

, Meppelink

, Brenkman

, Yodoi

, Holstege

, Burgering

. Redox-sensitive cysteines bridge p300/CBP-mediated acetylation and FoxO4 activity. Nat Chem Biol, 5:664–672. 2009.

16.

Dikalov

. Cross talk between mitochondria and NADPH oxidases. Free Radic Biol Med, 51:1289–1301. 2011.

17.

Eklund

, Kakar

. Recruitment of CREB-binding protein by PU.1, IFN-regulatory factor-1, and the IFN consensus sequence-binding protein is necessary for IFN-gamma-induced p67phox and gp91phox expression. J Immunol, 163:6095–6105. 1999.

18.

Forteza

, Salathe

, Miot

, Conner

. Regulated hydrogen peroxide production by Duox in human airway epithelial cells. Am J Respir Cell Mol Biol, 32:462–469. 2005.

19.

Franco

, Schoneveld

, Georgakilas

, Panayiotidis

. Oxidative stress, DNA methylation and carcinogenesis. Cancer Lett, 266:6–11. 2008.

20.

Fulton

. Nox5 and the regulation of cellular function. Antioxid Redox Signal, 11:2443–2452. 2009.

21.

Geiszt

, Witta

, Baffi

, Lekstrom

, Leto

. Dual oxidases represent novel hydrogen peroxide sources supporting mucosal surface host defense. FASEB J, 17:1502–1504. 2003.

22.

Ginabreda

, Cardoso

, Nobrega

, Ferreira

, Goncalves

, Vaisman

, Carvalho

. Negative correlation between thyroperoxidase and dual oxidase H2O2-generating activities in thyroid nodular lesions. Eur J Endocrinol, 158:223–227. 2008.

23.

Grasberger

, De Deken

, Mayo

, Raad

, Weiss

, Liao

, Refetoff

. Mice deficient in dual oxidase maturation factors are severely hypothyroid. Mol Endocrinol, 26:481–492. 2012.

24.

Grasberger

, Refetoff

. Identification of the maturation factor for dual oxidase. Evolution of an eukaryotic operon equivalent. J Biol Chem, 281:18269–18272. 2006.

25.

Grasberger

, Refetoff

. Genetic causes of congenital hypothyroidism due to dyshormonogenesis. Curr Opin Pediatr, 23:421–428. 2011.

26.

Harper

. The role of DUOX isozymes in the respiratory tract epithelium. Oxidants in Biology. Valacchi

, Davis

. Netherlands: Springer, 2008; 267–277.

27.

Hecker

, Vittal

, Jones

, Jagirdar

, Luckhardt

, Horowitz

, Pennathur

, Martinez

, Thannickal

. NADPH oxidase-4 mediates myofibroblast activation and fibrogenic responses to lung injury. Nat Med, 15:1077–1081. 2009.

28.

Hong

, Resnick

, Behar

, Wang

, Wands

, DeLellis

, Souza

, Spechler

, Cao

. Acid-induced p16 hypermethylation contributes to development of esophageal adenocarcinoma via activation of NADPH oxidase NOX5-S. Am J Physiol Gastrointest Liver Physiol, 299:G697–G706. 2010.

29.

Kang

, Zhang

, Chen

, Lin

, Wang

, Ou

. Nickel-induced histone hypoacetylation: the role of reactive oxygen species. Toxicol Sci, 74:279–286. 2003.

30.

Kang

, Zhang

, Kim

, Bae

, Hyun

. Epigenetic changes induced by oxidative stress in colorectal cancer cells: methylation of tumor suppressor RUNX3. Tumour Biol, 33:403–412. 2012.

31.

Kita

, Hikichi

, Hikami

, Tsuneyama

, Cui

, Osawa

, Ohnishi

, Mutoh

, Hoshino

, Bowlus

, Yamamoto

, Sugano

. Differential gene expression between flat adenoma and normal mucosa in the colon in a microarray analysis. J Gastroenterol, 41:1053–1063. 2006.

32.

Krause

. Tissue distribution and putative physiological function of NOX family NADPH oxidases. Jpn J Infect Dis, 57:S28–S29. 2004.

33.

Kuroda

, Ago

, Matsushima

, Zhai

, Schneider

, Sadoshima

. NADPH oxidase 4 (Nox4) is a major source of oxidative stress in the failing heart. Proc Natl Acad Sci U S A, 107:15565–15570. 2010.

34.

Lacroix

, Nocera

, Mian

, Caillou

, Virion

, Dupuy

, Filetti

, Bidart

, Schlumberger

. Expression of nicotinamide adenine dinucleotide phosphate oxidase flavoprotein DUOX genes and proteins in human papillary and follicular thyroid carcinomas. Thyroid, 11:1017–1023. 2001.

35.

Lassegue

, Griendling

. NADPH oxidases: functions and pathologies in the vasculature. Arterioscler Thromb Vasc Biol, 30:653–661. 2010.

36.

Lassegue

, Sorescu

, Szocs

, Yin

, Akers

, Zhang

, Grant

, Lambeth

, Griendling

. Novel gp91(phox) homologues in vascular smooth muscle cells: nox1 mediates angiotensin II-induced superoxide formation and redox-sensitive signaling pathways. Circ Res, 88:888–894. 2001.

37.

Laurent

, McCoy

3rd , Macina

, Liu

, Cheng

, Robine

, Papkoff

, Lambeth

. Nox1 is over-expressed in human colon cancers and correlates with activating mutations in K-Ras. Int J Cancer, 123:100–107. 2008.

38.

Lee

, Lee

, Lin

, Lee

, Lin

, Yang

. Activation and induction of cytosolic phospholipase A2 by IL-1beta in human tracheal smooth muscle cells: role of MAPKs/p300 and NF-kappaB. J Cell Biochem, 109:1045–1056. 2010.

39.

Liloglou

, Bediaga

, Brown

, Field

, Davies

. Epigenetic biomarkers in lung cancer. Cancer Lett, 2012. https://dx-doi-org.web.bisu.edu.cn/10.1016/j.canlet.2012.04.018.

40.

Lim

, Gu

, Kim

, Park

, Cho

, Park

, Jung

. Epigenetic changes induced by reactive oxygen species in hepatocellular carcinoma: methylation of the E-cadherin promoter. Gastroenterology, 135:2128–21402140.e1–e82008.

41.

, Qiu

, Huang

, Zou

, Hong

, Li

, Chen

, Yuan

. NADPH oxidase DUOX1 and DUOX2 but not NOX4 are independent predictors in hepatocellular carcinoma after hepatectomy. Tumour Biol, 32:1173–1182. 2011.

42.

Luxen

, Belinsky

, Knaus

. Silencing of DUOX NADPH oxidases by promoter hypermethylation in lung cancer. Cancer Res, 68:1037–1045. 2008.

43.

Luxen

, Noack

, Frausto

, Davanture

, Torbett

, Knaus

. Heterodimerization controls localization of Duox-DuoxA NADPH oxidases in airway cells. J Cell Sci, 122:1238–1247. 2009.

44.

Mandal

, Fu

, Levine

. REDOX regulation of IL-13 signaling in intestinal epithelial cells: usage of alternate pathways mediates distinct gene expression patterns. Cell Signal, 22:1485–1494. 2010.

45.

Min

, Lim

, Jung

. Downregulation of catalase by reactive oxygen species via hypermethylation of CpG island II on the catalase promoter. FEBS Lett, 584:2427–2432. 2010.

46.

Mitsushita

, Lambeth

, Kamata

. The superoxide-generating oxidase Nox1 is functionally required for Ras oncogene transformation. Cancer Res, 64:3580–3585. 2004.

47.

Morand

, Ueyama

, Tsujibe

, Saito

, Korzeniowska

, Leto

. Duox maturation factors form cell surface complexes with Duox affecting the specificity of reactive oxygen species generation. FASEB J, 23:1205–1218. 2009.

48.

Munson

, Fried

, Rowson

, Bonner

, Karumbaiah

, Diaz

, Courtneidge

, Knaus

, Brat

, Arbiser

, Bellamkonda

. Anti-invasive adjuvant therapy with imipramine blue enhances chemotherapeutic efficacy against glioma. Sci Transl Med, 4:127ra36. 2012.

49.

Nagai

, Betsuyaku

, Suzuki

, Nasuhara

, Kaga

, Kondo

, Nishimura

. Dual oxidase 1 and 2 expression in airway epithelium of smokers and patients with mild/moderate chronic obstructive pulmonary disease. Antioxid Redox Signal, 10:705–714. 2008.

50.

Niethammer

, Grabher

, Look

, Mitchison

. A tissue-scale gradient of hydrogen peroxide mediates rapid wound detection in zebrafish. Nature, 459:996–999. 2009.

51.

O'Hagan

, Wang

, Sen

, Destefano Shields

, Lee

, Zhang

, Clements

, Cai

, Van Neste

, Easwaran

, Casero

, Sears

, Baylin

. Oxidative damage targets complexes containing DNA methyltransferases, SIRT1, and polycomb members to promoter CpG Islands. Cancer Cell, 20:606–619. 2011.

52.

Ohye

, Sugawara

. Dual oxidase, hydrogen peroxide and thyroid diseases. Exp Biol Med (Maywood), 235:424–433. 2010.

53.

Pandey

, Patel

, Chen

, Qian

, Wang

, Barman

, Venema

, Stepp

, Rudic

, Fulton

. Expression and functional significance of NADPH oxidase 5 (Nox5) and its splice variants in human blood vessels. Am J Physiol Heart Circ Physiol, 302:H1919–H1928. 2012.

54.

Pelicano

, Carney

, Huang

. ROS stress in cancer cells and therapeutic implications. Drug Resist Updat, 7:97–110. 2004.

55.

Pendyala

, Gorshkova

, Usatyuk

, He

, Pennathur

, Lambeth

, Thannickal

, Natarajan

. Role of Nox4 and Nox2 in hyperoxia-induced reactive oxygen species generation and migration of human lung endothelial cells. Antioxid Redox Signal, 11:747–764. 2009.

56.

Quan

, Lim

, Jung

. Reactive oxygen species downregulate catalase expression via methylation of a CpG island in the Oct-1 promoter. FEBS Lett, 585:3436–3441. 2011.

57.

Quinlan

, Quatela

, Philips

, Settleman

. Activated Kras, but not Hras or Nras, may initiate tumors of endodermal origin via stem cell expansion. Mol Cell Biol, 28:2659–2674. 2008.

58.

Rathore

, Zheng

, Niu

, Liu

, Korde

, Ho

, Wang

. Hypoxia activates NADPH oxidase to increase [ROS]i and [Ca2+]i through the mitochondrial ROS-PKCepsilon signaling axis in pulmonary artery smooth muscle cells. Free Radic Biol Med, 45:1223–1231. 2008.

59.

Risch

, Plass

. Lung cancer epigenetics and genetics. Int J Cancer, 123:1–7. 2008.

60.

Salmeen

, Park

, Meyer

. The NADPH oxidases NOX4 and DUOX2 regulate cell cycle entry via a p53-dependent pathway. Oncogene, 29:4473–4484. 2010.

61.

Sancho

, Fabregat

. NADPH oxidase NOX1 controls autocrine growth of liver tumor cells through up-regulation of the epidermal growth factor receptor pathway. J Biol Chem, 285:24815–24824. 2010.

62.

Schilder

, Heiss

, Schachner

, Ziegler

, Reznicek

, Sorescu

, Dirsch

. NADPH oxidases 1 and 4 mediate cellular senescence induced by resveratrol in human endothelial cells. Free Radic Biol Med, 46:1598–1606. 2009.

63.

Shames

, Girard

, Gao

, Sato

, Lewis

, Shivapurkar

, Jiang

, Perou

, Kim

, Pollack

, Fong

, Lam

, Wong

, Shyr

, Nanda

, Olopade

, Gerald

, Euhus

, Shay

, Gazdar

, Minna

. A genome-wide screen for promoter methylation in lung cancer identifies novel methylation markers for multiple malignancies. PLoS Med, 3:e486. 2006.

64.

Shao

, Nadel

. Dual oxidase 1-dependent MUC5AC mucin expression in cultured human airway epithelial cells. Proc Natl Acad Sci U S A, 102:767–772. 2005.

65.

Sturrock

, Cahill

, Norman

, Huecksteadt

, Hill

, Sanders

, Karwande

, Stringham

, Bull

, Gleich

, Kennedy

, Hoidal

. Transforming growth factor-beta1 induces Nox4 NAD(P)H oxidase and reactive oxygen species-dependent proliferation in human pulmonary artery smooth muscle cells. Am J Physiol Lung Cell Mol Physiol, 290:L661–L673. 2006.

66.

Suh

, Arnold

, Lassegue

, Shi

, Xu

, Sorescu

, Chung

, Griendling

, Lambeth

. Cell transformation by the superoxide-generating oxidase Mox1. Nature, 401:79–82. 1999.

67.

Szatrowski

, Nathan

. Production of large amounts of hydrogen peroxide by human tumor cells. Cancer Res, 51:794–798. 1991.

68.

van der Vliet

. NADPH oxidases in lung biology and pathology: host defense enzymes, and more. Free Radic Biol Med, 44:938–955. 2008.

69.

Wang

, Zang

, Rosenfeld

, Schones

, Barski

, Cuddapah

, Cui

, Roh

, Peng

, Zhang

, Zhao

. Combinatorial patterns of histone acetylations and methylations in the human genome. Nat Genet, 40:897–903. 2008.

70.

Wenzel

, Mollnau

, Oelze

, Schulz

, Wickramanayake

, Muller

, Schuhmacher

, Hortmann

, Baldus

, Gori

, Brandes

, Munzel

, Daiber

. First evidence for a crosstalk between mitochondrial and NADPH oxidase-derived reactive oxygen species in nitroglycerin-triggered vascular dysfunction. Antioxid Redox Signal, 10:1435–1447. 2008.

71.

Wesley

, Bove

, Hristova

, McCarthy

, van der Vliet

. Airway epithelial cell migration and wound repair by ATP-mediated activation of dual oxidase 1. J Biol Chem, 282:3213–3220. 2007.

72.

Weyemi

, Lagente-Chevallier

, Boufraqech

, Prenois

, Courtin

, Caillou

, Talbot

, Dardalhon

, Al Ghuzlan

, Bidart

, Schlumberger

, Dupuy

. ROS-generating NADPH oxidase NOX4 is a critical mediator in oncogenic H-Ras-induced DNA damage and subsequent senescence. Oncogene, 31:1117–1129. 2012.

73.

Xiong

, Xiao

, Zhang

. Norepinephrine causes epigenetic repression of PKC{varepsilon} gene in rodent hearts by activating Nox1-dependent reactive oxygen species production. FASEB J, 26:2753–2763. 2012.

74.

, Linderholm

, Grasberger

, Harper

. DUOX2 bidirectional promoter polymorphisms confer differential immune responses in airway epithelia. Am J Respir Cell Mol Biol, 47:484–490. 2012.

75.

Yamaura

, Mitsushita

, Furuta

, Kiniwa

, Ashida

, Goto

, Shang

, Kubodera

, Kato

, Takata

, Saida

, Kamata

. NADPH oxidase 4 contributes to transformation phenotype of melanoma cells by regulating G2-M cell cycle progression. Cancer Res, 69:2647–2654. 2009.

76.

Zhang

, Lv

, Liu

, Zhu

, Su

, Wu

, Qi

, Wang

, Li

. HHMD: the human histone modification database. Nucleic Acids Res, 38:D149–D154. 2010.

77.

Zhu

, Yu

, Chen

, Ji

, Qian

, Han

. DNA hypermethylation of the NOX5 gene in fetal ventricular septal defect. Exp Ther Med, 2:1011–1015. 2011.