JXB Advance Access originally published online on May 12, 2006
Journal of Experimental Botany 2006 57(8):1633-1644; doi:10.1093/jxb/erj197
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
RESEARCH PAPER |
Effect of ascorbate and its oxidation products on H2O2 production in cell-suspension cultures of Picea abies and in the absence of cells
The Edinburgh Cell Wall Group, Institute of Molecular Plant Sciences, School of Biological Sciences, The University of Edinburgh, Daniel Rutherford Building, The King's Buildings, Edinburgh EH9 3JH, UK
To whom correspondence should be addressed. E-mail: S.Fry{at}ed.ac.uk
Received 26 January 2006; Accepted 14 March 2006
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Dehydroascorbate and traces of ascorbate were present apoplastically in living spruce (Picea abies) twigs. Since the proposed apoplastic ascorbate degradation pathway contains several steps that possibly generate H2O2, the effects of ascorbate and some of its degradation products were tested on apoplastic H2O2 concentrations in a cell culture of P. abies as a model and on non-enzymic H2O2 production in vitro. Ascorbate scavenged H2O2 in the culture medium of lignin-producing Picea cells and in spent and boiled spent medium; in the presence of Cu2+ or fresh medium, ascorbate led to the non-enzymic generation of H2O2. Preparations of dehydroascorbate (the initial oxidation-product of ascorbate), and diketogulonate (the hydrolysis-product of dehydroascorbate) induced H2O2 accumulation both non-enzymically and enzymically in Picea cell-suspensions. Paper electrophoresis showed that the dehydroascorbate and diketogulonate preparations contained several degradation products; some of these probably contributed to H2O2 production and/or scavenging in these experiments, and would also do so in vivo. These results indicate a complex ability of apoplastic ascorbate, dehydroascorbate, diketogulonate, and further products to modulate H2O2 concentrations, with potential consequences for the control of growth, development and lignification.
Key words: Apoplast, ascorbate, dehydroascorbate, diketogulonate, hydrogen peroxide, reactive oxygen species
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Vitamin C, L-ascorbic acid (AA), is a well-known antioxidant that efficiently scavenges free radicals, for example reactive oxygen species, formed in cellular metabolism or derived from the atmosphere (reviewed by Noctor and Foyer, 1998; Horemans et al., 2000; Smirnoff, 2005). In addition to its protective role against oxidative stress, AA is a co-factor of many enzymes and has a role in plant cell division and elongation (Chinoy, 1984; Asard et al., 2004). In many plant tissues AA is present at mM concentrations, predominantly as the singly ionized form (pKa1
4.2). A few per cent of the total of the plant's AA is found in the apoplast (the liquid permeating the cell walls; Castillo et al., 1987; Polle et al., 1990; Takahama, 1993a; Otter and Polle, 1994; Sánchez et al., 1997; Horemans et al., 2000; Hernandez et al., 2001; Veljovic-Jovanovic et al., 2001; Burkey et al., 2003; Córdoba-Pedregosa et al., 2003; Pedreira et al., 2004; Hu et al., 2005), making it an important antioxidant in the plant cell wall as well (Pignocchi and Foyer, 2003; Sanmartin et al., 2003; Jaspers et al., 2005). Many abiotic and biotic stresses lead to rapid changes in the concentration and/or redox status of the apoplastic AA, for example, wounding (Takahama, 1993a), salt stress (Hernandez et al., 2001), ozone treatment (Burkey et al., 2003; Jaspers et al., 2005), and water deficit (Hu et al., 2005). AA concentration also changes during the diurnal cycle (Burkey et al., 2003) and with normal development (Sánchez et al., 1997; Cordoba-Pedregosa et al., 2003; Pedreira et al., 2004). AA is oxidized either enzymically or non-enzymically to L-dehydroascorbate (DHA) via the monodehydro-L-ascorbate radical (Noctor and Foyer, 1998; Deutsch, 2000; Smirnoff, 2005). In the symplast an efficient enzymic regeneration of AA from DHA occurs via the AA–glutathione cycle (Foyer and Halliwell, 1976; Noctor and Foyer, 1998), making AA a major redox buffer in plants. Ascorbate oxidase (AAO), a cell wall-localized enzyme, regulates the redox status of AA in the apoplast: it oxidizes AA with the concomitant reduction of O2 to water (Pignocchi and Foyer, 2003; Sanmartin et al., 2003; Smirnoff, 2005). No clear evidence for an apoplastic AA regeneration system exists. However, different plasma membrane-associated carrier systems for AA and DHA translocation have been reported in several plant species (reviewed by Horemans et al., 2000). Preferential uptake of the oxidized form (DHA) into the cells for re-reduction has been described (Horemans et al., 1997; Kollist et al., 2001). Alternatively, as DHA is unstable in aqueous solutions, it can be further metabolized in the cell wall. The first step is the hydrolysis of DHA to form 2,3-diketo-L-gulonate (DKG), which is thought to be irreversible at physiological pH (Deutsch, 2000). In the apoplast of cultured rose cells, DHA is degraded via several intermediates to oxalate and L-threonate, some of these reactions possibly generating H2O2 (Green and Fry, 2005a, b).
Lignin is an important structural component of the cell wall of certain supportive and water-conducting cells. The final polymerization steps of lignin biosynthesis take place after the activation of monolignols to free radicals by the action of peroxidase +H2O2 and/or oxidase (laccase) +O2 followed by non-enzymic coupling of monolignol radicals to form the polymer, lignin (Boerjan et al., 2003; Ralph et al., 2004). Oxidative coupling of ferulates to form dehydrodiferulates in primary cell walls (and to some extent intra-protoplasmically: Fry et al., 2000) of gramineous monocots and of certain dicots (e.g. Chenopodiaceae) and of tyrosine residues of cell wall glycoproteins to form di-isodityrosine and related products are possibly mediated by peroxidases (Kerr and Fry, 2004; Fry 2004, Encina and Fry, 2005). All these cross-linking reactions in the cell wall are likely to lead to the restriction of cell elongation. In this context, it is interesting that a negative correlation was found between apoplastic H2O2 levels and the growth rate of pine hypocotyls, and that there was a positive correlation between the AA/H2O2 ratio and the growth rate (Pedreira et al., 2004).
AA interferes with the peroxidase-mediated oxidation of phenolics by reducing phenolic radicals back to the original phenolics, thereby scavenging H2O2 (Takahama, 1993a, b; Otter and Polle, 1994; Sanchez et al., 1997). It would therefore be predicted that, to allow lignification or diferulate formation to occur in the cell wall, apoplastic AA would first need to be completely oxidized to DHA. The reduced form (AA), however, only partially inhibited phenolic oxidation by covalently bound cell wall peroxidases from stems of Kalanchoë daigremontiana (Takahama, 1993a), young stems of Vigna angularis (Takahama, 1993b) or from lignifying needles of Picea abies (Otter and Polle, 1994). H2O2 content was shown to increase and AA content to decrease in the apoplastic fluid of pine hypocotyls with increasing hypocotyl age and along the hypocotyl axis, correlating with the reduction in growth (Pedreira et al., 2004).
In addition to its antioxidant character, the pro-oxidant nature of AA is well established. In the presence of transition metal ions, AA generates H2O2 non-enzymically (Dekker and Dickinson, 1940). H2O2 can then be used to generate extremely reactive hydroxyl radicals (•OH) via the Fenton reaction (Fry, 1998; Tabbì et al., 2001; Vreeburg and Fry, 2005). Hydroxyl radicals are hypothesized to have a role in cell-wall loosening by non-selectively breaking bonds within cell-wall polysaccharides, for example during root elongation (Schopfer, 2001; Rodriguez et al., 2002, 2004; Liszkay et al., 2004) or fruit ripening (Fry et al., 2001; Dumville and Fry, 2003; Vreeburg and Fry, 2005). In this connection, it is interesting that Pedreira et al. (2004) found a small increase in irreversible extension in response to exogenous 5 mM H2O2 in auxin-untreated pine hypocotyl segments (P <0.01 according to the data plotted in Fig. 2a of Pedreira et al., 2004). Likewise, Córdoba-Pedregosa et al. (2003) found higher apoplastic H2O2 concentrations in the apical 2 cm of onion roots (i.e. the portion containing the growing zone) than in the region 2–6 cm back from the tip. The apical 2 cm also had a lower AA/(AA+DHA) ratio in the apoplast than did the non-growing parts further back from the tip. Thus rapid cell expansion was associated here with a relatively oxidizing apoplastic environment (Córdoba-Pedregosa et al., 2003).
|
The present paper presents observations on AA, DHA, DKG, and apoplastic H2O2 production in cells of a conifer, the Norway spruce (Picea abies). Needles of Picea are among the plant materials already reported to possess both AA and DHA in the apoplast (Castillo et al., 1987; Polle et al., 1990; Otter and Polle, 1994), as are hypocotyls of another conifer, Pinus (Sánchez et al., 1997; Pedreira et al., 2004). Since the proposed apoplastic AA degradation pathway contains several steps that possibly generate H2O2 (Green and Fry, 2005a, b), the effects of AA and some of its degradation products were tested on apoplastic H2O2 levels. A cell culture of Picea abies that produces extracellular lignin (Simola et al., 1992; Brunow et al., 1998; Kärkönen et al., 2002) was used in in-vivo assays for H2O2. Peroxidases have an important role in the activation of monolignols for subsequent polymerization in this cell culture as the removal of endogenous H2O2 significantly decreased the extracellular lignin formation (Kärkönen et al., 2002; Koutaniemi et al., 2005). Non-enzymic responses in in-vitro assays were monitored by the use of fresh culture medium, cell-free spent culture medium, boiled spent culture medium, and CuSO4 (1 µM) solution. As some AA degradation products were observed to induce H2O2 generation in vivo, the endogenous AA and DHA content of the spent medium of spruce cell-suspensions and of the apoplast of spruce twigs were investigated during xylem differentiation. In this paper the two opposing effects of AA on H2O2 levels are confirmed: first, the scavenging of H2O2 in the culture medium of lignin-producing Picea cells and in spent and boiled spent medium; secondly, the non-enzymic generation of H2O2 in the presence of Cu2+ or fresh medium. The existence of degradation product(s) of AA that induce efficient H2O2 production (i) in a cell-dependent way, and (ii) in a non-enzymic reaction independent of cells is also reported. The data are discussed with reference to the biochemistry of the plant cell wall.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Chemicals
AAO, catalase, DHA, and xylenol orange (XO) were obtained from Sigma-Aldrich. AAO was dissolved (1 U µl–1) in 50 mM succinate (Na+, pH 5.6) supplemented with 0.05% BSA. DKG was prepared from the commercial DHA by alkali treatment (White and Krupka, 1965; Green, 2003): a stock of DHA (50 mM) was prepared in water, a 1.3-fold excess of NaOH was added and the mixture incubated at c. 20 °C for 6 min. The reaction was stopped by the addition of L-tartaric acid to a final pH (checked by pH paper) of
3.5–4.0.
Cell suspensions of spruce
Cultured cells of Picea abies Karst. (kindly provided by Professor emerita LK Simola, Professor TH Teeri, and Professor IA Kilpeläinen, University of Helsinki) were used as a model system for in vivo studies of H2O2 production. Callus was maintained as previously described (Simola et al., 1992; Kärkönen et al., 2002). For the experiment, c. 3 g callus (c. 2 weeks after subculturing) was transferred into 150 ml of liquid medium ‘5’ (Simola and Santanen, 1990) modified by omission of amino acids (except glycine), but supplemented with 10 mM tartrate (Na+, pH 4.5). Thus, the fresh medium (FM) contained: (i) Inorganic cations: Na+, 23 mM; 
, 10 mM; K+, 10 mM; Mg2+, 2 mM; Ca2+, 1 mM; Fe3+, 125 µM; Mn2+, 100 µM; Zn2+, 50 µM; Cu2+, 1 µM; Co2+, 0.5 µM; Ni2+, 0.1 µM. (ii) Inorganic anions: 
, 20 mM; Cl–, 7 mM; 
, 2.15 mM; 
, 2 mM; H3BO3, 200 µM; I–, 25 µM; 
, 1 µM. (iii) Organic components: sucrose, 58 mM; tartrate, 10 mM, pH 4.5; myo-inositol, 560 µM; EDTA, 125 µM; nicotinic acid, 4 µM; pyridoxine-HCl, 2.4 µM; thiamine-HCl, 0.3 µM; glycine, 26.6 µM; kinetin, 2.5 µM; 2,4-D, 0.5 µM.
The suspension was then aliquoted aseptically into 60-ml pots (Sterilin). The aliquots were incubated on a shaker (100 rpm) at 25 °C in an 18/6 h light/dark cycle (20 µmol m–2 s–1, Philips Master TL-D Super 80 18 W/830) for 24 h prior to the experiment so that the cells could acclimatize to suspension-culture conditions. The assays were conducted approximately 8–10 h after the start of the light period.
Elicitor preparations
Cell-wall fragments of a pathogenic fungus, Heterobasidion parviporum (= H. annosum of intersterility group S) B307 (Maijala et al., 2003) (a kind gift from Dr Pekka Maijala, Department of Applied Chemistry and Microbiology, University of Helsinki), were used as elicitor. The wall fragments were prepared according to Ayers et al. (1976) as modified by Schwacke and Hager (1992), suspended in water, autoclaved (120 °C, 20 min), freeze-dried and stored at –20 °C. For experiments, a defined dry weight of lyophilized material was resuspended in distilled water.
Detection of H2O2
H2O2 generation in vivo was studied after the addition of either AA or its breakdown products into a 5-ml cell suspension either without or with elicitor (5 µg ml–1). The concentration of H2O2 in the medium was measured at intervals (up to 3 h) by use of a xylenol orange (XO) assay (Bindschedler et al., 2001). This assay detects general hydroperoxides (Gay et al., 1999) and has been shown to give results comparable to those of the luminol method, an H2O2-specific assay, in Phaseolus vulgaris cell suspensions after elicitation (Bindschedler et al., 2001). When non-cell-mediated H2O2 production was being assayed, 5 ml of fresh nutrient medium (FM), spent medium (SM), boiled spent medium (BSM) or 1 µM CuSO4 was used instead of a cell suspension. At each time point, 100 µl of the medium (after sedimentation of the cells) or CuSO4 was sampled and immediately added to 1 ml of XO mixture (containing 125 µM XO, 100 mM D-sorbitol, 250 µM FeSO4, 250 µM (NH4)2SO4, 25 mM H2SO4). However, because AA interferes with the XO assay (probably by reducing Fe3+ back to Fe2+), all AA-containing samples were treated with AAO (1 U 100 µl–1) for 5 min before addition of the XO reagent; in control experiments, AAO did not affect H2O2 during a 5 min treatment (data not shown). All sample/XO mixtures were incubated for 40 min at room temperature before measurement of A560 against a blank prepared with 100 µl water+1 ml of XO mixture. Because of some darkening of the XO reagent mixture with time, and so that the background of the medium or CuSO4 solution could be subtracted, the A560 given by fresh nutrient medium (or 1 µM CuSO4)+XO at the beginning and at the end of the experiment was plotted against time. A correction factor obtained was subtracted from the A560 values given by samples+XO. CuSO4 was observed not to interfere with the XO assay. Experiments were performed at least in duplicate, with each treatment in triplicate. The pattern of the response was similar in each repetition, and the figures show the results of one representative experiment.
For comparison of independent experiments, a standard curve was prepared with different concentrations of H2O2 in the fresh nutrient medium or CuSO4 solution. A dilution series of H2O2 was prepared in water and a portion of this added to nutrient medium or CuSO4 immediately before addition to XO reagent.
Search for the active component in the nutrient medium responsible for induction of non-enzymic H2O2 production after DHA or DKG addition
Since commercial DHA and the DKG preparation induced rapid non-enzymic H2O2 production when added to fresh nutrient medium, the nutrient components were tested individually to indicate which one(s) induced DKG-dependent H2O2 production. Each nutrient salt was dissolved in water [at the same concentration as in medium ‘5’ (Simola and Santanen, 1990) and buffered to pH 4.5 with 10 mM tartrate (Na+). DKG (1 µmol)] was added to 1 ml of the buffered solution and incubated for 60 min at 25 °C, after which 100 µl of the solution was added to XO reagent.
Catalase treatment
Catalase (100 µg; 1100 U) in 10 µl water was added to 5 ml of cell suspension or medium at the same time as any AA, DHA, DKG, or elicitor (= time 0) when indicated.
UV absorption spectrum of the DKG preparation
The absorption spectrum of the DKG preparation (0.5 mM), without or with AAO treatment (4 U ml–1), in 50 mM succinate (Na+, pH 5.6) was scanned over the range of 190–390 nm with an Agilent 8453 spectrophotometer.
Analysis of the purity of DHA and DKG
DHA and DKG are unstable in aqueous solutions (Kimoto et al., 1993; Deutsch and Santhosh-Kumar, 1996; Nishikawa et al., 2001). For checking of the purity of these compounds, 5–200-µg portions of each compound were subjected to high-voltage electrophoresis at pH 6.5 (3.5 kV, 30 min) on Whatman 3MM paper as described by Fry (2000). The electrophoretogram was stained with silver nitrate.
Extraction of AA and DHA
For assessment of the physiological significance of the biochemical findings, AA and DHA were assayed in the apoplastic space of developing Picea wood and in the spent medium of Picea cell suspensions. Extraction of the extracellular washing fluid (EWF) was conducted according to Castillo et al. (1987) and Polle et al. (1990). The current year's twigs, with developing wood, were collected in Ruotsinkylä, South Finland, in mid-July (c. 09.30–10.00 h on a hot, sunny day) and transported to the laboratory on ice. Twigs of three individual trees were collected on two dates (one day between each) so the results are from six individuals. The bark and phloem were peeled off by hand and the secondary xylem was used for AA extraction. Twigs were cut into c. 1.0–1.5-cm lengths and briefly stored in distilled water. The material was blotted dry with paper tissue, weighed, and infiltrated for 10 min (2x5 min) with 40 mM acetate (Na+, pH 4.5) supplemented with 100 mM KCl in a water-flow vacuum. After blotting with tissue paper, the plant material was packed in 5-ml syringe barrels, which were installed into 15-ml centrifuge tubes already containing 50 µl 10% metaphosphoric acid (HPO3) to stabilize AA. EWF was eluted by centrifugation (5 min, 1000 g, 4 °C) and used for AA determination immediately.
After elution of the EWF, the remaining twigs were frozen in liquid nitrogen and stored at –80 °C until the next day when protoplasmic AA was extracted. For this, twigs were ground in liquid nitrogen in a mortar with a pestle. Four volumes of 5% HPO3 were added to the powder and the samples kept on ice for c. 20 min. The cell-wall debris was removed by centrifugation (5 min, 15 000 g, 4 °C), and the supernatants were transferred to new tubes and recentrifuged for 5 min before the AA analysis.
In assays for AA and DHA in the culture medium of Picea cells, non-buffered spent medium after 1 d in liquid culture was mixed with HPO3 (to 5%) and used for the AA determination.
Determination of AA and DHA
AA analysis was conducted with an AAO method modified from Takahama and Oniki (1992). AA has an absorbance maximum at 265 nm at neutral pH, whereas DHA has negligible absorbance in this region. The assay mixture contained (final volume 1 ml) 90 mM phosphate (Na+, pH 6.8), 38 mM succinate (Na+, in order to neutralize the sample containing HPO3), and 30 µl of extract, EWF or spent medium in 5% HPO3. After assay of the initial A265, AAO (2 U) was added to oxidize AA to DHA and A265 was remeasured. DHA was reduced back to AA by incubation with dithiothreitol (DTT, freshly prepared, to 14.8 mM; final pH 5.8) for 15 min at c. 20 °C and A265 was assayed again. With pure AA, 15 min was sufficient for complete back-reduction (data not shown). A standard curve was prepared with a dilution series of AA in 5% HPO3. From the changes in A265 value, the concentrations of AA and DHA were calculated.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Effect of ascorbate on H2O2 production
Traces of Cu2+ and other transition metals promote H2O2 production in the presence of AA (Dekker and Dickinson, 1940; Fry 1998). It was confirmed that an aerated solution of AA (250 µM) containing 1 µM CuSO4 generated H2O2 (at least 90 µM) in the absence of cells (Fig. 1). The effect of catalase verified the product as H2O2. Ten minutes' preincubation of AA [250 µM (Fig. 1) or 1 mM (data not shown)] with AAO totally abolished its effectiveness, indicating that AA itself was the active species in H2O2 generation, and that its initial oxidation product, DHA, was unable to mimic it.
|
, 
, 
etc.) = ascorbate; circles (• etc.) = dehydroascorbate; triangles (
, 
etc.) = diketogulonate; diamonds (
etc.) = exogenous H2O2; hexagons (
etc.) = no added oxidants/antioxidants.Fresh nutrient medium contains 1 µM Cu2+ and might therefore be expected to give yields of H2O2 in the presence of AA comparable to those shown in Fig. 1. When added to fresh nutrient medium, 100 µM AA rapidly induced some net production of H2O2 (Fig. 2) but much less than in pure 1 µM CuSO4 solution (Fig. 1). After the first 15 min, the H2O2 reached a steady-state concentration of 0.5–1.0 µM (Fig. 2 and data not shown). These results indicate that fresh medium contains a factor which, in the presence of AA, destroys H2O2: this destruction competed with the (AA+ Cu2+)-induced formation of H2O2. Possibilities could include the 25 µM iodide and 125 µM EDTA present in fresh medium. In spent or in boiled spent medium, AA functioned as an antioxidant, accelerating the removal of the endogenous H2O2 until it too reached a steady-state concentration of 0.5–1.0 µM (Fig. 2).
When AA (5–100 µM) was added to non-elicited spruce cell suspensions, the response was similar to that observed in spent or boiled spent medium: the steady-state extracellular H2O2 concentration routinely observed in water-treated control cultures (3.6±0.7 µM; mean ±SD of five experiments) was diminished by AA treatment (Fig. 3). This effect was probably partly due to the ability of AA to scavenge the free radicals (
–O•) formed from apoplastic phenolic compounds (–OH) by peroxidase +H2O2 (Takahama 1993a, b), thus regenerating the phenolics and constantly restoring their ability to react with any remaining H2O2. AA also promoted the disappearance of H2O2 from boiled spent medium about as effectively as in unboiled spent medium (Fig. 2). However, this was probably due to the heat-resistant peroxidase activity (c. 20% of the initial; data not shown; cf. Otter and Polle, 1994) still present after 5 min boiling. It was confirmed that coniferyl alcohol was secreted by these cells (data not shown) and that the addition of AA to cell suspensions caused exogenous H2O2 to disappear rapidly (Fig. 4). This scavenging effect was dependent on the AA concentration used (Fig. 3): at lower concentrations (5 µM AA), the H2O2 concentration returned to the value observed in water-treated control cell-suspensions, probably because 5 µM AA was fully oxidized during the 3-h experimental period, allowing H2O2 concentrations subsequently to build back up to their steady-state levels.
|
|
In the absence of AA, the Heterobasidion elicitor preparation induced a large burst of extracellular H2O2 after a brief lag period (Fig. 3). The peak occurred at 60–90 min, after which the H2O2 concentration started to decrease. In elicited cell-suspensions, 5 µM AA did not affect the timing or extent of this burst of H2O2 production (Fig. 3), probably because the large amounts of H2O2 formed were easily able to titrate out the limited amounts of AA. With 25 µM AA, however, there was evidence for an
8 min increase in the lag period (Fig. 3), comparable to the time taken for 25 µM H2O2 to be destroyed by AA in the presence of cells (Fig. 4). With 100 µM AA, the lag period for elicitation of net H2O2 production was increased to c. 30 min, possibly representing the time required to produce sufficient H2O2 to oxidize this high dose of AA; interestingly, H2O2 concentrations thereafter continued rising until at least 180 min post-elicitation (Fig. 3), suggesting that an oxidation-product of AA was able to enhance H2O2 production in vivo. The effects of two specific AA breakdown products were therefore tested: DHA and DKG.
Effect of dehydroascorbate on H2O2 production
When commercial DHA (1 mM) was incubated with 1 µM CuSO4 (Fig. 1) or with fresh nutrient medium in the absence of cells (Fig. 5), a gradual generation of H2O2 was observed. The formation of an XO-reactive product in the presence of commercial DHA was prevented by catalase, confirming the identity of H2O2 (Fig. 5). A 10-min pretreatment of the DHA with AAO abolished the subsequent non-enzymic formation of H2O2 in the presence of Cu2+ (Fig. 1). Since AAO-treated AA (i.e. in-situ prepared DHA) did not drive H2O2 production (Fig. 1), it is suggested that it was not DHA itself that induced H2O2 formation but that the commercial DHA preparation contained a small proportion of AA as an H2O2-generating impurity. The presence of an AA-like compound in the DHA preparation was also suggested by the promotion of H2O2 destruction when commercial DHA was added to spent medium (Fig. 5).
|
Nevertheless, commercial DHA (1 mM) did promote H2O2 accumulation when added to cell suspensions (Fig. 6), whereas AA had the opposite effect (Figs 2, 3). This observation indicates that a compound other than AA was responsible for the induction of H2O2 generation after DHA addition to the cells, and shows that the cells have an active role in using (commercial) DHA to produce H2O2. The latter conclusion was supported by that fact that DHA induced more H2O2 in the presence of cells (Fig. 6) than in cell-free 1 µM CuSO4 (Fig. 1) or fresh medium (Fig. 5). Boiled spent medium already contained some XO-reactive material (data not shown), whose identity as H2O2 was indicated by catalase sensitivity. Against this background of endogenous H2O2, it was not possible to detect any (additional) formation of H2O2 in boiled spent medium when supplied with commercial DHA (data not shown).
|
The process by which commercial DHA (1 mM) caused spruce cells to produce H2O2 was partially characterized. It required a short lag period (Fig. 6). Since the process did not occur in spent medium (Fig. 5), it clearly required the cells, not merely the enzymes of the spent medium. Accumulation of the XO-reactive product was blocked by catalase, but not by boiled catalase, confirming its identity as H2O2 (data not shown). NaN3 (100 µM) totally inhibited the DHA-induced increase in H2O2 concentration, and diphenylene iodonium (DPI; 5 µM) reduced it by
60–70% (Fig. 6), suggesting that the H2O2 production was enzymic, possibly catalysed by peroxidases and/or plasma membrane-bound NADPH oxidase. However, the inbitors used are unspecific, and further work is being carried out to characterize the enzymic systems involved (A Kärkönen, LK Simola, SC Fry, unpublished data). In elicited cell suspensions, commercial DHA usually had little quantitative effect on the H2O2 burst, although in some experiments (e.g. Fig. 6) the duration of the elicited peak of H2O2 accumulation was prolonged by the DHA. However, DHA did have a qualitative effect: 100 µM azide did not completely block (DHA+elicitor)-induced H2O2 accumulation (Fig. 6), unlike the situation with H2O2 production induced by elicitor alone (which was completely inhibited even by 20 µM azide; A Kärkönen, LK Simola, SC Fry, unpublished results) or by DHA alone (Fig. 6). DPI also appeared to be less inhibitory to H2O2 production induced by DHA+elicitor than to that induced by DHA alone (Fig. 6). These observations indicate that a mixture of DHA+elicitor was able to induce H2O2 production by a mechanism different from that (or those) evoked by either the elicitor or the DHA alone. The nature of the (DHA+elicitor)-induced mechanism remains unknown.
Effect of a diketogulonate preparation on H2O2 production
Addition of DKG to fresh medium evoked a rapid production of H2O2 (Fig. 5). Each component of the fresh culture medium (except Fe, sucrose, inositol, and growth regulators) was tested individually (data not shown): CuSO4 was the only one that catalysed non-enzymic H2O2 production after DKG feeding, as detected by the XO assay (Fig. 1). (The XO assay was not used in solutions containing Fe3+ since this is the species that reacts with XO; Gay et al., 1999.)
Addition of DKG to spent medium or boiled spent medium (Fig. 5), or to a live cell suspension (Fig. 7) also induced a rapid increase in H2O2 concentration. This was partly due to the fact that the DKG preparation already contained some H2O2 (XO assay; data not shown). However, the pre-existing H2O2 was strongly exceeded by that produced during the incubation with fresh or boiled spent medium. When added to fresh medium, DKG induced intense non-enzymic H2O2 formation (Fig. 5), which was clearly additional to any H2O2 already present in the DKG preparation. The identity of the H2O2 produced was verified by catalase addition (data not shown). DKG had much the same effect in boiled spent medium as in fresh medium (Fig. 5). In non-boiled spent medium (i.e. with secreted Picea enzymes still active), however, the high concentration of initially formed H2O2 gradually decreased (Fig. 5), probably because of the peroxidase activity and monolignols present in the spent medium (Kärkönen et al., 2002). The decrease was more rapid in some experiments than in others (data not shown), possibly because of variation in endogenous phenolic compounds or other H2O2 scavengers.
|
In fresh medium, the non-enzymic production of H2O2 by DKG (Fig. 5) was very rapid and had plateaued by 20 min, while in 1 µM CuSO4 it kept gradually rising almost linearly up to 90 min (Fig. 1). The plateauing is unexplained: it was probably not due to a steady state (where production=decomposition) since exogenous H2O2 was stable in the presence of fresh medium (data not shown).
The DKG preparation had a prominent peak of UV absorbance at 271–272 nm (at pH 5.6; 
780 M–1 cm–1 with respect to DKG molarity) probably attributable to an unidentified DKG degradation-product (Fig. 8). Interestingly, a 10 min pretreatment of DKG with AAO (before the addition of CuSO4) led to higher initial H2O2 concentrations than were obtained with non-AAO-treated DKG (Fig. 1). A dramatic decrease (removal of the peak) in the A271 of the DKG preparation was observed during AAO treatment (Fig. 8), probably indicating that the A271-compound was being converted to a new (non-UV-absorbing) compound that was responsible for this initial H2O2 production. The initial difference in the H2O2 concentrations formed by AAO-treated and -untreated DKG had disappeared by 90 min (Fig. 1), suggesting that the gradual non-enzymic degradation of AAO-untreated DKG had generated the same stimulatory metabolite.
|
In the case of non-elicited cell-suspensions, the immediately DKG-induced H2O2 (possibly H2O2 pre-existing in the DKG preparation) was transient, returning to control levels in 30–60 min (Fig. 7). The cell suspension's behaviour thus resembled that of spent medium (Fig. 5), suggesting that soluble extracellular materials played the major role in destroying H2O2 rather than cell-surface-located components.
When elicitor and DKG were added simultaneously, the immediately DKG-induced H2O2 was again seen; it was followed by a prolonged peak of H2O2 that remained high even when the H2O2 levels in elicitor-only controls had started to come down and those in DKG-only controls had returned to control values (Fig. 7). Catalase treatment confirmed the identity of H2O2 as the product formed in the presence of DKG+elicitor (data not shown). Thus, elicitor+DKG appeared to cause H2O2 production by a mechanism different from that/those triggered by either elicitor or DKG alone.
Active component(s) in dehydroascorbate and diketogulonate preparations inducing H2O2 production
It took at least 30 min to dissolve DHA (a 50 mM stock) in water. DHA is known to be unstable in water and non-enzymically hydrolyses to DKG, which is itself unstable (Kimoto et al., 1993; Deutsch and Santhosh-Kumar, 1996). This lability was confirmed by paper electrophoresis followed by silver staining (Fig. 9). The DHA preparation contained several breakdown products, DKG among others (paper electrophoretograms in other pHs show that it is not only DKG that co-migrates with the big spot); and the DKG preparation contained several further breakdown products including some that were not detectable in the DHA preparation, probably including the unknowns referred to by Green and Fry (2005a) as ‘C’ and ‘E’ (Fig. 9). These compounds were formed non-enzymically in aqueous solution at physiological pH, and will therefore probably also be generated from DHA and DKG in vivo. Work is underway to characterize the active compound(s) which contributed to H2O2 generation.
|
Natural occurrence of apoplastic ascorbate in Picea
For evaluation of the physiological significance of these biochemical findings, the AA and DHA content of developing Picea wood was assayed. Consistent results were obtained from six different trees. In general, the developing wood yielded c. 90 µl g–1 FW EWF, representing the xylem sap plus an eluate of the aqueous solution that permeates the cell walls. About 1.5% of the total (AA+DHA) was detected in the EWF, almost all in the oxidized form (ratio AA/(AA+DHA)=0.06). There was therefore little AA in the cell wall to interfere with lignification. The concentration of DHA in the EWF was
0.15 mM (i.e. 13.5±1.8 nmol g–1 FW tissue). In the protoplasm, by contrast, the majority of the AA+DHA was in the reduced form, the average ratio AA/(AA+DHA) being 0.77.
Spent medium of Picea cell suspensions contained no detectable AA or DHA (limit of detection 1 µM), indicating that these compounds were either not secreted or were rapidly reabsorbed by the cells or metabolized in the medium to further degradation products (e.g. those shown in Fig. 9; Green and Fry, 2005a, b).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Endogenous apoplastic ascorbate and dehydroascorbate
The concentrations of apoplastic AA+DHA observed here in lignifying spruce twigs (14.2±1.6 nmol g–1 FW tissue) are comparable with the levels previously reported in some other tissues. For example, apoplastic total AA+DHA in the current year's needles of spruce was 26 nmol g–1 FW) (Polle et al., 1990). Apoplastic fluid extracted from lignifying spruce needles contained a 5-fold lower AA+DHA concentration and a higher proportion of DHA than that from mature needles (1.17 mM and 5.26 mM with AA/(AA+DHA) ratios of 0.62 and 0.91, respectively: Otter and Polle, 1994). In Pinus hypocotyls, apoplastic AA was highest (2.8 nmol g–1 FW) up to 13 d after germination, decreasing to
1 nmol g–1 FW thereafter (Pedreira et al., 2004). Most investigators have found a much lower AA/(AA+DHA) ratio in the apoplast than in the symplast (Takahama, 1993a, 1998; Kollist et al., 2001; Veljovic-Jovanovic et al., 2001), although an exception was the report that apoplastic AA contributed about 80–90% of total apoplastic AA+DHA in young Pinus hypocotyls, decreasing to 65–70% after the growth of the hypocotyls had ceased (Sánchez et al., 1997). The present results with spruce twigs [AA/(AA+DHA)=0.06] are in this respect fairly typical. However, the possibility exists that during extraction of EWF some oxidation of AA occurred (Moldau et al., 1996; Sanmartin et al., 2003). It can be concluded that AA is secreted into the cell wall of developing xylem of spruce and a relatively high proportion of it is oxidized to DHA (as expected in a system with active lignin formation) and, in all probability, additional degradation products.
It was not possible to detect endogenous AA or DHA in the spent medium of spruce cell suspensions. Further work will be required to determine to what extent this observation is due to rapid turnover of a very small pool of AA, DHA and further degradation products. If cells of leaves, twigs, and roots secrete AA, it is likely that cultured cells also do, although such secreted AA would immediately be diluted into a large volume of culture medium making it difficult to detect.
Dehydroascorbate and other degradation products of ascorbate
Very soon after DHA has been dissolved in a neutral aqueous solution, some hydrolysis to DKG occurs (Bode et al., 1990; Kimoto et al., 1993). The appearance of a small proportion of AA (less than 1%) has also been demonstrated (Kimoto et al., 1993). It was found that DKG itself is also relatively unstable, yielding a range of incompletely characterized products. It might be suggested that since the preparations of DHA and DKG used in the present work were impure (Fig. 9), the present data do not establish the ability of these substances to produce and/or scavenge H2O2 in biological systems. However, it is likely that many or all of the products formed non-enzymically from DHA and DKG during the laboratory preparation of these substances would also be generated non-enzymically in the apoplast in vivo. Therefore, the data presented here can be assumed to provide information on the potential of AA, DHA, and DKA, and/or their respective degradation products, to contribute to the production and scavenging of H2O2 in the apoplast.
Although many of the reactions noted in the present work were non-enzymic, they were readily demonstrated at physiologically relevant pH values (4.5, typical of the apoplast in metabolically active plant cells) and room temperature. On this basis, these results do indeed support an extremely complex ability of apoplastic AA, DHA, DKG, and their further products to influence profoundly both the production and scavenging of H2O2, both in the presence of spruce cells and in cell-free culture media.
The production of H2O2 by AA, DHA and DKG at pH 4.5. was observed under certain circumstances DHA and DKG have previously been observed to accelerate the peroxidation of linoleic acid in neutral, but not in slightly acidic solutions (Takagi et al., 1988). A superoxide-scavenging agent, Tiron, suppressed linoleate peroxidation; catalase had no effect, suggesting that the superoxide radical was the active species initially generated during incubation with DHA and DKG. H2O2 generation was observed in a reaction mixture buffered to pH 4.5 even though DHA is thought to be more stable at more acidic conditions. Although H2O2 was the species detected in the present work, the possibility remains that a different reactive oxygen species, for example superoxide, was its precursor, forming H2O2 by dismutation.
Biological relevance of AA and its degradation products in the cell wall
AA's role as the major low molecular weight antioxidant in the apoplast is well established. The present results with AA are in accordance with the previous data and confirm the concentration-dependent antioxidant effect of AA in extracellular lignin-forming cell suspensions of spruce. The efficient removal of H2O2 in the culture medium by AA (Fig. 3) would interfere with peroxidase-mediated phenolic oxidation, thus reducing lignin formation, a situation similar to that when H2O2 was scavenged by KI (Nose et al., 1995; Kärkönen et al., 2002). Interestingly, the data about partial inhibition of phenolic oxidation by covalently bound cell wall peroxidases (Takahama, 1993a, b; Otter and Polle, 1994) suggest that some lignin polymerization could occur even in the presence of AA.
It was observed that further degradation products of AA can generate H2O2 in vivo (Figs 6, 7). Reported apoplastic concentrations of AA are in the millimolar range, with a higher percentage oxidation in the apoplast than in the cytoplasm. Even though DHA can be transported into the cytoplasm for re-reduction (Horemans et al., 2000), it is probable that some of it gets further degraded in the cell wall either enzymically or non-enzymically with the concomitant generation of H2O2 (Green and Fry, 2005a, b). This could have an effect, for example, during the growth of leaves or roots, since H2O2 can generate •OH radicals, which possibly have a role in cell wall loosening (Fry, 1998; Fry et al., 2001; Schopfer, 2001; Dumville and Fry, 2003; Rodríguez et al., 2002, 2004; Liszkay et al., 2004).
In addition, further DHA degradation products include a compound that inhibits peroxidase activity in vitro (A Kärkönen, SC Fry, unpublished results). The build-up of this inhibitory compound in the cell wall would decrease peroxidase-catalysed H2O2 scavenging and thus lead to the accumulation of H2O2 into the apoplast. As H2O2 is a signalling molecule mediating changes in gene expression or inducing direct physiological effects (Desikan et al., 2005) this can have a profound effect on the subsequent fate of the cells.
Future work will investigate the nature of the incompletely characterized degradation products of AA, DHA, and DKG, and their individual abilities to generate H2O2 in biological and biomimetic systems.
|
Acknowledgements |
|---|
AK thanks the Academy of Finland (Grant no. 105344 ) and Finnish Cultural Foundation for funding. We thank Professor emerita LK Simola, Professor TH Teeri, and Professor IA Kilpeläinen, University of Helsinki, for the permission to use the Picea abies callus as an experimental material. We thank Dr Pekka Maijala, University of Helsinki, for the gift of Heterobasidion parviporum.
SCF thanks the Biotechnology and Biological Sciences Research Council (BBSRC) for financial support.
|
Footnotes |
|---|
* Present address: Department of Applied Biology, PO Box 27 (Latokartanonkaari 7), FIN-00014 University of Helsinki, Finland.
|
Abbreviations |
|---|
AA, L-ascorbate anion and/or L-ascorbic acid; AAO, ascorbate oxidase; DHA, dehydro-L-ascorbate; DKG, 2,3-diketo-L-gulonate; EWF, extracellular washing fluid; XO, xylenol orange.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
In Asard H, May JM, Smirnoff N (Eds.). Vitamin C: function and biochemistry in animals and plants (2004) (Bios Scientific Publishers, London).
Ayers AR, Ebel J, Valent B, Albersheim P. (1976) Host–pathogen interactions. X. Fractionation and biological activity of an elicitor isolated from the mycelial walls of Phytophthora megasperma var. sojae. Plant Physiology 57:760–765.
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.[CrossRef][ISI]
Bode AM, Cunningham L, Rose RC. (1990) Spontaneous decay of oxidized ascorbic acid (dehydro-L-ascorbic acid) evaluated by high-pressure liquid chromatography. Clinical Chemistry 36:1807–1809.
Boerjan W, Ralph J, Baucher M. (2003) Lignin biosynthesis. Annual Review of Plant Biology 54:519–546.[CrossRef][Medline]
Brunow G, Ämmälahti E, Niemi T, Sipilä J, Simola LK, Kilpeläinen I. (1998) Labelling of a lignin from suspension cultures of Picea abies. Phytochemistry 47:1495–1500.[CrossRef]
Burkey KO, Eason G, Fiscus EL. (2003) Factors that affect leaf extracellular ascorbic acid content and redox status. Physiologia Plantarum 117:51–57.[CrossRef]
Castillo FJ, Miller PR, Greppin H. (1987) Extracellular biochemical markers of photochemical oxidant air pollution damage to Norway spruce. Experientia 43:111–115.
Chinoy JJ. (1984) The role of ascorbic acid in growth, differentiation and metabolism of plants (Nijhoff/Junk, The Hague).
Córdoba-Pedregosa M del C, Córdoba F, Villalba JM, González-Reyes JA. (2003) Differential distribution of ascorbic acid, peroxidase activity, and hydrogen peroxide along the root axis in Allium cepa L. and its possible relationship with cell growth and differentiation. Protoplasma 221:57–65.[CrossRef][ISI][Medline]
Dekker AO and Dickinson RG. (1940) Oxidation of ascorbic acid by oxygen with cupric ion as catalyst. Journal of the American Chemical Society 62:2165–2171.[CrossRef]
Desikan R, Hancock J, Neill S. (2005) Reactive oxygen species as signalling molecules. In Smirnoff N (Ed.). Antioxidants and reactive oxygen species in plants (Blackwell Publishing, Oxford) pp. 169–196.
Deutsch JC. (2000) Dehydroascorbic acid. Journal of Chromatography A 881:299–307.[CrossRef][ISI][Medline]
Deutsch JC and Santhosh-Kumar CR. (1996) Dehydroascorbic acid undergoes hydrolysis on solubilization which can be reversed with mercaptoethanol. Journal of Chromatography A 724:271–278.[CrossRef]
Dumville JC and Fry SC. (2003) Solubilization of tomato fruit pectins by ascorbate: a possible non-enzymic mechanism of fruit softening. Planta 217:951–961.[CrossRef][ISI][Medline]
Encina A and Fry SC. (2005) Oxidative coupling of a feruloyl-arabinoxylan trisaccharide (FAXX) in the walls of living maize cells requires endogenous hydrogen peroxide and is controlled by a low-Mr apoplastic inhibitor. Planta 223:77–89.[CrossRef][ISI][Medline]
Foyer CH and Halliwell B. (1976) The presence of glutathione and glutathione reductase in chloroplasts: a proposed role in ascorbic acid metabolism. Planta 133:21–25.[CrossRef]
Fry SC. (1998) Oxidative scission of plant cell wall polysaccharides by ascorbate-induced hydroxyl radicals. Biochemical Journal 332:507–515.
Fry SC. (2000) The growing plant cell wall: chemical and metabolic analysis 2nd edn. (Blackburn Press, Caldwell, New Jersey).
Fry SC. (2004) Oxidative coupling of tyrosine and ferulic acid residues, intra- and extra-protoplasmic occurrence, predominance of trimers and larger products, and possible role in inter-polymeric cross-linking. Phytochemistry Reviews 3:97–111.
Fry SC, Dumville JC, Miller JG. (2001) Fingerprinting of polysaccharides attacked by hydroxyl radicals in vitro and in the cell walls of ripening pear fruit. Biochemical Journal 357:729–735.[CrossRef][ISI][Medline]
Fry SC, Willis SC, Paterson AEJ. (2000) Intraprotoplasmic and wall-localized formation of arabinoxylan-bound diferulates and larger ferulate coupling-products in maize cell-suspension cultures. Planta 211:679–692.[CrossRef][ISI][Medline]
Gay C, Collins J, Gebicki JM. (1999) Hydroperoxide assay with the ferric–xylenol orange complex. Analytical Biochemistry 273:149–155.[CrossRef][ISI][Medline]
Green MA. (2003) Apoplastic ascorbate metabolism in rose cell-suspension cultures , PhD thesis, The University of Edinburgh, UK.
Green MA and Fry SC. (2005a) Vitamin C degradation in plant cells via enzymic hydrolysis of 4-O-oxalyl-L-threonate. Nature 433:83–87.[CrossRef][Medline]
Green MA and Fry SC. (2005b) Apoplastic degradation of ascorbate: novel enzymes and metabolites permeating the plant cell wall. Plant Biosystems 139:2–7.
Hernández JA, Ferrer MA, Jimenez A, Ros Barceló A, Sevilla F. (2001) Antioxidant systems and 
production in the apoplast of pea leaves. Its relation with salt-induced necrotic lesions in minor veins. Plant Physiology 127:817–831.
Horemans N, Asard H, Caubergs RJ. (1997) The ascorbate carrier of higher plant plasma membranes preferentially translocates the fully oxidized (dehydroascorbate) molecule. Plant Physiology 114:1247–1253.[Abstract]
Horemans N, Foyer CH, Potters G, Asard H. (2000) Ascorbate function and associated transport systems in plants. Plant Physiology and Biochemistry 38:531–540.[CrossRef]
Hu JF, Li GF, Gao ZH, Chen L, Ren HB, Jia WS. (2005) Regulation of water deficit-induced abscisic acid accumulation by apoplastic ascorbic acid in maize seedlings. Journal of Integrative Plant Biology 47:1335–1344.[CrossRef]
Jaspers P, Kollist H, Langebartels C, Kangasjärvi J. (2005) Plant responses to ozone. In Smirnoff N (Ed.). Antioxidants and reactive oxygen species in plants (Blackwell Publishing, Oxford) pp. 268–292.
Kärkönen A, Koutaniemi S, Mustonen M, Syrjänen K, Brunow G, Kilpeläinen I, Teeri TH, Simola LK. (2002) Lignification related enzymes in Picea abies suspension cultures. Physiologia Plantarum 114:343–353.[CrossRef][Medline]
Kerr EM and Fry SC. (2004) Extracellular cross-linking of xylan and xyloglucan in maize cell-suspension cultures: the role of oxidative phenolic coupling. Planta 219:73–83.[CrossRef][ISI][Medline]
Kimoto E, Tanaka H, Ohmoto T, Choami M. (1993) Analysis of the transformation products of dehydro-L-ascorbic acid by ion-pairing high-performance liquid chromatography. Analytical Biochemistry 214:38–44.[CrossRef][ISI][Medline]
Kollist H, Molday H, Oksanen E, Vapaavuori E. (2001) Ascorbate transport from the apoplast to the symplast in intact leaves. Physiologia Plantarum 113:377–383.[CrossRef][Medline]
Koutaniemi S, Toikka MM, Kärkönen A, Mustonen M, Lundell T, Simola LK, Kilpeläinen IA, Teeri TH. (2005) Characterization of basic p-coumaryl and coniferyl alcohol oxidizing peroxidases from a lignin-forming Picea abies suspension culture. Plant Molecular Biology 58:141–157.[CrossRef][ISI][Medline]
Liszkay A, van der Zalm E, Schopfer P. (2004) Production of reactive oxygen intermediates ( 
, H2O2, and •OH) by maize roots and their role in wall loosening and elongation growth. Plant Physiology 136:3114–3123.
Maijala P, Harrington TC, Raudaskoski M. (2003) A peroxidase gene family and gene trees in Heterobasidion and related genera. Mycologia 95:209–221.
Moldau H, Bichele I, Kollist H, Padu E. (1996) Enzymatic determination of ascorbic acid in leaf cell walls using acidic buffer during infiltration. Biologia Plantarum 38:229–236.
Nishikawa Y, Toyoshima Y, Kurata T. (2001) Identification of 3,4-dihydroxy-2-oxo-butanal (L-threosone) as an intermediate compound in oxidative degradation of dehydro-L-ascorbic acid and 2,3-diketo-L-gulonic acid in a deuterium oxide phosphate buffer. Bioscience, Biotechnology, and Biochemistry 65:1707–1712.[CrossRef][Medline]
Noctor G and Foyer CH. (1998) Ascorbate and glutathione, keeping active oxygen under control. Annual Review of Plant Physiology and Plant Molecular Biology 49:249–279.[CrossRef]