OUP user menu

NADPH oxidases: New kids on the block

Miklós Geiszt
DOI: http://dx.doi.org/10.1016/j.cardiores.2006.05.004 289-299 First published online: 15 July 2006

Abstract

Reactive oxygen species (ROS) play a pivotal role in many physiological processes including host defense, hormone biosynthesis, fertilization and cellular signaling. Altered production of ROS has been implicated in the development of immunodeficiency, hypothyroidism and cardiovascular pathologies. In the last few years, several enzymes were identified at the molecular level, which are now thought to be responsible for ROS production observed in diverse tissues. These enzymes show a high degree of homology to the phagocytic NADPH oxidase and are now designated the Nox family of NADPH oxidases. This review updates our knowledge on six new members of the Nox family: Nox1, Nox3, Nox4, Nox5, Duox1 and Duox2.

Keywords
  • NADPH oxidase
  • Nox
  • Duox
  • Superoxide
  • Reactive oxygen species (ROS)

1. Introduction

During the phagocytosis of foreign pathogens neutrophil granulocytes produce large quantities of ROS, which contributes to the killing of the invading microorganisms. In phagocytic cells, the ROS precursor, superoxide, is produced by the NADPH oxidase enzyme complex [1]. The phagocyte NADPH oxidase is dormant in resting cells but it becomes rapidly activated by several stimuli, including bacterial products and cytokines. The active enzyme complex catalyzes the transfer of one electron from NADPH to molecular oxygen, resulting in the formation of superoxide. The phagocytic NADPH oxidase (phox) is an enzyme complex composed of the membrane-bound cytochrome b558, three cytosolic factors (p47phox, p67phox, p40phox), and the small GTPase Rac2. During the activation of the phagocyte NADPH oxidase, cytosolic proteins translocate to the membrane initiating the production of superoxide. The molecular details of this process have been recently covered by excellent reviews [1,2], therefore it will not be discussed in this article. It was known for years that production of ROS is not restricted to phagocytic cells, but it is present in many cell types of the plant and animal kingdom [3]. The enzymatic basis of non-phagocytic ROS production was poorly understood for a long time, however, the recent discovery of gp91phox homologs revolutionized our understanding of ROS production. The new homologs along with gp91phox are now designated the Nox family of NADPH oxidases. The family has seven members, including Nox1, Nox2 (formerly known as gp91phox), Nox3, Nox4, Nox5, Duox1 and Duox2.

2 NADPH oxidase 1 (Nox1)

2.1 Molecular features and expression pattern of Nox1

Nox1 was the first homolog of gp91phox (now Nox2) to be identified [4]. Nox1 contains 564 amino acids and shows 56% identity to Nox2. Similarly to its phagocytic homolog, Nox1 contains six transmembrane domains and conserved motifs corresponding to binding sites of heme, flavin and NADPH (Fig. 1). Two alternatively spliced transcripts from the nox1 gene have been described; one of them is a genuine transcript that lacks exon 11 [5,6], but the other is a cloning artifact due to intramolecular template switching during the reverse transcriptase reaction [5,6]. Nox1 is highly expressed in the colon, but it is present in several other tissues and cells including smooth muscle, uterus, prostate, kidney, stomach and osteoclasts [4,7–9] (Table 1). As revealed by in situ hybridization experiments, Nox1 is expressed by epithelial cells of the colon [10,11]. Immunostaining experiments performed on guinea pig colon localized the Nox1 protein to the apical part of crypts [12]. Considering the short life span of colon epithelial cells, it is likely that Nox1 mRNA is synthesized during the maturation of epithelial cells and the protein functions in mature epithelial cells. We have a limited amount of information about the subcellular localization of the Nox1 protein. In vascular smooth muscle Nox1 localizes to the cell surface where it co-localizes with caveolin [13].

Fig. 1

Structure and membrane topology of Nox family NADPH oxidases. All Nox/Duox enzymes contain six membrane-spanning domains, two hemes and conserved motifs involved in NADPH and FAD binding. In addition to these structural features Nox5 contains four calcium-binding EF-hand motifs in its N-terminus, whereas Duox proteins contain an additional transmembrane domain, a peroxidase-like domain and two EF-hand motifs.

View this table:
Table 1

Tissue distribution and functions of Nox/Duox enzymes

EnzymeOther namesSite of expressionFunctionRegulators, protein interactions
Nox1Mox1, NOH-LColon, vascular smooth muscle, prostate, uterusHost defense? blood pressure regulationp22phox, NOXO1, NOXA1, Rac1
Nox2gp91phoxMyeloid cellsHost defense, signaling?p22phox, p47phox, p67phox, Rac1, Rac2
Nox3Inner ear, fetal tissuesOtoconia biosynthesisp22phox, NOXO1, Rac1
Nox4RenoxKidney, endothelium, osteoclastOxygen sensing? vasoregulation? signaling?p22phox
Nox5Lymph nodes, spleen, testisSignaling? capacitation??
Duox1 and 2Thox, Tox, LNOXThyroid, lung, salivary glands, gastrointestinal tractHormone synthesis, host defense? signaling??
  • Boldface indicates functions supported by genetic evidence.

2.2 Regulation of Nox1 activity

NIH3T3 fibroblasts transfected with Nox1 produced superoxide, suggesting that Nox1 is able to produce superoxide without the need of other regulatory proteins [4]. Other groups did not observe superoxide production in cells that heterologously expressed Nox1 [14–16]. When Nox1 was co-expressed with p47phox and p67phox, the cytosolic components of the phagocyte NADPH oxidase, Nox1-derived superoxide production was detected [14–16]. This observation suggested that in addition to the structural similarity, functional similarities also exist between the Nox2 and Nox1 systems. Searching in Genbank for potential p47phox homologs Geiszt and Leto have identified an novel homolog of p47phox, which was characterized by three groups and designated NAD(P)H oxidase organizer 1, or NOXO1 [14–16]. NOXO1 contains the functional domains of p47phox in the same configuration: an amino-terminal PX-domain, two tandem SH3-domains, and a carboxy-terminal, proline-rich motif serving as an SH3 domain binding site. A homolog of p67phox was also identified recently in colon epithelium [14–16]. Similar to p67phox, NOXA1 contains four amino-terminal TPR repeats that bind Rac1, a PB1 domain, and a single, carboxy-terminal SH3 domain that binds NOXO1. When NOXO1 and NOXA1 are co-expressed in various host cell lines that already contain Nox1, spontaneous [14,17] and PMA-induced superoxide production is observed [15,16]. In addition to these cytosolic regulators, the membrane protein p22phox also interacts with Nox1 [16,18–20]. Nox1-derived superoxide production is increased when p22phox is over-expressed in cells that already carry Nox1, NOXO1, and NOXA1 [16] and Nox1 expression promotes p22phox transport to the plasma membrane [20]. Furthermore, Nox1-based superoxide production is inhibited by p22phox mutations and p22phox specific siRNAs inhibit Nox1-based superoxide production in HEK293 cells, which endogenously express p22phox [21]. The Rac GTPase has an essential role in activation of the phagocyte NADPH oxidase [1]. The Nox1 system resembles its phagocyte counterpart in this aspect as well. In yeast two-hybrid system experiments, NOXA1 interacts with GTP-bound form of Rac1 [16]. Using Rac1 mutants and siRNA technique Ueyama et al. have recently demonstrated that Rac1 indeed supports the activity of Nox1 [20]. Helicobacter pylori lipopolysaccharide induced Nox1 activation in guinea pig gastric mucosal cells is dependent on Rac1 [22]. The findings described above illustrate the close functional similarities between Nox1- and Nox2-based NADPH oxidases.

2.3 Nox1 function in health and disease

The physiological function of Nox1 is currently unknown. It is likely that depending on the cell type where Nox1 is expressed the enzyme can serve diverse functions. The expression pattern of Nox1 and the high degree of structural and functional similarities between Nox1 and Nox2 suggest that Nox1 functions as a host defense enzyme in the colon [23]. In support with this view, treatment of guinea pig gastric pit cells with H. pylori LPS leads to Nox1 induction [8]. Furthermore, flagellin prepared from Salmonella enteridis stimulates ROS production in colon epithelial cells through TLR5 in a Nox1-dependent manner [12]. The stimulatory effect of interferon-gamma on Nox1 expression also argues for a role in mucosal immunity [11,24]. With the discovery of cytosolic regulators, it appears that Nox1 should no longer be viewed as a “low-output” oxidase and ROS produced by the colon oxidase might exert direct, microbicidal effect on intestinal pathogens. It is possible that disruption of the mucosal barrier by certain pathogens leads to local activation of the colon oxidase, which then helps to eliminate the invading pathogens. Based on the low level, Nox1-based ROS production in T84 cells, Kuwano et al. have suggested that Nox1 derived ROS function primarily as signaling molecules that enhance the production of proinflammatory cytokines [24]. ROS produced by the phagocyte NADPH oxidases have well-proven, direct microbicidal effects, but their role in signaling is also evident [25]. Such “dual” roles seem likely for the colon oxidase as well.

Originally Nox1 was described as an NADPH oxidase that stimulates mitogenesis when over-expressed in NIH 3T3 cells [4]. Experiments performed on other cell lines and colon tumor samples do not support this early observation. According to three independent studies on Nox1 expression in colon tumor samples, no positive correlation is observed between Nox1 expression levels and proliferation or malignancy [11,26,27]. Furthermore, the originally observed transforming effect of Nox1 is probably explained by accidental co-expression of an oncogenic Ras mutant [28].

Three recent studies suggest the involvement of Nox1 in the development of angiotensin II-induced hypertension [29–31]. Nox1-deficient mice show a blunted increase in blood pressure in response to angiotensin II [29,32], while smooth-muscle specific over-expression of Nox1 potentiates angiotensin II-induced hypertension [30]. Gavazzi et al. found decreased basal blood pressure in Nox1-deficient animals, while animals created by Matsuno et al. have normal basal blood pressure. Angiotensin II treatment induces increased oxidative stress in the vascular wall and this effect is reduced in Nox1-deficient mice. Angiotensin II induces vascular hypertrophy that is linked to development of higher blood pressure. Matsuno et al. found no effect of Nox1 deficiency on the development of angiotensin II-induced media hypertrophy, while in the experiments by Gavazzi et al., Nox1-deficient animals show a marked reduction in aortic media hypertrophy. The reason of this discrepancy is unclear, however different experimental protocols (age of animals, dose of angiotensin II) might be responsible. A role for angiotensin II in media hypertrophy response seems likely since transgenic mice over-expressing Nox1 in smooth muscle show increased media hypertrophy in response to angiotensin II [33]. Based on these studies, a role for Nox1 in the regulation of blood pressure seems likely, but several questions regarding the exact nature of this function remain to be answered. First of all, carefully controlled experiments should demonstrate that angiotensin II-induced increase in Nox1 expression in the vascular wall indeed occurs in smooth muscle cells. The results of experiments where Nox1 expression in transgenic mice is driven by the myosin heavy chain promoter should be interpreted carefully, since over-expression of a heme-containing enzyme at such high levels might have unwanted side effects on cellular signaling. Furthermore, earlier experiments studying Nox1 expression by sensitive PCR did not detect Nox1 expression in tissues with rich vascular supply [10], thus Nox1 is unlikely expressed in blood vessels in situ under physiological conditions. It is now widely accepted that Nox1 cannot produce ROS alone but requires additional proteins for its enzymatic activity [23]. The expression of p22phox and p47phox is well documented in vascular smooth muscle cells, but at this time no data support the expression of NOXO1 and NOXA1, the likely physiological regulators of Nox1. Since cytosolic proteins of the phagocyte NADPH oxidase can also support the enzymatic function of Nox1 [14–16], it is possible that these proteins represent physiological regulators of Nox1 in vascular smooth muscle cells.

Restenosis is a frequent complication of coronary angioplasty characterized by increased neointimal proliferation and elevated vascular ROS production. A Nox1 based NADPH oxidase seems to have an important role in this process since the expression of Nox1 and p22phox mRNAs were found to be increased in balloon-injured carotid arteries [34].

3 Nox3

3.1 Molecular features and expression pattern of Nox3

Nox3 is composed of 568 amino acids and shows 58% sequence identity to Nox2 [10] (Fig. 1). Originally, the Nox3 mRNA was not detected in adult tissues, but it was found in fetal tissues including lung, liver, kidney, and spleen [10,35]. The primary expression site for Nox3 is the inner ear, which was discovered through the genotypic analysis of the head tilt (het) mice that show balance defects due to mutation in the nox3 gene [36]. Sequence data in EST databases also pointed to the inner ear as a site of Nox3 expression, which was confirmed by RT-PCR analysis and in situ hybridization [37]. Within the inner ear, cochlear and vestibular sensory epithelia and the spiral ganglion express Nox3 mRNA [37], but the protein product has not yet been detected. The intracellular localization of Nox3 is currently not known.

3.2 Regulation of Nox3 activity

Nox3 forms a functional complex with p22phox, since Nox3 physically interacts with p22phox [38] and superoxide production of Nox3-transfected cells is dependent on the presence of p22phox [38]. In contrast to Nox1 and Nox2, Nox3 seems to be constitutively active, although cytosolic proteins can further stimulate its enzymatic activity [20,37–39]. Cytosolic components of the phagocyte oxidase effectively stimulate the enzymatic activity of Nox3 [20,37–39], but their physiological role is questionable, since p67phox is not detected in the inner ear [37]. Several groups have reported a stimulatory effect of NOXO1 on Nox3-dependent superoxide production [20,37–39]. According to the experiments of Bánfi et al., murine NOXO1 supports Nox3 only if it is co-expressed with NOXA1 [37]. On the contrary, three groups demonstrated that NOXO1 increased the activity of Nox3 in the absence of NOXA1 [20,38,39]. NOXO1 is a likely regulator of Nox3, since mutations in NOXO1 gene also lead to similar balance defects as those observed in the het mice [40,41]. Similarly to its role in the regulation of Nox1 and Nox2, Rac1 also seems to regulate the activity of Nox3 [20].

3.3 Nox3 function in health and disease

Genetic evidence suggests that Nox3 is essential for normal vestibular function. In the absence of Nox3 serious balance defects develops and otoconia are absent in the inner ear of the mutant animals [36]. Currently it is unclear how Nox3 would facilitate the formation of these structures. One possibility is that Nox3-derived ROS cross-link extracellular proteins through the formation of dityrosine bridges, leading to the formation of a protein precipitate, which later serves as nucleus in the calcification of otoconia. Bánfi et al. hypothesized that Nox3-derived ROS might contribute to the development of hearing loss in response to ototoxic drugs, like cisplatin [37].

4 Nox 4

4.1 Molecular features and expression pattern of Nox4

Nox 4 was originally described as Renox [42], which stands for renal oxidase, since Nox4 is most abundantly expressed in the kidney [42,43] (Table 1). Nox4 is a 578-amino acid protein with 39% homology to Nox2 (Fig. 1). In murine kidney, in situ hybridization experiments localized Nox4 mRNA expression to the renal cortex, where epithelial cells of proximal tubules showed high-level expression. We have observed a different expression pattern in human kidney, where Nox4 mRNA localized to medullary collecting ducts and also in epithelium on renal papillae (Geiszt et al., unpublished observations). Immunohistochemical studies also showed Nox4 expression in distal tubules of the human nephron [43]. Although glomeruli express relatively low Nox4 mRNA levels in comparison to other renal structures [42], Gorin et al. detected Nox4 mRNA in rat mesangial cells [44]. Nox4 mRNA was also found in many other tissues and cells including fetal liver, vascular endothelial cells, smooth muscle cells, murine osteoclasts, hematopoietic stem cells and adipocytes [42,45–47]. Using three different polyclonal, Nox4-specific antibodies, Kuroda et al. localized the Nox4 protein to the nucleus of human umbilical endothelial cells (HUVECs) [48]. In this work, the authors used the siRNA technique to demonstrate the specificity of their antibodies. Such control experiments should become a general practice that would enhance the credibility of reports exploring the subcellular localization of Nox proteins. In blood vessels, Nox4 is also present in smooth muscle cells, where it localizes to the endoplasmic reticulum and nucleus [13,49]. The intracellular localization of the Nox4 protein in kidney epithelial cells is currently unknown.

4.2 Regulation of Nox4 activity

The enzymatic activity of Nox4 was first demonstrated in experiments where Geiszt et al. detected constitutive superoxide production in Nox4-transfected NIH 3T3 fibroblasts [42]. Interestingly, HEK293 cells expressing Nox4 produce large amounts of hydrogen peroxide, but the authors did not detect superoxide release from the transfected cells [50]. This finding is probably explained by the localization of Nox4 to intracellular compartments, since superoxide produced intracellularly would dismutate into hydrogen peroxide, which would then diffuse readily to the extracellular space.

We have little information about the regulatory factors affecting Nox4 activity. Nox4 forms a molecular complex with p22phox [18,50] and the ROS production of Nox4-transfected cells is dependent on the expression of p22phox. Known cytosolic proteins did not affect the enzymatic activity of Nox4 [50]; furthermore, co-expression of mutant p22phox that does not bind the known Nox organizers did not diminish Nox4 activity [51]. These observations suggest that the Nox4-p22phox complex functions alone, but we cannot exclude the possibility of cooperation with some unknown endogenous proteins. Alterations in the expression level of the nox4 gene appear to provide an effective means for the regulation of Nox4-based ROS production. Hypoxia, for example, was shown to stimulate Nox4 expression in the murine kidney [52] and angiotensin II increased Nox4 mRNA level in A7r5 cells [53].

4.3 Functions of Nox4 in health and disease

The physiological function of Nox4 is currently unknown. It is perhaps not wise to pursue one specific role, since it appears that the function Nox4 may depend on its expression site. Nevertheless, the high expression level in the kidney suggests that the protein has some unique role in this organ. The expression pattern of Nox4 in the kidney is consistent with several possible, kidney-specific functions. We originally suggested that Nox4 functions as an oxygen sensor, which would regulate erythropoietin (EPO) synthesis in the kidney. In fact, Nox4 would be an ideally located sensor to regulate this process, since EPO synthesis in the murine kidney occurs in the proximal tubules [54] or in the close proximity of proximal tubules [55]. Currently we have no direct evidence to support this role, but several recent observations suggest that Nox4 might have a role in oxygen sensing [56–58].

ROS have an important role in the pathogenesis of diabetic nephropathy, but the specific sources of ROS were not identified. Two recent studies implicate Nox4 in this process. Etoh et al. showed increased expression of Nox4 and p22phox in the kidney of streptozotocin-induced diabetic rats [59]. They also showed that Nox4 and p22phox co-localized with 8-hydroy-deoxyguanosine (8-OHdG), which is a marker for ROS-induced DNA damage. A causative relationship between Nox4-derived ROS and diabetic nephropathy was shown by Gorin et al., who used antisense oligonucleotides to inhibit Nox4 expression [60]; this treatment effectively reduced ROS production and prevented the development of hypertrophy and increases in fibronectin expression. Nox4-derived ROS is also important in the development of mesangial cell hypertrophy in response to angiotensin II [44]. The authors proposed a model where Rac1 has a role in the signaling process by activating Nox4 in response to receptor-activated arachidonic acid release. This model requires further experimental evidence, since others found no effect of Rac1 on Nox4-derived ROS production [50].

Nox4 may have important roles in the cardiovascular system. According to expression studies performed on vascular endothelial cells, Nox4 seems to be the dominant ROS source in endothelial cells [61,62]. This localization is particularly exciting because it would mean that Nox4-produced superoxide could effectively antagonize the effect of nitric oxide (NO), produced in the same cell type. A recent publication [63] has located endogenous Nox4 to the nucleus of human umbilical endothelial cells (HUVECs). The nuclear fraction of HUVEC cells produced superoxide in an NADPH-dependent manner. Earlier reports already pointed to the nucleus as an intracellular site of ROS production in endothelial cells [64] and NADPH oxidase components were detected in nuclear fraction of endothelial cells [64] and B lymphocytes [65]. The nuclear localization of Nox4 suggests that it might regulate gene expression through production ROS. This suggestion is now supported by some experimental evidence [48], but the idea requires more rigorous testing. In blood vessels, Nox4 is also present in smooth muscle cells, where it localizes to the endoplasmic reticulum and nucleus [13,49]. 7-Ketocholesterol, a major oxysterol component in LDL, stimulates the expression of Nox4 in smooth muscle cells. ROS production by Nox4 may be responsible for the oxidative stress induced by 7-ketocholesterol [49]. Nox4 is also present in the heart, where cardiac fibroblasts express the enzyme [33]. TGF-beta stimulates the conversion of cardiac fibroblasts into myofibroblasts in a ROS-dependent manner. Nox4 is the likely source of oxidants in this process, since downregulation of Nox4 expression by siRNA inhibited both ROS production and the TGF-beta induced expression of smooth muscle actin [33]. The involvement of Nox4 in TGF-beta signaling was recently described in human pulmonary artery smooth muscle cells and in HUVECs as well [66,67].

Nox4 was suggested as a participant in insulin receptor signal transduction [68]. In many cells growth factors and insulin stimulates low-level hydrogen peroxide production. Hydrogen peroxide then inhibits tyrosine phosphatases, thus enhancing the tyrosine phosphorylation induced by the receptor agonists [69]. Recently it was shown that Nox4 would be involved in insulin-induced H2O2 production in 3T3-L1 adipocytes [68].

5 Nox5

5.1 Molecular features and expression pattern of Nox5

NADPH oxidase 5 (Nox5) was originally identified by Cheng et al. [35], who described a cDNA encoding a 565 amino acid protein with 27% identical to Nox2. Bánfi et al. have identified other products of the nox5 gene, Nox5α, β, γ and δ, which are larger proteins containing more than 700 amino acids [70]. These larger Nox5 isoforms contain a long intracellular N-terminus with Ca2+-binding EF-hand domains (Fig. 1). According to Northern-blot analysis Nox5 is mainly expressed in spleen, and testis and in fetal tissues [35,70] (Table 1). In RT-PCR experiments Nox5 mRNA was detected in several other tissues and cells including ovary, placenta, pancreas, vascular smooth muscle, bone marrow and uterus [35]. In situ hybridization data suggest that in testis pachytene spermatocytes express Nox5 [70]. In spleen Nox5 is expressed in the mantle zone surrounding the germinal centers and also in periarterial lymphoid sheaths [70]. These areas are rich in mature B-lymphocytes and T-lymphocytes, respectively. While other Nox/duox proteins are found in mice, rat and man, the mouse and rat genome does not contain the nox5 gene.

5.2 Regulation of Nox5 activity

The enzymatic activity of Nox5 was first demonstrated by Bánfi et al. who showed calcium-dependent ROS production in Nox5-transfected cell lines [70]. Calcium ions activate Nox5 by binding to the N-terminal EF-hand motifs (Fig. 1). Calcium binding changes the conformation of this domain enabling an intramolecular interaction between the N-terminal and C-terminal domains [71]. The regulation of Nox5 is probably not dependent on the presence of cytosolic regulators, although this question requires further investigation.

5.3 Nox5 function in health and disease

The physiological function of Nox5 is currently unknown. Several data support a key role for ROS in lymphocyte signaling [72]. The fact that Nox5 is the first NADPH oxidase, which is primarily found in lymphoid tissues, would suggest a role for Nox5 in lymphocyte signaling. On other hand in several lymphocytes subsets where stimulus induced ROS production is observed, Nox5 is not expressed [70]. Furthermore nox5 gene is not present in the mouse genome, arguing against a general role in lymphocyte signaling. Recent data suggest that under certain conditions Nox5 may have a role in signaling [73]. In hairy cell, which are mature malignant B cells, Nox5 seems to be the primary source of ROS and it regulates SHP-1 tyrosine phosphatase activity. The expression of Nox5 in developing spermatocytes is a very interesting observation since ROS has been implicated in the process of fertilization for many years in sea urchin eggs [74]. In sea urchins hydrogen peroxide is produced by the oocyte and it is used for the stabilization of the fertilization envelope by ovoperoxidase [74]. Since Nox5 is expressed in male gametocytes a similar role seems unlikely; on the other hand, Nox5 may have a regulatory role in spermatocytes.

6 Dual oxidases (Duox1 and Duox2)

6.1 The structure and expression pattern of Duox proteins

Dual oxidases were originally described as thyroid oxidases because they were first described in thyroid gland [75,76]. Edens et al. [77] reported the cloning of homologous sequences from C. elegans and suggested the Duox (Dual oxidase) nomenclature, based on the structural features of the proteins. Human Duox1 and Duox2 proteins contain 1551 and 1548 amino acids, respectively, and show 83% sequence similarity. The NADPH oxidase portion of the protein shows 47% similarity to gp91phox (Nox2). Duox proteins also have N-terminal peroxidase-homology domains, which show high similarity to other peroxidases (Fig. 1). Peroxidases are heme-containing proteins, however, the peroxidase-homology domains of Duox proteins probably do not bind heme, because they lack key amino acid residues essential for heme-binding that are present in highly conserved positions in all other peroxidases, including myeloperoxidase, lactoperoxidase and thyroperoxidase [78]. Between the peroxidase-homology and the NADPH oxidase domains are two EF-hand motifs, suggesting that calcium ions regulate their enzymatic activity.

Both Duox1 and Duox2 are expressed in the thyroid gland (Table 1). Immunostaining experiments located the proteins to the apical pole of thyrocytes [76]. Duox1 in also expressed in lung, pancreas, placenta, prostate and testis [79,80]. In situ hybridization and immunohistochemical experiments located Duox1 within the epithelial cells of airways [80,81]. The expression of Duox2 also seems to be more widespread as originally suggested. Beside the thyroid gland Duox2 is found in the gastrointestinal system including salivary glands, stomach, duodenum, colon and rectum [80,82,83]. In salivary glands terminal ducts show high expression of Duox2, while in the intestine epithelial cells express the protein, which is localized to the apical pole of the cells. Duox2 is also present in the airway epithelium [84].

6.2 Regulation of Duox activity

The presence of EF-hands in Duox proteins explains the stimulatory effect of calcium ions on ROS production in thyroid and airway epithelial cells [80,85,86]. Treatment of human bronchial epithelial cells with Duox1-specific antisense oligonucleotides suppresses calcium-induced ROS production thus proving that stimulated ROS production indeed reflects the enzymatic activity of Duox1 [80]. Beside the stimulatory effect of calcium, little is known about the regulation of Duox enzymes. A major obstacle in the investigation of Duox function was the absence of a cell model where the activity of heterologously expressed Duox enzymes could be studied [87]. Recently Ameziane-El-Hassani et al. have successfully reconstituted the enzymatic activity of Duox1 and Duox2 in HEK-293 cells and measured calcium-induced H2O2 generation in the particulate fraction of the cells [88]. An interesting feature of Duox expressing cells that H2O2 rather than superoxide is produced by them [74,80,85,86]. The explanation for this feature is unknown, however recent data suggest that post-translational modifications during the maturation process could be responsible for this unusual enzymatic activity [88].

Nox proteins, including Nox1, Nox2, Nox3 and Nox4, were shown to interact with p22phox, and in the case of Nox2, complex formation is an absolute requirement for the NADPH oxidase activity [1,16,18]. Duox proteins were shown to co-immunoprecipitate with p22phox, however this interaction does not seem to be essential for their enzymatic activity since (i) co-expression of p22phox does not affect the calcium-stimulated H2O2-producing activity in HEK-293 cells [88], (ii) C. elegans and Drosophila melanogaster express Duox enzymes, but their genomes do not appear to encode homologs of p22phox (Geiszt et al., unpublished observation), (iii) p22phox-deficient CGD patients do not have hypothyroidism.

Changes in gene expression levels probably also modify the ROS output of Duox-expressing cells. In dog thyrocytes, increasing the level of cAMP by forskolin stimulated the expression of Duox enzymes, but human thyroid cells did not respond to forskolin [76]. Human Duox1 and Duox2 promoters were shown to be active in differentiated thyroid cell lines (PC-Cl3 and FRTL5), however their activity was also observed in other cell types including Hela cells and Rat-2 fibroblasts [89]. In respiratory tract epithelial cells, Harper et al. detected increased Duox1 expression in response to Th2 cytokines, including IL-4 and IL-13, whereas the Th-1 cytokine IFN-gamma induced Duox2 expression [84].

6.3 Duox function in health and disease

In the thyroid gland Dual oxidases produce hydrogen peroxide, which is then utilized in the thyroperoxidase-mediated oxidation of iodide into reactive compounds. Duox proteins are present in the apical poles of thyroid cells exposed to the colloid of thyroid follicles, where they co-localize with thyroperoxidase [76]. This localization is consistent with their suggested role in hormone biosynthesis. Both Duox isoforms are present in the thyroid [75,76] but the reason of this apparent redundancy is currently unknown.

The physiological role for Duox2 in thyroid hormone biosynthesis was demonstrated by the identification of patients who have hypothyroidism due to mutations in the Duox2 gene. Moreno et al. [90], have identified one patient with a homozygous nonsense mutation in the Duox2 gene, deleting all functional domains of the protein. Three other patients had milder form of hypothyroidism caused by heterozygous mutations in the Duox2 gene [90]. The fact that Duox2 deficiencies result in hypothyroidism suggests that Duox1 and Duox2 each have non-redundant roles in hormone synthesis.

In lower species such as C. elegans and sea urchins Duox proteins participate in the modification of the extracellular matrix through cross-linking of proteins [79,91]. Cross-linking of proteins occurs through the formation of di- and trityrosine linkages. Worms injected with Duox-specific double-stranded RNA displayed phenotypes including blisters and other morphological defects resulting from defective cuticle biosynthesis [79]. The Duox-deficient animals had lower dityrosine content, proving a role for Duox in dityrosine formation.

During the fertilization of sea urchin eggs large amount of hydrogen peroxide is produced in a burst-like manner [74]. Hydrogen peroxide is then used by ovoperoxidase for stabilization of the fertilization envelope through the formation of dityrosine bridges. This mechanism helps to prevent polyspermy and provides a protective shell during embryogenesis. Wong et al. (2004) identified Udx1, a Duox isozyme in sea urchins S. purpuratus and L. variegatus [91]. Microinjection of an antibody, raised against the NADPH-binding region of Udx1, suppressed calcium ionophore-induced hydrogen peroxide production, suggesting that Udx1 is responsible for the respiratory burst of fertilization.

Dual oxidases might have a role in cellular signaling. Duox1-dependent expression of MUC5AC was observed in human airway epithelial cells and the authors suggested a role for protein kinase C in the regulation of Duox activity [92]. This study did not examine the role of calcium signal in Duox activation although calcium seems to be an essential regulator of enzyme activity. A study on lymphocyte signaling suggested that Duox1-derived ROS play a role in the amplification of early, B-cell receptor derived signals [93]. Although the authors suggest an impressive model for the role of ROS in B-cell signaling, the involvement of Duox1 in this process is less certain, since they detect Duox1 only in a leukaemic cell line, while tissues that are rich in B cells have tested negative for Duox1 expression [77].

Through the production of ROS, Duox proteins also have a role in host defense. This function was recently demonstrated in D. melanogaster, where inhibition of Duox function by RNA interference resulted in an impaired ability to eliminate bacteria [94]. Although no similar direct evidence exists in mammals, it is possible that Dual oxidases serve a role in host defense in mammals as well. Duox enzymes are present in epithelial cells of salivary gland ducts and along mucosal surfaces of the gastrointestinal tract, and major airways [80,83,86]. Based on this expression pattern Geiszt et al. have proposed a model where lactoperoxidase, abundant in milk, saliva, and bronchial secretions, uses H2O2 produced by Duox to oxidize thiocyanate to hypothiocyanite, which then serves as a potent oxidant effective against microbes [95,96]. Thus, the dual oxidases may represent the “missing arm” that would complete the lactoperoxidase-based microbicidal system that has been long recognized in many body fluids [97].

7 Conclusion

The discovery of novel NADPH oxidase isoforms represents an important advance in ROS research. Now it is clear that the expression of NADPH oxidases is more widespread than it was thought previously. The expression of these enzymes at sites where ROS production was not studied before expands our knowledge on the importance of ROS signaling. Genetic evidence now shows, that beside host defense, Nox/Duox enzymes are involved in thyroid hormone biosynthesis (Duox2), otoconia formation in the inner ear (Nox3) and regulation of blood pressure (Nox1). The physiological function of other Nox/Duox enzymes remains to be identified. Future experiments should explore the regulation of these novel enzymes and their role in disease pathogenesis.

Acknowledgements

Experimental work in the author's laboratory was financially supported by grants from the Hungarian Research Fund (OTKA 042573) and the Cystic Fibrosis Foundation. Miklos Geiszt is recipient of a Wellcome Trust International Senior Fellowship.

Footnotes

  • Time for primary review 25 days

References

  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
  36. [36]
  37. [37]
  38. [38]
  39. [39]
  40. [40]
  41. [41]
  42. [42]
  43. [43]
  44. [44]
  45. [45]
  46. [46]
  47. [47]
  48. [48]
  49. [49]
  50. [50]
  51. [51]
  52. [52]
  53. [53]
  54. [54]
  55. [55]
  56. [56]
  57. [57]
  58. [58]
  59. [59]
  60. [60]
  61. [61]
  62. [62]
  63. [63]
  64. [64]
  65. [65]
  66. [66]
  67. [67]
  68. [68]
  69. [69]
  70. [70]
  71. [71]
  72. [72]
  73. [73]
  74. [74]
  75. [75]
  76. [76]
  77. [77]
  78. [78]
  79. [79]
  80. [80]
  81. [81]
  82. [82]
  83. [83]
  84. [84]
  85. [85]
  86. [86]
  87. [87]
  88. [88]
  89. [89]
  90. [90]
  91. [91]
  92. [92]
  93. [93]
  94. [94]
  95. [95]
  96. [96]
  97. [97]
View Abstract