Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

The apoplastic oxidative burst in response to biotic stress in plants: a three‐component system

The apoplastic oxidative burst in response to biotic stress in plants: a three‐component system Abstract The oxidative burst, the generation of reactive oxygen species (ROS) in response to microbial pathogen attack, is a ubiquitous early part of the resistance mechanisms of plant cells. It has also become apparent from the study of a number of plant–pathogen interactions and those modelled by elicitor treatment of cultured cells that there may be more than one mechanism operating. However, one mechanism may be dominant in any given species. NADPH oxidases have been implicated in a number of systems and have been cloned and characterized. However, the enzyme system which is the major source of ROS in French bean (Phaseolus vulgaris) cells treated with a cell wall elicitor from Colletotrichum lindemuthianum, appears to be dependent on an exocellular peroxidase. The second component, the extracellular alkalinization, occurs as a result of the Ca2+ and proton influxes and the K+ efflux common to most elicitation systems as one of the earliest responses. The third component, the actual reductant/substrate, has remained elusive. The low molecular weight compound composition of apoplastic fluid was compared before and after elicitation. The substrate only becomes available some min after elicitation and can be extracted, so that by comparing the profiles by LC‐MS it has been possible to identify possible substrates. The mechanism has proved to be complex and may involve a number of low molecular weight components. Stimulation of H2O2 production was observed with saturated fatty acids such as palmitate and stearate without concomitant oxylipin production. This biochemical evidence is supported by immunolocalization studies on papillae forming at bacterial infection sites that show the peroxidase isoform present at sites of H2O2 production revealed by cerium chloride staining together with the cross‐linked wall proteins and callose and callose synthase. The peroxidase has been cloned and expressed in Pichia pastoris and has been shown to catalyse the oxidation reaction with the same kinetics as the purified enzyme. Furthermore, Arabidopsis plants transformed heterologously using the French bean peroxidase in antisense orientation have proved to be highly susceptible to bacterial and fungal pathogens. Thus it is possible that Arabidopsis is another species with the potential to mount an apoplastic oxidative burst and these transformed plant lines may be useful to identify the peroxidase that is responsible. Cell wall, fatty acids, French bean, hydrogen peroxide, oxidative burst, oxylipins, peroxidase, Phaseolus vulgaris L. DPI, diphenylene iodonium, FBP1, French bean peroxidase isoform 1, HR, hypersensitive response, HRP, horseradish peroxidase, PSI, photosystem I, PSII, photosystem II, ROS, reactive oxygen species. Introduction Reactive oxygen species (ROS) are rapidly produced in plants as a defence response to pathogen attack (Bolwell and Wojtaszek, 1997; Bolwell, 1999). There are a number of potential sources of ROS. It has been well recognized that protoplastic sources centred upon mitochondrial, chloroplastic or peroxisomal generating systems exist, but have mainly been studied in relation to abiotic stresses (Asada, 1999; del Rio et al., 2002). Moving towards the cell surface, in many species it is thought that the superoxide anion can be originated at the plasma membrane in response to biotic stresses from an NADPH oxidase (Lamb and Dixon, 1997) analogous to the system which occurs in neutrophils or from an NADH‐dependent superoxide synthase (Auh and Murphy, 1995). Superoxide is subsequently dismutated to hydrogen peroxide by superoxide dismutase. There are potential sources of apoplastic origin of hydrogen peroxide including peroxidases, amine oxidases and oxalate oxidases. In elicitor‐treated French bean cells hydrogen peroxide appears to originate from a cell wall peroxidase in a three component system requiring ion fluxes and leading to extracellular alkalinization and the release of a reductant (Robertson et al., 1995; Bolwell et al., 1995, 1999; Bolwell, 1996). The most widely studied potential mechanism is a system analogous to the mammalian NADPH oxidase system. Much of the evidence for this was based upon the inhibition of defence responses by DPI, a reasonably specific inhibitor of the mammalian oxidative burst. While there was a lack of direct biochemical characterization of an enzyme complex from plant cells (Bolwell, 1999), confirmation of the existence of the plant NADPH oxidase came from the molecular cloning of the plant analogue of gp91phox the core polypeptide of the mammalian enzyme, which also showed considerable homologies to ferric reductases of plants and fungi involved in iron uptake (Keller et al., 1998; Torres et al., 1998). Examination of the primary structure of the plant enzymes showed important regulatory differences to the mammalian enzyme, which is activated by cytosolic polypeptides. In the case of the plant protein there is an additional N‐terminus with EF‐hands suggesting direct activation by Ca2+ ions (Lamb and Dixon, 1997). Eight genes coding for the Arabidopsis homologues of mammalian NADPH oxidases (Atrobh A‐H) have been characterized. Dissecting the role of each of these in generating ROS in Arabidopsis has recently been advanced by reverse genetics (JDG Jones, personal communication). Transposon tagged lines were screened for insertions in all eight genes. Individual knockouts have no effect on the defence response. Crosses were performed and one double knockout was particularly significant. When challenged by Peronospora, this line was found by diaminobenzidine staining not to produce localized ROS accumulation, suggesting this class of protein makes a major contribution to ROS production. However, this line was still disease‐resistant. Two other potential mechanisms are also worthy of note, one because it deserves further exploration as a general mechanism and the second because it is another example of specialization in the ability to generate H2O2 in response to pathogen attack. Protoplast sources of ROS have, of course, been extensively studied in relation to abiotic stress, particularly from chloroplasts (Asada, 1999) and peroxisomes (del Rio et al., 2002). However, due to the restructuring of the cytoskeleton and transport of vesicles to the site of interaction of the pathogen with the host cell wall (McLusky et al., 1999), this redirection of cell material for papilla deposition also appears to be accompanied by larger organelles which are often observed in the vicinity. Potentially, chloroplasts, through PSI leakage, could generate superoxide and from PSII, singlet oxygen (Asada, 1999), while peroxisomes could generate superoxide from xanthine oxidase and three additional membrane‐bound polypeptides (del Rio, 2002). The specialized family of germin‐like proteins (Woo et al., 2000) contain the oxalate oxidases in cereals which can generate H2O2 either from oxalate under acidic conditions that might arise from vacuolar damage caused by pathogen attack. More recently they have been shown to have superoxide dismutase activity at more neutral pH amplifying a role in the generation of H2O2 under a wider range of cellular conditions. The major subject of this review, generation of ROS by apoplastic peroxidases has been studied in French bean. Critical to the formation of an hypothesis of an alternative source of hydrogen peroxide to NAD(P)H oxidases were the observations that the NADH/NAD ratio and ATP level transiently fall, at the same time as an increased oxygen uptake occurs, while the NADPH level remains constant (Robertson et al., 1995; Bolwell, 1996). In addition, the oxidative burst in bean cells was inhibited by KCN, whereas the burst is insensitive to cyanide in mammalian cells. Moreover, its relative insensitivity to diphenylene iodonium (DPI), a well‐known inhibitor of the NADPH‐oxidase, favours the hypothesis of an apoplastic peroxidase generating the ROS in French bean (Bolwell et al., 1998). Most haem proteins are capable of generating H2O2 at alkaline pH in a mechanism that involves the formation of compound III (FeII‐O‐O). In the case of horseradish peroxidase compound III can readily be reduced with a pH optimum of 8.5 by many reductants of which thiols are the most active. In the case of bean cells the particular peroxidase responsible for the oxidative burst has been purified from cell walls and has an Mr of 46000 and exhibits a pH optimum of 7.2 for the oxidation of cysteine as a model compound. Molecular cloning of this peroxidase isoform has been achieved and the cognate protein expressed in Pichia pastoris to model the reaction in vitro (Blee et al., 2001). Other systems that appear to be capable of this type of mechanism are lettuce (Bestwick et al., 1998), cotton (Martinez et al., 1998), onion (McLusky et al., 1999), and possibly Arabidopsis (Grant et al., 2000a). Component 1: the peroxidase Purification, molecular cloning and modelling The oxidative burst peroxidase was purified as an Mr 46000 isoform from walls of suspension‐cultured French bean cells (Zimmerlin et al., 1994). The cDNA was isolated using a strategy that involved the design of oligonucleotides corresponding to peptides close to the N‐ and C‐termini of the 46 000 kDa FBP1 and isolation of a number of PCR products. Sequencing was used to identify a clone that coded for an identical amino acid sequence corresponding to the sequence data obtained for FBP1 purified from cell walls of suspension‐cultured cells of French bean (Zimmerlin et al., 1994) together with the extensive additional sequences. Screening of a cDNA library corresponding to mRNA from elicitor‐induced cells resulted in the acquisition of a cDNA that contains sequence coding for the whole of the mature protein (Blee et al., 2001). FBP1 is cationic and differs from other peroxidases in its ability to generate H2O2 at relatively neutral pH. However, modelling its 3‐D structure using coordinates for barley, horseradish and peanut peroxidase (Fig. 1) did not show any striking feature that could account for these different properties in generating H2O2. However, multiple potential N‐linked glycosylation sites were observed on FBP1 and the surface charge was near neutral even though it migrates as a cationic protein in IEF. These glycosylation and surface charge differences among enzymes of a gene family could define subcellular location and substrate interactions and therefore further elucidate gene member roles in planta beyond expression patterns. Recombinant FBP1 was produced in Pichia pastoris and shown to catalyse the cysteine oxidation reaction at pH 7.2, which further confirmed FBP1 as the peroxidase responsible for the apoplastic oxidative burst (Blee et al., 2001). FBP1 was immunolocalized to infection sites during the interaction of non‐pathogenic strains of Xanthomonas campestris and French bean mesophyll cells. This interaction produces massive papillae aiding structural studies. The FBP1 co‐localized with H2O2 detected by cerium chloride staining in material surrounding the bacteria and nearby wall (Brown et al., 1998). Also co‐localization of other components of papillae, a cross‐linked anti‐microbial arabinogalactan protein and callose was demonstrated together with the callose synthase. This is in accordance with one hypothesis of peroxidase involvement and H2O2 generation being required for the construction of barriers at infection sites. It is also strong circumstantial microscopic evidence that FBP1 is responsible for H2O2 generation at sites of infection. Thus there is good biochemical and structural evidence for a peroxidase‐dependent oxidative burst in French bean cells in suspension culture and in leaves. Unfortunately French bean is not amenable to reverse genetics to explore this functionality further and to confirm this role of FBP1 in planta, so studies have been instigated in Arabidopsis. Fig. 1.  View largeDownload slide Contour mapping of the surface of the French bean peroxidase 1 (FBP1) responsible for the apoplastic oxidative burst. X‐ray crystallography co‐ordinates for barley peroxidase, horseradish peroxidase C, and peanut peroxidase were used to model the amino acid sequences of the predicted mature polypeptide for FBP1 as described previously (Blee et al., 2001). The molecules were all aligned similarly showing a view with the pore leading interiorly to the haem group located at the centre. The model displays surface charge; white, neutral; blue, positive; red, negative. Fig. 1.  View largeDownload slide Contour mapping of the surface of the French bean peroxidase 1 (FBP1) responsible for the apoplastic oxidative burst. X‐ray crystallography co‐ordinates for barley peroxidase, horseradish peroxidase C, and peanut peroxidase were used to model the amino acid sequences of the predicted mature polypeptide for FBP1 as described previously (Blee et al., 2001). The molecules were all aligned similarly showing a view with the pore leading interiorly to the haem group located at the centre. The model displays surface charge; white, neutral; blue, positive; red, negative. Transformation of French bean peroxidase in the antisense orientation into Arabidopsis One way of confirming the functionality of cognate proteins is to down‐regulate them through antisense or partial sense expression in transgenic plants. Arabidopsis plants transformed heterologously using the FBP1 in antisense orientation have proved to be highly susceptible to the DC3000 virulent strains of Pseudomonas syringae (Fig. 2) as well as avirulent strains of this bacterium and a number of fungal pathogens (GP Bolwell, KA Blee, F Ausubel, unpublished data). Work is currently underway trying to identify the endogenous Arabidopsis peroxidase that is down‐regulated. However, the highest scoring Arabidopsis thaliana peroxidase to the oxidative burst peroxidase FBP1 was Arabidopsis thaliana peroxidase A2 (ATPA2), accession X99952 at 61% amino acid identity, which shows leaf expression and has been implicated in lignification (Ostergaard et al., 2000). Overall, evidence could be obtained for a peroxidase‐dependent oxidative burst in Arabidopsis and indirectly support a role for FBP1‐like peroxidases in the oxidative burst. However, this would have to be reconciled with the emerging reverse genetics data for NADPH oxidases in Arabidopsis. In this context the avr‐mediated oxidative burst in Arabidopsis is DPI sensitive (Grant et al., 2000b) as is harpin‐induced ROS production in Arabidopsis cell cultures (Desikan et al., 1996). On the other hand, treatment of Arabidopsis cell cultures with a Fusarium‐derived elicitor, shows an oxidative burst which is even more insensitive to DPI than elicitation of French bean cells (data not shown). Work elsewhere has also provided evidence that both sources can operate in Arabidopsis (Grant et al., 2000a). Could the former phenomenon be due to an NADPH oxidase and the latter response of cultured cells to fungal elicitor be dependent upon an apoplastic peroxidase? In that case the NADPH oxidase would function in highly specific R gene–avr gene interactions while the apoplastic peroxidase system would be placed in the realm of responses to elicitor molecules, thought by some to represent a general defence response. This discrimination between the two types of bursts has been proposed for quite some time (Baker and Orlandi, 1995). Fig. 2.  View largeDownload slide Increased susceptibility of Arabidopsis leaves inoculated with the virulent DC3000 strain of Pseudomonas syringae. Growth of bacteria and inoculation conditions were as described earlier (Volko et al., 1998). (A) Col‐O, (B) antisense peroxidase line 1.1; (C) antisense peroxidase line 1.2. Leaves were innoculated on one half with 104 cfu bacteria, the other half with 1 mM MgCl2. Symptoms are pictured at 72 h after inoculation. Line 1.1 was more susceptible than 1.2. Fig. 2.  View largeDownload slide Increased susceptibility of Arabidopsis leaves inoculated with the virulent DC3000 strain of Pseudomonas syringae. Growth of bacteria and inoculation conditions were as described earlier (Volko et al., 1998). (A) Col‐O, (B) antisense peroxidase line 1.1; (C) antisense peroxidase line 1.2. Leaves were innoculated on one half with 104 cfu bacteria, the other half with 1 mM MgCl2. Symptoms are pictured at 72 h after inoculation. Line 1.1 was more susceptible than 1.2. Component 2: the pH change The pH change is absolutely essential Figure 3 shows these changes in pH using the Universal pH indicator. Apoplastic fluid can be readily extracted from cell walls (Bolwell et al., 1995) and culture medium aspirated at 2 min intervals following the addition of elicitor and beginning at t0 (left to right on the microtitre plate). The pH of both can be seen to change towards alkaline pH as reported previously (Bolwell et al., 1995) in a sequential manner. Staining of the cells in real time can be observed in an identical manner to that seen for the apoplastic fluid, validating methodologies for measuring changes in its composition. Previous work (Bolwell et al., 1995, 1999) used buffering conditions and ionophores to demonstrate that the oxidative burst was absolutely dependent upon extracellular alkalinization. The oxidative burst does not occur if the pH of the medium is held at 6.0 by strong buffering following elicitation. As far as the elicitor‐induced burst is concerned it follows the same pH dependence as the peroxidase‐catalysed oxidation of cysteine in vitro (Bolwell et al., 1995). As the luminol reaction is pH‐dependent these data have been confirmed using the xylenol orange method of detection (Bindschedler et al., 2001) which is pH‐independent (data not shown). Further work has indicated that some of the signal transduction involved in the activation of the apoplastic oxidative burst upstream of the extracellular alkalinization is involved in opening the ion channels (Bolwell et al., 1999; Bindschedler et al., 2001). Treatment of cultured bean cells with forskolin and dibutyryl cAMP enhances the production of active oxygen species in response to elicitor and one aspect of this is a premature alkalinization compared with normal elicitation. In mammals, it has been shown that both cAMP and cGMP can stimulate cyclic nucleotide‐gated cation channels leading to an influx of calcium and, probably, a similar system exists in plants since both cyclic nucleotides induced an increase of Ca2+ in tobacco (Volotovski et al., 1998). Moreover cyclic nucleotide‐gated cation channels have been cloned in Arabidopsis (Leng et al., 1999; Clough et al., 2000) and tobacco (Arazi et al., 2000). It is premature to identify whether these are responsible for the K+ efflux or H+ influx as well as the well‐studied Ca2+ influx (Grant et al., 2000b) but it is an attractive link. Unfortunately, although cAMP and its elevation in response to forskolin have been demonstrated in bean (Bolwell, 1992), together with activation of the oxidative burst by cholera toxin implicating G‐protein involvement in transducing an initial receptor ligand interaction (Bindschedler et al., 2001), there is contentious molecular identification of the transducing proteins. Fig. 3.  View largeDownload slide pH changes of (A) apoplastic fluid; (B) culture medium from elicited French bean cells determined colorimetrically using Universal indicator (range 5–8). The pH increases from 6 at t0min to 7.2 at t10min in the apoplast. Fig. 3.  View largeDownload slide pH changes of (A) apoplastic fluid; (B) culture medium from elicited French bean cells determined colorimetrically using Universal indicator (range 5–8). The pH increases from 6 at t0min to 7.2 at t10min in the apoplast. Component 3: the ‘reductant’ The release of the substrate, which logically would be a reductant, is probably also dependent upon the pH change. The major approach to its identification in French bean has been the subfractionation of apoplastic fluid that can support the chemiluminescence assay in the presence of HRP (Bolwell et al., 1999). Preliminary studies showed that some of the active material could be removed from apoplastic fluid by Dowex MR3, Amberlite IR‐120(H) and CM‐cellulose suggesting a positive charge at pH 5.5 (data not shown). If the compound was absorbed onto MR3 it could be eluted with 0.25 M NaCl, but it showed no spectral peaks characteristic of cysteine or glutathione. Indeed there were no significant peaks in the UV or visible ranges. However, it cannot be entirely hydrogen peroxide since there is at least a 2 min or greater delay between the appearance of the active material in the apoplast and the measurement of H2O2 in the medium with which it would be instantaneously exchangeable. Furthermore, the active material could be partitioned entirely into ethyl acetate in which it was preferentially soluble. GC/MS of active fractions (Bolwell et al., 1999) showed the presence of palmitate and stearate and these were extensively investigated for their possible involvement in the oxidative burst. Effect of fatty acids on the generation of H2O2 in vitro Stimulation of the generation of products that could be detected by chemiluminescence from the interaction of fatty acids with peroxidase in vitro had all the hallmarks of an enzyme reaction. There are precedents for the generation of H2O2 from saturated fatty acids by flavoproteins in β‐oxidation and by haem proteins in α‐oxidation and, specifically, by the cytochrome P450 from Candida (CYP52), which has a homologue in Arabidopsis (CYP86A1). In the apparent reaction catalysed by peroxidase, there was specificity in that only 14.0 (myristate) and longer chain length fatty acids showed activity. 12.0 (laurate) and lower chain length supported no activity whatsoever. There were differences in that the product of myristate, palmitate, 16‐OH palmitate, and stearate was completely destroyed by saturating levels of catalase (50 units) in the assay, while the product of the reaction with linoleic and linolenic acids was partially catalase‐insensitive. Although these in vitro experiments using horseradish peroxidase also showed a reaction with linoleic (18.2) and linolenic (18.3) acids, these were not extracted into active fractions separated from total apoplastic fluid following elicitation, although they were present in apoplastic fluid from unelicited cells (Fig. 4). The partial sensitivity to catalase (Table 1) suggested the production of lipid hydroperoxides, which can be detected by the luminol assay. Ethanol, used to dissolve the fatty acid or aid dispersal in the buffer at higher concentrations of fatty acid, gave some signal itself, but there was a substantial increase in the presence of fatty acid added up to 500 μM over the signal given by the solvent alone. Above that concentration there were solubility problems. The apparent reaction had a pH optimum of 7.2 (data not shown), the optimum seen for the oxidative burst in vivo, and displayed Michaelis‐Menton kinetics (Table 1). Although when the products were analysed by LC/GC‐MS there was evidence of some peaks containing ion fragments at m/z 73 and 75, which would contain at least COOH or OH, these were low in abundance. Furthermore, there was no evidence of ethyl esters, dimers, alkanes or alkenes, which might be expected for an oxidative reaction. The reaction was also tested using radiolabelled palmitate and stearate and both TLC and RP‐HPLC analysis indicated the substrate was unchanged (J‐P Salaun, GP Bolwell, unpublished data). Therefore the effect of fatty acids upon peroxidase in vitro seems to be a physical one, possibly through the formation of micelles, but why this effect should have the appearance, kinetically, of an enzyme reaction is unknown. Fig. 4.  View largeDownload slide Changes in apoplastic oxylipins over the period of the oxidative burst in French bean. Apoplastic fluid was extracted at 0, 14 and 26 min following elicitor treatment (Bolwell, 1996; Bolwell et al., 1999). Cells were harvested under gentle vacuum and washed with water. Twenty mM sodium acetate pH 6 was then allowed to infiltrate the cell wall and the apoplastic fluid collected under gentle vacuum and analysed for linoleic, linolenic and oxylipin content (Weichert et al., 1999). Levels of the various components are shown for 18.3 and its derivatives, 13‐HOT, 16‐HOT, 12‐HOT and 9‐HOT oxylipins and for 18.2 and its derivatives 13‐HOD and 9‐HOD oxylipins. Fig. 4.  View largeDownload slide Changes in apoplastic oxylipins over the period of the oxidative burst in French bean. Apoplastic fluid was extracted at 0, 14 and 26 min following elicitor treatment (Bolwell, 1996; Bolwell et al., 1999). Cells were harvested under gentle vacuum and washed with water. Twenty mM sodium acetate pH 6 was then allowed to infiltrate the cell wall and the apoplastic fluid collected under gentle vacuum and analysed for linoleic, linolenic and oxylipin content (Weichert et al., 1999). Levels of the various components are shown for 18.3 and its derivatives, 13‐HOT, 16‐HOT, 12‐HOT and 9‐HOT oxylipins and for 18.2 and its derivatives 13‐HOD and 9‐HOD oxylipins. Table 1.  Apparent kinetic properties of fatty acid stimulation of the production of H2O2 by peroxidases in vitro Production of hydrogen peroxide was measured in the standard luminol assay using 0.25 U of horseradish peroxidase (HRP) as described previously (Bolwell et al., 1995). All reactions had a pH optimum of 7.2. Fatty acid   Vmax (nmol−1 min−1 UHRP)   % Inhibition (in the presence of 50 U catalase)   Km (mM)   Laurate (12.0)   0    0  0  Myristate (14.0)   4.73  100  0.542  Palmitate (16.0)   5.53  100  0.234  16HO‐palmitate   1.35  100  0.211  Stearate (18.0)   6.34  100  0.447  Linoleic (18.2)  10.95   75  0.207  Linolenic (18.3)  33.92   66  0.758  Fatty acid   Vmax (nmol−1 min−1 UHRP)   % Inhibition (in the presence of 50 U catalase)   Km (mM)   Laurate (12.0)   0    0  0  Myristate (14.0)   4.73  100  0.542  Palmitate (16.0)   5.53  100  0.234  16HO‐palmitate   1.35  100  0.211  Stearate (18.0)   6.34  100  0.447  Linoleic (18.2)  10.95   75  0.207  Linolenic (18.3)  33.92   66  0.758  View Large Fatty acid metabolism in vivo The addition of fatty acids at several stages over the operation of the oxidative burst had no effect on the amplitude of the burst and very limited effect on timing, perhaps bringing it forward by about 1 min when added at t0 (data not shown). Since there was a suggestion that peroxidase could produce lipid hydroxides in vitro these were also looked for in vivo. The signature oxylipins (Weichert et al., 1999) for oxidation were not detected in apoplastic fluid over the period of the apoplastic burst in French bean cells (Fig. 4) and although linolenic and linoleic acids could be detected in apoplastic fluid before elicitation, they disappeared over the period of elicitation and a portion seems to be esterified, not oxidized, at this time. However, subsequent oxidation of fatty acids is known to occur in defence responses in French bean. There were increases in cytochrome P450‐dependent ω‐hydroxylase activity towards saturated fatty acids (Bolwell et al., 1997) over intermediate time periods (0–4 h post‐elicitation). It is also well established that later events in response to pathogens in French bean involve lipoxygenase and the production of oxylipins. Cis‐3‐hexenol and trans‐2‐hexenal were produced, commencing at about 12–15 h following inoculation of bean leaves with Pseudomonas syringae, indicating that a pathway via 13‐HOT was operational (Croft et al., 1993). Further searches for the ‘reductant’ The difficulty in establishing a precise role for fatty acids in the apoplastic burst in French bean cells is therefore unresolved at present and has prompted a further search for compounds in active fractions in the apoplastic fluid. Figure 5 shows the specific appearance of four compounds just before the peak of hydrogen peroxide production relative to the patterns of components from unelicited cells and following the completion of the burst. These were identified by GC‐MS as glycerol, malate, citrate, and succinate. Although individually these did not promote the luminol reaction with horseradish peroxidase in vitro, combinations of the carboxylic acids did give a reaction at about 20% of the standard compound, cysteine at 50 μM. Perhaps more significantly they could signify increased Redox metabolic activity or in the case of malate, the operation of the Redox reaction first proposed in the 1970s (Elstner and Heupel, 1976). However, their contribution to the apoplastic oxidative burst is unresolved at present. Certainly there is no obvious one component for a reductant and there are a number of components identified as being released, which can support H2O2 production in vitro by peroxidases. The possibility that complex mixtures of low molecular compounds stimulate dismutation of complex III from peroxidases exists and would be very difficult to resolve. Fig. 5.  View largeDownload slide GC‐MS of active apoplastic fractions from (A) unelicited cells, (B) elicited cells just before the oxidative burst and (C) elicited cells following termination of the oxidative burst (Bolwell et al., 1999). Cells were harvested under gentle vacuum and washed with water. Twenty mM sodium acetate pH 6 was then allowed to infiltrate the cell wall and the apoplastic fluid collected under gentle vacuum. The samples were then analysed by HPLC followed by GC‐MS (Bolwell et al., 1999). The peaks present only in the period just before the oxidative burst are (1) glycerol; (2) succinate; (3) malate; and (4) citrate. Fig. 5.  View largeDownload slide GC‐MS of active apoplastic fractions from (A) unelicited cells, (B) elicited cells just before the oxidative burst and (C) elicited cells following termination of the oxidative burst (Bolwell et al., 1999). Cells were harvested under gentle vacuum and washed with water. Twenty mM sodium acetate pH 6 was then allowed to infiltrate the cell wall and the apoplastic fluid collected under gentle vacuum. The samples were then analysed by HPLC followed by GC‐MS (Bolwell et al., 1999). The peaks present only in the period just before the oxidative burst are (1) glycerol; (2) succinate; (3) malate; and (4) citrate. Features of role of the ‘reductant’ In the absence of identification of the low molecular weight compound, features of the dependence of its role on other factors were investigated using the in vitro model compound, cysteine (Pichorner et al., 1992; Bolwell et al., 1995). Figure 6A shows that when cysteine was added to cells at t0, the oxidative burst is considerably enhanced, but with a time‐dependence similar to the normal elicitation curve rather than a continuous production of ROS, showing that cysteine can only be oxidized with the production of H2O2 when the elicitor induces suitable conditions. Figure 6B and C show that cysteine can be oxidized only when conditions for the oxidative burst are optimum. Stimulation is still observed when added at 10 min, but not markedly when added as the peak of H2O2 production is passing. Figure 7 shows that the response with cysteine is concentration‐dependent, with a maximum at 200 μM and with a sharp decline at higher concentrations. This sharp cut‐off is reminiscent of the shape of the curve with HRP, albeit with an optimum of around 500 μM in this case (Pichorner et al., 1992). However, these observations are conducive to believe that a reductant can only serve as a substrate during the time frame of conditions brought about by elicitor action. Circumstantially, this would implicate the pH change. Fig. 6.  View largeDownload slide Time dependence of the cysteine enhanced oxidative burst. (A) Cysteine added to a final concentration of 200 μM at zero time, (B) 200 μM added at 10 min, (C) 200 μM added at 16 min. Batches of suspension‐cultured cells were treated with elicitor and 1 ml samples taken periodically and hydrogen peroxide measured using the standard luminol assay (Bolwell et al., 1995). Cysteine was added at the time points shown and the control values at the peak reflect conditions without cysteine and the maximum chemiluminescence was similar in each experiment. Each experiment has been repeated at least three times and a typical result is shown. Note: enhancement is only seen during the onset of the burst and the correct pH conditions. Fig. 6.  View largeDownload slide Time dependence of the cysteine enhanced oxidative burst. (A) Cysteine added to a final concentration of 200 μM at zero time, (B) 200 μM added at 10 min, (C) 200 μM added at 16 min. Batches of suspension‐cultured cells were treated with elicitor and 1 ml samples taken periodically and hydrogen peroxide measured using the standard luminol assay (Bolwell et al., 1995). Cysteine was added at the time points shown and the control values at the peak reflect conditions without cysteine and the maximum chemiluminescence was similar in each experiment. Each experiment has been repeated at least three times and a typical result is shown. Note: enhancement is only seen during the onset of the burst and the correct pH conditions. Fig. 7.  View largeDownload slide Concentration dependence of the elicitor‐induced oxidative burst for added cysteine. Cysteine was added to various final concentrations at zero time to batches of suspension‐cultured cells and the oxidative burst measured following elicitation using the standard luminol assay (Bolwell et al., 1995). Note that cysteine enhances the production of hydrogen peroxide at all concentrations used, in a similar manner to that shown in Fig. 6A with a narrow range of hyper‐enhancement around 200 μM. Fig. 7.  View largeDownload slide Concentration dependence of the elicitor‐induced oxidative burst for added cysteine. Cysteine was added to various final concentrations at zero time to batches of suspension‐cultured cells and the oxidative burst measured following elicitation using the standard luminol assay (Bolwell et al., 1995). Note that cysteine enhances the production of hydrogen peroxide at all concentrations used, in a similar manner to that shown in Fig. 6A with a narrow range of hyper‐enhancement around 200 μM. Aspects of signalling Evidence has accumulated that the apoplastic oxidative burst involves a three component system, a peroxidase capable of generating H2O2 at neutral pH, a pH change in the apoplast towards alkalinity, and the generation or release of a reductant and/or other substrate(s). All these are interdependent and there are additional signal components to those already described. The earliest signalling events involved in the oxidative burst are beginning to be elucidated. The increase of cytosolic Ca2+, which occurs within s of elicitation, is thought to be a primary signal essential for the subsequent down‐stream events. Down‐stream events following the increase in cytosolic calcium include the production of ROS (Grant et al., 2000b) and the induction of defence‐related genes such as genes encoding PR proteins and phytoalexin production (Blume et al., 2000). In the case of an infection with an avirulent pathogen, this resulted in the establishment of a local programmed cell death (PCD) at the site of infection (Grant et al., 2000b). It is possible that Ca2+ has a direct effect on the NADPH oxidase, since there is evidence that the plant enzyme has an N‐terminal sequence with two Ca2+ binding EF‐hand motifs (Torres et al., 1998; Keller et al., 1998). However, the basis of the Ca2+‐dependence of the apoplastic oxidative burst is unresolved at present (Bindschedler et al., 2001), but it is most likely to be related to the release of the substrate/reductant. Beside the early calcium influx into the cytosolic compartment, a rapid efflux of K+ and Cl− and extracellular alkalinization of elicited cell cultures has also been observed (Scheel, 1998; Fellbrich et al., 2000). Extracellular alkalinization is of course essential for the apoplastic oxidative burst (Bolwell et al., 1995). Data obtained with forskolin (Bindschedler et al., 2001) circumstantially implcates cAMP in the alkalinization component and the role of cAMP as a secondary signal in plants (Assmann, 1995; Bolwell, 1995). More recently, there are the exciting reports of the identification and characterization of a new gene family in Arabidopsis which shares features with cyclic nucleotide‐gated channels from animals and inward‐rectifying K+ channels from plants (Kohler et al., 1999). A plant cyclic nucleotide‐gated cation channel, AtCNGC2, from Arabidopsis has been cloned and its function characterized (Leng et al., 1999). By analogy with similar animal cation channels, AtCNGC2 is suggested as being involved in signal transduction by allowing a flux of Ca2+, K+ and other ions in the presence of cAMP or cGMP. Furthermore, the Arabidopsis dnd1 gene (defence, no death) has been isolated from a mutant line that failed to produce a hypersensitive response to the avirulent Pseudomonas pathogens and this gene has been found to encode the same ion channel protein, AtCNGC2 (Clough et al., 2000). Arazi et al. have isolated a tobacco plasma channel protein with a high affinity for calmodulin and a highly conserved cyclic nucleotide‐binding domain (Arazi et al., 2000). A transgenic tobacco expressing a dominant‐acting calmodulin hyperactivates a calmodulin‐dependent NAD kinase leading to an increased and more rapid production of ROS, an increased alkalinization of the medium and the initiation of a more rapid PCD (Harding et al., 1997; Harding and Roberts, 1998). Taken together, there is growing evidence for the involvement of cAMP in signal transduction and cross‐talk between the various pathways of plants responding to the attack by avirulent pathogens including activation of the apoplastic oxidative burst. 1 To whom correspondence should be addressed. Fax: +441784434326. E‐mail: uhbc006@vms.rhbnc.ac.uk 2 Present address: Department of Biological Sciences, California State University, Chico, CA 95929‐0515, USA. 3 Present address: School of Biological Sciences, University of Manchester, Manchester M13 9PT, UK. 4 Present address: Institute of Biochemistry and Biophysics, PO Box 30, Kazan 420503, Russia. KAB was funded by the BBSRC (UK), SLG by the Leverhulme Trust and FM was the recipient of a NATO fellowship through the Royal Society (UK). We thank Dr Ivo Fuessner, Institute of Plant Genetics and Crop Plant Research, Gatersleben, Germany for performing the oxylipin GC/MS and Professor Jean‐Pierre Salaun, Roscoff, France, for providing standards and analysis of reactions involving radiolabelled fatty acids. References Arazi T, Kaplan B, Fromm H. 2000. A high affinity calmodulin‐binding site in a tobacco plasma membrane channel protein coincides with a characteristic element of cyclic‐nucleotide‐binding domains. Plant Molecular Biology  42, 591–601. Google Scholar Asada K. 1999. The water–water cycle in chloroplasts: scavenging of active oxygens and dissipation of excess photons. Annual Review of Plant Physiology and Plant Molecular Biology  50, 601–639. Google Scholar Assmann SM. 1995. Cyclic‐AMP as a 2nd messenger in higher‐plants—status and future prospects. Plant Physiology  108, 885–889. Google Scholar Auh CK, Murphy TM. 1995. Plasma membrane redox enzyme is involved in the synthesis of \({\mathrm{O}_{2}^{{\bullet}{-}}}\) and H2O2 by Phytophthora elicitor‐simulated rose cells. Plant Physiology  107, 1241–1247. Google Scholar Baker CJ, Orlandi EW. 1995. Active oxygen in plant pathogenesis. Annual Review of Phytopathology  33, 299–321. Google Scholar Bestwick CS, Brown IR, Mansfield JW. 1998. Localized change in peroxidase activity accompanies hydrogen peroxide generation during the development of a non‐host hypersensitive reaction in lettuce. Plant Physiology  118, 1067–1078. Google Scholar Bindschedler LV, Minibayeva F, Gardner SL, Gerrish C, Davies DR, Bolwell GP. 2001. Early signalling events in the apoplastic oxidative burst in suspension‐cultured French bean cells involve cAMP and Ca2+. New Phytologist  151, 185–194. Google Scholar Blee KA, Jupe SC, Richard G, Bolwell GP. 2001. Molecular identification and expression of the peroxidase responsible for the oxidative burst in French bean (Phaseolus vulgaris L.) and related members of the gene family. Plant Molecular Biology  47, 607–620. Google Scholar Blume B, Nürnberger T, Nass N, Scheel D. 2000. Receptor‐mediated increase in cytoplasmic free calcium required for activation of pathogen defense in parsley. The Plant Cell  12, 1425–1440. Google Scholar Bolwell GP. 1992. A role for phosphorylation in the down‐regulation of phenylalanine ammonia‐lyase in suspension‐cultured cells of French bean. Phytochemistry  31, 4081–4086. Google Scholar Bolwell GP. 1995. Cyclic AMP the reluctant messenger in plant cells. Trends in Biochemical Sciences  20, 492–495. Google Scholar Bolwell GP. 1996. The origin of the oxidative burst in plants. Biochemical Society Transactions  24, 438–442. Google Scholar Bolwell GP. 1999. Role of active oxygen species and NO in plant defence responses. Current Opinion in Plant Biology  2, 287–294. Google Scholar Bolwell GP, Blee KA, Butt VS, Davies DR, Gardner SL, Gerrish C, Minibayeva F, Rowntree EG, Wojtaszek P. 1999. Recent advances in understanding the origin of the apoplastic oxidative burst in plant cells. Free Radical Research  31, S137–145. Google Scholar Bolwell GP, Butt VS, Davies DR, Zimmerlin A. 1995. The origin of the oxidative burst in plants. Free Radical Research  23, 517–532. Google Scholar Bolwell GP, Davies DR, Gerrish C, Auh CK, Murphy TM. 1998. Comparative biochemistry of the oxidative burst produced by rose and French bean cells reveals two distinct mechanisms. Plant Physiology  116, 1379–1385. Google Scholar Bolwell GP, Gerrish C, Salaun J‐P. 1997. Changes in enzymes involved in suberization in elicitor‐treated French bean cells. Phytochemistry  45, 1351–1357. Google Scholar Bolwell GP, Wojtaszek P. 1997. Mechanisms for the generation of reactive oxygen species in plant defence—a broad perspective. Physiological and Molecular Plant Pathology  51, 347–366. Google Scholar Brown I, Trethowan J, Kerry M, Mansfield JW, Bolwell GP. 1998. Location of components of the oxidative cross‐linking of glycoproteins and callose synthesis in papillae formed during the interaction between non‐pathogenic strains of Xanthomonas campestris and French bean mesophyll cells. The Plant Journal  15, 333–343. Google Scholar Clough SJ, Fengler KA, I‐Ching Y, Lippok B, Smith RK, Bent AF. 2000. The Arabidopsis dnd1 ‘defense, not death’ gene encodes a mutated cyclic nucleotide gated ion channel. Proceedings of the National Academy of Sciences, USA  97, 9323–9328. Google Scholar Croft KPC, Juttner F, Sluzarenko AJ. 1993. Volatile products of the lipoxygenase pathway evolved from Phaseolus vulgaris (L.) leaves inoculated with Pseudomonas syringae pv. phaseolicola. Plant Physiology  101, 13–24. Google Scholar del Río LA, Corpas FJ, Sandalio LM, Palma JM, Gómez M, Barroso JB. 2002. Reactive oxygen species, antioxidant systems and nitric oxide in peroxisome. Journal of Experimental Botany  53, 1255–1272. Google Scholar Desikan R, Hancock JT, Coffey MJ, Neill SJ. 1996. Generation of active oxygen in elicited cells of Arabidopsis thaliana is mediated by a NADPH oxidase‐like enzyme. FEBS Letters  382, 13–17. Google Scholar Elstner EF, Heupel A. 1976. Formation of hydrogen peroxide by isolated cell walls from horseradish (Armoracia lapatholia). Planta  130, 175–180. Google Scholar Fellbrich G, Blume B, Brunner F, Hirt H, Kroj T, Ligterink W, Romanski A, Nürnberger T. 2000. Phytophthora parasitica elicitor‐induced reactions in cells of Petroselinum crispum. Plant Cell Physiology  41, 692–701. Google Scholar Grant JJ, Yun B‐W, Loake GJ. 2000a. Oxidative burst and cognate redox signaling reported by luciferase imaging: identification of a signal network that functions independently of ethylene, SA and Me‐JA but is dependent on MAPKK activity. The Plant Journal  24, 569–582. Google Scholar Grant M, Brown I, Adams S, Knight M, Ainslie A, Mansfield J. 2000b. The RPM1 plant disease resistance gene facilitates a rapid and sustained increase in cytosolic calcium that is necessary for the oxidative burst and hypersensitive cell death. The Plant Journal  23, 441–450. Google Scholar Harding SA, Oh SH, Roberts DM. 1997. Transgenic tobacco expressing a foreign calmodulin gene shows an enhanced production of active oxygen species. The EMBO Journal  16, 1137–1144. Google Scholar Harding SA, Roberts DM. 1998. Incompatible pathogen infection results in enhanced reactive oxygen and cell death responses in transgenic tobacco expressing a hyperactive mutant calmodulin. Planta  206, 253–258. Google Scholar Keller T, Damude HG, Werner D, Doerner P, Dixon RA, Lamb C. 1998. A plant homolog of the neutrophil NADPH oxidase gp91phox subunit gene encodes a plasma membrane protein with Ca2+ binding motifs. The Plant Cell  10, 255–266. Google Scholar Kohler C, Merkle T, Neuhaus G. 1999. Characterization of a novel gene family of putative cyclic nucleotide‐ and calmodulin‐regulated ion channels in Arabidopsis thaliana. The Plant Journal  18, 97–104. Google Scholar Lamb C, Dixon RA. 1997. The oxidative burst in plant disease resistance. Annual Review of Plant Physiology and Plant Molecular Biology  48, 251–275. Google Scholar Leng Q, Mercier RW, Yao W, Berkowitz GA. 1999. Cloning and first function characterization of a plant cyclic nucleotide‐gated cation channel. Plant Physiology  121, 753–761. Google Scholar Martinez C, Montillet JL, Bresson E, Angel JP, Dai GH, Daniel JF, Geiger JP. 1998. Apoplastic peroxidase generates superoxide anions in cells of cotton cotyledons undergoing the hypersensitive reaction to Xanthomonas campestris pv. malvacerarum race 18. Molecular Plant–Microbe Interactions  11, 1038–1047. Google Scholar McLusky SR, Bennett MH, Beale MH, Lewis MJ, Gaskin P, Mansfield JW. 1999. Cell wall alterations and localized accumulation of feruloyl 3′‐tyramine in onion epidermis at sites of attempted penetration by Botrytis allii are associated with actin polarization, peroxidase activity and suppression of flavonoid biosynthesis. The Plant Journal  17, 523–534. Google Scholar Ostergaard L, Teilum K, Mirza O, Petersen M, Welinder KG, Mundy J, Gajhede M, Hendriksen A. 2000. Arabidopsis ATPA2 peroxidase: expression and high‐resolution structure of a plant peroxidase with implications for lignification. Plant Molecular Biology  44, 231–243. Google Scholar Pichorner H, Couperus A, Korori SAA, Ebermann R. 1992. Plant peroxidase has a thiol oxidase function. Phytochemistry  31, 3371–3376. Google Scholar Robertson D, Davies DR, Gerrish C, Jupe SC, Bolwell GP. 1995. Rapid changes in oxidative metabolism as a consequence of elicitor treatment of suspension‐cultured cells of French bean (Phaseolus vulgaris L.). Plant Molecular Biology  27, 59–67. Google Scholar Scheel D. 1998. Resistance response physiology and signal transduction. Current Opinion in Plant Biology  1, 305–310. Google Scholar Torres MA, Onouchi H, Hamada S, Machida C, Hammond‐Kosack KE, Jones JD. 1998. Six Arabidopsis thaliana homologues of the human respiratory burst oxidase (gp91phox). The Plant Journal  14, 365–370. Google Scholar Volko SM, Boller T, Ausubel FM. 1998. Isolation of new Arabidopsis mutants with enhanced diseased susceptibility to Pseudomonas syringae by direct screening. Genetics  149, 537–548. Google Scholar Volotovski ID, Sokolovsky SG, Molchan OV, Knight MR. 1998. Second messengers mediate increases in cytosolic calcium in tobacco protoplasts. Plant Physiology  117, 1023–1030. Google Scholar Weichert H, Stenzel I, Berndt E, Wasternack C, Fuessner I. 1999. Metabolic profiling of oxylipins upon salicylate treatment in barley leaves—preferential induction of the reductase pathway by salicylate. FEBS Letters  464, 133–137. Google Scholar Woo EJ, Dunwell JM, Goodenough PW, Marvier AC, Pickersgill RW. 2000. Germin is a manganese containing homohexamer with oxalate oxidase and superoxide dismutase activities. Nature Structural Biology  11, 1036–1104. Google Scholar Zimmerlin A, Wojtaszek P, Bolwell GP. 1994. Synthesis of dehydrogenation polymers of ferulic acid with high specificity by a purified cell‐wall peroxidase from French bean (Phaseolus vulgaris L.). Biochemical Journal  94, 747–753. Google Scholar © Society for Experimental Biology http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Experimental Botany Oxford University Press

The apoplastic oxidative burst in response to biotic stress in plants: a three‐component system

Download PDF
10 pages

Loading next page...
 
/lp/oxford-university-press/the-apoplastic-oxidative-burst-in-response-to-biotic-stress-in-plants-gcPzXsw6tT

References (45)

Publisher
Oxford University Press
Copyright
© Society for Experimental Biology
ISSN
0022-0957
eISSN
1460-2431
DOI
10.1093/jexbot/53.372.1367
Publisher site
See Article on Publisher Site

Abstract

Abstract The oxidative burst, the generation of reactive oxygen species (ROS) in response to microbial pathogen attack, is a ubiquitous early part of the resistance mechanisms of plant cells. It has also become apparent from the study of a number of plant–pathogen interactions and those modelled by elicitor treatment of cultured cells that there may be more than one mechanism operating. However, one mechanism may be dominant in any given species. NADPH oxidases have been implicated in a number of systems and have been cloned and characterized. However, the enzyme system which is the major source of ROS in French bean (Phaseolus vulgaris) cells treated with a cell wall elicitor from Colletotrichum lindemuthianum, appears to be dependent on an exocellular peroxidase. The second component, the extracellular alkalinization, occurs as a result of the Ca2+ and proton influxes and the K+ efflux common to most elicitation systems as one of the earliest responses. The third component, the actual reductant/substrate, has remained elusive. The low molecular weight compound composition of apoplastic fluid was compared before and after elicitation. The substrate only becomes available some min after elicitation and can be extracted, so that by comparing the profiles by LC‐MS it has been possible to identify possible substrates. The mechanism has proved to be complex and may involve a number of low molecular weight components. Stimulation of H2O2 production was observed with saturated fatty acids such as palmitate and stearate without concomitant oxylipin production. This biochemical evidence is supported by immunolocalization studies on papillae forming at bacterial infection sites that show the peroxidase isoform present at sites of H2O2 production revealed by cerium chloride staining together with the cross‐linked wall proteins and callose and callose synthase. The peroxidase has been cloned and expressed in Pichia pastoris and has been shown to catalyse the oxidation reaction with the same kinetics as the purified enzyme. Furthermore, Arabidopsis plants transformed heterologously using the French bean peroxidase in antisense orientation have proved to be highly susceptible to bacterial and fungal pathogens. Thus it is possible that Arabidopsis is another species with the potential to mount an apoplastic oxidative burst and these transformed plant lines may be useful to identify the peroxidase that is responsible. Cell wall, fatty acids, French bean, hydrogen peroxide, oxidative burst, oxylipins, peroxidase, Phaseolus vulgaris L. DPI, diphenylene iodonium, FBP1, French bean peroxidase isoform 1, HR, hypersensitive response, HRP, horseradish peroxidase, PSI, photosystem I, PSII, photosystem II, ROS, reactive oxygen species. Introduction Reactive oxygen species (ROS) are rapidly produced in plants as a defence response to pathogen attack (Bolwell and Wojtaszek, 1997; Bolwell, 1999). There are a number of potential sources of ROS. It has been well recognized that protoplastic sources centred upon mitochondrial, chloroplastic or peroxisomal generating systems exist, but have mainly been studied in relation to abiotic stresses (Asada, 1999; del Rio et al., 2002). Moving towards the cell surface, in many species it is thought that the superoxide anion can be originated at the plasma membrane in response to biotic stresses from an NADPH oxidase (Lamb and Dixon, 1997) analogous to the system which occurs in neutrophils or from an NADH‐dependent superoxide synthase (Auh and Murphy, 1995). Superoxide is subsequently dismutated to hydrogen peroxide by superoxide dismutase. There are potential sources of apoplastic origin of hydrogen peroxide including peroxidases, amine oxidases and oxalate oxidases. In elicitor‐treated French bean cells hydrogen peroxide appears to originate from a cell wall peroxidase in a three component system requiring ion fluxes and leading to extracellular alkalinization and the release of a reductant (Robertson et al., 1995; Bolwell et al., 1995, 1999; Bolwell, 1996). The most widely studied potential mechanism is a system analogous to the mammalian NADPH oxidase system. Much of the evidence for this was based upon the inhibition of defence responses by DPI, a reasonably specific inhibitor of the mammalian oxidative burst. While there was a lack of direct biochemical characterization of an enzyme complex from plant cells (Bolwell, 1999), confirmation of the existence of the plant NADPH oxidase came from the molecular cloning of the plant analogue of gp91phox the core polypeptide of the mammalian enzyme, which also showed considerable homologies to ferric reductases of plants and fungi involved in iron uptake (Keller et al., 1998; Torres et al., 1998). Examination of the primary structure of the plant enzymes showed important regulatory differences to the mammalian enzyme, which is activated by cytosolic polypeptides. In the case of the plant protein there is an additional N‐terminus with EF‐hands suggesting direct activation by Ca2+ ions (Lamb and Dixon, 1997). Eight genes coding for the Arabidopsis homologues of mammalian NADPH oxidases (Atrobh A‐H) have been characterized. Dissecting the role of each of these in generating ROS in Arabidopsis has recently been advanced by reverse genetics (JDG Jones, personal communication). Transposon tagged lines were screened for insertions in all eight genes. Individual knockouts have no effect on the defence response. Crosses were performed and one double knockout was particularly significant. When challenged by Peronospora, this line was found by diaminobenzidine staining not to produce localized ROS accumulation, suggesting this class of protein makes a major contribution to ROS production. However, this line was still disease‐resistant. Two other potential mechanisms are also worthy of note, one because it deserves further exploration as a general mechanism and the second because it is another example of specialization in the ability to generate H2O2 in response to pathogen attack. Protoplast sources of ROS have, of course, been extensively studied in relation to abiotic stress, particularly from chloroplasts (Asada, 1999) and peroxisomes (del Rio et al., 2002). However, due to the restructuring of the cytoskeleton and transport of vesicles to the site of interaction of the pathogen with the host cell wall (McLusky et al., 1999), this redirection of cell material for papilla deposition also appears to be accompanied by larger organelles which are often observed in the vicinity. Potentially, chloroplasts, through PSI leakage, could generate superoxide and from PSII, singlet oxygen (Asada, 1999), while peroxisomes could generate superoxide from xanthine oxidase and three additional membrane‐bound polypeptides (del Rio, 2002). The specialized family of germin‐like proteins (Woo et al., 2000) contain the oxalate oxidases in cereals which can generate H2O2 either from oxalate under acidic conditions that might arise from vacuolar damage caused by pathogen attack. More recently they have been shown to have superoxide dismutase activity at more neutral pH amplifying a role in the generation of H2O2 under a wider range of cellular conditions. The major subject of this review, generation of ROS by apoplastic peroxidases has been studied in French bean. Critical to the formation of an hypothesis of an alternative source of hydrogen peroxide to NAD(P)H oxidases were the observations that the NADH/NAD ratio and ATP level transiently fall, at the same time as an increased oxygen uptake occurs, while the NADPH level remains constant (Robertson et al., 1995; Bolwell, 1996). In addition, the oxidative burst in bean cells was inhibited by KCN, whereas the burst is insensitive to cyanide in mammalian cells. Moreover, its relative insensitivity to diphenylene iodonium (DPI), a well‐known inhibitor of the NADPH‐oxidase, favours the hypothesis of an apoplastic peroxidase generating the ROS in French bean (Bolwell et al., 1998). Most haem proteins are capable of generating H2O2 at alkaline pH in a mechanism that involves the formation of compound III (FeII‐O‐O). In the case of horseradish peroxidase compound III can readily be reduced with a pH optimum of 8.5 by many reductants of which thiols are the most active. In the case of bean cells the particular peroxidase responsible for the oxidative burst has been purified from cell walls and has an Mr of 46000 and exhibits a pH optimum of 7.2 for the oxidation of cysteine as a model compound. Molecular cloning of this peroxidase isoform has been achieved and the cognate protein expressed in Pichia pastoris to model the reaction in vitro (Blee et al., 2001). Other systems that appear to be capable of this type of mechanism are lettuce (Bestwick et al., 1998), cotton (Martinez et al., 1998), onion (McLusky et al., 1999), and possibly Arabidopsis (Grant et al., 2000a). Component 1: the peroxidase Purification, molecular cloning and modelling The oxidative burst peroxidase was purified as an Mr 46000 isoform from walls of suspension‐cultured French bean cells (Zimmerlin et al., 1994). The cDNA was isolated using a strategy that involved the design of oligonucleotides corresponding to peptides close to the N‐ and C‐termini of the 46 000 kDa FBP1 and isolation of a number of PCR products. Sequencing was used to identify a clone that coded for an identical amino acid sequence corresponding to the sequence data obtained for FBP1 purified from cell walls of suspension‐cultured cells of French bean (Zimmerlin et al., 1994) together with the extensive additional sequences. Screening of a cDNA library corresponding to mRNA from elicitor‐induced cells resulted in the acquisition of a cDNA that contains sequence coding for the whole of the mature protein (Blee et al., 2001). FBP1 is cationic and differs from other peroxidases in its ability to generate H2O2 at relatively neutral pH. However, modelling its 3‐D structure using coordinates for barley, horseradish and peanut peroxidase (Fig. 1) did not show any striking feature that could account for these different properties in generating H2O2. However, multiple potential N‐linked glycosylation sites were observed on FBP1 and the surface charge was near neutral even though it migrates as a cationic protein in IEF. These glycosylation and surface charge differences among enzymes of a gene family could define subcellular location and substrate interactions and therefore further elucidate gene member roles in planta beyond expression patterns. Recombinant FBP1 was produced in Pichia pastoris and shown to catalyse the cysteine oxidation reaction at pH 7.2, which further confirmed FBP1 as the peroxidase responsible for the apoplastic oxidative burst (Blee et al., 2001). FBP1 was immunolocalized to infection sites during the interaction of non‐pathogenic strains of Xanthomonas campestris and French bean mesophyll cells. This interaction produces massive papillae aiding structural studies. The FBP1 co‐localized with H2O2 detected by cerium chloride staining in material surrounding the bacteria and nearby wall (Brown et al., 1998). Also co‐localization of other components of papillae, a cross‐linked anti‐microbial arabinogalactan protein and callose was demonstrated together with the callose synthase. This is in accordance with one hypothesis of peroxidase involvement and H2O2 generation being required for the construction of barriers at infection sites. It is also strong circumstantial microscopic evidence that FBP1 is responsible for H2O2 generation at sites of infection. Thus there is good biochemical and structural evidence for a peroxidase‐dependent oxidative burst in French bean cells in suspension culture and in leaves. Unfortunately French bean is not amenable to reverse genetics to explore this functionality further and to confirm this role of FBP1 in planta, so studies have been instigated in Arabidopsis. Fig. 1.  View largeDownload slide Contour mapping of the surface of the French bean peroxidase 1 (FBP1) responsible for the apoplastic oxidative burst. X‐ray crystallography co‐ordinates for barley peroxidase, horseradish peroxidase C, and peanut peroxidase were used to model the amino acid sequences of the predicted mature polypeptide for FBP1 as described previously (Blee et al., 2001). The molecules were all aligned similarly showing a view with the pore leading interiorly to the haem group located at the centre. The model displays surface charge; white, neutral; blue, positive; red, negative. Fig. 1.  View largeDownload slide Contour mapping of the surface of the French bean peroxidase 1 (FBP1) responsible for the apoplastic oxidative burst. X‐ray crystallography co‐ordinates for barley peroxidase, horseradish peroxidase C, and peanut peroxidase were used to model the amino acid sequences of the predicted mature polypeptide for FBP1 as described previously (Blee et al., 2001). The molecules were all aligned similarly showing a view with the pore leading interiorly to the haem group located at the centre. The model displays surface charge; white, neutral; blue, positive; red, negative. Transformation of French bean peroxidase in the antisense orientation into Arabidopsis One way of confirming the functionality of cognate proteins is to down‐regulate them through antisense or partial sense expression in transgenic plants. Arabidopsis plants transformed heterologously using the FBP1 in antisense orientation have proved to be highly susceptible to the DC3000 virulent strains of Pseudomonas syringae (Fig. 2) as well as avirulent strains of this bacterium and a number of fungal pathogens (GP Bolwell, KA Blee, F Ausubel, unpublished data). Work is currently underway trying to identify the endogenous Arabidopsis peroxidase that is down‐regulated. However, the highest scoring Arabidopsis thaliana peroxidase to the oxidative burst peroxidase FBP1 was Arabidopsis thaliana peroxidase A2 (ATPA2), accession X99952 at 61% amino acid identity, which shows leaf expression and has been implicated in lignification (Ostergaard et al., 2000). Overall, evidence could be obtained for a peroxidase‐dependent oxidative burst in Arabidopsis and indirectly support a role for FBP1‐like peroxidases in the oxidative burst. However, this would have to be reconciled with the emerging reverse genetics data for NADPH oxidases in Arabidopsis. In this context the avr‐mediated oxidative burst in Arabidopsis is DPI sensitive (Grant et al., 2000b) as is harpin‐induced ROS production in Arabidopsis cell cultures (Desikan et al., 1996). On the other hand, treatment of Arabidopsis cell cultures with a Fusarium‐derived elicitor, shows an oxidative burst which is even more insensitive to DPI than elicitation of French bean cells (data not shown). Work elsewhere has also provided evidence that both sources can operate in Arabidopsis (Grant et al., 2000a). Could the former phenomenon be due to an NADPH oxidase and the latter response of cultured cells to fungal elicitor be dependent upon an apoplastic peroxidase? In that case the NADPH oxidase would function in highly specific R gene–avr gene interactions while the apoplastic peroxidase system would be placed in the realm of responses to elicitor molecules, thought by some to represent a general defence response. This discrimination between the two types of bursts has been proposed for quite some time (Baker and Orlandi, 1995). Fig. 2.  View largeDownload slide Increased susceptibility of Arabidopsis leaves inoculated with the virulent DC3000 strain of Pseudomonas syringae. Growth of bacteria and inoculation conditions were as described earlier (Volko et al., 1998). (A) Col‐O, (B) antisense peroxidase line 1.1; (C) antisense peroxidase line 1.2. Leaves were innoculated on one half with 104 cfu bacteria, the other half with 1 mM MgCl2. Symptoms are pictured at 72 h after inoculation. Line 1.1 was more susceptible than 1.2. Fig. 2.  View largeDownload slide Increased susceptibility of Arabidopsis leaves inoculated with the virulent DC3000 strain of Pseudomonas syringae. Growth of bacteria and inoculation conditions were as described earlier (Volko et al., 1998). (A) Col‐O, (B) antisense peroxidase line 1.1; (C) antisense peroxidase line 1.2. Leaves were innoculated on one half with 104 cfu bacteria, the other half with 1 mM MgCl2. Symptoms are pictured at 72 h after inoculation. Line 1.1 was more susceptible than 1.2. Component 2: the pH change The pH change is absolutely essential Figure 3 shows these changes in pH using the Universal pH indicator. Apoplastic fluid can be readily extracted from cell walls (Bolwell et al., 1995) and culture medium aspirated at 2 min intervals following the addition of elicitor and beginning at t0 (left to right on the microtitre plate). The pH of both can be seen to change towards alkaline pH as reported previously (Bolwell et al., 1995) in a sequential manner. Staining of the cells in real time can be observed in an identical manner to that seen for the apoplastic fluid, validating methodologies for measuring changes in its composition. Previous work (Bolwell et al., 1995, 1999) used buffering conditions and ionophores to demonstrate that the oxidative burst was absolutely dependent upon extracellular alkalinization. The oxidative burst does not occur if the pH of the medium is held at 6.0 by strong buffering following elicitation. As far as the elicitor‐induced burst is concerned it follows the same pH dependence as the peroxidase‐catalysed oxidation of cysteine in vitro (Bolwell et al., 1995). As the luminol reaction is pH‐dependent these data have been confirmed using the xylenol orange method of detection (Bindschedler et al., 2001) which is pH‐independent (data not shown). Further work has indicated that some of the signal transduction involved in the activation of the apoplastic oxidative burst upstream of the extracellular alkalinization is involved in opening the ion channels (Bolwell et al., 1999; Bindschedler et al., 2001). Treatment of cultured bean cells with forskolin and dibutyryl cAMP enhances the production of active oxygen species in response to elicitor and one aspect of this is a premature alkalinization compared with normal elicitation. In mammals, it has been shown that both cAMP and cGMP can stimulate cyclic nucleotide‐gated cation channels leading to an influx of calcium and, probably, a similar system exists in plants since both cyclic nucleotides induced an increase of Ca2+ in tobacco (Volotovski et al., 1998). Moreover cyclic nucleotide‐gated cation channels have been cloned in Arabidopsis (Leng et al., 1999; Clough et al., 2000) and tobacco (Arazi et al., 2000). It is premature to identify whether these are responsible for the K+ efflux or H+ influx as well as the well‐studied Ca2+ influx (Grant et al., 2000b) but it is an attractive link. Unfortunately, although cAMP and its elevation in response to forskolin have been demonstrated in bean (Bolwell, 1992), together with activation of the oxidative burst by cholera toxin implicating G‐protein involvement in transducing an initial receptor ligand interaction (Bindschedler et al., 2001), there is contentious molecular identification of the transducing proteins. Fig. 3.  View largeDownload slide pH changes of (A) apoplastic fluid; (B) culture medium from elicited French bean cells determined colorimetrically using Universal indicator (range 5–8). The pH increases from 6 at t0min to 7.2 at t10min in the apoplast. Fig. 3.  View largeDownload slide pH changes of (A) apoplastic fluid; (B) culture medium from elicited French bean cells determined colorimetrically using Universal indicator (range 5–8). The pH increases from 6 at t0min to 7.2 at t10min in the apoplast. Component 3: the ‘reductant’ The release of the substrate, which logically would be a reductant, is probably also dependent upon the pH change. The major approach to its identification in French bean has been the subfractionation of apoplastic fluid that can support the chemiluminescence assay in the presence of HRP (Bolwell et al., 1999). Preliminary studies showed that some of the active material could be removed from apoplastic fluid by Dowex MR3, Amberlite IR‐120(H) and CM‐cellulose suggesting a positive charge at pH 5.5 (data not shown). If the compound was absorbed onto MR3 it could be eluted with 0.25 M NaCl, but it showed no spectral peaks characteristic of cysteine or glutathione. Indeed there were no significant peaks in the UV or visible ranges. However, it cannot be entirely hydrogen peroxide since there is at least a 2 min or greater delay between the appearance of the active material in the apoplast and the measurement of H2O2 in the medium with which it would be instantaneously exchangeable. Furthermore, the active material could be partitioned entirely into ethyl acetate in which it was preferentially soluble. GC/MS of active fractions (Bolwell et al., 1999) showed the presence of palmitate and stearate and these were extensively investigated for their possible involvement in the oxidative burst. Effect of fatty acids on the generation of H2O2 in vitro Stimulation of the generation of products that could be detected by chemiluminescence from the interaction of fatty acids with peroxidase in vitro had all the hallmarks of an enzyme reaction. There are precedents for the generation of H2O2 from saturated fatty acids by flavoproteins in β‐oxidation and by haem proteins in α‐oxidation and, specifically, by the cytochrome P450 from Candida (CYP52), which has a homologue in Arabidopsis (CYP86A1). In the apparent reaction catalysed by peroxidase, there was specificity in that only 14.0 (myristate) and longer chain length fatty acids showed activity. 12.0 (laurate) and lower chain length supported no activity whatsoever. There were differences in that the product of myristate, palmitate, 16‐OH palmitate, and stearate was completely destroyed by saturating levels of catalase (50 units) in the assay, while the product of the reaction with linoleic and linolenic acids was partially catalase‐insensitive. Although these in vitro experiments using horseradish peroxidase also showed a reaction with linoleic (18.2) and linolenic (18.3) acids, these were not extracted into active fractions separated from total apoplastic fluid following elicitation, although they were present in apoplastic fluid from unelicited cells (Fig. 4). The partial sensitivity to catalase (Table 1) suggested the production of lipid hydroperoxides, which can be detected by the luminol assay. Ethanol, used to dissolve the fatty acid or aid dispersal in the buffer at higher concentrations of fatty acid, gave some signal itself, but there was a substantial increase in the presence of fatty acid added up to 500 μM over the signal given by the solvent alone. Above that concentration there were solubility problems. The apparent reaction had a pH optimum of 7.2 (data not shown), the optimum seen for the oxidative burst in vivo, and displayed Michaelis‐Menton kinetics (Table 1). Although when the products were analysed by LC/GC‐MS there was evidence of some peaks containing ion fragments at m/z 73 and 75, which would contain at least COOH or OH, these were low in abundance. Furthermore, there was no evidence of ethyl esters, dimers, alkanes or alkenes, which might be expected for an oxidative reaction. The reaction was also tested using radiolabelled palmitate and stearate and both TLC and RP‐HPLC analysis indicated the substrate was unchanged (J‐P Salaun, GP Bolwell, unpublished data). Therefore the effect of fatty acids upon peroxidase in vitro seems to be a physical one, possibly through the formation of micelles, but why this effect should have the appearance, kinetically, of an enzyme reaction is unknown. Fig. 4.  View largeDownload slide Changes in apoplastic oxylipins over the period of the oxidative burst in French bean. Apoplastic fluid was extracted at 0, 14 and 26 min following elicitor treatment (Bolwell, 1996; Bolwell et al., 1999). Cells were harvested under gentle vacuum and washed with water. Twenty mM sodium acetate pH 6 was then allowed to infiltrate the cell wall and the apoplastic fluid collected under gentle vacuum and analysed for linoleic, linolenic and oxylipin content (Weichert et al., 1999). Levels of the various components are shown for 18.3 and its derivatives, 13‐HOT, 16‐HOT, 12‐HOT and 9‐HOT oxylipins and for 18.2 and its derivatives 13‐HOD and 9‐HOD oxylipins. Fig. 4.  View largeDownload slide Changes in apoplastic oxylipins over the period of the oxidative burst in French bean. Apoplastic fluid was extracted at 0, 14 and 26 min following elicitor treatment (Bolwell, 1996; Bolwell et al., 1999). Cells were harvested under gentle vacuum and washed with water. Twenty mM sodium acetate pH 6 was then allowed to infiltrate the cell wall and the apoplastic fluid collected under gentle vacuum and analysed for linoleic, linolenic and oxylipin content (Weichert et al., 1999). Levels of the various components are shown for 18.3 and its derivatives, 13‐HOT, 16‐HOT, 12‐HOT and 9‐HOT oxylipins and for 18.2 and its derivatives 13‐HOD and 9‐HOD oxylipins. Table 1.  Apparent kinetic properties of fatty acid stimulation of the production of H2O2 by peroxidases in vitro Production of hydrogen peroxide was measured in the standard luminol assay using 0.25 U of horseradish peroxidase (HRP) as described previously (Bolwell et al., 1995). All reactions had a pH optimum of 7.2. Fatty acid   Vmax (nmol−1 min−1 UHRP)   % Inhibition (in the presence of 50 U catalase)   Km (mM)   Laurate (12.0)   0    0  0  Myristate (14.0)   4.73  100  0.542  Palmitate (16.0)   5.53  100  0.234  16HO‐palmitate   1.35  100  0.211  Stearate (18.0)   6.34  100  0.447  Linoleic (18.2)  10.95   75  0.207  Linolenic (18.3)  33.92   66  0.758  Fatty acid   Vmax (nmol−1 min−1 UHRP)   % Inhibition (in the presence of 50 U catalase)   Km (mM)   Laurate (12.0)   0    0  0  Myristate (14.0)   4.73  100  0.542  Palmitate (16.0)   5.53  100  0.234  16HO‐palmitate   1.35  100  0.211  Stearate (18.0)   6.34  100  0.447  Linoleic (18.2)  10.95   75  0.207  Linolenic (18.3)  33.92   66  0.758  View Large Fatty acid metabolism in vivo The addition of fatty acids at several stages over the operation of the oxidative burst had no effect on the amplitude of the burst and very limited effect on timing, perhaps bringing it forward by about 1 min when added at t0 (data not shown). Since there was a suggestion that peroxidase could produce lipid hydroxides in vitro these were also looked for in vivo. The signature oxylipins (Weichert et al., 1999) for oxidation were not detected in apoplastic fluid over the period of the apoplastic burst in French bean cells (Fig. 4) and although linolenic and linoleic acids could be detected in apoplastic fluid before elicitation, they disappeared over the period of elicitation and a portion seems to be esterified, not oxidized, at this time. However, subsequent oxidation of fatty acids is known to occur in defence responses in French bean. There were increases in cytochrome P450‐dependent ω‐hydroxylase activity towards saturated fatty acids (Bolwell et al., 1997) over intermediate time periods (0–4 h post‐elicitation). It is also well established that later events in response to pathogens in French bean involve lipoxygenase and the production of oxylipins. Cis‐3‐hexenol and trans‐2‐hexenal were produced, commencing at about 12–15 h following inoculation of bean leaves with Pseudomonas syringae, indicating that a pathway via 13‐HOT was operational (Croft et al., 1993). Further searches for the ‘reductant’ The difficulty in establishing a precise role for fatty acids in the apoplastic burst in French bean cells is therefore unresolved at present and has prompted a further search for compounds in active fractions in the apoplastic fluid. Figure 5 shows the specific appearance of four compounds just before the peak of hydrogen peroxide production relative to the patterns of components from unelicited cells and following the completion of the burst. These were identified by GC‐MS as glycerol, malate, citrate, and succinate. Although individually these did not promote the luminol reaction with horseradish peroxidase in vitro, combinations of the carboxylic acids did give a reaction at about 20% of the standard compound, cysteine at 50 μM. Perhaps more significantly they could signify increased Redox metabolic activity or in the case of malate, the operation of the Redox reaction first proposed in the 1970s (Elstner and Heupel, 1976). However, their contribution to the apoplastic oxidative burst is unresolved at present. Certainly there is no obvious one component for a reductant and there are a number of components identified as being released, which can support H2O2 production in vitro by peroxidases. The possibility that complex mixtures of low molecular compounds stimulate dismutation of complex III from peroxidases exists and would be very difficult to resolve. Fig. 5.  View largeDownload slide GC‐MS of active apoplastic fractions from (A) unelicited cells, (B) elicited cells just before the oxidative burst and (C) elicited cells following termination of the oxidative burst (Bolwell et al., 1999). Cells were harvested under gentle vacuum and washed with water. Twenty mM sodium acetate pH 6 was then allowed to infiltrate the cell wall and the apoplastic fluid collected under gentle vacuum. The samples were then analysed by HPLC followed by GC‐MS (Bolwell et al., 1999). The peaks present only in the period just before the oxidative burst are (1) glycerol; (2) succinate; (3) malate; and (4) citrate. Fig. 5.  View largeDownload slide GC‐MS of active apoplastic fractions from (A) unelicited cells, (B) elicited cells just before the oxidative burst and (C) elicited cells following termination of the oxidative burst (Bolwell et al., 1999). Cells were harvested under gentle vacuum and washed with water. Twenty mM sodium acetate pH 6 was then allowed to infiltrate the cell wall and the apoplastic fluid collected under gentle vacuum. The samples were then analysed by HPLC followed by GC‐MS (Bolwell et al., 1999). The peaks present only in the period just before the oxidative burst are (1) glycerol; (2) succinate; (3) malate; and (4) citrate. Features of role of the ‘reductant’ In the absence of identification of the low molecular weight compound, features of the dependence of its role on other factors were investigated using the in vitro model compound, cysteine (Pichorner et al., 1992; Bolwell et al., 1995). Figure 6A shows that when cysteine was added to cells at t0, the oxidative burst is considerably enhanced, but with a time‐dependence similar to the normal elicitation curve rather than a continuous production of ROS, showing that cysteine can only be oxidized with the production of H2O2 when the elicitor induces suitable conditions. Figure 6B and C show that cysteine can be oxidized only when conditions for the oxidative burst are optimum. Stimulation is still observed when added at 10 min, but not markedly when added as the peak of H2O2 production is passing. Figure 7 shows that the response with cysteine is concentration‐dependent, with a maximum at 200 μM and with a sharp decline at higher concentrations. This sharp cut‐off is reminiscent of the shape of the curve with HRP, albeit with an optimum of around 500 μM in this case (Pichorner et al., 1992). However, these observations are conducive to believe that a reductant can only serve as a substrate during the time frame of conditions brought about by elicitor action. Circumstantially, this would implicate the pH change. Fig. 6.  View largeDownload slide Time dependence of the cysteine enhanced oxidative burst. (A) Cysteine added to a final concentration of 200 μM at zero time, (B) 200 μM added at 10 min, (C) 200 μM added at 16 min. Batches of suspension‐cultured cells were treated with elicitor and 1 ml samples taken periodically and hydrogen peroxide measured using the standard luminol assay (Bolwell et al., 1995). Cysteine was added at the time points shown and the control values at the peak reflect conditions without cysteine and the maximum chemiluminescence was similar in each experiment. Each experiment has been repeated at least three times and a typical result is shown. Note: enhancement is only seen during the onset of the burst and the correct pH conditions. Fig. 6.  View largeDownload slide Time dependence of the cysteine enhanced oxidative burst. (A) Cysteine added to a final concentration of 200 μM at zero time, (B) 200 μM added at 10 min, (C) 200 μM added at 16 min. Batches of suspension‐cultured cells were treated with elicitor and 1 ml samples taken periodically and hydrogen peroxide measured using the standard luminol assay (Bolwell et al., 1995). Cysteine was added at the time points shown and the control values at the peak reflect conditions without cysteine and the maximum chemiluminescence was similar in each experiment. Each experiment has been repeated at least three times and a typical result is shown. Note: enhancement is only seen during the onset of the burst and the correct pH conditions. Fig. 7.  View largeDownload slide Concentration dependence of the elicitor‐induced oxidative burst for added cysteine. Cysteine was added to various final concentrations at zero time to batches of suspension‐cultured cells and the oxidative burst measured following elicitation using the standard luminol assay (Bolwell et al., 1995). Note that cysteine enhances the production of hydrogen peroxide at all concentrations used, in a similar manner to that shown in Fig. 6A with a narrow range of hyper‐enhancement around 200 μM. Fig. 7.  View largeDownload slide Concentration dependence of the elicitor‐induced oxidative burst for added cysteine. Cysteine was added to various final concentrations at zero time to batches of suspension‐cultured cells and the oxidative burst measured following elicitation using the standard luminol assay (Bolwell et al., 1995). Note that cysteine enhances the production of hydrogen peroxide at all concentrations used, in a similar manner to that shown in Fig. 6A with a narrow range of hyper‐enhancement around 200 μM. Aspects of signalling Evidence has accumulated that the apoplastic oxidative burst involves a three component system, a peroxidase capable of generating H2O2 at neutral pH, a pH change in the apoplast towards alkalinity, and the generation or release of a reductant and/or other substrate(s). All these are interdependent and there are additional signal components to those already described. The earliest signalling events involved in the oxidative burst are beginning to be elucidated. The increase of cytosolic Ca2+, which occurs within s of elicitation, is thought to be a primary signal essential for the subsequent down‐stream events. Down‐stream events following the increase in cytosolic calcium include the production of ROS (Grant et al., 2000b) and the induction of defence‐related genes such as genes encoding PR proteins and phytoalexin production (Blume et al., 2000). In the case of an infection with an avirulent pathogen, this resulted in the establishment of a local programmed cell death (PCD) at the site of infection (Grant et al., 2000b). It is possible that Ca2+ has a direct effect on the NADPH oxidase, since there is evidence that the plant enzyme has an N‐terminal sequence with two Ca2+ binding EF‐hand motifs (Torres et al., 1998; Keller et al., 1998). However, the basis of the Ca2+‐dependence of the apoplastic oxidative burst is unresolved at present (Bindschedler et al., 2001), but it is most likely to be related to the release of the substrate/reductant. Beside the early calcium influx into the cytosolic compartment, a rapid efflux of K+ and Cl− and extracellular alkalinization of elicited cell cultures has also been observed (Scheel, 1998; Fellbrich et al., 2000). Extracellular alkalinization is of course essential for the apoplastic oxidative burst (Bolwell et al., 1995). Data obtained with forskolin (Bindschedler et al., 2001) circumstantially implcates cAMP in the alkalinization component and the role of cAMP as a secondary signal in plants (Assmann, 1995; Bolwell, 1995). More recently, there are the exciting reports of the identification and characterization of a new gene family in Arabidopsis which shares features with cyclic nucleotide‐gated channels from animals and inward‐rectifying K+ channels from plants (Kohler et al., 1999). A plant cyclic nucleotide‐gated cation channel, AtCNGC2, from Arabidopsis has been cloned and its function characterized (Leng et al., 1999). By analogy with similar animal cation channels, AtCNGC2 is suggested as being involved in signal transduction by allowing a flux of Ca2+, K+ and other ions in the presence of cAMP or cGMP. Furthermore, the Arabidopsis dnd1 gene (defence, no death) has been isolated from a mutant line that failed to produce a hypersensitive response to the avirulent Pseudomonas pathogens and this gene has been found to encode the same ion channel protein, AtCNGC2 (Clough et al., 2000). Arazi et al. have isolated a tobacco plasma channel protein with a high affinity for calmodulin and a highly conserved cyclic nucleotide‐binding domain (Arazi et al., 2000). A transgenic tobacco expressing a dominant‐acting calmodulin hyperactivates a calmodulin‐dependent NAD kinase leading to an increased and more rapid production of ROS, an increased alkalinization of the medium and the initiation of a more rapid PCD (Harding et al., 1997; Harding and Roberts, 1998). Taken together, there is growing evidence for the involvement of cAMP in signal transduction and cross‐talk between the various pathways of plants responding to the attack by avirulent pathogens including activation of the apoplastic oxidative burst. 1 To whom correspondence should be addressed. Fax: +441784434326. E‐mail: uhbc006@vms.rhbnc.ac.uk 2 Present address: Department of Biological Sciences, California State University, Chico, CA 95929‐0515, USA. 3 Present address: School of Biological Sciences, University of Manchester, Manchester M13 9PT, UK. 4 Present address: Institute of Biochemistry and Biophysics, PO Box 30, Kazan 420503, Russia. KAB was funded by the BBSRC (UK), SLG by the Leverhulme Trust and FM was the recipient of a NATO fellowship through the Royal Society (UK). We thank Dr Ivo Fuessner, Institute of Plant Genetics and Crop Plant Research, Gatersleben, Germany for performing the oxylipin GC/MS and Professor Jean‐Pierre Salaun, Roscoff, France, for providing standards and analysis of reactions involving radiolabelled fatty acids. References Arazi T, Kaplan B, Fromm H. 2000. A high affinity calmodulin‐binding site in a tobacco plasma membrane channel protein coincides with a characteristic element of cyclic‐nucleotide‐binding domains. Plant Molecular Biology  42, 591–601. Google Scholar Asada K. 1999. The water–water cycle in chloroplasts: scavenging of active oxygens and dissipation of excess photons. Annual Review of Plant Physiology and Plant Molecular Biology  50, 601–639. Google Scholar Assmann SM. 1995. Cyclic‐AMP as a 2nd messenger in higher‐plants—status and future prospects. Plant Physiology  108, 885–889. Google Scholar Auh CK, Murphy TM. 1995. Plasma membrane redox enzyme is involved in the synthesis of \({\mathrm{O}_{2}^{{\bullet}{-}}}\) and H2O2 by Phytophthora elicitor‐simulated rose cells. Plant Physiology  107, 1241–1247. Google Scholar Baker CJ, Orlandi EW. 1995. Active oxygen in plant pathogenesis. Annual Review of Phytopathology  33, 299–321. Google Scholar Bestwick CS, Brown IR, Mansfield JW. 1998. Localized change in peroxidase activity accompanies hydrogen peroxide generation during the development of a non‐host hypersensitive reaction in lettuce. Plant Physiology  118, 1067–1078. Google Scholar Bindschedler LV, Minibayeva F, Gardner SL, Gerrish C, Davies DR, Bolwell GP. 2001. Early signalling events in the apoplastic oxidative burst in suspension‐cultured French bean cells involve cAMP and Ca2+. New Phytologist  151, 185–194. Google Scholar Blee KA, Jupe SC, Richard G, Bolwell GP. 2001. Molecular identification and expression of the peroxidase responsible for the oxidative burst in French bean (Phaseolus vulgaris L.) and related members of the gene family. Plant Molecular Biology  47, 607–620. Google Scholar Blume B, Nürnberger T, Nass N, Scheel D. 2000. Receptor‐mediated increase in cytoplasmic free calcium required for activation of pathogen defense in parsley. The Plant Cell  12, 1425–1440. Google Scholar Bolwell GP. 1992. A role for phosphorylation in the down‐regulation of phenylalanine ammonia‐lyase in suspension‐cultured cells of French bean. Phytochemistry  31, 4081–4086. Google Scholar Bolwell GP. 1995. Cyclic AMP the reluctant messenger in plant cells. Trends in Biochemical Sciences  20, 492–495. Google Scholar Bolwell GP. 1996. The origin of the oxidative burst in plants. Biochemical Society Transactions  24, 438–442. Google Scholar Bolwell GP. 1999. Role of active oxygen species and NO in plant defence responses. Current Opinion in Plant Biology  2, 287–294. Google Scholar Bolwell GP, Blee KA, Butt VS, Davies DR, Gardner SL, Gerrish C, Minibayeva F, Rowntree EG, Wojtaszek P. 1999. Recent advances in understanding the origin of the apoplastic oxidative burst in plant cells. Free Radical Research  31, S137–145. Google Scholar Bolwell GP, Butt VS, Davies DR, Zimmerlin A. 1995. The origin of the oxidative burst in plants. Free Radical Research  23, 517–532. Google Scholar Bolwell GP, Davies DR, Gerrish C, Auh CK, Murphy TM. 1998. Comparative biochemistry of the oxidative burst produced by rose and French bean cells reveals two distinct mechanisms. Plant Physiology  116, 1379–1385. Google Scholar Bolwell GP, Gerrish C, Salaun J‐P. 1997. Changes in enzymes involved in suberization in elicitor‐treated French bean cells. Phytochemistry  45, 1351–1357. Google Scholar Bolwell GP, Wojtaszek P. 1997. Mechanisms for the generation of reactive oxygen species in plant defence—a broad perspective. Physiological and Molecular Plant Pathology  51, 347–366. Google Scholar Brown I, Trethowan J, Kerry M, Mansfield JW, Bolwell GP. 1998. Location of components of the oxidative cross‐linking of glycoproteins and callose synthesis in papillae formed during the interaction between non‐pathogenic strains of Xanthomonas campestris and French bean mesophyll cells. The Plant Journal  15, 333–343. Google Scholar Clough SJ, Fengler KA, I‐Ching Y, Lippok B, Smith RK, Bent AF. 2000. The Arabidopsis dnd1 ‘defense, not death’ gene encodes a mutated cyclic nucleotide gated ion channel. Proceedings of the National Academy of Sciences, USA  97, 9323–9328. Google Scholar Croft KPC, Juttner F, Sluzarenko AJ. 1993. Volatile products of the lipoxygenase pathway evolved from Phaseolus vulgaris (L.) leaves inoculated with Pseudomonas syringae pv. phaseolicola. Plant Physiology  101, 13–24. Google Scholar del Río LA, Corpas FJ, Sandalio LM, Palma JM, Gómez M, Barroso JB. 2002. Reactive oxygen species, antioxidant systems and nitric oxide in peroxisome. Journal of Experimental Botany  53, 1255–1272. Google Scholar Desikan R, Hancock JT, Coffey MJ, Neill SJ. 1996. Generation of active oxygen in elicited cells of Arabidopsis thaliana is mediated by a NADPH oxidase‐like enzyme. FEBS Letters  382, 13–17. Google Scholar Elstner EF, Heupel A. 1976. Formation of hydrogen peroxide by isolated cell walls from horseradish (Armoracia lapatholia). Planta  130, 175–180. Google Scholar Fellbrich G, Blume B, Brunner F, Hirt H, Kroj T, Ligterink W, Romanski A, Nürnberger T. 2000. Phytophthora parasitica elicitor‐induced reactions in cells of Petroselinum crispum. Plant Cell Physiology  41, 692–701. Google Scholar Grant JJ, Yun B‐W, Loake GJ. 2000a. Oxidative burst and cognate redox signaling reported by luciferase imaging: identification of a signal network that functions independently of ethylene, SA and Me‐JA but is dependent on MAPKK activity. The Plant Journal  24, 569–582. Google Scholar Grant M, Brown I, Adams S, Knight M, Ainslie A, Mansfield J. 2000b. The RPM1 plant disease resistance gene facilitates a rapid and sustained increase in cytosolic calcium that is necessary for the oxidative burst and hypersensitive cell death. The Plant Journal  23, 441–450. Google Scholar Harding SA, Oh SH, Roberts DM. 1997. Transgenic tobacco expressing a foreign calmodulin gene shows an enhanced production of active oxygen species. The EMBO Journal  16, 1137–1144. Google Scholar Harding SA, Roberts DM. 1998. Incompatible pathogen infection results in enhanced reactive oxygen and cell death responses in transgenic tobacco expressing a hyperactive mutant calmodulin. Planta  206, 253–258. Google Scholar Keller T, Damude HG, Werner D, Doerner P, Dixon RA, Lamb C. 1998. A plant homolog of the neutrophil NADPH oxidase gp91phox subunit gene encodes a plasma membrane protein with Ca2+ binding motifs. The Plant Cell  10, 255–266. Google Scholar Kohler C, Merkle T, Neuhaus G. 1999. Characterization of a novel gene family of putative cyclic nucleotide‐ and calmodulin‐regulated ion channels in Arabidopsis thaliana. The Plant Journal  18, 97–104. Google Scholar Lamb C, Dixon RA. 1997. The oxidative burst in plant disease resistance. Annual Review of Plant Physiology and Plant Molecular Biology  48, 251–275. Google Scholar Leng Q, Mercier RW, Yao W, Berkowitz GA. 1999. Cloning and first function characterization of a plant cyclic nucleotide‐gated cation channel. Plant Physiology  121, 753–761. Google Scholar Martinez C, Montillet JL, Bresson E, Angel JP, Dai GH, Daniel JF, Geiger JP. 1998. Apoplastic peroxidase generates superoxide anions in cells of cotton cotyledons undergoing the hypersensitive reaction to Xanthomonas campestris pv. malvacerarum race 18. Molecular Plant–Microbe Interactions  11, 1038–1047. Google Scholar McLusky SR, Bennett MH, Beale MH, Lewis MJ, Gaskin P, Mansfield JW. 1999. Cell wall alterations and localized accumulation of feruloyl 3′‐tyramine in onion epidermis at sites of attempted penetration by Botrytis allii are associated with actin polarization, peroxidase activity and suppression of flavonoid biosynthesis. The Plant Journal  17, 523–534. Google Scholar Ostergaard L, Teilum K, Mirza O, Petersen M, Welinder KG, Mundy J, Gajhede M, Hendriksen A. 2000. Arabidopsis ATPA2 peroxidase: expression and high‐resolution structure of a plant peroxidase with implications for lignification. Plant Molecular Biology  44, 231–243. Google Scholar Pichorner H, Couperus A, Korori SAA, Ebermann R. 1992. Plant peroxidase has a thiol oxidase function. Phytochemistry  31, 3371–3376. Google Scholar Robertson D, Davies DR, Gerrish C, Jupe SC, Bolwell GP. 1995. Rapid changes in oxidative metabolism as a consequence of elicitor treatment of suspension‐cultured cells of French bean (Phaseolus vulgaris L.). Plant Molecular Biology  27, 59–67. Google Scholar Scheel D. 1998. Resistance response physiology and signal transduction. Current Opinion in Plant Biology  1, 305–310. Google Scholar Torres MA, Onouchi H, Hamada S, Machida C, Hammond‐Kosack KE, Jones JD. 1998. Six Arabidopsis thaliana homologues of the human respiratory burst oxidase (gp91phox). The Plant Journal  14, 365–370. Google Scholar Volko SM, Boller T, Ausubel FM. 1998. Isolation of new Arabidopsis mutants with enhanced diseased susceptibility to Pseudomonas syringae by direct screening. Genetics  149, 537–548. Google Scholar Volotovski ID, Sokolovsky SG, Molchan OV, Knight MR. 1998. Second messengers mediate increases in cytosolic calcium in tobacco protoplasts. Plant Physiology  117, 1023–1030. Google Scholar Weichert H, Stenzel I, Berndt E, Wasternack C, Fuessner I. 1999. Metabolic profiling of oxylipins upon salicylate treatment in barley leaves—preferential induction of the reductase pathway by salicylate. FEBS Letters  464, 133–137. Google Scholar Woo EJ, Dunwell JM, Goodenough PW, Marvier AC, Pickersgill RW. 2000. Germin is a manganese containing homohexamer with oxalate oxidase and superoxide dismutase activities. Nature Structural Biology  11, 1036–1104. Google Scholar Zimmerlin A, Wojtaszek P, Bolwell GP. 1994. Synthesis of dehydrogenation polymers of ferulic acid with high specificity by a purified cell‐wall peroxidase from French bean (Phaseolus vulgaris L.). Biochemical Journal  94, 747–753. Google Scholar © Society for Experimental Biology

Journal

Journal of Experimental Botany – Oxford University Press

Published: May 15, 2002

There are no references for this article.