Antioxidant and cytoprotective responses to redox stress

13

Biochem. Soc. Symp. 71, 157–176 (Printed in Great Britain) © 2004 The Biochemical Society

Antioxidant and cytoprotective responses to redox stress Joanne Mathers*1, Jennifer A. Fraser*1, Michael McMahon*, Robert D.C. Saunders†, John D. Hayes* and Lesley I. McLellan*2 *Biomedical Research Centre, University of Dundee, Ninewells Hospital and Medical School, Dundee DD1 9SY, U.K., and †Department of Biological Sciences, The Open University, Walton Hall, Milton Keynes MK7 6AA, U.K.

Abstract Aerobic cells produce reactive oxygen species as a consequence of normal cellular metabolism, and an array of antioxidant systems are in place to maintain the redox balance. When the redox equilibrium of the cell is upset by pro-oxidant environmental stimuli, adaptive responses to the redox stress take place, which can result in up-regulation of antioxidant proteins and detoxification enzymes. Over the past few years, it has become apparent that members of the CNC (cap ‘n’ collar)-basic leucine zipper family of transcription factors are principal mediators of defensive responses to redox stress. In mammals, the CNC family members nuclear factor-erythroid 2 p45-related factors 1 and 2 (Nrf1 and Nrf2) have been shown to be involved in the transcriptional up-regulation of cytoprotective genes including those encoding glutamate cysteine ligase, NAD(P)H:quinone oxidoreductase, glutathione S-transferases and aldo–keto reductases. An evolutionarily conserved system exists in Caenorhabditis elegans, and it is possible that Drosophila melanogaster may also utilize CNC transcription factors to induce antioxidant genes in response to pro-oxidant chemicals. The advent of microarray and proteomic technologies has advanced our understanding of the gene batteries regulated by oxidative insult, but has highlighted the complexity of gene regulation by environmental factors. This review focuses on the antioxidant response to environmental stress, and the impact that microarrays and proteomics have made in this field.

1These 2To

authors contributed equally to this chapter. whom correspondence should be addressed (e-mail [email protected]). 157

158

J. Mathers et al.

Introduction ROS (reactive oxygen species) are formed as by-products from normal cellular metabolism, and the redox balance is maintained by a range of intracellular antioxidants. ROS are also produced in response to environmental stimuli such as UV light, -rays, pathogens and drugs. In addition, extracellular signalling molecules such as the cytokines interleukin 1 and tumour necrosis factor generate ROS [1]. The importance of ROS as subcellular messengers is becoming increasingly apparent. Changes in the redox status of the intracellular environment can activate multiple signalling pathways, altering gene transcription, the cell cycle and apoptosis. It is, however, unclear how specificity is achieved in redox-dependent signalling, and the exact mechanisms by which oxidants regulate this process are only just starting to be elucidated. The role of ROS in cell signalling pathways is the subject of several excellent recent reviews [1–6], and here we will concentrate primarily on the gene-regulatory effects of thiol-reactive chemicals, which disturb the antioxidant status of the cell.

ROS, thiol-reactive agents and cellular antioxidants It is not within the scope of this article to give a full description of all antioxidant systems and types of ROS. This is covered comprehensively by Halliwell and Gutteridge [7], but a brief outline of some key points is given below. Oxygen metabolism in aerobic organisms is associated with the generation of numerous ROS that can cause extensive damage to major cellular constituents such as membrane lipids, DNA and proteins [7]. ROS encompass a variety of diverse chemical species including superoxide (O2• ), the extremely unstable and reactive hydroxyl radical (OH•) and the relatively long-lived H2O2. The majority of intracellular ROS production occurs in the mitochondria, and it has been estimated that this organelle converts 1–2% of O2 molecules consumed into superoxide anions [5]. Superoxide can also be generated by NADPH oxidase and cytochrome P450 reductase; the former is attracting increasing interest for its role in cell signalling [1,2,8]. Superoxide dismutases (SODs) convert O2• to H2O2, which in turn can be reduced to water by several different antioxidant systems (discussed below). Alternatively, H2O2 can generate highly reactive and damaging oxidants such as hydroxyl radical (in the presence of transition metals) or ferryl iron and protein-bound free radicals (in the presence of ferric haem proteins). Normally, endogenously produced ROS are largely scavenged by a diverse and multifaceted antioxidant defence system that functions to eliminate oxidative insult. SODs catalyse the dismutation of O2• to H2O2 and oxygen [7]. The SODs use various metal ions at their active sites in the dismutation of O2• . As their name suggests, the copper–zinc SOD isoenzymes (CuZnSOD), including the predominantly cytoplasmic homodimeric SOD1 and homotetrameric extracellular SOD (EC-SOD or SOD3), have a copper and a zinc atom in each active site. Whereas the copper acts directly in the dismutation of

© 2004 The Biochemical Society

Antioxidant and cytoprotective responses to redox stress

159

O2• by undergoing alternate oxidation and reduction, the zinc functions to stabilize the enzyme. Another type of SOD has manganese at the active site. Homotetrameric manganese SOD (MnSOD or SOD2) is found predominantly in the mitochondria and catalyses the same reaction as CuZnSODs. The importance of MnSOD in normal mitochondrial function is illustrated by the fact that mice with a targeted deletion of MnSOD die within 3 weeks post-partum, and show severe mitochondrial damage in cardiac tissue [9]. H2O2 produced by the SODs can be detoxified by catalases or peroxidases. In mammals, catalase activity appears to be usually confined to the peroxisomes, suggesting that it is unlikely that catalase contributes to the detoxification of H2O2 generated at extra-peroxisomal sites such as the mitochondria [7]. Thiol-dependent peroxidases play a major role in H2O2 and organic peroxide reduction. These enzymes include the selenium-dependent glutathione peroxidase family and the Prxs (peroxiredoxins) [10,11]. Certain members of the GST (glutathione S-transferase) superfamily of enzymes can also reduce organic peroxides, although the major detoxification role for GST is in the conjugative metabolism of electrophilic agents [10]. Glutathione is the most abundant non-protein thiol in most aerobic cells and makes a major contribution to the redox status of the cell. It is an important scavenger of free radicals, as well as serving as a substrate for glutathione peroxidases and GST. Treatment of cells with thiol-reactive agents [e.g. DEM (diethyl maleate)] depletes glutathione and upsets the redox balance. This results in oxidative stress (defined by Sies [12] as a “disturbance in the pro-oxidant/antioxidant balance in favour of the former, leading to potential damage”), although not necessarily the same kind of oxidative stress experienced by cells subjected to an increase in superoxide or H2O2. It is worth noting here that agents such as DEM can also react with unprotonated cysteine residues in certain proteins (e.g. [13]). This has the potential to generate a kind of cellular stress distinct from that of the redox balance, and generates a layer of complexity in deciphering the mechanisms underlying the adaptive responses to thiol-reactive agents.

Adaptive responses to oxidative stress Adaptive responses to oxidative stress are complex and differ depending on cell type and the nature of the stress. Important cellular factors in determining the response to pro-oxidant stimuli will be the presence or absence of redox sensors, signalling relay systems or transcription factors that mediate responses. The antioxidant status of the cell, as well as the levels of detoxification or metabolic enzymes, are also likely to have an impact on the response to stress. The type of oxidant will influence the response, as will the dose and duration of stimulus. Oxidative stimuli impact on phosphorylation-dependent signalling pathways, and can modulate the activity of redox-sensitive transcription factors such as AP-1 (activator protein 1) or nuclear factor B [5]. Although the mechanisms by which alterations in the redox equilibrium stimulate different © 2004 The Biochemical Society

160

J. Mathers et al.

cellular responses are not precisely understood, recent evidence suggests that oxidants can impinge on signalling by altering enzyme activity through direct oxidation or by influencing redox-dependent protein–protein interactions [1]. The former is exemplified by the finding that protein phosphatases which contain a reactive cysteine at the catalytic site (protein tyrosine phosphatases or dual-specificity phosphatases) can be oxidized and inactivated [14–17]. Redoxdependent inactivation of phosphatases has the potential to impact on downstream phosphorylation events, and it has recently been shown that oxidative stress activates PI 3-kinase (phosphoinositide 3-kinase)-dependent signalling by the inactivation of the protein tyrosine phosphatase PTEN (phosphatase and tensin homologue) [18]. The redox-dependent regulation of protein–protein interactions also appears to have an important role in the cellular response to oxidative stress [6]. For example, ASK (apoptosis signal-regulating kinase) 1, which activates JNK (c-Jun N-terminal protein kinase) and p38 signalling cascades, can be inhibited through its interaction with thioredoxin [19]. The interaction between thioredoxin and ASK1 was found to be dependent on the redox status of thioredoxin, where thioredoxin only bound to ASK1 under reducing conditions. The model proposed is that oxidative stress can induce dissociation of thioredoxin from ASK1, which would permit activation of ASK1-dependent signalling pathways. GSTP1 has also been shown to be involved in the regulation of JNK activity by redox-dependent protein–protein interactions. It was shown that association of GSTP1 with JNK occurred in non-stressed cells and inhibited JNK activity [20]. Treatment of cells with H2O2 caused dissociation of GSTP1 and activation of JNK activity. More recently, it has been shown that the nuclear localization of the Yap1 transcription factor in Saccharomyces cerevisiae is regulated by its interaction with a thioredoxin peroxidase, Gpx3 [21,22]. In S. cerevisiae, Yap1 is a critical mediator of the adaptive response to H2O2, and regulates the expression of stress-response genes, which include those encoding antioxidant enzymes. Yap1 contains a nuclear export signal, which is sterically obstructed under conditions of oxidative stress due to conformational changes in the protein caused by disulphide bond formation. This causes accumulation of Yap1 in the nucleus. Delaunay et al. [21] showed that the formation of the disulphide bond is mediated by Gpx3 in response to H2O2 treatment. It is unclear whether similar regulatory mechanisms could occur in mammalian cells. As discussed earlier, cells can respond to oxidative stress in different ways. Of particular interest to us is the adaptation to redox stress that involves the induction of antioxidant and detoxification enzymes. This antioxidant response to sub-lethal stress is relevant to cancer chemoprevention, as well as prevention of other degenerative diseases associated with aging [23,24]. The principal players in the transcriptional up-regulation of cytoprotective genes in response to environmental stress are members of the CNC (cap ‘n’ collar) family of transcription factors, and the cytoplasmic regulator Keap1.



161

The role of the ARE (antioxidant-responsive element) and CNC transcription factors in regulation of antioxidant and detoxification genes The ARE A significant contribution to cell defence against oxidative insult is mediated through transcriptional activation of antioxidant and detoxification genes via a cis-acting enhancer known as the ARE. This element was discovered by Pickett and co-workers [25] more than a decade ago, and since then ARE consensus sequences have been identified in the 5 flanking regions of many antioxidant and cytoprotective genes [23]. The binding consensus sequence of the ARE has been elucidated (5-RTGAYNNNGCR-3) and comparison with other sequences reveals that it shares similarity with the DNA-recognition sequences of NF-E2 (nuclear factor-erythroid 2; 5-RTGASTCAGCR-3) and v-Maf (5-TGCTGACTCAGCA-3) [23]. NF-E2 belongs to the CNC bZIP (basic leucine zipper) family of transcription factors which includes NF-E2 p45-related factors 1, 2 and 3 (Nrf1, Nrf 2 and Nrf3) as well as the more distantly related transcriptional repressors Bach1 and Bach2 (reviewed in [23,26]). Unlike NF-E2, which is expressed specifically in haematopoietic tissues, Nrf1 and Nrf2 expression is more ubiquitous and both are widely expressed in many tissues [27]. The widespread tissue-distribution profile of Nrf1 and Nrf2, and the striking similarity of the NF-E2 and v-Maf DNA-binding motif with the ARE led to the hypothesis that CNC transcription factors may regulate antioxidant and cytoprotective gene transcription through the ARE [23]. The majority of studies have implicated Nrf2 as the key mediator of ARE-dependent activation of antioxidant gene expression. Transient transfection of Nrf2 into HepG2, RL34 or Hepa1c1c7 cells induces reporter gene activity from constructs derived from the human and mouse NQO [NAD(P)H:quinone oxidoreductase] 1 gene promoters [28,29]. By contrast, expression of dominant-negative Nrf2 inhibits endogenous Nrf2 function and blocks both basal and tBHQ (t-butylhydroquinone)-inducible activity of NQO1 ARE-driven reporter activity in astrocytes [30]. Moreover, Nrf2 binding to ARE-containing promoters from a variety of antioxidant genes has been demonstrated [29,31,32]. Unequivocal evidence for the role of Nrf2 in regulation of cellular protection against oxidative stress has come from targeted disruption of the mouse Nrf2 gene. While mice lacking Nrf2 are viable and fertile, Western blot analyses of hepatic and intestinal tissues from these mice showed diminished levels of NQO1 and Alpha and Mu classes of GST [27,32]. This was correlated with a reduction in GST and NQO1 enzyme activity [27]. Further analyses showed an impaired ability of Nrf2-deficient mice to induce NQO1 and Alpha and Mu classes of GST following administration of several agents capable of mediating ARE-dependent gene induction [27]. The data suggest that expression of antioxidant and cytoprotective genes under normal and oxidatively stressed conditions is regulated by Nrf2 via the ARE.


162

J. Mathers et al.

Glutathione levels are significantly reduced in the livers of Nrf2-deficient mice [33]. Comparison of fibroblasts from wild-type and Nrf2/ mice showed that the decrease in glutathione was correlated with a dramatic reduction in levels of GCL (glutamate cysteine ligase) transcripts. GCL catalyses the first, and rate-limiting, step in glutathione synthesis and is composed of a catalytic (GCLC) and regulatory (GCLM) subunit. The 5 region of human GCLC contains four ARE consensus sequences. The ARE positioned 3.1 kb upstream of the transcriptional start site, known as ARE4, has been shown to be important for basal and inducible expression of GCLC in response to -naphthoflavone [34]. Indeed, studies have shown that Nrf2 in combination with small Maf proteins was capable of binding the ARE4 under basal and inducible conditions [31]. GCLM contains an ARE consensus sequence close to the transcriptional start site [35,36], although work in our laboratory did not find that it was a functional ARE in HepG2 cells treated with tBHQ [37]. This was in contrast to work from Mulcahy’s group [38], who found this consensus sequence to function as an ARE in HepG2 cells treated with -naphthoflavone. Mulcahy and colleagues had performed a different type of mutational analysis to that of Galloway and McLellan [37], and they later went on to show that the conflicting data could be explained by the fact that GCLM contains a variant ARE in its promoter region [39]. The regulation of GCLM by a Nrf2-dependent mechanism is also supported by in vivo studies. Induction and basal expression of both GCLM and GCLC is reduced in tissues of Nrf2/ mice [27,40]. Futhermore, endogenous expression of both GCLM and GCLC can be rescued by transient expression of Nrf2 in Nrf2/ mouse embryonic fibroblasts [33], indicating that Nrf2 is a relevant factor in transcription of glutathione-associated genes. Reduced resynthesis of glutathione in Nrf2-deficient macrophages has also been correlated with decreased activation of the cysteine xc- transporter [41]. This transporter maintains the intracellular cysteine pool and therefore can modulate glutathione synthesis. In the absence of Nrf2, macrophages were unable to induce xc- activity and maintain GSH synthesis following DEM-mediated oxidative stress. Nrf2-deficient macrophages also displayed loss of basal and inducible haem oxygenase 1, stress protein A170 and Prx MSP23 expression in response to oxidative insult from a wide variety of agents [41]. Oligonucleotide microarray analyses have been used to identify the global changes in gene expression mediated by Nrf2 by comparing the expression patterns of genes in Nrf2+/+ and Nrf2/ cells. The basal expression of 137 genes was altered in Nrf2-deficient cortical astrocytes [30] and of the 207 genes induced following treatment with tBHQ, 181 were dependent on the presence of Nrf2. Nrf2-dependent genes were categorized as detoxification genes (including those encoding NQO1, GST and aldehyde dehydrogenase), antioxidants and genes involved in regulating the reducing potential, growth and defence/immune response and inflammation. Nrf2 was found to regulate the expression of several glutathione- or thioredoxin-dependent genes including the GCL subunits, GST family members, glutathione peroxidase 4, glutathione reductase and Prx I. The other antioxidant genes shown to be under the regulation of Nrf2 (including SOD, catalase and thioredoxin) function at different © 2004 The Biochemical Society


163

levels in cytoprotection. Regulation of several enzymes, including G6PDH (glucose-6-phosphate dehydrogenase), involved in the supply of reducing equivalents was also found to be Nrf2-dependent. From this analysis, it is clear that Nrf2 is capable of up-regulating several sets of genes in a co-ordinated manner, to prevent the accumulation of compounds capable of cellular damage and to redress the cellular redox status. A more extensive microarray analysis was carried out on the livers of Nrf+/+ and Nrf/ mice [42] following treatment with the chemopreventive agent, dithiolethione. This study identified several more clusters of genes regulated by Nrf2. Here it was shown that genes encoding enzymes involved in xenobiotic metabolism (e.g. cytochrome P450s, aldehyde and carbonyl reductases), chaperone and stress-responsive proteins and the ubiquitin/proteosomal system are regulated by Nrf2. A separate microarray study showed that other genes involved in xenobiotic metabolism including those encoding multidrug resistance protein 1, UDP-glucuronosyl transferases, aflatoxin aldehyde reductase and aldo-keto reductases are also regulated by Nrf2 in mouse intestine [43]. Taken together, these studies indicate that Nrf2 regulation extends beyond the primary control of cellular stress into secondary protective mechanisms such as recognition, repair and degradation of damaged cellular components. These findings imply that Nrf2 provides a more global mechanism of cytoprotection which incorporates cellular responses to a variety of forms of stress. Indeed, recent findings show that loss of Nrf2 sensitizes macrophages to cell death following exposure to endoplasmic reticulum stress and the unfolded protein response [44]. Nrf1 and Nrf2 are expressed ubiquitously, although their expression patterns are not identical [27]. It is therefore possible that in tissues where Nrf1 is predominantly expressed over Nrf2, the former is responsible for driving basal and inducible antioxidant gene expression. Indeed, analysis of fibroblasts obtained from mice with a targeted disruption of Nrf1 revealed that the absence of Nrf1 increased cell sensitivity to oxidative stress induced by paraquat, cadmium and diamide [45]. The increase in sensitivity was associated with increased oxidative burden due to decreased intracellular glutathione content. The decrease in intracellular glutathione was correlated with lowered basal expression of Gclm and glutathione synthetase transcripts [45], and sensitivity to oxidative stress was associated with a decreased ability to induce transcription of Gclm and glutathione synthetase following administration of paraquat. Overexpression of Nrf1 in COS cells was shown to elevate intracellular glutathione levels and transactivate a luciferase reporter construct containing 4.2 kb of the 5 flanking region of human GCLC [46]. Electrophoretic mobility shift assays have shown that the ARE4 in this region can also be bound by combinations of Nrf1 and small Maf proteins G and K [46]. These analyses suggest that Nrf1 may also contribute to cellular protection from oxidative stress by regulating cellular capacity to synthesize glutathione and antioxidant gene transcription via the ARE. Nrf1 and Nrf2 are, however, functionally distinct, as targeted disruption of Nrf1 causes embryonic lethality [47], and recent evidence suggests that Nrf2 is the more important protein in AREdriven gene expression [48].


164

J. Mathers et al.

It is clear that Nrf2, and probably Nrf1, are key factors in the regulation of many genes involved in cellular defence against oxidative stress. In the absence of Nrf2, the lack of induction of protective enzymes sensitizes cells and tissues to the cytotoxic effects of agents such as paraquat and DEM. The importance of Nrf2 is underscored by the dramatic increase in sensitivity of Nrf2/ mice to gastric tumours following administration of benzo(a)pyrene and the inability of chemopreventative agents such as oltipraz to protect against benzo(a)pyrene-induced chemical carcinogenesis in mice lacking Nrf2 [49]. Regulation of Nrf2 activity The actin-binding protein Keap1 negatively regulates Nrf2 by controlling its stability and subcellular localization [50,51]. In the unstressed cell, Nrf2 interacts with the Kelch domain of Keap1 through its N-terminal Neh2 domain [51] and, as a result, Nrf2 is sequestered in the cytoplasm. Following administration of thiol-reactive chemicals such as DEM, and in response to oxidative stress, the Keap1–Nrf2 complex is disrupted. Dissociation from Keap1 causes the rapid accumulation of Nrf2 in the nucleus [51] where it associates with small Maf transcription factors and mediates ARE-dependent gene expression. Keap1 is a cysteine-rich protein and it has been proposed that by virtue of its sensitivity to oxidative modification, it may act as a redox sensor so that changes in the intracellular reducing environment could be transduced to Nrf2 via modification of Keap1. It has been suggested that oxidative modification may cause conformational changes in Keap1 which disrupt the interaction with Nrf2, thus negating Keap1 repression [52]. A recent study has shown that oxidative modification of Keap1 in response to treatment with indoles could be mediated by activation of NADPH oxidase activity [53]. Targeted deletion of Keap1 showed that, in Keap1/ mice, Nrf2 constitutively accumulates in the nucleus and drives high-level expression of both GCL subunits, NQO1 and Prx [54]. Furthermore, macrophages from these animals were unresponsive to oxidative insult and expression of these genes could not be elevated in response to treatment with DEM. These data show that without Keap1, regulation of Nrf2-dependent gene expression is largely abolished, and the ability of cells to further up-regulate antioxidant and detoxification genes in response to oxidative stress is diminished. Several protein kinase-dependent signalling pathways have also been reported to regulate ARE-driven gene expression, and include the MAPKs (mitogen-activated protein kinases), PI 3-kinase, PKC (protein kinase C) and the endoplasmic reticulum-resident kinase, PERK (protein kinase R-like endoplasmic reticulum kinase) (Figure 1). In several instances, this has been demonstrated through the use of chemical inhibitors of these pathways. For example, pretreatment with the PI 3-kinase inhibitor wortmannin reduced binding to the ARE in nuclear extracts from H4IIE cells and prevented induction of rat GSTA2 mRNA [55]. Transient overexpression of members of the MAPK pathways has been shown to trigger transcription of ARE-dependent reporter activity in HepG2 cells [56]. This is thought to occur in an Nrf2dependent manner as a dominant-negative form of Nrf2 abolished this effect and prevented MAPK-dependent induction of endogenous haem oxygenase. © 2004 The Biochemical Society


165

Environmental stress

Perturbation of redox homoeostasis

PI3K

JNK ERK p38

ER PKC Protein unfolding

Akt

SH

SH S

Keap1

SH

PERK

T Nrf2

SR

SH

SR

Keap1

S P

P

S

T Nrf2 ARE

Maf

S

GSTA2 NQO1 GCLC GCLM HO-1

Figure 1 Potential pathways leading to the activation of Nrf2. Nrf2 regulates the expression of several antioxidant genes. Upon treatment with agents that generate environmental stress and perturb the intracellular redox status, the interaction between Keap1 and Nrf2 becomes disrupted, permitting Nrf2 to translocate to the nucleus and enhance transcription. The exact molecular mechanism triggering the disruption of Nrf2 and Keap1 has not been defined, although several studies have suggested that Nrf2 may become phosphorylated at serine and threonine residues by certain kinases, including PI 3-kinase (PI3K), PKC, PERK and members of the MAPK family. Keap1 is also thought to be a target in this response, and oxidation or covalent modification of cysteine residues within Keap1 has been proposed to destabilize its interaction with Nrf2 and promote dissociation of the Nrf2–Keap1 complex. ER, endoplasmic reticulum; ERK, extracellular-signal-regulated kinase; HO, haem oxygenase. © 2004 The Biochemical Society

166

J. Mathers et al.

The precise contribution of the various MAPK pathways to ARE-dependent gene regulation is, however, unclear as there are conflicting reports in the literature. For example, p38 was shown to negatively regulate basal and tBHQ-inducible expression of NQO1 [56], but positively regulate inducible expression of GCL subunits in response to pyrrolidine dithiocarbamate [57]. Exactly how members of the MAPK family transmit signals to modulate ARE transcription is unknown, but it should be borne in mind that certain experimental strategies such as overexpression of MAPK family members may cause cellular dysfunction, with oxidative stress as a possible side effect. Nrf2, however, has been shown to be directly phosphorylated by PKC and PERK in vitro [44,58]. Phosphorylation was correlated with increased nuclear localization of Nrf2 in response to stress and was prevented by catalytically inactive PERK or inhibition of PKC using staurosporine. PKC-independent phosphorylation occurs at the Neh2 domain of Nrf2, which is known to mediate interactions with Keap1 [59]. It is thought that phosphorylation disrupts the association of Nrf2 and Keap1, as it was shown that Keap1 is less able to coimmunoprecipitate with PKC-phosphorylated Nrf2 than unmodified Nrf2 [59]. Keap1 also contains several putative phosphorylation sites but it is unknown whether it is also targeted by kinases during oxidative stress. In addition, Bach1, a negative regulator of ARE-driven gene expression, appears to be regulated by phosphorylation [60]. It is clear that a variety of different signals stimulated by various forms of stress converge on the Nrf2–Keap1 complex. What determines the specificity and importance of each signalling pathway is not yet known but it is likely to be cell type-specific and depend on the relative abundance of the various factors in the signalling networks. Evolutionary conservation of CNC–Keap 1 gene regulation Although most of the work on the Nrf2–Keap1 pathway has been performed in mammalian systems, recent publications suggest that similar pathways may exist in other organisms. Orthologues of both Nrf2 and Keap1 have been identified in Danio rerio (zebrafish) and characterized in vitro by Kobayashi et al. [61]. Treatment of zebrafish larvae with tBHQ induced the expression of several antioxidant and detoxification genes (gstp1, nqo1 and gclc) [61]. Their expression was proposed to be Nrf2-dependent, as overexpression of nrf2 in D. rerio embryos enhanced expression levels of gstp1, nqo1 and gclc in the absence of oxidative stress. Futhermore, Nrf2 was also shown to be essential for the inducible expression of gstp1, as targeted knock-down of nrf2 abolished gstp expression [61]. Like the mammalian Nrf2 protein, the zebrafish Nrf2 interacts with and is repressed by Keap1. Repression of Nrf2 by Keap1 is mediated by cytoplasmic sequestration and occurs by binding through the Neh2 and Kelch domains. Indeed, specific mutation of the ETGE motif in the Neh2 domain of Nrf2 from zebrafish abolished Keap1-mediated repression of Nrf2 and permitted nuclear accumulation of the mutant Nrf2 form [61]. Moreover, overexpression of Keap1 was unable to repress gstp1 expression in zebrafish embryos co-overexpressing the E82G mutant Nrf2 [61]. These findings highlight the importance of Nrf2–Keap1 in regulating the cellular response



167

600 M DEM

450 M DEM

300 M DEM

150 M DEM

DMSO

Recombinant GCLC

to oxidative stress in zebrafish and indicate that Nrf2–Keap1 is a highly conserved regulator of cellular protection in vertebrates. An Nrf2-like signalling pathway also appears to be conserved in nematodes. Recent reports suggest that the transcription factor SKN-1 (skinhead-1) is important in regulation of oxidant signalling in Caenorhabditis elegans in a manner similar to Nrf2 [62]. SKN-1 is a distant relative of Nrf2 and the two proteins share identity at the CNC domain in the N-terminus. Furthermore, the consensus sequence of the SKN-1 DNA-binding motif is similar to that of Nrf2. The SKN-1 DNA-binding sites have been identified in the 5 flanking regions of several C. elegans antioxidant genes including gcl, gst, nqo and sod [62]. In C. elegans, expression of a gcl promoter-driven reporter construct was dramatically increased in response to paraquat. This was abolished in SKN-1deficient worms, as was basal expression of the gclc reporter gene, suggesting that SKN-1 mediates both constitutive and inducible expression of gclc. Moreover SKN-1 mutants are more sensitive to oxidative stress than wild-type worms [62]. These findings suggest that SKN-1 may contribute to cellular protection from oxidative stress by regulating antioxidant gene transcription in a manner similar to Nrf2 and the ARE. Unlike Nrf2, SKN-1 binds DNA as a monomer and does not require additional factors such as small Maf proteins. A functional homologue of Keap1 has not yet been identified in C. elegans. SKN1 does, however, share sequence similarity with a portion of the Neh2 domain of Nrf2 and is shown to accumulate in the nucleus following oxidative stress in a manner similar to Nrf2 [62]. Bioinformatic and genetic analyses suggest that a system analogous to the Nrf2–Keap1 antioxidant regulators may also be present in Drosophila melanogaster. The CNC locus (CG17894) encodes three transcripts and protein isoforms, Cnc-A, Cnc-B and Cnc-C [63]. The cnc-A and cnc-C transcripts are expressed ubiquitously in Drosophila embryos whereas expression of the cnc-B transcript is restricted to head and mandibular structures. The cnc-A, cncB and cnc-C transcripts encode proteins containing 533, 805 and 1296 amino

100kDa

75kDa

Figure 2 Treatment of Drosophila S2 cells with DEM induces expression of the catalytic subunit of GCL. Drosophila S2 cells were treated with DM (or DMSO as a control) for 16 h. Total cell lysates (30 g of protein) were subject to Western blotting using antiserum raised against the catalytic subunit of Drosophila GCL (GCLC). © 2004 The Biochemical Society

168

J. Mathers et al.

acids respectively and are 100% identical at the C-terminus. This region is also highly similar to the CNC-bZIP domain of NE-F2 p45. Work has shown that Cnc-B acts as a transcriptional repressor during Drosophila development and mutation of cnc-B leads to lethal head and mandibular defects [63]. A Maf orthologue (Maf-S) was identified as a Cnc-B-binding protein via yeast twohybrid analysis [64] and mutation of Maf-S recapitulates the embryonic defects formed by cnc-B mutations, suggesting Maf-S is a functional partner of Cnc-B. The DNA-binding element of the Cnc-B/Maf-S heterodimer has been mapped [64] and comparison shows that it is identical to the DNA-binding motif of TCF11 (transcription factor 11)/Nrf1/MafG, and very similar to the consensus binding site of Nrf2 and SKN-1. Moreover, Cnc-A and Cnc-B can both bind this element in combination with Maf-S. Drosophila appears to possess regulatory systems capable of inducing antioxidant gene expression, as GCLC levels can be increased in Drosophila S2 cells following exposure to oxidative insult generated by tBHQ [65] and DEM (Figure 2). Although the biological function of Cnc-A and Cnc-C has not been characterized, it is possible that they may regulate antioxidant gene expression in vivo. Alignment of several Nrf2 sequences reveals that the ETGE motif and highly hydrophobic domain of the Neh2 domain of Nrf2 are present in the Nterminal region of Cnc-C [61] but not Cnc-A and Cnc-B. This motif has been shown to be essential for Keap1-mediated sequestration of Nrf2 in the cytoplasm. This, in combination with the identification of a hypothetical gene encoding a Drosophila Keap1 orthologue in the Drosophila genome sequence (CG3962), and the observation that Cnc proteins form complexes with Maf-S which bind NF-E2-like elements, is highly suggestive that Cnc forms (in particular Cnc-C) may mediate Nrf2-like functions in vivo. Functional analyses of the Cnc protein isoforms will be required to confirm this and such analyses are currently under way in our laboratory.

DNA microarrays and proteomics to study the antioxidant response to stress As detailed earlier in this review, redox regulation is now known to be an important mechanism in signalling and the activation and regulation of transcription factors [1]. To fully understand the complexity of cellular redox regulation, it is essential the targets of such pathways are identified both at the gene and protein level. Over the past decade this goal has become more feasible with the emergence of the ‘-omics’ technologies of genomics [66] and proteomics [67]. The era of genomics, the study of complete genomes, has exploded following completion of sequencing of the entire genomes of many experimental animal systems, including budding yeast (S. cerevisiae) [68], nematode (C. elegans) [69] and fruit flies (D. melanogaster) [70]. With the imminent completion of the Human Genome Project [71] there is now a wealth of genetic knowledge freely available in the public domain; this information has been greatly exploited in microarray technology, which is emerging as one of the most powerful techniques for analysing gene regulation and function. In this technique, © 2004 The Biochemical Society


169

mRNA extracted from cells or tissue is reverse-transcribed to cDNA with the incorporation of a fluorescent tag. This is then hybridized to a complementary single-stranded DNA sequence immobilized on an array chip. Quantification of the amount of each mRNA is then determined from the intensities of fluorescence observed. Living organisms constantly sense and adapt to their redox environment, inducing batteries of genes to maintain their redox homoeostasis. As microarray technology allows quantitative comparison between treated and untreated samples (e.g. between normoxic and oxidatively stressed samples) it has proven itself a valuable technique in providing an insight into cellular redox regulation. Alternative gene splicing and post-translational modification of proteins means that a single gene can give rise to a diversity of functional protein products. This level of complexity is beginning to be addressed in the field of ‘proteomics’, the analysis of genomic complements of proteins. In particular, 2D-PAGE (two-dimensional PAGE), in conjunction with MS for protein identification [72,73], has emerged as a widely used technology in the analysis of the effects of oxidative insult on global protein expression. Several recent studies have employed either a genomic microarray or a proteomic 2D-PAGE approach to unravel the complex picture of cellular responses to oxidative stress [43,74–79]. Numerous stress-generating agents including H2O2, diamide, tBHQ and paraquat have been used to treat a wide range of species, including S. cerevisiae, C. elegans and D. melanogaster, and mammalian cell lines (see Table 1). From these studies, it appears that certain categories of genes and proteins are consistently induced in response to redox stress. Perhaps rather unsurprisingly, one category consistently exhibiting enhanced levels of expression following oxidative stress treatment are the antioxidant defence enzymes. Over the numerous studies carried out, virtually every known class of antioxidant is represented, including thioredoxins, Prxs, SODs, catalases, glutathione peroxidase, glutathione synthetase and GCL. Detoxifying enzymes, including numerous members of the GST family, were also found to be induced in several studies (see Table 1). The induction of GSTs occurred in all species investigated, including yeast, Drosophila and mammals, and with all oxidative agents tested. It is important to note that most studies only investigate the expression profile of one oxidative stress-inducing agent in a single cell type. Relatively few groups have attempted comparative studies to investigate whether distinct oxidative stress treatments evoke common or disparate gene-expression profiles. One such comparative microarray analysis in MCF7 cells exposed to H2O2, menadione or tBHQ concluded that both common and distinct genes were induced by the different oxidants. Common to all three treatments was the up-regulation of antioxidant enzymes including glutathione peroxidase and GCLM. Distinctions were, however, apparent in induction of hsp (heat-shock protein) genes. Menadione or tBHQ caused induction of Hsp-coding genes to a similar extent, whereas comparable levels of induction were not observed in H2O2-treated cells [75].


Organism

Drosophila MCF7 cells

Mouse

Mouse

HUVECs

Yeast

HUVECs

Experimental approach

Microarray Microarray

Microarray


Microarray

2D-PAGE

2D-PAGE

2D-PAGE

Paraquat

H2O2

H2O2/tBHQ

tBHQ

Sulphoraphane

Paraquat H2O2, menadione, tBHQ

Inducing agent GSTD1, GST6b, arsenite-translocating ATPase GPX2, GLRX, Hsp70, ubiquitin conjugation and ligase enzymes NQO1, GST, GCL, UGT, G6PDH, Hsp40, 6-phosphogluconate dehydrogenase NQO1, various GSTs, GCLC, Prx isoforms, G6PDH, Nrf2 Prx I–III, Hsp27, glyceraldehyde-3-phosphate dehydrogenase Cytochrome c peroxidase, SOD, thioredoxin1/2, TRX reductase, glutathione reductase, various Hsps, proteosome subunits, G6PDH, enolase, alcohol dehydrogenase Prx I, Prx II, Prx III, DnaJ

Examples of genes/proteins induced

[77]

[78] [76,79]

[30]

[75] [43]

[74]

Reference

Table 1 Genes and proteins induced by redox stress. Abbreviations: GPX, glutathione peroxidase; GLRX, glutaredoxin; UGT, UDP-glucuronosyltransferase; HUVECs, human umbilical endothelial vein cells; TRX, thioredoxin.

170 J. Mathers et al.


171

Stimulus-specific responses are known to occur in Escherichia coli following exposure to either H2O2 or superoxide radicals (O2•). This distinction is explained by two partially overlapping regulons, one responsive to H2O2 and the other to O2•. The differential response to oxidants is regulated by the redox-sensitive transcription factors OxyR and SoxRS, which are regulated by H2O2 and O2• respectively [80,81]. In yeast, Yap-1 transcription factors play an important role in regulating the response to oxidative stress [82,83], and the use of 2D-PAGE allowed Godon et al. [79] to identify the ‘H2O2 stimulon’ in S. cerevisiae. This study identified 115 proteins that are induced by H2O2 in S. cerevisiae [79]. Subsequent work showed that Yap1 mediates the induction of at least 32 proteins in response to oxidative stress [76]. Of these, 15 were shown to require the presence of the Skn7 transcription factor. Another category of enzymes consistently induced by oxidative insult are metabolic enzymes including NAD(P)H-regenerating enzymes. NAD(P)H functions as an indirect antioxidant by regenerating reduced glutathione (GSH) from its oxidized form (GSSG) in a process catalysed by NADPH-dependent glutathione reductase. Recently, however, it has been proposed that NAD(P)H may also be capable of acting as a direct antioxidant [84]. A major endogenous source of NADPH is the pentose phosphate pathway, which acts to generate appropriate levels of ribose 5-phosphate and erythrose 4-phosphate, precursors of purines and aromatic amino acids respectively. Within the oxidative branch of this pathway, NADPH is generated as a by-product of reactions catalysed by G6PDH and gluconate-6-phosphate dehydrogenase, and mutations in these enzymes cause sensitivity to H2O2 in yeast [85]. This antioxidant function is conserved in mammals, as mouse embryonic stem cells lacking G6PDH displayed an elevated sensitivity to H2O2 [86]. The link between metabolic pathways and ROS dynamics is further substantiated by the finding that switching carbon sources in yeast from glucose to oleate dramatically increases transcription of the antioxidant enzymes, thioredoxin, glutathione peroxidase and cytosolic SOD [87]. Cellular redox state is an important factor in maintaining protein structure and function. The accumulation of abnormal or denatured proteins in a cell in response to various experimental treatments, including oxidative stress, can potentially be toxic to the cell. Hsps function as molecular chaperones to promote protein renaturation, prevent further aggregation and denaturation and to facilitate proteolytic degradation of damaged proteins [88]. It is therefore not surprising that numerous proteomic and microarray investigations have observed induction of Hsps in response to oxidative insults (see Table 1). This response appears to be highly conserved in all species investigated, including bacteria, yeast and mammals, and occurs in response to most oxidative stress-generating agents tested. Hsp activity is not only regulated at a transcriptional level. It appears that the redox environment can directly regulate activity. Under reducing conditions in vitro the reactive cysteine residues of Hsp33 are co-ordinated to a zinc atom, rendering the enzyme inactive. However, under oxidizing conditions the zinc atom is released, intramolecular disulphide bonds are formed and the protein exhibits potent chaperone activity [89]. © 2004 The Biochemical Society

172

J. Mathers et al.

Our own work using both a microarray approach and a 2D-PAGE proteomics approach to investigate the impact of the thiol-depleting agent DEM on gene and protein expression in Drosophila S2 cells has observed induction of genes and proteins belonging to each of the major categories discussed above (J. Mathers, R.A. Akhtar, J.Z. Bone, L.I. McLellan and R.D.C. Saunders, unpublished work). Post-translational modifications such as phosphorylation, acetylation and glycosylation modulate the activity of many eukaryotic proteins. Analysis of such modifications can provide clues about the biological function and regulation of a protein [90]. Proteomic studies together with MS are beginning to emerge as common methodologies to investigate redox-dependent post-translational modifications of proteins, providing an indispensable insight into the mechanistic basis behind redox signalling. For example, phosphorylation changes the charge of a protein, which is then apparent on a two-dimensional gel. In a recent study, non-reducing two-dimensional electrophoresis was used to identify proteins in human T-lymphocytes that are glutathionylated following oxidative stress (diamide treatment or H2O2 exposure) [91]. Glutathionylation, the formation of mixed disulphides between glutathione and the cysteines of some proteins, has been postulated as a mechanism through which protein function can be regulated by redox status [92]. It can also occur by reaction of proteins with S-nitrosoglutathione, which is generated from NO [92,93]. Proteins which have been reported to be glutathionylated include redox enzymes such as Prx I, heat-shock proteins Hsp60 and Hsp70 and a regulator of protein folding, protein disulphide isomerase [91]. The Prx antioxidants are a family of thiol-dependent peroxidases that reduce H2O2 and alkyl hydroperoxides to water and alcohol. Using 2D-PAGE to compare proteins in human umbilical endothelial vein cells before and after exposure to H2O2 variants of Prx I and Pr III were generated within 30 min of exposure to 100 M H2O2. These variants exhibited altered migration consistent with a decreased isoelectric point, suggestive of oxidation [78]. Subsequent 2D-PAGE analyses in combination with MS studies determined that the shift in isoelectric point of Prx I following exposure to H2O2 [94] or tumour necrosis factor [95] was due to oxidation of the catalytic cysteine residue to sulphenic acid, resulting in inactivation of the enzyme. More recently it has been shown using 2D-PAGE in combination with activity assays that the inactivated sulphinic form of Prx I produced in HeLa, A594 and W136 cells following exposure to H2O2 can be reduced back to the catalytically active thiol form [96]. It cannot be disputed that both genomic microarray and proteomic 2DPAGE have allowed considerable advances to be made in the understanding of the complex cellular response to redox status. The next challenge must lie in furthering these advances to understand the mechanisms behind the redox regulation of signalling pathways.



173

The Biotechnology and Biological Sciences Research Council (grant numbers ERA16290 and G15091, to support J.M. and J.A.F.) and the Association for International Cancer Research (to support M.M.) are gratefully acknowledged for financial support.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31.

Finkel, T. (2003) Curr. Opin. Cell Biol. 15, 247–254 Finkel, T. (2000) FEBS Lett. 476, 52–54 Allen, R.G. and Tresini, M. (2000) Free Radical Biol. Med. 28, 463–499 Hancock, J.T., Desikan, R. and Neill, S.J. (2001) Biochem. Soc. Trans. 29, 345–350 Kamata, H. and Hirata, H. (1999) Cell. Signaling. 11, 1–14 Adler, V., Yin, Z., Tew, K.D. and Ronai, Z. (1999) Oncogene 18, 6104–6111 Halliwell, B. and Gutteridge, J.M.C. (1999) in Free Radicals in Biology and Medicine, Oxford University Press, Oxford Irani, K. and Goldschmidt Clermont, P.J. (1998) Biochem. Pharmacol. 55, 1339–1346 Lebovitz, R.M., Zhang, H., Vogel, H., Cartwright, Jr, J., Dionne, L., Lu, N., Huang, S. and Matzuk, M.M. (1996) Proc. Natl. Acad. Sci. U.S.A. 93, 9782–9787 Hayes, J.D. and McLellan, L.I. (1999) Free Radical Res. 31, 273–300 Wood, Z.A., Schroder, E., Robin Harris, J. and Poole, L.B. (2003) Trends Biochem. Sci. 28, 32–40 Sies, H. (1991) in Oxidative Stress II. Oxidants and Antioxidants, Academic Press, London Castillo, E.A., Ayte, J., Chiva, C., Moldon, A., Carrascal, M., Abian, J., Jones, N. and Hidalgo, E. (2002) Mol. Microbiol. 45, 243–254 Xu, D., Rovira, I.I. and Finkel, T. (2002) Dev. Cell 2, 251–252 Meng, T.C., Fukada, T. and Tonks, N.K. (2002) Mol. Cell 9, 387–399 Savitsky, P.A. and Finkel, T. (2002) J. Biol. Chem. 277, 20535–20540 Lee, S.R., Yang, K.S., Kwon, J., Lee, C., Jeong, W. and Rhee, S.G. (2002) J. Biol. Chem. 277, 20336–20342 Leslie, N.R., Bennett, D., Lindsay, Y.E., Stewart, H., Gray, A. and Downes, C.P. (2003) EMBO J. 22, 5501–5510 Saitoh, M., Nishitoh, H., Fujii, M., Takeda, K., Tobiume, K., Sawada, Y., Kawabata, M., Miyazono, K. and Ichijo, H. (1998) EMBO J. 17, 2596–2606 Adler, V., Yin, Z., Fuchs, S.Y., Benezra, M., Rosario, L., Tew, K.D., Pincus, M.R., Sardana, M., Henderson, C.J., Wolf, C.R., Davis, R.J. and Ronai, Z. (1999) EMBO J. 18, 1321–1334 Delaunay, A., Pflieger, D., Barrault, M.B., Vinh, J. and Toledano, M.B. (2002) Cell 111, 471–481 Georgiou, G. (2002) Cell 111, 607–610 Hayes, J.D. and McMahon, M. (2001) Cancer Lett. 174, 103–113 Finkel, T. and Holbrook, N.J. (2000) Nature (London) 408, 239–247 Rushmore, T.H., Morton, M.R. and Pickett, C.B. (1991) J. Biol. Chem. 266, 11632–11639 Motohashi, H., O’Connor, T., Katsuoka, F., Engel, J.D. and Yamamoto, M. (2002) Gene 294, 1–12 McMahon, M., Itoh, K., Yamamoto, M., Chanas, S.A., Henderson, C.J., McLellan, L.I., Wolf, C.R., Cavin, C. and Hayes, J.D. (2001) Cancer Res. 61, 3299–3307 Venugopal, R. and Jaiswal, A.K. (1996) Proc. Natl. Acad. Sci. U.S.A. 93, 14960–14965 Nioi, P., McMahon, M., Itoh, K., Yamamoto, M. and Hayes, J.D. (2003) Biochem. J. 374, 337–348 Lee, J.M., Calkins, M.J., Chan, K., Kan, Y.W. and Johnson, J.A. (2003) J. Biol. Chem. 278, 12029–12038 Wild, A.C., Moinova, H.R. and Mulcahy, R.T. (1999) J. Biol. Chem. 274, 33627–33636


174 32.

33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53.

54. 55. 56. 57. 58. 59. 60.

J. Mathers et al. Itoh, K., Chiba, T., Takahashi, S., Ishii, T., Igarashi, K., Katoh, Y., Oyake, T., Hayashi, N., Satoh, K., Hatayama, I., Yamamoto, M. and Nabeshima, Y. (1997) Biochem. Biophys. Res. Commun. 236, 313–322 Chan, J.Y. and Kwong, M. (2000) Biochim. Biophys. Acta 1517, 19–26 Mulcahy, R.T., Wartman, M.A., Bailey, H.H. and Gipp, J.J. (1997) J. Biol. Chem. 272, 7445–7454 Galloway, D.C., Blake, D.G., Shepherd, A.G. and McLellan, L.I. (1997) Biochem. J. 328, 99–104 Galloway, D.C., Blake, D.G. and McLellan, L.I. (1999) Biochim. Biophys. Acta 1446, 47–56 Galloway, D.C. and McLellan, L.I. (1998) Biochem. J. 336, 535–539 Moinova, H.R. and Mulcahy, R.T. (1998) J. Biol. Chem. 273, 14683–14689 Erickson, A.M., Nevarea, Z., Gipp, J.J. and Mulcahy, R.T. (2002) J. Biol. Chem. 277, 30730–30737 Chanas, S.A., Jiang, Q., McMahon, M., McWalter, G.K., McLellan, L.I., Elcombe, C.R., Henderson, C.J., Wolf, C.R., Moffat, G.J., Itoh, K. et al. (2002) Biochem. J. 365, 405–416 Ishii, T., Itoh, K., Takahashi, S., Sato, H., Yanagawa, T., Katoh, Y., Bannai, S. and Yamamoto, M. (2000) J. Biol. Chem. 275, 16023–16029 Kwak, M.K., Wakabayashi, N., Itoh, K., Motohashi, H., Yamamoto, M. and Kensler, T.W. (2003) J. Biol. Chem. 278, 8135–8145 Thimmulappa, R.K., Mai, K.H., Srisuma, S., Kensler, T.W., Yamamoto, M. and Biswal, S. (2002) Cancer Res. 62, 5196–5203 Cullinan, S.B., Zhang, D., Hannink, M., Arvisais, E., Kaufman, R.J. and Diehl, J.A. (2003) Mol. Cell. Biol. 23, 7198–7209 Kwong, M., Kan, Y.W. and Chan, J.Y. (1999) J. Biol. Chem. 274, 37491–37498 Myhrstad, M.C., Husberg, C., Murphy, P., Nordstrom, O., Blomhoff, R., Moskaug, J.O. and Kolsto, A.B. (2001) Biochim. Biophys. Acta 1517, 212–219 Chan, J.Y., Kwong, M., Lu, R., Chang, J., Wang, B., Yen, T.S.B. and Kan, Y.W. (1998) EMBO J. 17, 1779–1787 Leung, L., Kwong, M., Hou, S., Lee, C. and Chan, J.Y. (2003) J. Biol. Chem., 278, 48112–48119 Ramos Gomez, M., Kwak, M.K., Dolan, P.M., Itoh, K., Yamamoto, M., Talalay, P. and Kensler, T.W. (2001) Proc. Natl. Acad. Sci. U.S.A. 98, 3410–3415 McMahon, M., Itoh, K., Yamamoto, M. and Hayes, J.D. (2003) J. Biol. Chem. 278, 21592–21600 Itoh, K., Wakabayashi, N., Katoh, Y., Ishii, T., Igarashi, K., Engel, J.D. and Yamamoto, M. (1999) Genes Dev. 13, 76–86 Dinkova Kostova, A.T., Holtzclaw, W.D., Cole, R.N., Itoh, K., Wakabayashi, N., Katoh, Y., Yamamoto, M. and Talalay, P. (2002) Pro. Natl. Acad. Sci. U.S.A. 99, 11908–11913 Sekhar, K.R., Crooks, P.A., Sonar, V.N., Friedman, D.B., Chan, J.Y., Meredith, M.J., Starnes, J.H., Kelton, K.R., Summar, S.R., Sasi, S. and Freeman, M.L. (2003) Cancer Res. 63, 5636–5645 Wakabayashi, N., Itoh, K., Wakabayashi, J., Motohashi, H., Noda, S., Takahashi, S., Imakado, S., Kotsuji, T., Otsuka, F., Roop, D.R. et al. (2003) Nat. Genet. 35, 238–245 Kang, K.W., Ryu, J.H. and Kim, S.G. (2000) Mol. Pharmacol. 58, 1017–1025 Yu, R., Chen, C., Mo, Y.Y., Hebbar, V., Owuor, E.D., Tan, T.H. and Kong, A.N. (2000) J. Biol. Chem. 275, 39907–39913 Zipper, L.M. and Mulcahy, R.T. (2000) Biochem. Biophys. Res. Commun. 278, 484–492 Huang, H.C., Nguyen, T. and Pickett, C.B. (2000) Proc. Natl. Acad. Sci. U.S.A. 97, 12475–12480 Huang, H.C., Nguyen, T. and Pickett, C.B. (2002) J. Biol. Chem. 277, 42769–42774 Suzuki, H., Tashiro, S., Sun, J., Doi, H., Satomi, S. and Igarashi, K. (2003) J. Biol. Chem., 278, 49246–49253


Antioxidant and cytoprotective responses to redox stress 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75.

76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91.

92. 93.

175

Kobayashi, M., Itoh, K., Suzuki, T., Osanai, H., Nishikawa, K., Katoh, Y., Takagi, Y. and Yamamoto, M. (2002) Genes Cells 7, 807–820 An, J.H. and Blackwell, T.K. (2003) Genes Dev. 17, 1882–1893 McGinnis, N., Ragnhildstveit, E., Veraksa, A. and McGinnis, W. (1998) Development 125, 4553–4564 Veraksa, A., McGinnis, N., Li, X., Mohler, J. and McGinnis, W. (2000) Development 127, 4023–4037 Fraser, J.A., Saunders, R.D. and McLellan, L.I. (2002) J. Biol. Chem. 277, 1158–1165 Zhou, H., Roy, S., Schulman, H. and Natan, M.J. (2001) Trends Biotechnol. 19, S34–S39 Phizicky, E., Bastiaens, P.I., Zhu, H., Snyder, M. and Fields, S. (2003) Nature (London) 422, 208–215 Goffeau, A., Barrell, B.G., Bussey, H., Davis, R.W., Dujon, B., Feldmann, H., Galibert, F., Hoheisel, J.D., Jacq, C., Johnston, M. et al. (1996) Science 274, 546, 563–567 The C. elegans Sequencing Consortium (1998) Science 282, 2012–2018 Adams, M.D., Celniker, S.E., Holt, R.A., Evans, C.A., Gocayne, J.D., Amanatides, P.G., Scherer, S.E., Li, P.W., Hoskins, R.A., Galle, R.F. et al. (2000) Science 287, 2185–2195 Burley, S.K., Almo, S.C., Bonanno, J.B., Capel, M., Chance, M.R., Gaasterland, T., Lin, D., Sali, A., Studier, F.W. and Swaminathan, S. (1999) Nat. Genet. 23, 151–157 Govorun, V.M. and Archakov, A.I. (2002) Biochemistry (Moscow) 67, 1109–1123 Cutler, P. (2003) Proteomics 3, 3–18 Zou, S., Meadows, S., Sharp, L., Jan, L.Y. and Jan, Y.N. (2000) Proc. Natl. Acad. Sci. U.S.A. 97, 13726–13731 Chuang, Y.Y., Chen, Y., Gadisetti, Chandramouli, V.R., Cook, J.A., Coffin, D., Tsai, M.H., DeGraff, W., Yan, H., Zhao, S., Russo, A., Liu, E.T. and Mitchell, J.B. (2002) Cancer Res. 62, 6246–6254 Lee, J., Godon, C., Lagniel, G., Spector, D., Garin, J., Labarre, J. and Toledano, M.B. (1999) J. Biol. Chem. 274, 16040–16046 Mitsumoto, A., Nakagawa, Y., Takeuchi, A., Okawa, K., Iwamatsu, A. and Takanezawa, Y. (2001) Free Radical Res. 35, 301–310 Mitsumoto, A., Takanezawa, Y., Okawa, K., Iwamatsu, A. and Nakagawa, Y. (2001) Free Radical Biol. Med. 30, 625–635 Godon, C., Lagniel, G., Lee, J., Buhler, J.M., Kieffer, S., Perrot, M., Boucherie, H., Toledano, M.B. and Labarre, J. (1998) J. Biol. Chem. 273, 22480–22489 Storz, G. and Imlay, J.A. (1999) Curr. Opin. Microbiol. 2, 188–194 Pomposiello, P.J. and Demple, B. (2001) Trends Biotechnol. 19, 109–114 Toone, W.M. and Jones, N. (1999) Curr. Opin. Genet. Dev. 9, 55–61 Toone, W.M., Morgan, B.A. and Jones, N. (2001) Oncogene 20, 2336–2346 Kirsch, M. and De Groot, H. (2001) FASEB J. 15, 1569–1574 Juhnke, H., Krems, B., Kotter, P. and Entian, K.D. (1996) Mol. Gen. Genet. 252, 456–464 Pandolfi, P.P., Sonati, F., Rivi, R., Mason, P., Grosveld, F. and Luzzatto, L. (1995) EMBO J. 14, 5209–5215 Koerkamp, M.G., Rep, M., Bussemaker, H.J., Hardy, G.P., Mul, A., Piekarska, K., Szigyarto, C.A., De Mattos, J.M. and Tabak, H.F. (2002) Mol. Biol. Cell 13, 2783–2794 Ruddock, L.W. and Klappa, P. (1999) Curr. Biol. 9, R400–R402 Jakob, U., Muse, W., Eser, M. and Bardwell, J.C. (1999) Cell 96, 341–352 Mann, M. and Jensen, O.N. (2003) Nat. Biotechnol. 21, 255–261 Fratelli, M., Demol, H., Puype, M., Casagrande, S., Eberini, I., Salmona, M., Bonetto, V., Mengozzi, M., Duffieux, F., Miclet, E. et al. (2002) Proc. Natl. Acad. Sci. U.S.A. 99, 3505–3510 Klatt, P. and Lamas, S. (2000) Eur. J. Biochem. 267, 4928–4944 Padgett, C.M. and Whorton, A.R. (1995) Am. J. Physiol. 269, C739–C749


176 94. 95. 96.

J. Mathers et al. Yang, K.S., Kang, S.W., Woo, H.A., Hwang, S.C., Chae, H.Z., Kim, K. and Rhee, S.G. (2002) J. Biol. Chem. 277, 38029–38036 Rabilloud, T., Heller, M., Gasnier, F., Luche, S., Rey, C., Aebersold, R., Benahmed, M., Louisot, P. and Lunardi, J. (2002) J. Biol. Chem. 277, 19396–19401 Woo, H.A., Chae, H.Z., Hwang, S.C., Yang, K.S., Kang, S.W., Kim, K. and Rhee, S.G. (2003) Science 300, 653–656
