Production of reactive oxygen species (ROS) is a widely reported response of plants to wounding. However, the nature of enzymes responsible for ROS production and metabolism in the apoplast is still an open question. We identified and characterized the proteins responsible for the wound-induced production and detoxification of ROS in the apoplast of wheat roots (Triticum aestivum L.). Compared to intact roots, excised roots and leachates derived from them produced twice the amount of superoxide (O2•−). Wounding also induced extracellular peroxidase (ECPOX) activity mainly caused by the release of soluble peroxidases with molecular masses of 37, 40 and 136 kD. Peptide mass analysis by electrospray ionization–quadrupole time-of-flight–tandem mass spectrometry (ESI–QTOF–MS/MS) following lectin affinity chromatography of leachates showed the presence of peroxidases in unbound (37 kD) and bound (40 kD) fractions. High sensitivity of O2•−-producing activity to peroxidase inhibitors and production of O2•− by purified peroxidases in vitro provided evidence for the involvement of ECPOXs in O2•− production in the apoplast. Our results present new insights into the rapid response of roots to wounding. An important component of this response is mediated by peroxidases that are released from the cell surface into the apoplast where they can display both oxidative and peroxidative activities.
Wound stress caused by various biotic and abiotic factors is common in plants and can lead to the loss of tissues and organs, and facilitate penetration by pathogens. Plants respond to wounding by activating defence systems that can repair damaged tissues and/or deter pathogens. It is well established that production of ROS is one of the universal responses of various plants to wounding (Orozco-Cárdenas, Narvaez-Vasquez & Ryan 2001; Ross, Küpper & Jacobs 2006). ROS production is also induced by other mechanical stresses, for example, shaking cultured soya and tobacco cells in hypoosmotic media (Legendre et al. 1993; Cazaléet al. 1998) and bombarding potato tubers with small metal particles (Johnson, Doherty & Croy 2003). Many of the genes that code the enzymes involved in ROS production and their metabolism are activated by wounding. In Arabidopsis thaliana, these include genes encoding respiratory burst oxidase homolog D; putative l-ascorbate oxidase; glutathione reductase; glutathione S-transferase; and peroxidase ATP21, ATP15a and ATP24a (Cheong et al. 2002). In the apoplast, peroxidases are an important class of enzymes responsible for the stress-induced formation and degradation of ROS (Bindschedler et al. 2006; Fecht-Christoffers et al. 2006). Apart from their role in H2O2 detoxification, some apoplastic peroxidases are capable of oxidative activity that can generate superoxide anion radicals (O2•−) or H2O2 at physiological pHs. In the mechanisms described in reviews of Ros Barceló (2000), Bolwell et al. (2002), Kawano (2003) and Almagro et al. (2008), the oxidative function of peroxidases requires the presence of appropriate reductant and alkalinization of apoplastic pH. Plants have a great diversity of peroxidases, and these include class III peroxidases (EC 220.127.116.11), to which apoplastic peroxidases belong. Analysis of the Arabidopsis genome revealed more than 73 full-length class III peroxidase genes (Tognolli et al. 2002), while in the rice genome 138 peroxidase genes were found (Passardi, Penel & Dunand 2004). Moreover, many more isoforms can be generated by post-transcriptional and post-translational modifications (Van Engelen et al. 1991), and their activity influenced by various environmental and endogenous stimuli. One role of cell wall peroxidase-induced radical formation is lignin polymerization and suberin formation (Hiraga et al. 2001), important for wound healing. It is also known that ROS in the apoplast could be directly toxic for invading pathogens (Bindschedler et al. 2006), and play roles in intracellular signal transduction (Gechev et al. 2006). Thus, peroxidase-mediated ROS metabolism appears to play a key role in the response of plants to wounding.
Stress not only modulates the activity of peroxidases but also stimulates their release. The cell wall has an enormous capacity to retain various peroxidase isoforms. Using Catharanthus roseus cells and protoplasts, Mera et al. (2003) showed that only 4% of the total peroxidase activity was secreted into the growth medium by cultured cells, while in protoplasts 45% was secreted, implying that in normal cells the wall retains most of the peroxidases that can be secreted. However, following stress, these cell wall-bound peroxidases can be released into the apoplast (Sgherri, Quartacci & Navari-Izzo 2007). While the precise mechanism of their release is unknown, it seems likely that it enables plants to provide targeted ROS production, essential for the initial stress response.
In our earlier work, we discovered that following excision of the roots from wheat seedlings, the rates of O2•− production by the roots greatly increased and were correlated with changes in the membrane potential and ion fluxes (Minibayeva, Kolesnikov & Gordon 1998). Inhibitors of peroxidases (e.g. KCN, NaN3), and to much lesser extent inhibitors of flavin-containing enzymes (e.g. diphenylene iodonium), reduced O2•− production (Minibayeva et al. 1998). Wounding and various effectors, such as salicylic acid, carbonic acids, Mn2+ and detergents, increased peroxidase activity in the leachates (referred to hereafter as ECPOX), which was correlated with an increase in O2•− production (Minibayeva et al. 2001). Later work showed that wounding caused the release of several ECPOX isoforms (Minibayeva, Mika & Lüthje 2003). Taken together, these results suggest that apoplastic peroxidases are involved in O2•− production in roots. The aim of the present work was to characterize the peroxidases released following wounding by studying their dual ability to metabolize classical peroxidase substrates and H2O2, but also to produce O2•−. We also aimed to determine their electrophoretic mobility and to carry out peptide mass analysis of partially purified peroxidases.
MATERIALS AND METHODS
Wheat (Triticum aestivum L. cv. Ljuba; Niva Tatarstana, Kazan, Russia) seedlings were grown hydroponically in 0.25 mm CaCl2 at 22 °C for 5 d. Intact seedlings were incubated with only their roots in solution to avoid enzyme release from the rest of the plant. Roots (250 mg) were stressed by excising them from the seedlings and further incubating with gentle agitation in 3 mL of 0.25 mm CaCl2 (pH 7.0) for 1 h.
In the experiments in which the sources of the released peroxidases were studied, excised roots were subjected to a ‘triple washing’ that involved changing the incubation solution three times at 5 min intervals. The roots were then further incubated without changing the solution for 90 min (total incubation time: 105 min). To displace ionically bound enzymes from the cell wall, the heavy metal gadolinium [1 mm Gd(NO3)3] from the lanthanide family was applied to the triple washed roots. To study the possibility of the de novo synthesis and secretion of wound-induced peroxidases from the cytosol, the inhibitors of protein synthesis 8.5 µm cyclohexamide and 40 µm actinomycin D, and the inhibitors of protein secretion 8.8 µm brefeldin A and 50 µm monensin were applied to the triple washed roots. In this series of experiments, enzyme activities and potassium release were measured by sampling the incubation solution after 30, 60 and 90 min.
The seedlings used for protein purification were grown in a hydroponic medium containing 1 mm CaCl2, 1 mm KCl and 1 mm MgCl2 for 5 d in the dark at 25 °C.
Extracellular O2•− production was estimated using two independent colorimetric assays, and the identity of the radical was confirmed by ESR. In the first assay, the O2•−-dependent oxidation of 1 mm epinephrine (Sigma, St Louis, MO, USA; pH 7.0) by roots and leachates was followed at A480 (ε = 4.02 mm−1 cm−1) (Takeshige & Minakami 1979). In the second assay, the rates of O2•− production by purified proteins were measured using the tetrazolium salt 2,3-bis(2-methoxy-4-nitro-5-sulphophenyl)-2H-tetrazolium-5-carboxanilide (XTT) (Sigma) (A470; ε = 21.6 mm−1 cm−1) (Able, Guest & Sutherland 1998). The assay mixture contained 50 mm phosphate buffer (pH 7.0), 0.2 mm XTT, 0.2 mm NADH (Sigma) and sample aliquot. The specificity of the assays was confirmed by the inhibition of O2•− production by 250 units mL−1 SOD by up to 75, 95 and 100% in roots, leachates and purified proteins, respectively. In ESR measurements of O2•− production, the spin trap Tiron (4,5-dihydroxy-1,3-benzene-disulphonic acid disodium salt; Sigma) was used (McRae, Baker & Thompson 1982; Beckett et al. 2003). After removing the plant material, the pH of the incubation solution containing 50 mm Tiron was adjusted to 8.5, and ESR signals from Tiron semiquinone radicals were recorded with a radiospectrometer RE 1306 (Smolensk, Russia) at a modulation of 100 kHz with amplitude of 1.0 G and a time constant of 0.3 s.
H2O2 concentrations were measured using the xylenol orange assay (A560) (Gay & Gebicki 2000). Working reagent contained 0.1 mL of reagent A comprising 25 mm FeSO4, 25 mm (NH4)2SO4 and 2.5 m H2SO4, and 10 mL of reagent B comprising 125 µm xylenol orange (Sigma) and 100 mm sorbitol. Assays contained 0.1 mL of the sample solution and 3 mL of the working reagent. Specificity of the assay was confirmed by the inhibition of H2O2 production by 500 units mL−1 catalase. The rates of exogenous breakdown of 500 µm H2O2 by roots and leachates were measured. The sensitivity of H2O2 breakdown to the inhibitor of catalase ATZ (Sigma) and the inhibitor of peroxidase NaN3 was tested. After pre-incubation of roots or leachates with and without inhibitor for 15 min, an aliquot of H2O2 was added to give a final concentration of 500 µm. The concentration of H2O2 that remained in the solution was measured after 30 min. The pH of all solutions was adjusted to 7.0.
Enzyme activity assays
Peroxidase activity was usually measured as the oxidation of guaiacol in the presence of H2O2. The assay solution (1 mL) contained 25 mm sodium acetate buffer (pH 5.0), 25 µL of sample, 8.8 mm H2O2 and 8.26 mm guaiacol (A470; ε = 26.6 mm−1 cm−1). Rates were corrected using appropriate blank controls for any spontaneous oxidation of substrates. Other substrates tested were 0.1 mmo-dianisidine (A460; ε = 30.0 mm−1 cm−1), 0.5 mm 2,2′-azino-bis(3-ethylbenzthiazoline-6-sulphonic acid) (ABTS) (A405; ε = 36.8 mm−1 cm−1), 0.1 mm coniferyl alcohol (A265; ε = 7.5 mm−1 cm−1), 0.1 mmp-coumaric acid (A310; ε = 16.6 mm−1 cm−1), 0.1 mm ferulic acid (A310; ε = 16.6 mm−1 cm−1) and indole-3-acetic acid (IAA) (A261; ε = 3.2 mm−1 cm−1). Phenolic compounds were dissolved in 50% (v/v) dimethyl sulphoxide (DMSO), which resulted in a final concentration of 0.5% (v/v) DMSO in the assay solution.
Activity of MDH in leachates was assessed by the reduction of 0.4 mm oxalate acetate by 25 µL aliquot of sample in 50 mm tricine buffer, pH 8.5. After addition of 0.2 mm NADH, decrease in A340 was measured.
Cytosolic contamination was assessed by measuring the activity of G6PDH (Löhr & Waller 1974). The leakage of cytosolic enzymes into leachates was quantified by measuring the proportion of G6PDH released into the medium following stress.
Potassium release was measured using a flame photometer Phlapho 41 (Carl Zeiss, Jena, Germany).
Proteins in leachates were precipitated by (NH4)2SO4 (30–80%) and centrifuged at 23 500 g for 20 min. The resulting pellet was resuspended in 25 mm sodium acetate buffer, pH 5.0, and desalted in the same buffer using PD-10 columns (GE-Healthcare, Freiburg, Germany). ECPOXs were purified by successive Con A affinity and SEC on a high-performance liquid chromatography (HPLC) system (ÄKTA; GE-Healthcare) at 4 °C.
For affinity chromatography, Con A Sepharose column (7 × 35 mm; GE-Healthcare) was equilibrated with 20 mm Tris buffer, pH 7.4, 0.5 m NaCl. Bound proteins were eluted with 20 mmα-methyl-d-glucopyronaside (M-αd-GP) in the linear gradient (0–100%). Flow-through fractions (2.5 mL) and fractions with eluted proteins (1 mL) were collected at a flow rate of 0.2 mL min−1. Fractions, which displayed high guaiacol peroxidase activity, were subjected to further separation by SEC.
For SEC, after buffer exchange using PD-10 columns with 50 mm phosphate buffer (pH 7.0), 0.15 m NaCl, 1 mm ethylenediaminetetraacetic acid (EDTA), 1% glycerol proteins were concentrated using Centricons (YM-10; Millipore, Bedford, NY, USA). Samples of 500 µL were loaded on a Superdex 200 column (HR 10/30; GE-Healthcare), and proteins were eluted with phosphate buffer with a flow rate of 0.5 mL min−1. The fraction size was varied by software to be between 0.75 and 0.5 mL depending on absorbance (λ = 280 nm). Estimates of the molecular masses of native peroxidases were calculated using a semilogarithmic plot of the molecular mass values for the calibration proteins (Bio-Rad, Munich, Germany) against the elution volumes.
Electrophoretic study of ECPOXs
Electrophoretic studies were carried out using concentrated leachates and purified peroxidases from wounded roots. Samples were concentrated using Centricons and ‘Ultrapure-MC’ concentrators (Millipore), further separated by modified native sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS–PAGE) on 11% (w/v) polyacrylamide gels (Laemmli 1970) and stained either by 6.3 mm TMB (Thomas, Ryan & Levin 1976; Mika & Lüthje 2003) or by incubation in 20 mm guaiacol in 250 mm acetate buffer containing 10% (v/v) glycerol, pH 5.0. Almost immediately after the addition of 1 mm H2O2, orange bands appeared at the position of the peroxidases. For in-gel staining of O2•− production, gels were pre-equilibrated in 50 mm phosphate buffer, pH 7.4, containing 0.1 mm MgCl2, 1 mm CaCl2 and 10% (v/v) glycerol for 30 min. Following this, they were incubated in the dark in the same buffer containing 0.5 mm NBT (Sigma) and 0.4 mm NADH. Blue formazan bands appeared at the position of O2•− production (López-Huertas et al. 1999). Control gels were incubated in the absence of an electron donor. Gels were also stained with the ‘Silver Plus’ stain (Bio-Rad). Gels were run with molecular mass markers (‘Broad Range’; Bio-Rad) that were heated with mercaptoethanol before loading and stained using Coomassie brilliant blue G250 (Sigma).
Protein concentrations were measured according to Bradford (1976) with bovine serum albumin (BSA) as the standard.
Mass spectrometry and sequence analysis
Protein bands were cut out after native SDS–PAGE and haem or silver staining, and following trypsin digestion proteins were extracted with 50% acetonitrile/5% formic acid. Extracts were dried in vacuum concentrator, redissolved in 5% methanol/5% formic acid, desalted on a C18 mZipTip (Millipore, Schwalbach, Germany) and eluted with 1 mL 60% methanol/5% formic acid. Eluates were mixed with matrix solution (saturated solution of α-cyano-4-coumaric acid in 30% acetonitrile/water/0.1% trifluoroacetic acid) and analysed by nano-electrospray mass spectrometry (ESI–MS/MS) as described by Mika, Buck & Lüthje (2008). ESI–MS/MS was performed with a hybrid tandem mass spectrometer (QTOF II; Micromass, Manchester, UK). Mass spectral analysis was performed with a MALDI mass spectrometer REFLEX IV (Bruker, Bremen, Germany) with time-of-flight detector. MS/MS spectra obtained by collision-induced fragmentation of the peptides were evaluated both manually and by Mascot MS/MS ion search algorithm (Matrix Sciences, London, UK). Protein sequences were extracted from the most recent versions of available databases (http://www.ncbi.nlm.nih.gov, http://peroxidase.isb-sib.ch).
The rate of extracellular O2•− production in excised roots was almost double compared to that of intact roots, and also in leachates derived from these roots (Table 1). Cytoplasmic contamination of leachates from wounded roots, estimated by the activity of cytoplasmic marker G6PDH, was only 1.4 ± 1.2%, indicating that these increases in extracellular O2•−-producing activity were not simply the result of the leakage of enzymes from cytoplasm at the wounded site. Production of O2•− in both intact and wounded roots was confirmed by ESR using the non-permeable spin trap Tiron. Wounding increased the amplitude of the ESR spectrum of the Tiron semiquinone radical formed (Fig. 1). Exogenous application of the non-permeable electron donors NADH, NADPH and ferrocyanide to excised roots further increased extracellular O2•− production. Production of O2•− by both excised roots and, to a greater extent, leachates was strongly inhibited by 0.1 mm KCN (Table 2).
Table 1. Extracellular O2•− production in roots and leachates, and guaiacol peroxidase activity in the leachates derived from intact and excised roots after 1 h incubation of roots in 0.25 mm CaCl2 (pH 7.0)
O2•− production (µmol g−1 FW min−1)
Peroxidase activity (µmol g−1 FW min−1)
Values are given ± standard deviation, n = 3.
4.6 ± 0.1
6.3 ± 0.1
87.9 ± 15.3
8.2 ± 0.3
10.3 ± 0.2
2307 ± 14.2
Table 2. Extracellular production of O2•− in roots and leachates after 1 h incubation of excised roots with exogenous electron donors, with and without 0.1 mm KCN. Values are given ± standard deviation, n = 9
O2•− production (%)
Excised roots incubated in 0.25 mm CaCl2 were used as a control.
Excision of roots from seedlings caused alkalinization of the incubation medium by ca. 0.5 unit within 1 h (data not shown), and also stimulated the release of peroxidase (Table 1) and MDH activities into the extracellular solution. Interestingly, the ECPOX activity of wounded roots was 26 times higher than that of intact roots, a much greater increase compared to the wound-induced increase in O2•− production. The activity of MDH in the leachates derived from wounded roots was readily measurable (about 30 µmol mg−1 protein min−1), although it was undetectable in intact roots.
Excised roots were subjected to a ‘triple washing’ by changing the incubation solution three times at 5 min intervals. The production of O2•− and peroxidase activity in the leachates rapidly declined to low levels (Fig. 2a). During further incubation of roots without changing the solution for 90 min, extracellular O2•− production and peroxidase activity recovered significantly (Fig. 2a). The K+ concentration of the leachate remained almost constant, indicating that K+ leakage from the washed roots was very low. Application of 1 mm Gd(NO3)3 to washed roots sharply increased the amount of extracellular O2•− and peroxidase activity (by 500 and 800%, respectively). This was not accompanied by the K+ leakage (Fig. 2b).
Incubation of excised roots with exogenously supplied 1 mm H2O2 caused a rapid but transient increase in the rate of extracellular O2•− production (Fig. 3). The rate of O2•− production initially increased seven times, dropped to half of this value within the next 10 min and then returned to the initial level after 60 min.
The rates of exogenous detoxification of 500 µm H2O2 by: (1) excised roots pre-incubated in 0.25 mm CaCl2 for 1 h (‘unwashed roots’); (2) roots from (1) placed into new incubation medium (‘washed roots’); and (3) leachates from (1) were measured. Excised unwashed roots completely metabolized exogenous H2O2 within 30 min (Fig. 4). Interestingly, not only unwashed roots (1), but also washed roots (2) and leachates (3) metabolized ca. 50% of exogenous H2O2 during the same time. H2O2 detoxification was not sensitive to the catalase inhibitor ATZ, while the peroxidase inhibitor NaN3 strongly inhibited this reaction (Fig. 5). Breakdown of H2O2 by leachates (3) was more sensitive to NaN3 (inhibited by 86%) than breakdown by washed (by 66%) and unwashed (by 71%) roots.
Proteins in leachates derived from intact and wounded roots were precipitated with 30–80% (NH4)2SO4 and then purified by affinity chromatography. In samples from intact roots, almost no guaiacol peroxidase activity was bound to the Con A Sepharose column (Fig. 6), while in samples from wounded roots, total guaiacol peroxidase activity was equally divided between bound and unbound proteins. Bound proteins with the highest peroxidase activity were eluted by 20 mm M-αd-GP (Fig. 6).
All fractions containing guaiacol peroxidase activity after Con A chromatography were further purified by SEC. Proteins derived from the intact roots showed a single peak of peroxidase activity that eluted at 15.2 mL (Fig. 7a). Proteins derived from the wounded roots that did not bind to the Con A column were also present in a single peak eluted at 15.8 mL (Fig. 7b), while putatively glycosylated proteins (i.e. those that bound to the Con A column) eluted as two peaks at 12.9 and 15.2 mL (Fig. 7c). Using standard proteins, we estimated that the ECPOX from intact roots had a molecular mass of 40 kD, while the mass of ECPOX from wounded roots (ECPOX 1) that did not bind to Con A was 37 kD. The masses of the two glycosylated ECPOXs were 136 kD (ECPOX 2) and 40 kD (ECPOX 3) for the first and second peaks, respectively.
ECPOXs readily oxidized guaiacol and artificial substrates such as o-dianisidine, and to a lesser extent ABTS (Table 3). Natural substrates such as phenolic acids and alcohols were oxidized in the following order of increasing rate of metabolism: p-coumaric acid < ferulic acid < coniferyl alcohol (Table 3). The exception was high molecular mass glycosylated ECPOX 2, which oxidized ferulic acid relatively more slowly. The rates of oxidation of coniferyl alcohol by the leachates and purified peroxidases, especially by ECPOX 1, were very high. IAA was metabolized by all peroxidases at very low rates.
Table 3. Peroxidase substrate specificity in the leachates from excised roots and purified proteins
Rates of oxidation (µmol µg−1 protein min−1)
Enzyme activities were measured in the presence of 8.8 mm H2O2 and concentrations of common peroxidase substrates as indicated. Data are given as means ± standard deviation, n = 3. For the wavelengths used and molar extinction coefficients, see Materials and methods.
n.d., not detected.
14.8 ± 0.8
508 ± 44
104 ± 12
160 ± 24
96 ± 20
32 ± 0
28 ± 4
12.2 ± 0.2
836 ± 68
44 ± 4
284 ± 32
1.2 ± 0
880 ± 0
12 ± 0
8 ± 0
9.4 ± 0.2
580 ± 44
324 ± 60
376 ± 36
18.6 ± 1.8
1032 ± 88
104 ± 16
416 ± 24
47.8 ± 1.6
3096 ± 84
820 ± 44
828 ± 196
28 ± 8
32 ± 4
28 ± 4
Classical peroxidase inhibitors such as KCN and NaN3 almost completely inhibited the activities of ECPOXs (Table 4), suggesting the presence of a haem prosthetic group. ECPOX activities in the leachates were almost unaffected by 0.1 mm MnCl2 (data not shown) and highly stimulated by 0.5 mm MnCl2. However, purified peroxidases were not significantly affected by 0.5 mm MnCl2 (Table 4). Application of 1 mm CaCl2 only slightly stimulated the activity of the leachate and the high molecular mass ECPOX 2, while the activity of the lower molecular mass peroxidase isoforms, especially ECPOX 1, greatly increased.
Table 4. Influence of inhibitors and effectors on peroxidase activity in the leachates from excised roots and purified proteins
Relative peroxidase activity (%)
Enzyme activities were measured in the presence of 8.26 mm guaiacol and 8.8 mm H2O2 at pH 5.0. Absolute rates of peroxidase activity in the control correspond to the rates of metabolism of 8.26 mm guaiacol given in Table 3. Data are given as means of the relative activities ± standard deviation, n = 3.
100 ± 6
100 ± 9
100 ± 24
100 ± 7
2 ± 0
1 ± 0
3 ± 1
4 ± 0
4 ± 0
11 ± 2
160 ± 7
111 ± 9
69 ± 18
107 ± 3
126 ± 5
666 ± 40
105 ± 13
158 ± 25
Purified ECPOXs were tested for their ability to produce superoxide radicals, measured by XTT reduction. Superoxide production by purified ECPOX from intact roots was not detectable. ECPOX 1 from wounded roots had the highest rates of XTT reduction, when expressed on a protein unit basis, while rates for the glycosylated proteins were lower (Table 5).
Table 5. Superoxide production by partially purified ECPOXs from intact and wounded roots
Peroxidase isoforms on the basis of their binding to Con A
Molecular mass (kD)
XTT reduction (µmol mg−1 protein min−1)
Data are given as means of the activities ± standard deviation, n = 3.
n.d., not detected.
Con A-unbound ECPOX
Con A-unbound ECPOX 1
9.7 ± 1.8
Con A-bound ECPOX 3
4.3 ± 0.0
Con A-bound ECPOX 2
3.0 ± 0.2
Following electrophoresis of concentrated non-purified leachates, peroxidase visualization by both guaiacol and the haem-binding phenolic compound TMB revealed two main bands with molecular masses of ca. 37 and 40 kD, and a feint band with a molecular mass of ca. 28 kD, corresponding to the bands with same molecular masses visualized by silver staining (Fig. 8). For purified proteins, estimation of their molecular masses by SEC was further confirmed by modified SDS–PAGE. In the Con A-unbound fractions, peroxidases with molecular masses of 37 and 28 kD were identified following TMB staining (Fig. 8). In the Con A-bound fractions, TMB staining showed a single band with a molecular mass of 40 kD (Fig. 8). Isoelectric focusing of purified peroxidases followed by TMB staining revealed a single broad band corresponding to a pI of ca. 7.6 (data not shown). None of the bands separated under reducing conditions induced by dithiothreitol (DTT) (data not shown), indicating that these proteins are monomers.
Identities of the proteins were confirmed as peroxidases using nano-electrospray mass spectrometry analysis. A single peptide was identified from the Con A-unbound 37 kD protein band. The unique ECPOX could not be identified. A BLAST search at NCBI demonstrated that the peptide has a highly conserved sequence found in at least two wheat peroxidases (Table 6). Both peroxidases were suggested to be haem-containing proteins with two Ca2+-binding sites (Baga, Chibbar & Kartha 1995). A third peroxidase was suggested, but the E-value was lower compared to the others. None of these proteins appears to be glycosylated. Two peptides were identified from the 28 kD protein band (Table 6). BLAST searches with these sequences indicated a cytosolic ascorbic peroxidase. This protein was identified by amino acid sequence similarity with sequences from rice (Oryza sativa L.). A search in the peroxidase database demonstrated that both peptides could be found in TaAPX05. A single peptide was identified from the Con A-bound 40 kD protein band (Table 6). This peptide has been reported as a wheat EST (BE406480). An additional BLAST search with the full-length sequence of this EST indicated a peroxidase, but the unique protein could not be identified. ECPOX 2, a putative glycosylated isoform with a molecular mass of 136 kD, was not detectable following electrophoresis and visualization using silver, guaiacol or TMB staining.
Table 6. Peptide sequences of peroxidases after Con A chromatography of concentrated leachate from wounded roots
Experimental relative molecular mass (kD)
ESI–MS/MS spectra obtained by collision-induced fragmentation of the peptides were evaluated both manually and by Mascot MS/MS ion search algorithm.
Con A-unbound ECPOX
Con A-unbound ECPOX
Con A-bound ECPOX
Wounding of roots stimulates extracellular O2•− production and induces release of peroxidases
It has become clear that despite the traditional view of ROS as toxic molecules, a so-called ‘physiological poison’, temporary accumulation of ROS is a universal reaction of organisms to stress. As discussed in the Introduction, this accumulation probably has many roles, including pathogen defence, and in higher plants synthesis of protective polyphenolics such as lignin and suberin. In addition, the roles of ROS as second messengers involved in the regulation of cell development, adaptation and programmed cell death are well established (for reviews, see Laloi, Apel & Danon 2004; Gechev et al. 2006; Foyer & Noctor 2009). Here, we show that excision induces a burst of apoplastic O2•− in excised wheat roots and leachates derived from them (Table 1; Fig. 1). In plants, sources of extracellular ROS production include plasma membrane NAD(P)H dehydrogenases or oxidases, amine oxidases, the non-haem Fe-containing enzyme lipoxygenase and cell wall peroxidases (Mika et al. 2004; Sicilia et al. 2005; Bindschedler et al. 2006; Cona et al. 2006; Sagi & Fluhr 2006). However, the specific contribution of each source to the oxidative burst in plants in response to different stimuli is still a matter of debate. In our earlier publications (Minibayeva et al. 1998, 2001), we demonstrated that the sensitivity of O2•− production by wounded roots to flavoprotein inhibitors was rather low, implying that NAD(P)H oxidase-like enzymes do not make a major contribution to the total production of extracellular O2•− by roots. Results from the present study provide new insights into the rapid response of roots to wounding, and underline the importance of apoplastic peroxidases. The high sensitivity of O2•− production to KCN by both roots and leachates (Table 2), wound-induced release of peroxidases into the extracellular solution (Table 1) and capability of purified ECPOXs to synthesize O2•− (Table 5) provide evidence that apoplastic peroxidases in roots function not only in their conventional mode by detoxifying H2O2, but can be a major source of wound-induced ROS in the apoplast.
Wound-induced release of O2•− and peroxidase activity apparently did not occur simply as passive release from the cut surface or through damaged plasma membranes, because leachates contained only trace amount of cytoplasmic contamination and K+. Furthermore, extracellular O2•− synthesis and peroxidase activity were depleted by washing, but recovered significantly when roots were further incubated without changing the solution (Fig. 2a). The mechanisms of this recovery are intriguing. One can speculate that it might occur because of the release of more tightly bound isoforms from the cell wall and plasma membrane, or the secretion of existing or newly synthesized peroxidases from the cytoplasm. The latter seems unlikely to happen, because recovered activity was not diminished by incubating roots with inhibitors of protein secretion (brefeldin A and monensin) and synthesis (actinomycin D and cyclohexamide) (data not shown). These results are consistent with the findings of Nair & Showalter (1996) that wound-induced stimulation of cell wall peroxidases in carrot roots was the result of the activation of pre-existing isoforms rather than de novo synthesis of proteins. The large increase in peroxidase activity and O2•− production in the leachates that occurred following the treatment of roots with the heavy metal gadolinium (Fig. 2b) suggested that many released peroxidases are ionically bound to the cell wall. Lanthanide metals cannot readily penetrate the plasma membrane, but rather bind to the cell surface from which they can displace proteins and ions. Thus, data presented here suggest that wounding induces the controlled, progressive release of pre-existing peroxidase isoforms into solution, and stimulates their capacity to produce O2•−.
Characterization of stress-induced peroxidases
Initial characterization of peroxidases released following wounding was based on their ability to metabolize classical peroxidase substrates (Table 3) and sensitivity to typical peroxidase inhibitors such as KCN and NaN3 (Tables 2 & 4). Both Ca2+ and Mn2+ strongly stimulated the ECPOX activity in the leachate (Table 4), typical for many peroxidases (Hiraga et al. 2001). Calcium probably maintains the protein conformation, and Mn2+ is a cofactor involved in the regulation of enzyme activity (Van Huystee, Rodriguez Marañón & Wan 1996). Typically for peroxidases, ECPOXs in the leachate were found to be diverse. Two major isoforms with molecular masses of ca. 37 and 40 kD (Fig. 8) were visualized following TMB (haem group) and guaiacol (activity) staining of separated by native PAGE proteins. Importantly, these bands corresponded to the bands visualizing O2•−, suggesting that these ECPOXs contribute to O2•− production. Moreover, the patterns of TMB staining confirmed the presence in the leachates some higher and lower molecular mass proteins with haem groups, probably corresponding to catalases and ascorbic (class I) peroxidases. Other workers have reported that the apoplast of higher plants may contain both class I and class III peroxidases (e.g. de Pinto & De Gara 2004).
Our earlier work had demonstrated that wound-induced ECPOXs are sensitive to lectins, implying that they are glycosylated (Minibayeva et al. 2003). As a first step in purification, we therefore initially separated them using a Con A column. Con A binds reversibly α-d-mannosyl, α-d-glycosyl or similar residues of glycoproteins, whereas other sugar residues will not bind to the column. Therefore, it should be noted that some peroxidases which did not bind to Con A column may have been glycosylated, but with other sugar residues. However, for simplicity, in the following discussion peroxidases that did not bind to a Con A column are here referred to as ‘non-glycosylated’. Interestingly, while intact roots released only a non-glycosylated isoform, wounding stimulated the activity of both non-glycosylated and glycosylated peroxidases (Figs 6 & 8). Peptide mass analysis confirmed that a peptide from non-glycosylated 37 kD peroxidase has a highly conserved sequence found in at least two wheat peroxidases, while 28 kD isoform has homology with the ascorbic peroxidase from rice (Table 6). Glycosylated 40 kD peroxidase contained a peptide that has been reported as a wheat peroxidase EST (Table 6). SEC analysis of the glycosylated forms suggested the presence of an additional isoform with a molecular mass of 136 kD (Fig. 7). This isoform never appeared in PAGE gels, possibly because it is a complex or oligomer that breaks down into its constituent monomers or loses activity following electrophoresis.
Three partially purified wound-induced ECPOX isoforms found here differed in their sensitivity to inhibitors and effectors, and ability to metabolize various substrates. The most active isoform was the 37 kD non-glycosylated peroxidase, ECPOX 1 (Tables 3 & 4). Interestingly, this isoform had the similar molecular mass as the isoform from intact roots, but its activity was much higher (Fig. 6). Comparing the two other wound-induced glycosylated peroxidases, 40 kD ECPOX 3 tended to be more sensitive to effectors and showed a higher affinity for substrates than 136 kD ECPOX 2 (Tables 3 & 4). Thus, the wound-induced stimulation of the release of the non-glycosylated peroxidase isoform and the induction of the release of glycosylated peroxidases with different sensitivity to substrates and effectors suggested that all these isoforms may play different roles in the wounding response. According to Cosio & Dunand (2008), the diversity of processes catalysed by peroxidases, as well as the great number of their genes, suggests the existence of functional specialization of the members of this protein family.
Dual roles of ECPOXs in ROS metabolism in the apoplast
Direct evidence that extracellular ECPOXs can produce O2•− radicals came from experiments with partially purified wound-induced peroxidases (Table 5). All three peroxidase isoforms [the 37 kD non-glycosylated peroxidase (ECPOX 1) and glycosylated peroxidases (ECPOX 2 and 3)] demonstrated the capacity to produce O2•− (Table 5). In their ‘oxidative cycle’, peroxidases form O2•− using H2O2 as a substrate, for example, when participating in lignin biosynthesis (Halliwell 1978). In the present study, adding 1 mm H2O2 to wounded roots caused a rapid but transient increase in extracellular O2•− production (Fig. 3), suggesting that in wheat roots oxidative activities of peroxidases can be modulated by the supply of H2O2. One can speculate that cooperativity of redox enzymes in the apoplast in the production and detoxification of ROS may occur. For example, apoplastic amine oxidases and oxalate oxidases can produce H2O2, which can be used by peroxidases to produce O2•− (Cona et al. 2006).
Data presented here show that ECPOXs from wheat roots not only produce O2•−, but can also efficiently metabolize exogenously supplied H2O2. Both excised roots and leachates derived from them rapidly reduced the concentration of exogenously supplied H2O2 (Fig. 4). Peroxidases were almost certainly responsible for breaking down the H2O2, because breakdown was sensitive to the peroxidase inhibitor NaN3 but not the catalase inhibitor ATZ (Fig. 5). However, these data do not exclude the involvement of catalases in the H2O2 metabolism in the apoplast of higher plants (Vanacker, Carver & Foyer 1998; Agostini et al. 2002). The ability of roots to rapidly break down H2O2 will protect them from the potentially damaging consequences of the pro-oxidative activities of peroxidases and other apoplastic enzymes.
Among essential conditions that can switch peroxidases from an ROS breaking down to an ROS-producing mode, the alkalinization of the apoplast and the presence of appropriate reductants are important factors (Bolwell et al. 1999). Firstly, an increase in the pH of the medium following excision and incubation of roots for 1 h may provide the necessary condition for apoplastic peroxidases to shift to oxidizing cycle. Secondly, exogenous applications of a reductant such as NADH stimulated apoplastic O2•− production (Table 2). A potential source of NADH in the apoplast is MDH (Córdoba-Pedregosa et al. 1998; Hadži-TaškovićŠukalovićet al. 1999), and in our experiments wounding caused a release of proteins possessing MDH activity. A model for O2•− production by apoplastic peroxidases involving MDH was proposed by Elstner & Heupel (1976), but as yet no reliable evidence exists for the occurrence of NADH in the apoplast. Other potential apoplastic peroxidase reductants include naturally occurring phenolic acids and phenolics, and root leachates could rapidly metabolize these compounds (Table 3). Such phenolics can play roles in both the oxidative and peroxidative cycles of cell wall peroxidases (Hadži-TaškovićŠukalović, Vuletić & Vučinić 2005). It therefore seems likely that the secretion or apoplastic synthesis of appropriate reductants is an important regulatory mechanism for determining whether peroxidases work to produce or to detoxify ROS.
The present study gives new insights into the rapid response of roots to wounding. Our results provide strong evidence that an important component of this response is mediated by pre-existing peroxidases that are released from the cell surface into the apoplast where they can display both oxidative and peroxidative activities. Many factors, such as supply of reductants, shifts in pH and cooperativity with other redox enzymes, can probably switch peroxidases between synthesizing and breaking down ROS, providing the concentrations and molecular species of activated oxygen needed during the early stages of the response of roots to stress. Some peroxidase isoforms may have been selected because they are particularly effective at producing ROS, while others are better suited to ROS removal. It is also possible that some peroxidases can display dual functions during stress, and are efficient in both peroxidative and oxidative cycles. These peroxidase-mediated changes in ROS metabolism enable plants to respond rapidly to wounding and may assist defence processes. Current findings may provide the basis for future manipulation of wound response by the creation of transgenic plants with altered levels of peroxidase gene expression and activities.
This study was supported by the RFBR (No. 09-04-01394), Program of RF president for the Leading Scientific Schools (No. 5492.2008.4). F.M. had a visiting research fellowship sponsored by Deutsche Forschungsgemeinschaft (DFG 436 RUS 17/123/02). We also gratefully acknowledge the financial support of the South African–Russian bilateral agreement for scientific collaboration. The authors appreciate helpful discussions with Angela Mika (University of Hamburg, FRG).