Cross-talk between calcium and reactive oxygen species originated from NADPH oxidase in abscisic acid-induced antioxidant defence in leaves of maize seedlings

Authors

  • M. JIANG,

    1. College of Life Science, Nanjing Agriculture University, Nanjing, People's Republic of China and
    2. Department of Biology, Hong Kong Baptist University, Kowloon Tong, Hong Kong, People's Republic of China
    Search for more papers by this author
  • J. ZHANG

    Corresponding author
    1. Department of Biology, Hong Kong Baptist University, Kowloon Tong, Hong Kong, People's Republic of China
    Search for more papers by this author

J. Zhang. Fax: 852 3411 5995; e-mail: jzhang@hkbu.edu.hk

ABSTRACT

The signal interactions between calcium (Ca2+) and reactive oxygen species (ROS) originated from plasma membrane NADPH oxidase in abscisic acid (ABA)-induced antioxidant defence were investigated in leaves of maize (Zea mays L.) seedlings. Treatment with ABA led to significant increases in the activity of plasma membrane NADPH oxidase, the production of leaf O2, and the activities of several antioxidant enzymes such as superoxide dismutase (SOD), catalase (CAT), ascorbate peroxidase (APX) and glutathione reductase (GR). However, such increases were blocked by the pretreatment with Ca2+ chelator EGTA or Ca2+ channel blockers La3+ and verapamil, and NADPH oxidase inhibitors such as diphenylene iodonium (DPI), imidazole and pyridine. Treatment with Ca2+ also significantly induced the increases in NADPH oxidase activity, O2 production and the activities of antioxidant enzymes, and the increases were arrested by pretreatment with the NADPH oxidase inhibitors. Treatment with oxidative stress induced by paraquat, which generates O2, led to the induction of antioxidant defence enzymes, and the up-regulation was suppressed by the pretreatment of Ca2+ chelator and Ca2+ channel blockers. Our data suggest that a cross-talk between Ca2+ and ROS originated from plasma membrane-bound NADPH oxidase is involved in the ABA signal transduction pathway leading to the induction of antioxidant enzyme activity, and Ca2+ functions upstream as well as downstream of ROS production in the signal transduction event in plants.

Abbreviations
ABA

abscisic acid

APX

ascorbate peroxidase

ASC

ascorbate

CAT

catalase

DPI

diphenylene iodonium

EDTA

ethylenediaminetetraacetic acid

EGTA

ethylene glycol-bis(α-amino ethyl ether)-N,N,N′,N′-tetraacetic acid

GR

glutathione reductase

NBT

nitro blue tetrazolium

PMSF

phenylmethylsulfonyl fluoride

PVP

polyvinylpyrrolidone

ROS

reactive oxygen species

SOD

superoxide dismutase

XTT

sodium,3′-[1-[phenylamino-carbonyl]-3,4-tetrazolium]-bis(4-methoxy-6-nitro) benzenesulfonic acid hydrate.

INTRODUCTION

Plants respond and adapt to environmental stresses by altering their cellular metabolism and invoking various defence mechanisms. One important regulator of plant responses to abiotic stresses is the phytohormone abscisic acid (ABA). ABA is involved in abiotic stresses such as drought, low temperature, high temperature, and salinity, and plays important roles in the regulation of plant responses to these stressed conditions (Shinozaki & Yamaguchi-Shinozaki 1997; Xiong et al. 2001; Larkindale & Knight 2002).

Three lines of evidence indicate that one mode of ABA action is related to the oxidative stress in plant cells. First, ABA causes an increased generation of reactive oxygen species (ROS) such as O2 and H2O2 (Guan, Zhao & Scandalios 2000; Pei et al. 2000; Jiang & Zhang 2001; Murata et al. 2001; Zhang et al. 2001). Second, ABA induces the expression of antioxidant genes and enhances the capacity of antioxidant defence systems, including enzymatic and non-enzymatic constituents (Bueno et al. 1998; Guan & Scandalios 1998a, b; Bellaire et al. 2000; Guan et al. 2000; Jiang & Zhang 2001, 2002a). Third, treatment with higher concentrations of ABA results in cellular oxidative damage (Bueno et al. 1998; Jiang & Zhang 2001). ROS plays an important intermediary role in the ABA signal transduction pathway leading to the induction of antioxidant defence systems (Guan et al. 2000; Jiang & Zhang 2002b).

The source of ROS induced by ABA is gaining attention. Plasma membrane-bound NADPH oxidase, which transfers electrons from cytoplasmic NADPH to O2 to form O2, followed by dismutation of O2 to H2O2, may be involved in ABA signal transduction in plant cells. In ABA-induced stomatal closing of guard cells of Arabidopsis, the requirement of cytosolic NADPH, ABA-induced ROS production, and the inhibitory effects of diphenylene iodonium (DPI), a well-known inhibitor of neutrophil NADPH oxidase, suggest that NADPH oxidase contributes to early ABA signal transduction (Pei et al. 2000; Murata et al. 2001). Our recent study, using two-phase fractionated plasma membrane extracts and several widely used neutrophil NADPH oxidase inhibitors, such as DPI, imidazole and pyridine, has demonstrated that NADPH oxidase is involved in ABA-induced ROS production and ABA-enhanced antioxidant defence systems (Jiang & Zhang 2002b). In addition to NADPH oxidase, the light reaction in chloroplasts also contributes to ABA-induced production of ROS (Zhang et al. 2001).

Ca2+ has also been shown to be involved in ABA signal transduction in plant cells. ABA stimulates the increases in cytosolic Ca2+ by inducing both Ca2+ influx from the extracellular space and Ca2+ release from intracellular stores (Pei et al. 2000; Murata et al. 2001). In ABA-induced stomatal closing of Arabidopsis, ABA-induced H2O2 production and the H2O2-activated Ca2+ channels are important components of ABA signal transduction (Pei et al. 2000; Murata et al. 2001). These observations indicate that Ca2+ is in the downstream of ROS production in ABA signalling. However, a striking feature of the plant NADPH oxidase homologues is the presence of two Ca2+-binding EF hand motifs, suggesting Ca2+ may play an important role in the regulation of NADPH oxidase activity (Keller et al. 1998; Grant & Loake 2000; Sagi & Fluhr 2001). Ca2+ may regulate NADPH oxidase activity by activating the gp91phox subunit of NADPH oxidase directly, or indirectly via phosphorylation, following the Ca2+-mediated activation of a specific Ca2+-dependent protein kinase, or activating the production of NADPH via NAD kinase regulated by calmodulin (Grant & Loake 2000; Sagi & Fluhr 2001; Neill et al. 2002a). These observations imply that Ca2+ is a signal that functions in the upstream of ROS in plant cells. However, whether there exists such a mode in ABA-induced signal transduction is unclear. Furthermore, no single study has investigated the inter-relationship between Ca2+ and ROS originated from NADPH oxidase in ABA-induced antioxidant defence in plants.

In the present study, we investigated the effects of treatments with ABA, Ca2+ and oxidative stress induced by paraquat (PQ), which generates O2 (Alscher, Erturk & Heath 2002) on plasma membrane NADPH oxidase activity, leaf O2 production, and the activities of several antioxidant enzymes such as superoxide dismutase (SOD), catalase (CAT), ascorbate peroxidase (APX) and glutathione reductase (GR) in leaves of maize seedlings. Meanwhile, we also investigated the effects of pretreatments with the Ca2+ chelator EGTA, which chelates extracellular Ca2+ (Gong et al. 1998b; Blume et al. 2000), and the plasma membrane Ca2+ channel blockers La3+ and verapamil, which inhibit the influx of extracellular Ca2+ into cells across plasma membrane (Gong, Li & Chen 1998a; Pei et al. 2000), and several NADPH oxidase inhibitors such as DPI, which binds to the flavoprotein component of the NADPH oxidase complex, imidazole and pyridine, which bind to the b-type cytochrome component (Auh & Murphy 1995; Van Gestelen, Asard & Caubergs 1997; Orozco-Cárdenas, Narváez-Vásquez & Ryan 2001), on ABA- or Ca2+- or paraquat-induced NADPH oxidase activity, O2 production, and the activities of antioxidant enzymes. The objective of the present study was to elucidate the inter-relationship between Ca2+ and ROS that originated from NADPH oxidase in ABA signal transduction leading to the up-regulation of antioxidant enzyme activity in plants.

MATERIALS AND METHODS

Plant material and treatments

Seeds of maize (Zea mays L., cv. Yidan 22; from Xiangtan, China) were sown in trays of sands in a greenhouse at a temperature of 22–28 °C, photosynthetic active radiation (PAR) of 400 µmol m−2 s−1 (enhanced with high-pressure sodium lamps) and photoperiod of 14/10 h (day/night), and watered daily. When the second leaf was fully expanded, the plants were collected and used for all investigations.

The plants were excised at the base of the stem, rinsed with distilled water, and the cut ends of stems were placed in beakers wrapped with aluminium foil containing 250 mL distilled water, EGTA (1, 5, 10 mm), La3+ (5 mm), verapamil (1 mm), DPI (10, 50, 100 µm), imidazole (20 mm), and pyridine (20 mm) solutions for 4 h, respectively, and then exposed to ABA (100 µm) or Ca2+ (10 mm) or paraquat (1 µm) treatment for 8 h at 25 °C with a continuous light intensity of 200 µmol m−2 s−1 (PAR). The plants treated with distilled water under the same conditions during the whole period served as controls for the above. After treatment of detached maize plants, the second leaves were sampled and immediately frozen under liquid nitrogen, and then stored at −80 °C until analysis, except for those needed for O2 determination and isolation of plasma membranes, which were conducted immediately after sampling.

Isolation of plasma membrane

Leaf plasma membranes were isolated using the two-phase aqueous polymer partition system (Larsson, Widell & Kjellbom 1987). Samples were homogenized in four volumes of the extraction buffer (50 mm Tris-HCl, pH 7.5, 0.25 m sucrose, 1 mm ASC, 1 mm EDTA, 0.6% PVP, and 1 mm PMSF). The homogenate was filtered through four layers of cheesecloth, and the resulting filtrate was centrifuged at 10 000 g for 15 min. Microsomal membranes were pelleted from the supernatant by centrifugation at 50 000 g for 30 min. The pellet was suspended in 0.33 m sucrose, 3 mm KCl, and 5 mm potassium phosphate, pH 7.8. The plasma membrane fraction was isolated by adding the microsomal suspension to an aqueous two-phase polymer system to give a final composition of 6.2% (w/w) Dextran T500, 6.2% (w/w) PEG 3350, 0.33 m sucrose, 3 mm KCl, and 5 mm potassium phosphate, pH 7.8. Three successive rounds of partitioning yielded the final upper phase. The upper phase produced was diluted five-fold in Tris-HCl dilution buffer (10 mm, pH 7.4) containing 0.25 m sucrose, 1 mm EDTA, 1 mm dithiothreitol, 1 mm ASC and 1 mm PMSF. The fractions were centrifuged at 120 000 g for 30 min. The pellets were then re-suspended in Tris-HCl dilution buffer and used immediately for further analysis. All procedures were carried out at 4 °C. Protein content of plasma membranes was determined according to the method of Bradford (1976) with BSA as standard.

In order to assess the purity of the plasma membrane preparation, the activities of marker enzymes such as vanadate-sensitive ATPase (plasma membrane), cytochrome c oxidase (mitochondrial membrane), and NADPH-cytochrome c reductase (endoplasmic reticulum) were assayed as described by Briskin, Leonard & Hodges (1987). The data (Table 1) indicate that the final membrane fraction was highly enriched in plasma membranes and depleted in intracellular membranes.

Table 1.  Specific activities (nmol min−1  mg−1 protein) of marker enzymes in the microsomal (MF) and plasma membrane (U3) fractions from leaves of maize seedlings
Marker enzymeMFU3U3/MF
  1. Data are means ±SE (n = 3).

Vanadate-sensitive ATPase 28.6 ± 1.3223.7 ± 11.67.82
Cytochrome c oxidase299.4 ± 9.6 44.0 ± 3.10.15
NADPH-cytochrome c reductase 11.3 ± 0.9  3.2 ± 0.20.28

Determination of NADPH oxidase activity of plasma membranes

The NADPH-dependent O2-generating activity in isolated plasma membrane vesicles was determined by following the reduction of XTT by O2 (Sagi & Fluhr 2001). The assay mixture of 1 mL contained 50 mm Tris-HCl buffer (pH 7.5), 0.5 mm XTT, 100 µm NADPH and 15–20 µg of membrane proteins. The reaction was initiated with the addition of NADPH, and XTT reduction was determined at 470 nm. Corrections were made for background production in the presence of 50 units SOD. Rates of O2 generation were calculated using an extinction coefficient of 2.16 × 104 m−1 cm−1.

Determination of leaf O2 production

Production of O2 was measured as described by Able, Guest & Sutherland (1998) by monitoring the reduction of XTT in the presence of O2, with some modifications. Leaves (1 g) were homogenized with 5 mL of 50 mm Tris-HCl buffer (pH 7.5) and centrifuged at 5000 g for 10 min. The reaction mixture (1 mL) contained 50 mm Tris-HCl buffer (pH 7.5), 50 µg leaf supernatant proteins and 0.5 mm XTT. The reduction of XTT was determined at 470 nm for 5 min. Corrections were made for the background absorbance in the presence of 50 units SOD. The production rate of O2 was calculated using an extinction coefficient of 2.16 × 104 m−1 cm−1.

Enzyme assays

Frozen leaf segments (0.5 g) were homogenized in 10 mL of 50 mm potassium phosphate buffer (pH 7.0) containing 1 mm EDTA and 1% PVP, with the addition of 1 mm ASC in the case of the APX assay. The homogenate was centrifuged at 15 000 g for 20 min at 4 °C and the supernatant was immediately used for the following enzyme assays.

Total SOD activity was assayed by monitoring the inhibition of photochemical reduction of NBT according to the method of Giannopolitis & Ries (1977). One unit of SOD activity was defined as the amount of enzyme that required to cause 50% inhibition of the reduction of NBT as monitored at 560 nm. CAT activity was assayed by measuring the rate of decomposition of H2O2 at 240 nm, as described by Aebi (1984). APX activity was measured by monitoring the decrease in absorbance at 290 nm as ASC was oxidized, as described by Nakano & Asada (1981). GR activity was measured by following the change in A340 as oxidized glutathione (GSSG)-dependent oxidation of NADPH, according to the method of Schaedle & Bassham (1977).

Statistical analysis

The results presented were the mean of six replicates. Means were compared by one-way analysis of variance and Duncan's multiple range test at 5% level of significance.

RESULTS

Involvement of Ca2+ and NADPH oxidase in ABA-induced antioxidant enzyme activity

The ABA treatment led to significant increases in the activity of plasma membrane NADPH oxidase (Fig. 1a), the production rate of leaf O2 (Fig. 1b), and the activities of antioxidant enzymes, SOD (Fig. 2a), CAT (Fig. 2b), APX (Fig. 2c), and GR (Fig. 2d), in leaves of maize seedlings. The activities of NADPH oxidase, SOD, CAT, APX and GR increased by 60, 24, 32, 32 and 25%, respectively, and the production rate of leaf O2 increased by 42%, when compared with the control values. Pretreatments with the Ca2+ chelator EGTA and the Ca2+ channel blockers La3+ and verapamil, and the NADPH oxidase inhibitors DPI, imidazole and pyridine substantially reduced the increases induced by ABA treatment in the activity of NADPH oxidase (Fig. 1a), the production rate of leaf O2 (Fig. 1b), and the activities of these antioxidant enzymes (Fig. 2a–d) in leaves of maize seedlings, and the inhibition of EGTA and DPI on these parameters exhibited a dose-dependent manner.

Figure 1.

Effects of pretreatments with the Ca2+ chelator EGTA and the Ca2+ channel blockers La3+ and verapamil, and the NADPH oxidase inhibitors DPI, imidazole and pyridine on the activity of NADPH oxidase (a) and the production rate of leaf O2(b) in the leaves of maize plants exposed to ABA treatment. Plants were excised at the base of the stems and were pretreated with distilled water, EGTA (1, 5, 10 mm), La3+ (La, 5 mm), verapamil (Ver, 1 mm), DPI (10, 50, 100 µm), imidazole (IM, 20 mm), and pyridine (PY, 20 mm) for 4 h, respectively, and then exposed to ABA (100 µm) treatment for 8 h. The plants treated with distilled water under the same conditions during the whole period served as controls for the above. Values are means ± SE (n = 6). Means denoted by the same letter did not significantly differ at P < 0.05 according to Duncan's multiple range test.

Figure 2.

Effects of pretreatments with the Ca2+ chelator EGTA and the Ca2+ channel blockers La3+ and verapamil, and the NADPH oxidase inhibitors DPI, imidazole and pyridine on the activities of SOD (a), CAT (b), APX (c), and GR (d) in the leaves of maize plants exposed to ABA treatment. Plant treatments and statistical analyses are the same as for Fig. 1.

Involvement of NADPH oxidase in Ca2+-induced antioxidant enzyme activity

In order to determine further whether ABA-induced increases in the activity of NADPH oxidase, ROS production and the activities of antioxidant enzymes are related to the action of Ca2+, we investigated the effects of different concentrations of Ca2+ on these parameters in leaves of maize seedlings. Treatments with 4–10 mm Ca2+ led to significant increases in the activity of NADPH oxidase (Fig. 3a) and the production rate of leaf O2 (Fig. 3b) in a dose-dependent manner. Treatment with 10 mm Ca2+ increased the activity of NADPH oxidase and the production rate of leaf O2 by 86 and 102%, respectively, compared with the control values. In vitro experiments showed that the addition of different concentrations of Ca2+ (0.5–2 mm) in the assay reaction medium for NADPH oxidase activity significantly enhanced the activity of NADPH oxidase in a dose-dependent pattern (Fig. 4a). The higher concentration of Ca2+ (2.5 mm) led to a slight reduction in the activity of NADPH oxidase, compared with that of 2 mm Ca2+. The induction by Ca2+ was fully blocked by the addition of the Ca2+ chelator EGTA (Fig. 4a). However, the addition of ABA (20–100 µm) in the assay reaction medium did not affect the activity of NADPH oxidase (Fig. 4b). As with the activity of NADPH oxidase and the production of O2, treatments with different concentrations of Ca2+ (4–10 mm) also significantly enhanced the activities of antioxidant enzymes, SOD (Fig. 5a), CAT (Fig. 5b), APX (Fig. 5c) and GR (Fig. 5d) in a dose-dependent manner. Treatment with 10 mm Ca2+ increased the activities of SOD, CAT, APX and GR by 28, 50, 41 and 30%, respectively, compared to the control values. Treatment with 2 mm Ca2+ did not affect the activity of NADPH oxidase (Fig. 3a) and the production rate of leaf O2 (Fig. 3b), and also did not affect the activities of antioxidant enzymes (Fig. 5a–d) in leaves of maize seedlings.

Figure 3.

Effects of different concentrations of Ca2+ treatments on the activity of NADPH oxidase (a) and the production rate of leaf O2 (b) in the leaves of maize plants. Plants were excised at the base of the stems and were treated with different concentrations of Ca2+ (2–10 mm) for 8 h. The plants treated with distilled water for 8 h served as the control. Statistical analyses are the same as for Fig. 1.

Figure 4.

Effects of Ca2+ (a) and ABA (b) at different concentrations added in the assay reaction medium on the activity of NADPH oxidase in the control leaves of maize plants. Ca2+, Ca2++ EGTA (10 mm), and ABA were added 5 min before the reduction of XTT was initiated. Statistical analyses are the same as for Fig. 1.

Figure 5.

Effects of different concentrations of Ca2+ treatments on the activities of SOD (a), CAT (b), APX (c), and GR (d) in the leaves of maize plants. Plants were excised at the base of the stems and were treated with different concentrations of Ca2+ (2–10 mm) for 8 h. The plants treated with distilled water for 8 h served as the control. Statistical analyses are the same as for Fig. 1.

In order to confirm the Ca2+-induced increases in the activities of antioxidant enzymes are related to the activity of NADPH oxidase and the production of ROS, we investigated the effects of pretreatment with NADPH oxidase inhibitors such as DPI, imidazole and pyridine on these parameters in leaves of maize seedlings. Pretreatments with the NADPH oxidase inhibitors almost fully blocked the Ca2+-induced increases in the activity of NADPH oxidase (Fig. 6a) and the production of leaf O2 (Fig. 6b), and reduced the enhancement induced by Ca2+ treatment in the activities of SOD (Fig. 7a), CAT (Fig. 7b), APX (Fig. 7c) and GR (Fig. 7d) in leaves of maize seedlings.

Figure 6.

Effects of pretreatments with the NADPH oxidase inhibitors DPI, imidazole and pyridine on the activity of NADPH oxidase (a) and the production rate of leaf O2 (b) in the leaves of maize plants exposed to Ca2+ treatment. Plants were excised at the base of the stems and were pretreated with distilled water, DPI (100 µm), imidazole (IM, 20 mm), and pyridine (PY, 20 mm) for 4 h, respectively, and then exposed to Ca2+ (10 mm) treatment for 8 h. The plants treated with distilled water under the same conditions during the whole period served as controls for the above. Statistical analyses are the same as for Fig. 1.

Figure 7.

Effects of pretreatments with the NADPH oxidase inhibitors DPI, imidazole and pyridine on the activities of SOD (a), CAT (b), APX (c), and GR (d) in the leaves of maize plants exposed to Ca2+ treatment. Plant treatments are as described in Fig. 6. Statistical analyses are the same as for Fig. 1.

Involvement of Ca2+ in oxidative stress-induced antioxidant enzyme activity

In order to determine whether ABA-induced increases in the activities of antioxidant enzymes results from the increased generation of ROS, we investigated the effects of oxidative stress induced by paraquat, which generates O2 and its decomposition product H2O2, on the activities of antioxidant enzymes. Meanwhile, the effects of pretreatments with the Ca2+ chelator EGTA and Ca2+ channel blockers La3+ and verapamil on the paraquat-induced activities of antioxidant enzymes were also investigated so as to test whether Ca2+ is involved in ROS-induced antioxidant enzyme activity in leaves of maize seedlings. Figure 8 shows that oxidative stress led to significant increases in the activities of SOD (Fig. 8a), CAT (Fig. 8b), APX (Fig. 8c) and GR (Fig. 8d) in leaves of maize seedlings. Treatment with 1 µm paraquat enhanced the activities of SOD, CAT, APX and GR by 26, 42, 41 and 28%, respectively, compared with the control values. Pretreatments with the Ca2+ chelator EGTA and Ca2+ channel blockers La3+ and verapamil completely blocked paraquat-induced increases in the activities of antioxidant enzymes in leaves of maize seedlings.

Figure 8.

Effects of pretreatments with the Ca2+ chelator EGTA, and the Ca2+ channel blockers La3+ and verapamil on the activities of SOD (a), CAT (b), APX (c), and GR (d) in the leaves of maize plants exposed to paraquat treatment. Plants were excised at the base of the stems and were pretreated with distilled water, EGTA (10 mm), La3+ (La, 5 mm), verapamil (Ver, 1 mm) for 4 h, respectively, and then exposed to paraquat (PQ, 1 µm) treatment for 8 h. The plants treated with distilled water under the same conditions during the whole period served as controls for the above. Statistical analyses are the same as for Fig. 1.

DISCUSSION

It has been documented that ABA can induce increases in cytosolic Ca2+ (Allen et al. 2000; Pei et al. 2000; Murata et al. 2001) and ROS production (Guan et al. 2000; Pei et al. 2000; Jiang & Zhang 2001, 2002a, b; Murata et al. 2001; Zhang et al. 2001), and enhance the capacity of antioxidant defence systems in plants (Guan & Scandalios 1998a, b; Bellaire et al. 2000; Guan et al. 2000; Jiang & Zhang 2001, 2002a, b). Although it has been shown that ROS plays an important intermediary role in ABA-induced antioxidant defence (Guan et al. 2000; Jiang & Zhang 2002b), the inter-relationship between Ca2+ and ROS in ABA-induced antioxidant defence is unclear. In Arabidopsis, ABA-induced ROS production triggers the influx of Ca2+ and the increase in cytosolic Ca2+, which induces stomatal closing (Pei et al. 2000; Murata et al. 2001). However, in some biotic and abiotic stresses, these stresses trigger a Ca2+ influx, and the increased cytosolic Ca2+ stimulates the production of ROS, which induces the physiological response (Chen & Li 2001; Yang & Poovaiah 2002). Moreover, Ca2+ has a signal function upstream as well as downstream of ROS in plant responses to pathogens (Lamb & Dixon 1997; Bowler & Fluhr 2000).

In the present study, three lines of evidence indicate that a cross-talk between Ca2+ and ROS is involved in ABA-induced increases in the activities of antioxidant enzymes in leaves of maize seedlings. First, the increases induced by ABA treatment in the ROS production and the activities of antioxidant enzymes SOD, CAT, APX and GR were substantially reduced by the pretreatments with the Ca2+ chelator EGTA and the Ca2+ channel blockers La3+ and verapamil, and the NADPH oxidase inhibitors DPI, imidazole and pyridine, respectively (Figs 1 & 2). Second, Ca2+ treatment induced increases in the production of ROS and the activities of antioxidant enzymes in a dose-dependent manner (Figs 3 & 5), and the increases were almost completely blocked by the pretreatments with the NADPH oxidase inhibitors (Figs 6 & 7). Third, the increases induced by oxidative stress in the activities of antioxidant enzymes were fully arrested by the pretreatments with the Ca2+ chelator and the Ca2+ channel blockers (Fig. 8). Although the absolute specificity of each inhibitor used in this study can always be questioned, the similar results obtained with the different types of inhibitors of NADPH oxidase and calcium, together with the dose effect of EGTA and DPI inhibition, undoubtedly provide unequivocal evidence for the involvement of both Ca2+ and ROS in the ABA signalling. Our data suggest that Ca2+ acts in the upstream as well as downstream of ROS production in the ABA signal transduction pathway leading to the induction of the antioxidant defence system, and the signal interactions between Ca2+ and ROS play a pivotal role in ABA-induced antioxidant defence in plants.

ROS is inevitably produced in higher plant cells during normal metabolism. Biotic and abiotic stresses often lead to an increased generation of ROS. The possible sources of ROS generation include plasma membrane NADPH oxidase, cell wall peroxidases, amine oxidase, oxalate oxidase, flavin-containing oxidases, and the mitochondria, chloroplasts and peroxisomes (Grant & Loake 2000; Neill et al. 2002a; Vranová, Inzé & Breusegem 2002). Our recent study has demonstrated that NADPH oxidase is involved in ABA-induced ROS production and ABA-enhanced antioxidant defence systems, including enzymatic and non-enzymatic constituents (Jiang & Zhang 2002b). In this paper, our data indicate further that the communication between Ca2+ and NADPH oxidase may play a central role in ABA-enhanced antioxidant defence in plants. Ca2+ not only induced an increase in the activity of NADPH oxidase in plants (Fig. 3), but in vitro the addition of Ca2+ in the assay reaction medium also significantly enhanced the activity of NADPH oxidase (Fig. 4), indicating a direct regulation by Ca2+. Other studies have also confirmed that plant NADPH oxidase, unlike mammalian gp91phox, can produce O2 in the absence of additional cytosolic components (Sagi & Fluhr 2001; Simon-Plas, Elmayan & Blein 2002; Torres, Dangl & Jones 2002), and can be regulated directly by Ca2+ (Sagi & Fluhr 2001). The pretreatments with the Ca2+ chelator EGTA and the Ca2+ channel blockers La3+ and verapamil almost completely suppressed the ABA-induced increases in the activity of NADPH oxidase, the production of ROS, and the activities of antioxidant enzymes (Figs 1 & 2), suggesting that Ca2+ plays a pivotal role in the regulation of NADPH oxidase activity and antioxidant enzyme activity in the ABA signal transduction.

The mechanism that Ca2+ regulates antioxidant defence is still open. It has been shown that Ca2+ binds to calmodulin (CaM), a ubiquitous calcium-binding protein, and the Ca2+/CaM complex stimulate the activities of antioxidant enzymes such as CAT (Yang & Poovaiah 2002) and SOD (Gong & Li 1995). However, an increase in cytosolic Ca2+ mediated by H2O2 also brings about a reduction in the activity of SOD in tobacco (Price et al. 1994). In the present study, our results showed that Ca2+-induced increases in the activity of NADPH oxidase, the production of ROS, and the activities of antioxidant enzymes were almost fully blocked by the pretreatments with the NADPH oxidase inhibitors DPI, imidazole and pyridine (Figs 6 & 7), suggesting that Ca2+-stimulated ROS production, which originates mainly from NADPH oxidase, contributes to the induction of antioxidant enzyme activity in plant cells. On the other hand, Ca2+ overload, which causes toxic levels of ROS production and results in cellular oxidative damage (Chen & Li 2001), may be a causative factor of the reduction in the activity of antioxidant enzyme activity.

Plasma membrane NADPH oxidase is thought to use cytosolic NADPH to reduce O2 at the apoplastic membrane face (Lamb & Dixon 1997; Grant & Loake 2000; Neill, Desikan & Hancock 2002b; Pastori & Foyer 2002). H2O2 produced by the dismutation of O2 may be transported from the apoplast to the cytosol through water channels (aquaporins; Neill et al. 2002b; Pastori & Foyer 2002). H2O2 generated in chloroplasts, mitochondria and peroxisomes can also move into cytosol (Neill et al. 2002b; Shigeoka et al. 2002). The cytosolic H2O2 directly triggers local signal transduction events, and then induces gene expression. Abundant evidence has been shown that ROS, especially H2O2, are involved in the cellular signalling process as secondary messengers to induce a number of genes and proteins involved in stress defences, including SOD, CAT, APX, GR, glutathione peroxidase, guaiacol peroxidase, glutathione-S-transferase and pathogenesis–related (PR) protein (Levine et al. 1994; Prasad et al. 1994; Lamb & Dixon 1997; Karpinski et al. 1999; Morita et al. 1999; Desikan et al. 2001; Neill et al. 2002a; Vranováet al. 2002). However, the mechanisms by which this process occurs remain to be elucidated. In the present study, the data showed that the increases in the activities of antioxidant enzymes induced by oxidative stress by paraquat, which binds to the thylakoid membrane of the chloroplasts and transfers the electrons to O2 in a chain reaction that causes continuous formation of O2 (Alscher et al. 2002), were fully blocked by the pretreatments with the Ca2+ chelator and the Ca2+ channel blockers, indicating that Ca2+ is a stringent requirement for the ROS-induced antioxidant enzyme activity. However, neither signalling pathway(s) nor transcription factors and promoter elements specific for the redox regulation have been identified in plants to date (Neill et al. 2002a; Vranováet al. 2002). Furthermore, no calcium-dependent protein kinases have been shown are regulated by ROS (Neill et al. 2002a). Therefore, we cannot give an exact explanation for the requirement of Ca2+ in the oxidative stress-induced antioxidant defence. Further investigations are needed on this matter.

In conclusion, our data indicate that both Ca2+ and ROS originated from plasma membrane NADPH oxidase are involved in ABA signal transduction, and a cross-talk between Ca2+ and ROS plays a pivotal role in ABA-induced antioxidant defencce. ABA triggers a Ca2+ influx, and the increased cytosolic Ca2+ stimulates the production of ROS by plasma membrane-bound NADPH oxidase, which results in the induction of antioxidant enzyme activity mediated by cytosolic Ca2+. Our results suggest the existence of intracellular networks rather than linear pathways in ABA signal transduction leading to the induction of antioxidant defence in plants.

ACKNOWLEDGMENTS

We are grateful for grants obtained from the Faculty Research Grants of Hong Kong Baptist University, the Research Grants Council of Hong Kong University Grants Council and Area of Excellence for Plant and Fungal Biotechnology.

Received 29 November 2002; received in revised form 17 December 2002; accepted for publication 19 December 2002

Ancillary