- Copyright © 1998 American Society of Plant Physiologists
Abstract
Cultured cells of rose (Rosa damascena) treated with an elicitor derived from Phytophthora spp. and suspension-cultured cells of French bean (Phaseolus vulgaris) treated with an elicitor derived from the cell walls of Colletotrichum lindemuthianum both produced H2O2. It has been hypothesized that in rose cells H2O2is produced by a plasma membrane NAD(P)H oxidase (superoxide synthase), whereas in bean cells H2O2 is derived directly from cell wall peroxidases following extracellular alkalinization and the appearance of a reductant. In the rose/Phytophthoraspp. system treated with N,N-diethyldithiocarbamate, superoxide was detected by a N,N′-dimethyl-9,9′-biacridium dinitrate-dependent chemiluminescence; in contrast, in the bean/C. lindemuthianum system, no superoxide was detected, with or withoutN,N-diethyldithiocarbamate. When rose cells were washed free of medium (containing cell wall peroxidase) and then treated with Phytophthora spp. elicitor, they accumulated a higher maximum concentration of H2O2 than when treated without the washing procedure. In contrast, a washing treatment reduced the H2O2 accumulated by French bean cells treated with C. lindemuthianum elicitor. Rose cells produced reductant capable of stimulating horseradish (Armoracia lapathifolia) peroxidase to form H2O2but did not have a peroxidase capable of forming H2O2 in the presence of reductant. Rose and French bean cells thus appear to be responding by different mechanisms to generate the oxidative burst.
An oxidative burst is a common response of plant cells to physical or biological stress. The production of ROS such as superoxide and H2O2 has been noted when plants are challenged with particular viral, bacterial, or fungal pathogens (Mehdy, 1994; Low and Merida, 1996; Wojtaszek, 1997; Bolwell and Wojtaszek, 1998). It is generally considered a component of the hypersensitive response, an acute defensive syndrome, and is often highly specific for particular genotypes of pathogen and host, but nonspecific interactions can also be demonstrated. The oxidative burst may be related to subsequent events in the challenged cells, such as programmed cell death, although the exact nature of the relationship is unclear. Free radical oxidation of plasma membrane lipids, induced by ROS, may kill cells directly. Alternatively, superoxide (Jabs et al., 1996) or H2O2 (Levine et al., 1994) may serve as signals leading indirectly to mortality. However, in at least one case it has been shown that an oxidative burst by itself is not sufficient to trigger programmed cell death (Glazener et al., 1996).
The oxidative burst is often a very rapid response, occurring within seconds in some systems, such as cultured cells of French bean (Phaseolus vulgaris) and soybean (Bolwell et al., 1995). In other systems, such as rose (Rosa damascena) cultured cells (Arnott and Murphy, 1991), it may be delayed for minutes or hours, but in general it is thought not to require de novo protein synthesis. Thus, it involves the activation of pre-existing enzymes.
The source of the ROS is under study. There are several hypotheses to explain the appearance of H2O2 in the medium of cultured cells and the apoplastic fluid of whole-plant tissues. The earliest, proposed to explain the origin of H2O2 needed for formation of lignin in developing xylem of horseradish (Armoracia lapathifolia), involves the reduction of O2to superoxide by phenolic and NAD. radicals produced by peroxidase (Yamazaki and Yokota, 1973; Elstner and Heupel, 1976; Gross et al., 1977; Halliwell, 1978). In this model the source of electrons in the apoplast is said to be malate, exported across the plasma membrane by a malate/oxalacetate carrier and used to reduce NAD+ by apoplastic malate dehydrogenase. H2O2 is formed by the dismutation of superoxide.
A second hypothesis, proposed for the oxidative burst induced in French bean cultured cells by Colletotrichum lindemuthianum cell wall elicitor, involves an apoplastic peroxidase in a more direct way (Bolwell et al., 1995). The O2-heme complex of peroxidase is reduced to compound III by reductants exported from the cell. Under the proper conditions, i.e. elevated pH, the complex is effectively hydrolyzed to release H2O2. In this model the source of electrons has not been identified, but the release of a reductant from elicited cells has been observed.
A third hypothesis, proposed for the oxidative burst induced in rose cultured cells by Phytophthora spp. elicitor (Auh and Murphy, 1995) and for other systems, does not involve peroxidase. Rather, a trans-plasma membrane superoxide synthase (NAD(P)H oxidase) transfers electrons from NADH or NADPH in the cytoplasm to O2 to form superoxide. H2O2 is formed by the dismutation of superoxide. Superoxide synthesis has been observed in purified preparations of plasma membrane from several plant systems (Vianello and Macrı́, 1989; Qiu et al., 1994; Doke and Miura, 1995; Murphy and Auh, 1996; Van Gestelen et al., 1997).
The experimental systems used to derive the second and third hypotheses involved relatively similar systems: cultured parenchymatous cells challenged with elicitors derived from fungal (or protistan) cell walls. However, not all of the same experiments were used to develop the hypotheses. We believed it was important to determine whether the differences in interpretation are due to fundamental differences in the origin of H2O2 or to differences in approach. The objective of the present study was to compare the two systems, French bean and rose, to resolve this question.
MATERIALS AND METHODS
Chemicals
Luminol, lucigenin, and horseradish (Armoracia lapathifolia) peroxidase (EC 1.11.1.7) were obtained from Sigma. DDC was from Sigma or Fisher Scientific and DPI was from Cookson Chemicals, Ltd. (Southampton, UK) or Calbiochem.
Cells and Elicitor
Cells of rose (Rosa damascena Mill. cv Gloire de Guilan) were derived and grown in a suspension culture as previously described (Murphy et al., 1979). Elicitor was derived fromPhytophthora cinnamomea or Phytophthora megasperma, as described by Auh and Murphy (1995), and used at a final concentration of 25 or 15 μg mL−1 Glc equivalents, respectively. Derivation and maintenance of cell cultures of French bean (Phaseolus vulgaris L.) and preparation of elicitor from Colletotrichum lindemuthianum were as described previously (Dixon and Lamb, 1979). Elicitor was used at a final concentration of 30 μg mL−1 Glc equivalents.
Determination of the Oxidative Burst
Lucigenin Assay for Superoxide
The accumulation of superoxide in the cell medium was measured by lucigenin-dependent chemiluminescence. The assay was conducted in a total volume of 2 mL by placing 0.2 mL of cell suspension and 0.2 mL of 1 mm lucigenin in 0.1 m Gly-NaOH buffer (pH 9.0) containing 1 mm EDTA and 1 mm sodium salicylate. The SOD inhibitor Na-DDC was added to the cell suspensions at 1 mm to block the dismutation of superoxide to H2O2 by SOD. The chemiluminescence was detected in the luminometer, which detects real-time luminosity, or by using the single-channel mode in a scintillation spectrometer. This procedure has been considered a specific assay of superoxide (Corbisier et al., 1987), but recent reports (Liochev and Fridovich, 1997; Vásques-Vivar et al., 1997) indicate that reduced lucigenin can react with O2to produce superoxide. Thus, enzymatic or nonenzymatic reactions that reduce lucigenin can give a false-positive reading.
Luminol Assay for H2O2
In rose cells, luminol-dependent chemiluminescence was detected in a total volume of 1 mL by combining 0.2 mL of cell-suspension medium and 0.01 mL of 1 mm luminol solution in 50 mm Tris buffer (pH 8.0) in a scintillation vial. Following the addition of 0.01 mL of 13 mm K3Fe(CN)6 the scintillation vial was immediately placed in a scintillation spectrometer (model LS8000, Beckman) and chemiluminescence was detected on single-channel mode. Counts were reported every 12 s for 36 s and the last value was used. Earlier experiments with this system (Auh and Murphy, 1995) used a similar technique but relied on the addition of horseradish peroxidase together with the substrate H2O2 for the oxidation of luminol. In French bean cells, the oxidative burst was routinely measured using a luminometer fitted with a rotating cuvette holder and an injection port (model 1250, LKB Wallac, Broma, Sweden). A 1-mL sample of suspension culture was loaded into the luminometer cuvette and continually stirred; 200 μL of 1 mm luminol solution was then injected rapidly into the cuvette. Real-time luminosity was recorded every second beginning from 10 s before injection until the level returned to background. The time between sampling the suspension culture and injecting the luminol was always less than 1 min. In some experiments, specified in “Results,” a scintillation counter was used to allow a more exact comparison between the rose and French bean systems. The luminometer measures chemiluminescence immediately after the addition of luminol and with stirred cells, whereas there is a delay of several seconds using the liquid-scintillation counter. However, over the time scale of measurements of the oxidative burst in cells, the latter delay is not significant. In each case, the assay was calibrated with a solution containing a known concentration of H2O2.
Determination of Superoxide Production by NADH Oxidase and H2O2 Production by Peroxidases
Superoxide production by a partially solubilized enzyme preparation from rose cell plasma membrane was measured by Cytc reduction. A reaction mixture contained buffer (20 mm Tris-Cl, pH 7.5, and 3 mm MgCl2) to give a total volume of 0.5 mL, 100 μm NADH, 0.02% (w/v) Triton X-100, 100 μm Cyt c, DPI in concentrations as noted in Figure 4, and 0.1 to 0.6 μg of enzyme preparation protein. Change inA 550 was measured over the 1st min in a DU-640 spectrophotometer (Beckman). The data were adjusted by subtracting the background rate of reduction obtained in the presence of 40 units of SOD.
A, H2O2 accumulation. Inhibition by DPI of the luminol-dependent chemiluminescence of rose cells elicited with cell wall extract from Phytophthoraspp. and of French bean cells elicited with cell wall extract fromC. lindemuthianum. B, Enzyme activity. Inhibition by DPI of the synthesis of superoxide by rose cell NADH oxidase and of H2O2 (luminol-dependent chemiluminescence) by horseradish peroxidase in the presence of 0.5 mm Cys. The points indicate means (bars are ses) from at least three independent experiments.
Measurements of H2O2production were determined for purified peroxidases in the luminometer, the operation of which was as described above. The reactions contained 1 mL of stirred buffer to which 3 μL of peroxidase containing a known number of units (e.g. 3 units, as specified by Sigma for horseradish peroxidase) was added. The buffers used were at the pH optimum for the requisite peroxidase. Borate buffer, pH 8.5, was used for horseradish peroxidase (type XII, Sigma) and phosphate buffer, pH 7.2, was used for French bean peroxidase 1 (FBP1) from French bean. Two-hundred microliters of 1 mm luminol solution was then added and the initial burst of chemiluminescence was measured. Reductant such as Cys (0.5 mm) was then added and the subsequent chemiluminescence monitored.
RESULTS
Comparison of the Effectiveness of Elicitors on Rose and French Bean Cells in Generating H2O2
Both systems were capable of inducing rapid production of H2O2. For rose cells andPhytophthora spp. elicitor, peak concentrations of about 20 μm were reached at about 45 to 120 min after the initial addition of elicitor. In French bean peak concentrations of about 130 μm H2O2 were reached 8 to 16 min after the addition of C. lindemuthiamumelicitor. When C. lindemuthiamum elicitor was added to rose cells, the cells produced variable, but generally weak, levels of H2O2: the luminol-dependent chemiluminescence indicated a peak of about 7 μm H2O2 at 30 to 60 min (Fig.1A). Phytophthora spp. elicitor applied to French bean cells gave a peak response at 16 min that was on average less than 15% of the peak response seen at 8 min with C. lindemuthiamum elicitor applied to the same batch of cells (Fig. 1B). Thus, reciprocal experiments gave similar results in both systems: heterologous elicitors induced H2O2 production but at reduced levels compared with the homologous elicitors. In general,Phytophthora spp. elicitor stimulated a slower response andC. lindemuthiamum elicitor stimulated a more rapid response.
Accumulation of H2O2(luminol-dependent chemiluminescence) in the medium of rose (A) and French bean (B) cells elicited by cell wall extract fromPhytophthora spp. or C. lindemuthianum. Representative plots are shown. For rose cells, the peak H2O2 concentrations were 20 ± 8 μm (mean ± se; n = 4) with Phytophthora spp. elicitor and 7 ± 2 μm (n = 4) with C. lindemuthianum elicitor. For French bean cells, the peak H2O2 concentrations were 133 ± 39 (n = 5) with C. linedmuthianumelicitor and 16 ± 3 (n = 3) withPhytophthora spp. elicitor.
The Effect of Using Washed Cells with Homologous Elicitor
In rose cells the appearance of H2O2 was increased by an initial washing of the cells three times with a solution containing 1 mm CaCl2 and 0.1 mm KCl (Qian et al., 1993). The increase was ascribed to the loss of peroxidases (or catalase) that consumed the H2O2. In contrast, prewashing French bean cells reduced the subsequent luminol-dependent chemiluminescence (Fig. 2). The contrasting effect of washing the cells suggests that the two cell species produce H2O2 in different ways and that the French bean system requires a component that is lost in the washing step.
Effects of washing on the luminol-dependent chemiluminescence of rose cells elicited by cell wall extract fromPhytophthora spp. and of French bean cells elicited by cell wall extract from C. lindemuthianum. The data for rose cells were calculated fromQian et al. (1993). Results are means ± se (n = 3). For rose cells, values were normalized to the mean with washed cells; for French bean cells, values were normalized to the mean with unwashed cells.
Comparative Effects of Elicitors on Superoxide Production by Rose and French Bean Cells
The presumptive accumulation of superoxide, measured by lucigenin-dependent chemiluminescence in the presence of DDC, was readily detected in rose cells when elicited withPhytophthora spp. (Auh and Murphy, 1995). In contrast, a significant accumulation of superoxide was never observed in French bean cells treated with either C. lindemuthianum orPhytophthora spp. elicitor, either with or without DDC (data not shown).
The Effect of KCN and DPI on Cell-Generated Superoxide and H2O2
Peroxidases are exquisitely sensitive to micromolar concentrations of KCN (Saunders et al., 1964), whereas flavin-dependent enzymes, including the mammalian NADPH oxidase, are strongly inhibited by DPI (Cross and Jones, 1986). These two inhibitors played a strong role in the interpretation of the mechanism of H2O2 generation in the French bean and rose systems (Auh and Murphy, 1995; Bolwell et al., 1995). We expected that low concentrations of KCN would inhibit synthesis of H2O2 in French bean cells more strongly than in rose cells. However, it was not possible to make such a distinction, apparently because cells detoxified KCN over the short time in which they synthesized H2O2. Figure3 illustrates this point. The effect of KCN on the accumulation of H2O2 by French bean cells was substantially less 15 min after the addition of elicitor than 10 min after the addition.
Effect of KCN on luminol-dependent chemiluminescence of French bean cells, elicited by cell wall extract from C. lindemuthianum and assayed for 10 or 15 min after the addition of elicitor. The points indicate means (bars are ses) from two to three independent experiments.
DPI inhibited H2O2accumulation induced by Phytophthora spp. (orC. lindemuthianum) elicitor in rose cells with 50% inhibition at 2 μm, whereas in the French bean system induced with C. lindemuthianum elicitor, 50% inhibition required about 40 μm (Fig.4A). Over the time course of the oxidative burst (up to 150 min), 15 μm DPI inhibited respiration of rose cells by less than 20% and 100 μm inhibited respiration by less than 30%; the respiration of French bean cells was increased 12 and 43% by 15 and 100 μm DPI, respectively. This indicates that metabolically generated reducing power was available for the reduction of O2, and energy was available for signal cascades initiated by elicitor. At longer times, DPI inhibited O2 uptake more strongly in rose cells (in one experiment, more than 90% after an overnight incubation) but not in French bean cells (in one experiment, cells treated overnight with 100 μm DPI had higher respiration rates than untreated controls).
The contrasting sensitivities of rose and French bean cells were matched by the sensitivities of plasma membrane NADH oxidase and peroxidase. DPI inhibited the production of superoxide by a rose plasma membrane NADH oxidase more than 50% at 0.1 μm and of H2O2 by peroxidase, either from horseradish (pH optimum 8.5) or FBP1 (pH optimum 7.2), in the presence of the reducing agent Cys, approximately 50% at concentrations up to 240 μm (Fig. 4B).
The Induced Appearance of Reducing Agent
The synthesis of H2O2by French bean is stimulated by a reducing agent released from cells challenged with C. lindemuthianum elicitor (Bolwell et al., 1995). Rose cells also released an agent capable of stimulating H2O2 synthesis by horseradish peroxidase, more when treated with C. lindemuthianum elicitor than with Phytophthora spp. elicitor (Fig. 5). However, the extracellular peroxidases collected in medium or high-salt eluate from rose cultures did not produce detectable H2O2 when provided with Cys as a reducing agent. In fact, peroxidases or catalases in the medium and eluate from rose cultures reduced the accumulation of H2O2 produced by horseradish peroxidase plus Cys (TableI).
Induction by cell wall extracts ofPhytophthora spp. (Ph. elicitor) andC. lindemuthianum (C. elicitor) of secretion from rose cells of an agent that stimulates H2O2 synthesis by horseradish peroxidase. The data represent means from three independent experiments. By analysis of variance, variation among beakers was shown to be significant at the 2% confidence level; variation among sampling times was significant at the 0.2% level. Beaker 3 (C. elicitor) at 30 min was significantly different from all other samples (Student'st test, 5% level).
Synthesis of H2O2 by peroxidase plus Cys
The Effect of DDC on the Detection of Superoxide and H2O2
DDC is known as a chelator of Cu ions and an inhibitor of the Cu/Zn isozyme of SOD (Heikkila et al., 1976; Kelner and Alexander, 1986). It was added to rose cells in the expectation that the inhibition of SOD would allow the accumulation of superoxide (observed through lucigenin-dependent chemiluminescence) and inhibit the accumulation of H2O2(luminol-dependent chemiluminescence) during an oxidative burst, assuming that H2O2 was formed by dismutation from superoxide. These expectations were fulfilled (Auh and Murphy, 1995). In the French bean cell system, DDC abolished luminol-dependent chemiluminescence; however, it did not potentiate the detection of superoxide by lucigenin-dependent chemiluminescence. The lack of production of superoxide in the French bean system, in contrast to results in the rose system, supports the contention that the oxidative bursts in the two systems have different origins. However, further experiments suggested that DDC reacts directly with H2O2: the addition of H2O2 to DDC caused a rapid change in the UV absorption spectrum of DDC (Fig.6). This implies that DDC removes H2O2 from solution. Thus, the role of DDC in the stimulation of lucigenin-dependent chemiluminescence and inhibition of luminol-dependent chemiluminescence must be reinterpreted. Specifically, the inhibition of luminol-dependent chemiluminescence by DDC cannot be interpreted to mean that the production of H2O2 proceeds by the dismutation of superoxide.
Reaction of 0.1 mm DDC with 4 mm H2O2. The peaks at 258 and 283 nm represent DDC in the absence of H2O2; the spectrum for 4 mm H2O2 alone is also shown. The curves identified as 10 s, 3 min, and 10 min show the spectrum of the DDC sample at the indicated times after the addition of H2O2.
DISCUSSION
Comparative biochemical studies of the oxidative burst produced by rose and French bean cells have revealed two distinct mechanisms. Both systems ultimately produced H2O2 in response to elicitor treatment, but the intermediate formation of superoxide was not as significant in French bean as in rose cells.
Differences in the two systems include (a) the kinetics of the oxidative burst and the peak concentrations of H2O2 that accumulated, (b) the effect of washing the cells, (c) the detection of superoxide in the presence of DDC, (d) the effect of DPI in inhibiting the oxidative burst, and (e) the ability of extracellular peroxidases to generate H2O2 in the presence of reducer.
Evidence related to the detection of superoxide in the medium of elicited cells requires some qualification. As noted in “Materials and Methods,” the reduction of lucigenin can lead to false-positive indications for the presence of superoxide (Liochev and Fridovich, 1997; Vásques-Vivar et al., 1997). Thus, the detection of superoxide (lucigenin-dependent luminescence) detected in the media of elicited rose cells by Auh and Murphy (1995) could actually have been produced by reducing substances released from the cells. However, the presence of DDC-insensitive SOD in the apoplast of French bean cells could have prevented a positive indication of superoxide. The presence of superoxide in the apoplast is an important issue (Jabs et al., 1996), but one that will require further investigation.
The finding that different mechanisms are responsible for the production of H2O2 in different elicited systems may relate to systems other than the ones we studied. The synthesis of H2O2 by tomato cells treated with elicitor from Cladosporium fulvum is very sensitive to submicromolar concentrations of DPI (K. Hammond-Kosack, personal communication), whereas synthesis of H2O2 by lettuce cells treated with Pseudomonas syringae pv phaseolicolawas considered to be more sensitive to azide and KCN than to DPI (Bestwick et al., 1997). The rapid production of H2O2 by soybean cells treated with a Verticilium dahliae elicitor was inhibited 60% by 6 μm KCN (Apostol et al., 1989). Allan and Fluhr (1997) reported that ROS induced in tobacco epidermal tissue by different agents arose by different mechanisms: the elicitor cryptogein stimulated an oxidative burst, measured as oxidation of dichlorofluorescein, that was sensitive to DPI but not to exogenous catalase, and added Arg stimulated oxidation of dichlorofluorescein that was sensitive to catalase but not to DPI.
The effects of DDC on the accumulation of H2O2 could not be used to distinguish the two mechanisms. DDC blocked the detection of H2O2 (luminol-dependent chemiluminescence) by reacting rapidly with H2O2. This effect may also have been responsible, at least in part, for the finding that DDC allowed an accumulation of superoxide. H2O2 oxidizes ferrous ion, which in turn oxidizes superoxide. Since iron is a required component of plant cell culture media, we would expect ferrous/ferric ions to be present at the cation-exchange sites of the cell walls.
It is a common observation that evidence from inhibitor studies is plagued with uncertainties, and our results reinforce that statement. DPI inhibits peroxidase-mediated generation of H2O2, as well as NAD(P)H-oxidase-mediated generation of superoxide, an observation also made by Deme et al. (1994) and Dwyer et al. (1996). Thus, a careful study of the concentration dependence of inhibition by DPI is needed if such inhibition is to be used to distinguish peroxidase- and flavoprotein-mediated synthesis of H2O2.
ACKNOWLEDGMENTS
The authors are grateful to Han Vu and Thuyhuong Nguyen, University of California, Davis, for their excellent technical assistance.
Footnotes
-
↵1 This work was supported in part by the U.S. Department of Agriculture National Research Initiative-Cooperative State Research Service (grant no. 94-37100-0788) to T.M.M.
-
↵2 Present address: Department of Plant Pathology, North Carolina State University, Raleigh, NC 27695.
-
↵* Corresponding author; e-mail tmmurphy{at}ucdavis.edu; fax 1–916–752–5410.
Abbreviations:
- DDC
- N,N-diethyldithiocarbamate
- DPI
- diphenyleneiodonium
- lucigenin
- N,N′-dimethyl-9,9′-biacridium dinitrate
- luminol
- 5-amino-2,3-dihydro-1,4-phthalazinedione
- ROS
- reactive oxygen species
- SOD
- superoxide dismutase
- Received November 5, 1997.
- Accepted December 4, 1997.