INTRODUCTION

Plants activate a series of defence mechanisms following challenges by potential pathogens. One of the earliest characterised events is the oxidative burst, during which reactive oxygen species such as H2O2 and O2 are rapidly generated1, 2. H2O2 production is a key response, as it may orchestrate several processes including cell wall rigidification, transcription of defence-related genes and hypersensitive (programmed) cell death3,4,5,6. The oxidative burst can also be induced by microbial elicitors such as harpin, fungal peptides and oligogalacturonic acid [OGA, (C6H11O10)n]1, 7. OGA, a well-studied elicitor, is derived from plant cell walls. It induces defence responses including the oxidative burst in various species7,8,9,10 as well as stomatal closure via the stimulation of H2O2 synthesis11.

Calcium is a ubiquitous second messenger and changes in cytosolic calcium concentration ([Ca2+]cyt) are known to mediate many signalling processes in higher plants12,13,14. Much evidence has accumulated that increases in [Ca2+]cyt are key to subsequent signalling during plant responses to challenges from pathogens and elicitors15. Expression of the aequorin gene from Aequorea aequorea in the cytoplasm of plant cells provides a means for ac-curate, non-invasive quantification of changes in [Ca2+]cyt16. When reconstituted with coelenterazine, aequorin acts as a bioluminescent indicator of [Ca2+]cyt. Because of the pioneering work of Knight et al16, aequorin technology has been widely applied in plants to report changes in [Ca2+]cyt in response to abiotic stimuli.

In this paper, the relationship between [Ca2+]cyt, the oxidative burst and defence gene expression following challenges with OGA was investigated in Arabidopsis thaliana. Changes in [Ca2+]cyt were monitored in transgenic Arabidopsis seedlings expressing the calcium reporter protein aequorin16 and inhibitors of calcium fluxes and the oxidative burst were used to manipulate various responses. Our results indicate that increases in cytosolic calcium and H2O2 are both required for OGA-induced defence gene transcription and expression.

MATERIALS AND METHODS

Arabidopsis transformation and Southern analysis

Arabidopsis thaliana plants (ecotype Landsberg erecta) were grown in pots containing a 1:1 mixture of sand and field soil in a greenhouse under a 16 h light/8 h dark cycle at 25 °C. The plasmid pMAQ 2.416, 17 was obtained from Dr Heather Knight (University of Oxford, Oxford, UK). pMAQ 2.4 contains the aequorin coding sequence under the control of the CaMV 35S promoter and OCS terminator sequences. The plasmid was mobilized to Agrobacterium tumefaciens strain LBA4404 by electroporation (Bio-Rad, Hercules, CA, USA)18. Arabidopsis plants (5 W old) were transformed with A. tumefaciens containing pMAQ2.4 via vacuum infiltration and grown in a greenhouse for the collection of seeds as described by Andrew Bent (http://www.Arabidopsis.org). These seeds were then screened on 0.8% (w/v) agar-solidified Murashige and Skoog (MS) medium supplemented with 50 μg/ml kanamycin, under a 16 h photoperiod17. Seedlings that grew normally on kanamycin for 3 W were selected as experimental material. Transformation was confirmed by Southern analysis. Genomic DNA from non-transformed and transformed seedlings was digested with different restriction enzymes, fractionated by agarose electrophoresis, blotted onto nylon membrane and hybridized with a dig-labelled aequorin gene probe. The aequorin (aeq) gene probe was labeled and hybridization detected using a Dig DNA labeling and detection kit (CAT NO. 1093657, Roche Diagnostics GmbH, Mannheim, Germany) according to the manufacturer's instructions.

Chemicals and elicitor

Diphenyleneiodonium (DPI), EGTA, LaCl3, pyridine, quinacrine, imidazole and Ruthenium Red were obtained from Sigma Chemical Co. (USA). They were made up in DMSO as stock solutions. An equivalent amount of DMSO solvent was used as control treatments. Inhibitors were used at concentrations of the same order of magnitude as those reported in Tang and Smith19 and Orozaco-Car-denas et al5. OGA was purified from acid hydrolysate of citrus pectin (Sigma Chemical Co, USA) as described by Nothnagel et al20 and Legendre et al7. It was subsequently dialyzed against distilled water using dialysis tubing with a cut-off of approximately 500 D. The prepared OGA was adjusted to a final concentration of 1 mg/ml of galacturonic acid equivalents (Gal equiv) with distilled water, as determined by the method of Blumenkrantz and Asboe-Hansen21. In inhibitor experiments, seedlings were pre-treated with different inhibitors for 30 min, followed by OGA treatment and then harvested at the indicated times for H2O2 and [Ca2+]cyt assay.

Measurement of H2O2

Seedlings were incubated in Tris-HCl buffer (10 mM, pH 6.0) with OGA at various concentrations, or pre-treated for 30 min in Tris-HCl buffer containing pharmacological agents at various concentrations, and then OGA was added to give a final concentration of 40 μg Gal equiv/ml. At various times seedlings were harvested by flash-freezing in liquid nitrogen and stored at −70°C until subsequent extraction. Frozen leaves (0.2 g) were ground to powder under liquid nitrogen and homogenized with 1 ml of 0.2 M HClO4 at 4°C. The extract was held on ice for 5 min and then centrifuged at 10,000 g for 10 min at 4°C. The supernatant was collected and either processed immediately or quick-frozen at −70°C until further analysis. The concentration of H2O2 was measured as described by Chen et al22.

Measurement of [Ca2+]cyt

Changes in [Ca2+]cyt were measured non-invasively by luminometry of intact seedlings of transgenic Arabidopsis expressing aequorin. Reconstitution of aequorin was performed in vivo essentially as described in Knight et al17 by floating seedlings on water containing 2.5 μM coelenterazine in the dark overnight at 21°C. Experiments were performed by putting a single seedling in a plastic luminometer cuvette containing 0.2 ml Tris-HCl buffer (10 mM, pH 6.0), and then placing the cuvette in a luminometer (FG-200, Shanghai, China). After injecting with the appropriate volume of OGA stock solution (1 mg/ml), luminescence counts were recorded immediately and calibrated against Ca2+concentrations as described by Blume et al23.

Northern-blot analysis

Total RNA was extracted and subjected to Northern analysis as described in Desikan et al24 and Orozaco-Cardenas et al5. Total RNA (10 μg) extracted from seedlings was fractionated by electrophoresis on 1.4% agarose gels with formaldehyde and blotted onto nitrocellulose membranes. Prehybridization and hybridization were carried out as recommended in the Dig DNA labeling and detection kit (Roche Diagnostics GmbH, Mannheim, Germany). The probes encoding chalcone synthase (CHS, No. 177N23T7), glutathione S-transferase (GST, No. 103G1T7), phenylalanine ammonia-lyase (PAL, No. 82B7T7) and actin (ACT, No. 172E20T7) were all obtained from the Arabidopsis Biological Resource Center (ABRC; Columbus, OH 43210 USA). The probe for pathogen related protein 1 (PR-1) was obtained from Professor W. Broekaert in Belgium. ACT probe was used as the loading control. Hybridisation was detected as per the manufacturer's recommendations. For inhibitor experiments, seedlings were incubated in liquid MS medium for 6 h with shaking (50 rpm) in the growth cabinet, with or without pre-treatment of inhibitors prior to RNA extraction.

SDS-PAGE and Western-blot analysis

Arabidopsis seedlings (2 g wet weight) were ground in a mortar and pestle before adding 2 ml of extraction buffer (125 mM Tris-Cl, 7.75% [w/v] SDS, 10% [v/v] mercaptoethanol, pH 7.0) and grinding further. The mixture was transferred to a centrifuge tube, raised to room temperature, and centrifuged in a swing-out rotor at 3,500 g for 10 min. The supernatant containing 30 μg total protein was mixed with 6×(v/v) sample buffer (0.1 M Tris-Cl, 12% [w/v] SDS, 9% [v/v] glycerol, 60% [v/v] mercaptoethanol, pH 6.8) and was separated by SDS-PAGE. After electrophoresis, the separated proteins were transferred to a Hybond-C Extra nitrocellulose membrane (Amersham Pharmacia Biotech, Sydney) using a Multiphor II semi-dry blotting apparatus (Amersham Pharmacia Biotech, Sydney) according to the manufacturer's instructions. Hybrids were detected using the BM Chemiluminescence Blotting Substrate POD system (Roche). Rabbit antibodies against Arabidopsis thaliana PR-1(1:1000), CHS(1:1000), GST(1:1500) and PAL(1:500) were used.

RESULTS

Arabidopsis transformation

F1 seeds from aequorin-transformed plants were plated on MS + kanamycin agar medium and after 3 weeks kanamycin-tolerant seedlings were selected. Southern analysis indicated that the plants were sucessfully transformed, and contained 3-6 copies of the aequorin gene (Fig 1A). Northern blot analysis demonstrated that the inserted aequorin gene was expressed normally (Fig 1B). To confirm the presence of the active aequorin protein, transgenic seedlings were exposed to 100 mM calcium and aequorin luminescence was determined as described in Materials and Methods. A sharp and transient increase in luminescence induced by extracellular calcium was observed only on the transformed seedlings (Fig 2).

Figure 1
figure 1figure 1

Transformation of Arabidopsis with aequorin. (A) Genomic DNA was extracted from non-transformed (C) and transformed seedlings and 30 μg genomic DNA were digested by various restriction enzymes and fractionated by agarose electrophoresis. The gel was then blotted onto nylon membrane and the membrane hybridized with a dig-labelled aequorin gene probe. (B) Total RNA (20 μg) from non-transformed (N.T.) and transformed (T.) seedlings was subjected to Northern analysis using a dig-labelled aequorin gene probe. aeq, aequorin transcript.

Figure 2
figure 2

Extracellular calcium triggers a transient increase in luminescence only in Arabidopsis seedlings expressing recombinant aequorin protein. Seedlings transformed with aequorin (T.) or non-transformed (N.T.) were exposed to 100 mM Ca2+and luminescence determined using a luminometer.

Increases in cytosolic calcium induced by OGA are modulated by inhibitors of calcium fluxes but not by inhibitors of the oxidative burst

OGA at 40 μg Gal equiv/ml induced a rapid and substantial transient increase in [Ca2+]cyt that peaked within ca. 15 sec (Fig 3A). This response to OGA was dose-dependent: an OGA concentration as low as 10 μg Gal equiv/ml induced a detectable increase in cytosolic calcium, which was saturated at OGA concentrations above 50 μg Gal equiv/ml (Fig 3B). Pre-treatment with the calcium chelator EGTA or the calcium channel blockers LaCl3 or Ruthenium Red (Fig 3C) suppressed substantially the OGA-induced increase in [Ca2+]cyt. However, pre-treatment with the NADPH oxidase inhibitors DPI and pyridine (Fig 3C) had no effect on the transient increase in [Ca2+]cyt caused by OGA treatment.

Figure 3
figure 3figure 3figure 3

Effects of inhibitors of Ca2+flux and NADPH oxidase on OGA-induced transient increases in [Ca2+]cyt in Arabidopsis seedlings. (A) Increase in [Ca2+]cyt induced by OGA at 40 μg Gal equiv/ml. OGA, oligogalacturonic acid treatment; Con, control. Experiments were repeated at least six times. (B) Dose-response for maximum increase in [Ca2+]cyt against OGA concentration. (C) The effects of calcium flux inhibitors (L, 10 μM LaCl3; R, 100 μM Ruthenium Red; E, 1mM EGTA) and NADPH oxidase inhibitors (D, 10 μM DPI; P, 10 mM pyridine) on the OGA- (O, 40 μg Gal equiv/ml) induced increase in [Ca2+]cyt. C, Control. Seedlings were pre-incubated for 30 min prior to OGA challenge. Data in B and C represents the means ± SE from six experiments and was analysed by one-way ANOVA followed by Tukey's test. Different symbols indicate significant differences between treatments (p < 0.05).

OGA-induced oxidative burst is blocked by inhibitors of NADPH oxidase

OGA-induced H2O2 accumulation occurred in a bi-phasic fashion (Fig 4A). The initial increase by 40 μg Gal equiv/ml OGA peaked at ca. 5 min. The H2O2 content then declined again to control values. A second, much larger increase in H2O2 content, which peaked at ca. 60 min and declined thereafter, was observed (Fig 4A). Similar to the effect on [Ca2+]cyt, the response of H2O2 production to OGA was also dose-dependent and saturated at 50 μg Gal equiv/ml (Fig 4B). OGA-induced H2O2 production was effectively eliminated by pre-treatment with the NADPH oxidase inhibitors quinacrine, DPI, pyridine and imidazole (Fig 4C). Pre-treatment with the calcium channel blockers LaCl3, Ruthenium Red and, to a much lesser extent, with the calcium chelator EGTA, also inhibited the OGA-induced oxidative burst (Fig 4C).

Figure 4
figure 4figure 4figure 4

Effects of inhibitors of calcium flux and NADPH oxidase on the OGA-induced oxidative burst in Arabidopsis seedlings (A) Time-course of H2O2 accumulation induced by 40 μg Gal equiv/ml OGA. (B) Dose-response (60 min). (C) The effects of calcium flux inhibitors (L, 10 μM LaCl3; R, 100 μM Ruthenium Red; E, 1mM EGTA) and NADPH oxidase inhibitors (I, 10 mM imidazole; Q, 100 μM quinacrine; D, 10 μM DPI; P, 10 mM pyridine) on OGA- (O, 40 μg Gal equiv/ml) induced increases in H2O2 accumulation (60 min). C, Control. Seedlings were pre-incubated for 30 min prior to OGA challenge. Data represents the means ± SE from six experiments and was analysed by one-way ANOVA followed by Tukey's test. Different symbols indicate significant differences between treatments (p < 0.05).

Induction of defence gene transcription by OGA

OGA treatment for 6 h at 40 μg Gal equiv/ml increased the transcription of CHS, PR-1, GST and PAL, compared to the control. Treatment with the NADPH oxidase inhibitors DPI and pyridine inhibited the OGA-induced increases in the transcription of these genes. Treatment with the calcium flux inhibitors EGTA, LaCl3 and Ruthenium Red also reduced substantially the induction of gene expression by OGA compared to OGA treatment alone (Fig 5).

Figure 5
figure 5figure 5

Suppression of OGA-induced defence gene transcription of PR-1, CHS, GST and PAL by various inhibitors. Seedlings were pre-treated in buffer or various inhibitors (L, 10 μM LaCl3; R, 100 μM Ruthenium Red; E, 1 μM EGTA; D, 10 μM DPI; P, 10 mM pyridine) for 30 min prior to exposure to OGA (40 μg Gal equiv/ml). (A), RNA was then extracted after 6 h of OGA treatment (40 μg Gal equiv/ml) and the resulting blots hybridized to dig-labelled PR-1, CHS, GST or PAL gene probes. Blots were subsequently stripped and re-probed with an actin (ACT) probe to confirm equal RNA loading. (B) shows the relative abundance of gene transcription from (A). The maximum RNA amount was taken as 1.0.

Accumulation of defence gene products induced by OGA

Similar to its effects on gene transcription, OGA treatment for 36 h at 40 μg Gal equiv/ml also induced the accumulation of PR-1, CHS, PAL and GST proteins. DPI, pyridine, EGTA, LaCl3 and Ruthenium Red pretreatments for 30 min inhibited the OGA-induced accumulation of PR-1, CHS, PAL and GST (Fig 6).

Figure 6
figure 6figure 6

Accumulation of defence gene products induced by OGA. (A)Total protein was extracted after 36 h of OGA treatment (40 μg Gal equiv/ml) and Western blotting was performed with anti-PR-1, anti-CHS, anti-GST, and anti-PAL antibodies. C, control; O, OGA treatment; O+L, pretreated with LaCl3 for 30 min following with OGA; O+R, pretreated with RR for 30 min following with OGA; O+E, pretreated with EGTA following with OGA; O+D, pretreated with DPI for 30 min following with OGA; O+P, pretreated with pyridine following with OGA. (B) shows the relative abundance of protein accumulation from (A). The maximum protein amount was taken as 1.0.

DISCUSSION

Plants transformed with a cDNA encoding the Ca2+-sensitive luminescent protein aequorin provide a simple, non-invasive means of measuring [Ca2+]cyt in whole plants. Many new signals initiating rapid changes in [Ca2+]cyt have subsequently been detected with this technology, including the mechanical signals of touch and wind, salt/drought, heat shock, and osmotic stress. Such technology has also been used to study pathogen- or pathogen-derived elicitor -induced calcium signalling, but the effects of endogenous elicitors (such as OGA, derived from the plant cell wall) have not been analysed. Thus in order to determine the effects of the endogenous elicitor OGA on calcium fluxes and the oxidative burst in Arabidopsis, transgenic seedlings expressing the calcium reporter protein aequorin were generated. Southern and Northern analyses (Fig 1A, B), coupled to the induction of luminescence by extracellular calcium (Fig 2), demonstrated that these seedlings were suitable for such studies.

OGA induces the oxidative burst in tobacco, tomato and Commelina communis7, 11, 25. In the present study, OGA not only strongly induced H2O2 production, it also induced a substantial increase in [Ca2+]cyt. Furthermore, OGA-stimulated increases in [Ca2+]cyt and H2O2 production were rapid and saturable, suggesting the existence of a receptor mediating OGA effects (Fig 3A, B; Fig 4A, B). There is considerable evidence that the elicitor- and pathogen-induced oxidative bursts in Arabidopsis are mediated by Atrboh proteins, homologues of the gp91 sub-unit of mammalian NADPH oxidase26. Several Atrboh genes have been cloned from Arabidopsis27, 28, 29. Elicitor- or pathogen-induced oxidative bursts are inhibited by known inhibitors of NADPH oxidase4, 24 and H2O2 accumulation is also much reduced in Atrboh knock-out mutants 26, although the possibility that alternative H2O2 -generating systems operate in Arabidopsis must be borne in mind30. Nevertheless, in the present study, known inhibitors of plasma membrane NADPH oxidase (DPI or pyridine) did eliminate the OGA-induced oxidative burst in Arabidopsis seedlings (Fig 4C), suggesting that plasma membrane NADPH oxidase was responsible for the OGA-induced oxidative burst. In addition, the OGA-stimulated rapid and substantial increase of cytsolic calcium was inhibited by EGTA (a calcium chelator), LaCl3 (a plasma membrane calcium channel blocker) and Ruthenium Red (an intracellular membrane calcium channel blocker), but not by DPI or pyridine (inhibitors of the oxidative burst) (Fig 3C). These data indicate that the oxidative burst is not required for the increases in [Ca2+]cyt. Since both Ruthenium Red and LaCl3 inhibited OGA-induced increase in cytosolic calcium, it is likely that calcium fluxes from both extracellular and intracellular pools are required for calcium increases. EGTA, which chelates extracellular calcium ions, also blocked OGA-induced increased [Ca2+]cyt and H2O2 production, but the effect was not as substantial as those of LaCl3 or Ruthenium Red. Cessna and Low10 reported that extracellular calcium was not essential for OGA-stimulated [Ca2+]cyt increases in Nicotiana tabacum suspension cultures. On the other hand, Desikan et al31 reported that the calcium ionophore ionomycin stimulated, and EGTA inhibited, the oxidative burst in Arabidopsis suspension cultures induced by the bacterial elicitor harpin32. It may be that different calcium pools are utilized in different species or in response to different stimuli, a suggestion previously made by Cessna and Low10. However, whatever the source of calcium for the increase in [Ca2+]cyt is, this increase in cytosolic calcium is necessary for the subsequent oxidative burst. This is evident from the observation that inhibitors of calcium fluxes reduced both the oxidative burst and calcium increase, whereas the oxidative burst inhibitors completely suppressed H2O2 accumulation but had no effect on [Ca2+]cyt. Grant et al (2000) also provided evidence that increased [Ca2+]cyt precedes and is required for the oxidative burst in Arabidopsis plants 15. The downstream processes mediating calcium responses have yet to be clearly resolved, but are likely to involve post-translational modifications of enzymes such as protein kinases and protein phosphatases, and even the NADPH oxidase itself28.

OGA has previously been reported to induce defence responses in soybean, tomato and tobacco cells7, 11, 25. Cytosolic calcium and H2O2 are both known to modulate defence gene expression following exposure of plant cells to various stresses6, 14. The data reported here show that OGA similarly induces defence responses in seedlings of Arabidopsis thaliana, including the transcription of PR-1, CHS, GST, PAL genes and accumulation of PR-1, CHS, PAL and GST proteins. Induction of these genes and proteins is inhibited by treatments which reduces the increases in [Ca2+]cyt and the oxidative burst, indicating that both calcium and H2O2 were required for OGA-activated defence gene expression.