Oxidative stress and ozone: perception, signalling and response


Margarete Baier. E-mail: margarete.baier@uni-bielefeld.de


The primary site of ozone interaction with plant cells is the extracellular matrix where ozone challenges the antioxidant protection of the cells. Accordingly, ozone sensitivity generally correlates with the ascorbate status of the apoplast, which is an important signal initiation point. In addition, ozone sensing takes place by covalent modification of redox-sensitive components of the plasma membrane, for example ion channels like the plasma membrane Ca2+-channels. Subsequent intracellular signal transduction is an intriguing network of hormone, Ca2+ and MAPK signalling pathways, significantly overlapping with oxidative burst-induced pathogen signalling. Comparison of recent transcriptome analysis revealed that in addition to genes generally induced by all kinds of oxidative stress, for example, transcripts for PR-proteins and most antioxidant enzymes, approximately one-third of the responsive transcripts are ozone specific, indicating jasmonic acid, salicylic acid and ethylene-independent redox signalling triggered by extracellular redox sensing.


2-Cys peroxiredoxin-A


abscisic acid


1-aminocyclopropane-1-carboxylic acid


reduced glutathione


jasmonic acid


mitogen-activated protein kinase


reactive nitrogen species


reactive oxygen species


salicylic acid


ultra violet.


Plants face a continuously changing environment. In order to complete their life cycle and maintain a sufficient level of fitness to reproduce, they need to adapt to abiotic and biotic conditions. Periods of deviation from optimum are sensed in order to anticipate severe stress and to activate hardening processes. Two basic mechanisms allow for sensing of such environmental cues. First, sensors and receptors have evolved to directly monitor physical or chemical parameters (sensors) or to catch transient chemical signals. Alternatively, response reactions are triggered indirectly by modified metabolic activities. One particular element of metabolic imbalance often involved in adaptation processes is the stimulated production of reactive oxygen species (ROS) (Van Breusegem et al. 2001). ROS oxidize cellular constituents such as lipids, proteins and nucleic acids and can initiate radical chain reactions. An intricate network of defence and repair mechanisms counteracts these oxidation reactions. Imbalances between ROS generation and safe detoxification represent metabolic states that frequently are referred to as oxidative stress. In order to understand oxidative signalling, the chemical type of ROS and the subcellular site of generation are of importance.

In biological systems, oxidative stress results from the presence of elevated levels of oxidizing agents that are able to abstract electrons from essential organic molecules and disturb cellular functions. The most commonly encountered oxidizing molecules in the cell are reactive oxygen species (ROS) that derive from the abundantly available and rather inert dioxygen (O2) (Elstner 1990). Direct transfer of one electron to O2 produces the superoxide anion radical (O2·−) that subsequently acts as oxidant in single electron transfer reactions to form hydrogen peroxide (H2O2), hydroxyl radical (OH·) and finally H2O. Due to its exceedingly high reactivity OH· will react within diffusion distance with any organic molecule and is unlikely to serve as a signalling molecule whereas such function is assumed for O2·−, H2O2, O2 as well as reactive nitrogen species such as peroxynitrite (ONOO·–) (De Gara, de Pinto & Tommasi 2003; Neill, Desikan & Hancock 2003; Dietz 2003).

In photosynthetic cells, excess light affects the intracellular redox homeostasis and can evoke oxidative stress similar to O3. In both cases (i) O2·− and H2O2 are discussed as signalling molecules; (ii) the signals are produced distant from the nucleus, where transcriptional regulation takes place and information has to be transferred across membranes; and (iii) the low molecular weight antioxidants ascorbate and glutathione generally participate in the regulation of the cellular redox homeostasis. The major difference between sensing oxidative stress caused by atmospheric O3 and intracellularly produced ROS is that in plasmatic compartments the thiol system is employed and could sense ROS information at the site of production (Pastori & Foyer 2002; Dietz 2005). In contrast in the secretory pathway and the apoplast, thiols are mainly in the oxidized disulphide form (Vitale & Denecke 1999), which makes direct sensing of oxidative stress by thiol/disulphide proteins impossible.

This review summarizes aspects of oxidative stress, how they might be sensed, as well as how and which specific or general responses are induced. The immense body of literature available on all aspects of ROS in plants thwarts any attempt to cover this topic comprehensively. Instead, in the following, selected and recent results of physiological responses to oxidative stress focusing on ozone-induced adaptive mechanisms will be reviewed.


Sensing ozone by its reaction products

At the cellular level, in leaves of ozone-sensitive plants acute exposure to high ozone can induce chlorosis and necrotic lesions whereas accelerated leaf senescence has been observed for chronic exposure to low ozone levels (Schlagnhaufer et al. 1995; Rao, Koch & Davis 2000a; Rao, Lee & Davis 2002; Mahalingam et al. 2003). As a mechanism for these ozone-induced damages the generation of reactive oxygen species as superoxide and hydrogen peroxide by ozone degradation in the apoplast has been proposed.

Ozone sensitivity is generally correlated with the ascorbate status of the leaf tissue (Conklin & Barth 2004), which was genetically confirmed by isolation of ascorbic acid biosynthetic mutants on the basis of their increased sensitivity to O3 fumigation (Conklin et al. 1999). The antioxidant ascorbate accumulates to millimolar concentrations in leaf apoplasts and may scavenge there significant amounts of ozone (Noctor & Foyer 1998; Plöchl et al. 2000). Accordingly, when for instance in an ozone-sensitive genotype of Raphanus sativus L. the l-ascorbate content was increased by supplying the biosynthetic precursor l-galactono-1,4-lactone, tolerance to ozone was increased (Maddison et al. 2002). However, using quite high concentrations of 300 p.p.b. O3, Luwe, Takahama & Heber (1993) observed a very intriguing time-dependent relationship between oxidation of extracellular ascorbate and subsequent oxidation of the intracellular glutathione pool, while the cellular ascorbate redox state was unaltered during fumigation. Ascorbate regeneration was tightly coupled to GSH within the cell and transport activity was indicated to replenish the reduced ascorbate pool in the apoplast. By the glutathione-dependent regeneration mechanism, the extracellular ascorbate oxidation affected the intracellular thiol signature.

Experimental work strongly indicates that ascorbate is not the only signal initiation point in ozone sensing. As demonstrated by Jacob & Heber (1998) using oxidation-sensitive fluorescence dyes, even co-infiltration of leaves with 10 m m ascorbic acid did not prevent ascorbic-independent oxidation reactions. The reactivity of ozone is too high to be selective. From recent work several potential pathways of redox-dependent signalling can be deduced (Fig. 1): Using aequorin as an intracellular Ca2+-indicator Clayton et al. (1999) showed a specific increase in the intracellular Ca2+-concentration if atmospheric ozone reached concentrations above 70 p.p.b. It indicates oxidative activation of Ca2+ channels similar to the response to ABA-induced H2O2 (Kwak et al. 2003; Suhita et al. 2004; Dietrich et al. 2004). In accordance with this hypothesis, short-term exposure to O3 decreased the stomata aperture (McAinsh et al. 2002). In terms of signalling, the Ca2+ channel would function as ROS sensor. Thereby, ROS and free radicals not only can affect stomatal ion conductance but also ion channels from other tissues. Consequently, redox-dependent alterations in ion conductance might be a more general signalling pathway under oxidative stress [Fig. 1, (1)]. It is tempting to assume that ROS interact directly with plasma membrane bound receptors and trigger downstream events in the cytosol. This mechanism has tentatively been described for the animal system, where ROS activates the epidermal growth factor receptor to undergo Tyr phosphorylation (Esposito et al. 2003) [Fig. 1, (2)]. Furthermore, nonenzymatic or lipoxygenase-mediated break down of lipids [(3)] (Munnik & Meijer 2001), ROS, in particular H2O2 as diffusible messenger [(4)] (Laloi, Apel & Danon 2004), modulation of cytosolic ascorbate [(5)] and glutathione [(6)] relations, respectively (Noctor & Foyer 1998; Gomez et al. 2004) are well established regulatory and signalling compounds and may represent other routes of O3 triggered signalling from the site of chemical reaction of O3 in the apoplast or plasma membrane to the cytosol.

Figure 1.

Model of possible O3-dependent signalling pathways: (1) Oxidative activation of Ca2+-channels triggers Ca2+ influx into the cytosol, (2) yet unknown receptors sense oxidative stimuli in the apoplast and transduce the information to the cytosol, (3) enzymatic or non-enzymatic oxidation of unsaturated lipid to peroxides may initiate lipid-derived signalling, (4) H2O2 generated in the apoplast may diffuse into the cytosol, (5) extracellularly oxidized ascorbate is regenerated in the cytosol at the expense of (6) glutathione; both low molecular weight antioxidants are known to affect metabolic activities and gene expression. Finally, without reacting in the apoplast ozone may diffuse across the cell membrane, generate reactive oxygen species (ROS) in the symplast that trigger biochemical and genetic responses (7).

Oxidative burst

Recent studies presented evidence that ozone activates an oxidative burst and may induce a hypersensitive response (HR) similar to defence responses of plants to pathogen infections (see Kangasjärvi et al. 2005 for details). Involvement of the plasmamembrane NADPH oxidase in ozone-triggered ROS accumulation has been, for example, shown in the Arabidopsis thaliana mutant rcd1 (radical-induced cell death 1; Overmyer et al. 2000). In this work, application of the NADPH oxidase inhibitor diphenylene iodonium inhibited ROS accumulation and reduced leaf damage of the rcd1 mutant indicating that ozone initiates active cellular ROS production (Overmyer et al. 2000). In addition, in leaves of tobacco cultivar Bel W3 Pasqualini et al. (2003) observed an ozone-induced oxidative burst accompanied by transient increase of transcript levels of the hypersensitive response marker pathogenesis-related-1a. The tobacco leaves displayed increased protease activity and chromatin condensation indicating that ROS-induced programmed cell death is initiated by ozone-induced oxidative stress (Pasqualini et al. 2003). As another example, in transgenic tobacco plants with reduced catalase activity an active cell death programme was triggered by changes in H2O2 homeostasis and involvement of a signalling cascade leading to an NADPH oxidase-dependent burst has been suggested (Dat et al. 2003).


Downstream oxidative signalling involved in mediating the ozone responses within the cell is a two-step response process comprised of (i) fast-signalling events including a rapid change in the Ca2+ signature and protein phosphorylation, and (ii) the activation of auxiliary signalling pathways, which stabilize and amplify the primary signal (Fig. 2).

Figure 2.

Control of protein phosphorylation cascades. Oxidative bursts induce Ca2+-efflux, which activates protein phosphorylation. MAPK induce ethylene production, which stimulates MAPK-signalling. SA promotes ethylene, while JA inhibits SA-induction of ethylene biosynthesis. In feed-forward loops MAPKs induce expression of signalling kinases, which than increase the phosphorylation activity. Antagonists are phosphatases like PP2C, which can be redox-regulated themselves.

Oxidatively induced Ca2+-signals rapidly activate protein phosphorylation

Ca2+-release by oxidative activation of redox-sensitive Ca2+-channels (Clayton et al. 1999; McAinsh et al. 2002) causes rapid changes of the protein phosphorylation pattern (Agrawal, Rakwal & Iwahashi 2002a, Agrawal et al. 2002b) (Fig. 2). Signal duration is controlled by Ca2+-pumps, for example, in the ER-membrane (Sze et al. 2000) and induction of Ca2+-binding proteins (Agrawal et al. 2002c).

Mitogen-activated protein kinase (MAPK)-cascades play key roles in the change of the phosphorylation pattern (Samuel, Miles & Ellis 2000; Morris 2001). For example in rice seedlings a 66-kDa ERK-type MAPK is one of the earliest phosphorylated proteins. Ozone-induced phosphorylation stabilizes the enzyme (Agrawal et al. 2002b), which increases the kinase activity. Studies with the protein phosphatase inhibitors cantharidin, endothall and okadaic acid revealed that in addition transcription, as, for example, of the MAPK OsBMWK1, is stimulated by protein phosphorylation (Agrawal et al. 2003a). The MAPK-signal is controlled by redox-sensitive protein phosphatases 2C (PP2C), like the alpha-alpha PP2C MP2C, which is induced under oxidative conditions and regulates signal transduction specificity by supporting SIMK signalling via specific inhibition of the kinase SAMK (Meskiene et al. 2003).

Phytohormones control the signal

The plant hormones jasmonic acid (JA), salicylic acid (SA), ethylene and ABA control the ozone response. Thereby, in ABA-signal transduction the ABA-antagonistic PP2Cs ABI1 and ABI2 are inhibited by H2O2 (Meinhard & Grill 2001; Meinhard, Rodriguez & Grill 2002). Therefore, under oxidative stress the MAPK-mediated ABA signal is strengthened. With ABA-signalling leading to H2O2 generation (Jiang et al. 2003), the redox-sensitivity of the PP2C exponentially amplifies the primary signal. In this context it should be noted that several ABA responsive genes were activated in the Arabidopsis mutant vtc1 (Pastori et al. 2003), which has only 30% of wild-type ascorbate. This observation indicates that the threshold of the redox regulation of ABA-signal induction is low. A putative initiation site for ABA signal induction is the violaxanthin cycle. It is activated, if photosynthetic electron transport demands for energy dissipation (Demmig-Adams & Adams 1996). Low availability of reduced ascorbate arrests it in the epoxide status (Neubauer & Yamamoto 1994) supporting biosynthesis of the ABA precursor xanthoxin (Marin et al. 1996).

Regulation of the induction and spreading of oxidative stress symptoms by the phytohormones ethylene, SA and JA has been reported in several recent studies (Rao et al. 2002; Vahala et al. 2003; Moeder et al. 2002). In the Arabidopsis mutant rcd1, for instance, exogenous ethylene promoted superoxide-dependent cell death whereas exogenous application of methyl jasmonate inhibited its propagation. However, jasmonates, as well as ethylene, are involved in lesion containment (Overmyer et al. 2000). Based on studies of the O3-sensitive, JA-insensitive Arabidopsis mutant jar1 and the O3-tolerant, ethylene-insensitive ein2 mutant, Tuominen et al. (2004) hypothesized that early accumulation of ethylene stimulates spreading of cell death and suppresses protection by JA. Late accumulation of JA, however, inhibits the ethylene pathway and the propagation of cell death (Tuominen et al. 2004).

JA, SA and ethylene are second messengers in oxidative signal transduction. These phytohormones can mimic O3 (Agrawal et al. 2003b) and are generally induced by oxidative bursts (Watanabe, Seo & Saki 2001). Jasmonic acid, salicylic acid and ethylene regulate the strength of the primary signal by triggering MAPK cascades (Fig. 2; Agrawal et al. 2003a, b; Jiang et al. 2003; Guo & Ecker 2004). In general, ethylene and SA amplify the oxidative signal, while JA constricts ozone-induced damage (Watanabe et al. 2001; Langebartels et al. 2002; Kangasjärvi et al. 2005). Physiological analysis of the Arabidopsis mutant oji1 (ozone-sensitive and jasmonate-insensitive) (Kanna et al. 2003) demonstrated that JA signalling antagonizes ozone-induced induction of ACC synthase-catalysed ethylene biosynthesis (Bae et al. 1996) and analysis of NahG and npr1 plants, which are impaired in SA signalling, revealed that ozone-induced ethylene production depends on SA (Rao et al. 2002).

Biosynthesis of JA and SA are controlled by plastid lipid metabolism (Kachroo et al. 2003), lipid peroxidation and by oxidative inactivation of antioxidant enzymes like ascorbate peroxidases and peroxiredoxins (Mano et al. 2001; König et al. 2002). In parallel photosynthetic imbalances stimulate energy dissipation processes such as the violaxanthin cycle, which forms the precursors of ABA biosynthesis (Marin et al. 1996). It can be suggested that in oxidative signalling chloroplast metabolism, and photosynthesis in particular, control the signal strength and the integration of environmental parameters not only by photo-oxidative ROS generation (Baier & Dietz 1999; Chang et al. 2004), but also by biosynthesis of second messengers. Therefore, even for O3, which attacks plant cells from the outside, within the plant cell general oxidative signal transduction pathways including chloroplast-to-nucleus signalling are likely to be involved in the regulation of plant responses.

Complexity, redundancy and specificity in oxidative signal transduction

In the transmission of secondary signals like JA, SA, ethylene and ABA MAPK are involved. Arabidopsis thaliana encodes 10 MAPKKKK, 80 MAPKKK, 10 MAPKK and 23 MAPK (Jonak et al. 2002), which form complex signalling networks with synergistic and antagonistic links (Fig. 3). The ‘double jeopardy’ of MAPK-signalling was described in tobacco lines over-expressing or suppressing salicylate-induced protein kinase (SIPK), as both manipulations led to higher ozone sensitivity (Samuel & Ellis 2002). SIPK, like the wound induced MAPK WIPK (Seo, Sano & Ohashi 1999), is activated by the MAPKK MEK2 (Yang, Liu & Zhang 2001). During ozone exposure, activation of SIPK was prolonged in over-expressing lines, while activation of WIPK occurred in SIPK-suppressed lines.

Figure 3.

Transmission of oxidative signal takes place via at least three MAPK signalling cascades. MPK3, MPK4 and MPK6 are generally stress induced, while MPK1 and MPK2 mediate the ABA response which is redox controlled by ROS activation of antagonistic PP2C and by MPK3. MPK6 action, which transmits the ethylene signal, inhibits MPK4 activation, while MPK3 stimulates ABA-signalling.

In Arabidopsis thaliana, the MAP kinases MPK3, MPK4 and MPK6 are activated by various abiotic stresses (Ichimura et al. 2000; Kovtun et al. 2000) and might be central elements of oxidative signal transduction (Fig. 3). MPK6 and MPK3, which are the Arabidopsis homologues of SIPK and WIPK, respectively, are activated by the MAP kinase kinases MKK4 and MKK5 (Asai et al. 2002) and the MAPKKK ANP1 (Kovtun et al. 2000). In MPK6-silenced plants MPK4 is activated (Menke et al. 2004) indicating that MPK6 suppresses MPK4 activation. As MPK4 is under control of MKK1 (Huang et al. 2000), Menke et al. (2004) suggested that in the response to wounding two MAPK cascades act in parallel. In addition, SA-independent MAPK-regulated signal transduction takes place. For example, Jiang et al. (2003) showed recently that the SA-insensitive ABA-MAPK is mediated by the MAPKK MEK1/2. MPK3 over-expressing Arabidopsis are more sensitive to ABA during germination (Lu et al. 2002), demonstrating that there is cross-talk between the MPK3- and ABA-signalling cascades. Ethylene signalling, which is often antagonistic to ABA-signalling (Ghassemian et al. 2000), is transmitted through the SIMK-homologue MPK6 (Guo & Ecker 2004).

The complexity and dynamics of MAPK-regulated oxidative signalling becomes obvious by comparison of promoter regulation of the cytosolic ascorbate peroxidase APx2 and the nuclear encoded chloroplast 2-Cys peroxiredoxin 2CPA. The promoters of the two peroxidases both respond to ABA and to redox signals (Fryer et al. 2003; Baier, Ströher & Dietz 2004). APx2 is synergistically induced by ABA and oxidative stress (Fryer et al. 2003), while 2CPA is oxidatively induced by wounding and photo-oxidative stress via one MAPKK, but strongly suppressed by ABA via an antagonistically responding MAPKK (Baier et al. 2004). Like with 2CPA regulation, the wounding response of Apx2 cannot be mimicked by JA (Baier et al. 2004; Chang et al. 2004), which indicates transmission by other secondary messengers with a specific impact of ABA.


New insights into the responses to oxidative stress signalling and the specificity of secondary messengers have been provided recently by PCR-based suppression subtractive hybridization (Mahalingam et al. 2003), transcriptome analysis investigating responses to oxidative stress in wild-type plants (Desikan et al. 2001; Tamaoki et al. 2003a) and the analysis of transgenic plants and mutants with various genetic backgrounds (Vranova et al. 2002; Tamaoki et al. 2003b). Based on comparison of transcriptome patterns in plants treated with O3 (Tamaoki et al. 2003a), H2O2, sublethal doses of methyl viologen (Vranova et al. 2002) or pathogens (Durrant et al. 2000; Maleck et al. 2000), it has been estimated that oxidative stress affects approximately 2% of the plant transcriptome (Tamaoki et al. 2003a).

Characterizing the transcriptional responses of Arabidopsis to, for example, ozone and to pathogen treatment Mahalingam et al. (2003) distinguished between shared responses of ozone and pathogen exposure and transcripts specifically regulated by ozone. By hybridization of Arabidopsis-cDNA-macroarrays Tamaoki et al. (2003a) identified 205 ozone-responsive transcripts after 12 h exposure to O3 (200 nL L−1) and comprehensively compared the signalling pathways of ethylene, JA and SA on ozone-responsive gene expression. Approximately 75% of the ozone-responsive transcripts were induced and 48 of 205 genes were suppressed by O3. Among the 109 transcripts with known functions, 33 are involved in metabolism like monodehydroascorbate reductase, glutaredoxin and pyruvate kinase, 24 in cellular organization and biogenesis, 25 in cell rescue/defence (e.g. glutathione S-transferase, PR4, l-ascorbate peroxidase), 11 in signal transduction like cyclophilin ROC7, six in energy (e.g. chlorophyll a/b-binding protein and the gamma subunit of the mitochondrial F1-ATPase), five in protein synthesis and degradation, three in transcription and two in transport. This study corresponds to results of previous studies of genes for which activation or suppression by O3 has been described. In response to O3 exposure, transcript levels of cytosolic ascorbate peroxidase (Willekens et al. 1994; Kubo et al. 1995), lipoxygenase (Maccarrone, Veldink & Vliegenhardt 1992), blue copper-binding protein (Langebartels et al. 2000), glutathione S-transferase, PR proteins and catalases (Lim, Woo & Nam 2003) were elevated, whereas transcript levels of chlorophyll a/b-binding protein (Miller, Arteca & Pell 1999), rbcS (Glick et al. 1995; Miller et al. 1999), and Cu/Zn superoxide dismutase (Kliebenstein, Monde & Last 1998) were reduced. Additionally, elevated transcript levels of enzymes such as glutathione S-transferase, glutaredoxin, cyclophilin ROC7 and ascorbate peroxidase indicate the involvement of thiol-/disulphide transitions and thiol-dependent redox regulation in acclimation to O3 exposure. With a proteomics approach Agrawal et al. (2002c) studied the effects of ozone exposure in rice. The ozone treatment resulted, for example, in decreased accumulation of photosynthetic proteins whereas induced accumulation was, for example, observed for an ascorbate peroxidase, superoxide dismutase, and pathogenesis-related (PR) proteins. By contrast to acute treatments, chronic exposure to ozone in the field caused a smaller proportion of the Arabidopsis thaliana transcriptome to be altered, with five times more genes down-regulated than up-regulated (Miyazaki et al. 2004).

As exposure to high levels of O3 induces the synthesis of ethylene, JA and SA and activates the corresponding signalling pathways (Sharma & Davis 1997; Rao et al. 2000a, b; Kangasjärvi et al. 2005), the regulation of 157 O3-induced genes was analysed in signal transduction deficient mutants as the ethylene-insensitive ein2, JA-resistant jar1 and SA-insensitive npr1. While approximately 50% of the transcripts were inhibited in ein2 and jar1, only 20% of the transcripts were inhibited in npr1 (Tamaoki et al. 2003a), indicating that JA and ethylene signalling are more crucial for the induction than SA signalling at least under these particular conditions. Of the O3-induced genes with function in cell rescue and defence such as l-ascorbate peroxidase, glutathione S-transferase and hevein-like protein (PR4) about 70% were controlled by ethylene and JA and were suppressed by SA, suggesting that ethylene and JA signalling are important for the induction of defence gene expression as outlined above.

In the study presented by Tamaoki et al. (2003a) 16 genes were identified whose expression was induced by all three signalling molecules. However, about 30% of 157 induced O3-responsive transcripts were independent of ethylene, JA and SA pathways indicating control by other still unknown factors. By the comparison of 205 ozone-responsive transcripts in response to drought, salinity, UV-B, low temperature, high temperature and acid rain as ROS (Tamaoki et al. 2004) found that the transcriptional responses to O3 and UV-B stress are similar, but distinct from the responses to the other stresses. It may be concluded that O3 and UV-B share common features via the generation of reactive oxygen species (Mackerness et al. 1999). The comparison of the transcript regulation patterns has provided important clues to the specificity of oxidative stress responses and allows differentiating O3 induced responses from regulation in context of photosynthesis and other sources of ROS. Detailed functional characterization of transcripts and proteins identified in transcriptomic and proteomic studies will provide further information on physiological responses to oxidative stress in plants.

In plant cells, scavenging of reactive oxygen species is mediated by antioxidant metabolites, for example, as ascorbate, glutathione, and tocopherols as well as by detoxifying enzymes, for example, as superoxide dismutase, ascorbate peroxidase, peroxiredoxin and catalase (reviewed in Mittler 2002; Neill, Desikan & Hancock 2002; Dietz 2003). Recently, in a transgenic approach, increased tolerance to oxidative stress that correlated with decreased membrane damage has been shown in tobacco over-expressing Chlamydomonas glutathione peroxidase (Yoshimura et al. 2004). As another mechanism for intracellular ROS-scavenging accumulation of the antioxidant thylakoid membrane-bound polyamine putrescine was induced by enhanced ozone concentrations in ozone-tolerant, but not in ozone-sensitive tobacco (Navakoudis et al. 2002). Consequently, a regulatory role of polyamines for the adaptation of the photosynthetic apparatus has been suggested (Navakoudis et al. 2002). Another antioxidant compound isoprene has a role in quenching H2O2 and decreased lipid peroxidation of membranes (Loreto et al. 2001; Loreto & Velikova 2001).


Recent analysis of transcriptome patterns in wild-type plants and mutants has expanded our knowledge about responses that are induced by oxidative stress. However, in contrast to the experimental approaches, where, for example, the ozone concentration was suddenly and strongly increased, in nature, during a vegetation period, the maximum daily ozone concentrations peak several times a month over periods of weeks and induce hardening of the plants (Luwe 1996). Therefore, one can assume that most responses described in short-term fumigation experiments are acute stress reactions. To understand oxidative signalling in general, however, we will have to distinguish the intracellular oxidative sensing mechanisms from the complex networks observed in short-term stress experiments. Defining the ‘switch board’ of oxidative signalling and the thresholds of particular signalling pathways will challenge us in the coming years. Besides introducing more and more defined transcriptomic approaches, analysis of promoter responses in physiological contexts may guide us to the cellular redox signalling networks. In this context, recently investigated examples are the analysis of the APx2, PetE1 and 2CPA promoters in Arabidopsis thaliana (Oswald et al. 2001; Fryer et al. 2003; Chang et al. 2004; Baier et al. 2004). Thereby, the APx2 promoter responds to photosynthetically produced ROS (Chang et al. 2004), the PetE1 promoter is controlled by the redox state of the plastoquinone pool as long as the concentration of sugars is low (Oswald et al. 2001), and the activity of the 2CPA promoter correlates, like that of the thioredoxin redox regulation system (Schürmann & Jacquot 2000), with the electron pressure on ferredoxin and the redox state of the [NADP+ + NADPH] pool (Baier et al. 2004). While forward genetic screens for mutants affected in PetE1 promoter regulation led to isolation of mutants defect in the transmission of the plant hormone abscisic acid (ABA) (Huijser et al. 2000), physiological analysis of reporter gene plants showed that ABA, which can cause oxidative bursts (Kwak et al. 2003) and whose transmission is redox regulated (Meinhard & Grill 2001; Meinhard et al. 2002), is a modulator of redox-active promoters like that of APx2 (Fryer et al. 2003) and 2CPA (Baier et al. 2004). Constitutive induction of APx2 promoter activity in rax1 (Ball et al. 2004), demonstrates that glutathione-regulated thiol switches can play an important role in transmission of ROS signals. Future work such as the analysis of the rimb-mutants, which were screened for lower 2CPA promoter activity and show impairment of several redox-regulated genes relative to the redox state of the ascorbate pool (Heiber et al. 2004), will point to further regulators of intracellular redox signalling networks which adjust nuclear gene expression to physiological variation of the redox status.


The authors’ work that is summarized in this review was supported by the Deutsche Forschungsgemeinschaft (FOR387, Di346/6; Ba 2011/2)