Nitrate Efflux Is an Essential Component of the Cryptogein Signaling Pathway Leading to Defense Responses and Hypersensitive Cell Death in Tobacco

Abstract

There is much interest in the transduction pathways by which avirulent pathogens or derived elicitors activate plant defense responses. However, little is known about anion channel functions in this process. The aim of this study was to reveal the contribution of anion channels in the defense response triggered in tobacco by the elicitor cryptogein. Cryptogein induced a fast nitrate (NO₃⁻) efflux that was sensitive to anion channel blockers and regulated by phosphorylation events and Ca²⁺ influx. Using a pharmacological approach, we provide evidence that NO₃⁻ efflux acts upstream of the cryptogein-induced oxidative burst and a 40-kD protein kinase whose activation seems to be controlled by the duration and intensity of anion efflux. Moreover, NO₃⁻ efflux inhibitors reduced and delayed the hypersensitive cell death triggered by cryptogein in tobacco plants. This was accompanied by a delay or a complete suppression of the induction of several defense-related genes, including hsr203J, a gene whose expression is correlated strongly with programmed cell death in plants. Our results indicate that anion channels are involved intimately in mediating defense responses and hypersensitive cell death.

INTRODUCTION

Inoculation of resistant plants with potential pathogenic microorganisms triggers complex and integrated defense mechanisms, including active oxygen species (AOS) production, reinforcement of the cell wall, and transcriptional activation of defense-related genes that encode phytoalexin biosynthetic enzymes and pathogenesis-related proteins (Fritig et al., 1998). In many cases, these defense responses are manifested by a rapid, localized cell death termed the hypersensitive response (HR), which results in the formation of necrotic lesions around the infection sites (Hammond-Kosack and Jones, 1996). The HR is believed to help plant defense by limiting the spread of microorganisms and may represent a form of programmed cell death (PCD) (Lam et al., 2001). In addition, the distal uninfected parts of the plants usually develop systemic acquired resistance (SAR), which leads to a broad range of resistance against diverse pathogens (Ryals et al., 1996). Both the HR and SAR are regulated by a complex network of signaling molecules, including AOS, salicylic acid, nitric oxide, and jasmonic acid (Enyedi et al., 1992; Alvarez et al., 1998; Wendehenne et al., 2001).

Plant defense responses are initiated by the direct or indirect recognition of microorganism-derived molecules called elicitors (Ebel and Cosio, 1994). Purified elicitors usually mimic avirulent pathogen attacks when applied to a given plant; thus, they provide an excellent tool to elucidate the mechanisms of plant defense. Cryptogein is a 10-kD proteinaceous elicitor purified from the culture filtrates of the oomycete Phytophthora cryptogea, an avirulent pathogen of tobacco (Ricci, 1997).

Cryptogein belongs to a family of homologous proteinaceous elicitors, termed elicitins, that are secreted by all Phytophthora species analyzed to date (Ricci, 1997). Upon application to tobacco, cryptogein triggers a HR-like response, elicits the accumulation of defense-related genes, and induces acquired resistance to the black shank–causing agent Phytophthora parasitica var nicotianae (Keller et al., 1996a). Salicylic acid was shown to be essential for cryptogein-induced SAR but not for the HR-like necrosis response (Keller et al., 1996b).

By studying the effects of cryptogein on tobacco cell suspensions, it has been possible to characterize early events implicated as transduction components in the elicitor induction of defense responses. These include cryptogein-specific binding to high-affinity sites in the plasma membrane (Wendehenne et al., 1995; Bourque et al., 1999), protein phosphorylation (Viard et al., 1994; Lecourieux-Ouaked et al., 2000), Ca²⁺ influx (Tavernier et al., 1995), K⁺ efflux (Blein et al., 1991), activation of a plasma membrane NADPH oxidase responsible for AOS production, cytosol acidification, and, at least in part, extracellular medium alkalinization (Pugin et al., 1997), activation of the pentose phosphate pathway (Pugin et al., 1997), mitogen-activated protein kinase (MAPK) activation (Lebrun-Garcia et al., 1998, 2002), disruption of the microtubular cytoskeleton (Binet et al., 2001), nitric oxide production (Foissner et al., 2000), and induction of defense-related genes (Suty et al., 1995).

Although the exact sequence and relationships between these events are not understood fully, we identified protein phosphorylation followed by Ca²⁺ influx as the earliest steps (Tavernier et al., 1995). These steps also appear to be required for cryptogein-induced late reactions, including phytoalexin synthesis (Tavernier et al., 1995) and cell death (Binet et al., 2001). Interestingly, cell death triggered by cryptogein (but also by other elicitins) is regulated independently of the oxidative burst (Dorey et al., 1999; Rustérucci et al., 1999; Binet et al., 2001).

Current evidence supports the notion that plasma membrane anion channels are essential components of early signal transduction processes in plants. Numerous stimuli, including abscisic acid (Ward et al., 1995), auxin (Zimmermann et al., 1994; Thomine et al., 1997), blue light (Cho and Spalding, 1996), and abiotic stresses (Lewis et al., 1997; Cazalé et al., 1998), rapidly activate plasma membrane anion channels, which, because of the outward-directed anion gradients across the plasma membrane, drive passive effluxes from the cytoplasm into the extracellular space. A combination of pharmacological and biophysical methods indicates that one function of these anion channels might be to initiate or amplify plasma membrane depolarization, which in turn may activate Ca²⁺ voltage-dependent channels and/or K⁺ channels (Ward et al., 1995).

Several lines of evidence suggest the involvement of similar electrical signaling processes in plant cells challenged by avirulent pathogens. Plasma membrane depolarization and Cl⁻ efflux are among the earliest signaling events detectable in elicitor-treated parsley and tobacco cells (Nürnberger et al., 1994; Pugin et al., 1997; Zimmermann et al., 1998). Moreover, anion channel antagonists have been shown to interfere with early and late elicitor- or pathogen-induced responses such as Ca²⁺ influx (Ebel et al., 1995), AOS production (Jabs et al., 1997; Rajasekhar et al., 1999), MAPK activation (Ligterink et al., 1997), phytoalexin synthesis (Ebel et al., 1995; Jabs et al., 1997), and HR development (Levine et al., 1996). Collectively, these studies emphasize the important role of anion channels in plant defense against pathogens. Consistent with these studies, Lacomme and Roby (1999) recently reported the identification of an early HR-induced cDNA that encodes a protein showing similarities to mitochondria voltage-dependent gated anion channels, a family of channels involved in the release of cytochrome c during apoptosis in mammals.

Previously, we demonstrated that application of cryptogein to tobacco cells induces plasma membrane depolarization and Cl⁻ efflux, both of which occur after a similar lag period of ∼5 min (Pugin et al., 1997). This result led us to hypothesize that plasma membrane anion channel activation could be a key step in the cryptogein-induced signaling cascade. To address this hypothesis, in the present study, we first analyzed the effects of different classes of anion channel antagonists on cryptogein-induced anion efflux. Then, we undertook a detailed biochemical and molecular analysis of the relationships between cryptogein-induced anion channel activation and other elicitor-mediated events, in particular Ca²⁺ influx, kinase activation, AOS production, defense-related gene activation, and hypersensitive cell death.

Our results support a model in which cryptogein-induced anion channel activation constitutes an early branch point in the signaling pathway leading to the activation of inducible defense responses. These findings clearly complement an emerging body of evidence indicating that membrane channel–related proteins are involved intimately in mediating HR cell death.

RESULTS

Cryptogein Induces a Nitrate Efflux Inhibited by Various Anion Channel Antagonists

Changes in external nitrate (NO₃⁻) concentration were measured in cryptogein-treated tobacco cell suspensions. Indeed, high NO₃⁻ permeability seems a general feature of plant plasma membrane anion channels (Schmidt and Schroeder, 1994; Frachisse et al., 1999, 2000; Barbier-Brygoo et al., 2000). In contrast to control cells, cryptogein triggered a significant NO₃⁻ efflux that occurred after a lag period of 5 min (Figure 1A). After 60 min, the NO₃⁻ extracellular content reached a maximum, corresponding to 4.8 ± 0.15 μmol/g fresh cells.

Figure 1.

Cryptogein-Induced NO₃⁻ Efflux.

(A) Time course of NO₃⁻ efflux induced by 25 nM cryptogein. After cryptogein addition to tobacco cell suspensions, external medium aliquots were taken at the times indicated. NO₃⁻ content was determined using a colorimetric assay. Open circles indicate control cells, and closed circles indicate cells treated with 25 nM cryptogein. Results are from one of five representative experiments. Values shown are means ± se of triplicate assays. FW, fresh weight.

(B) Time course of variations in intracellular NO₃⁻ concentrations during cryptogein treatment. At the indicated period of cryptogein (25 nM) treatment, cells were harvested and analyzed for NO₃⁻ concentrations using a colorimetric assay. White bars indicate control cells, and black bars indicate cells treated with 25 nM cryptogein. Results are from one of three representative experiments. Values shown are means ± se of triplicate assays.

(C) Effect of increasing concentrations of cryptogein on NO₃⁻ efflux. The external concentration was determined after a 90-min treatment with cryptogein as described in (A). Values were obtained by subtraction of the NO₃⁻ content in the external medium of control cells from the NO₃⁻ content in the external medium of cryptogein-treated cells. Results are from one of three representative experiments. Values shown are means ± se of duplicate assays.

Because the cellular NO₃⁻ concentration was 7.1 ± 0.3 mM, the cryptogein-triggered efflux of NO₃⁻ represented a loss of internal NO₃⁻ content of ∼60% (Figure 1B). NO₃⁻ efflux increased with increasing concentrations of cryptogein up to 25 nM, but higher elicitor concentrations did not result in further significant NO₃⁻ efflux (Figure 1C). The elicitor concentration required to induce 50% NO₃⁻ efflux was ∼4 nM. This result is well correlated with previous findings showing that cryptogein binds specifically to tobacco plasma membrane binding sites, with a K_d value of 2 nM (Wendehenne et al., 1995; Bourque et al., 1999).

To determine whether cryptogein-induced NO₃⁻ efflux was mediated by anion channels, a series of structurally unrelated anion channel antagonists were tested. Table 1 summarizes the effects of these compounds on NO₃⁻ efflux measured 30 min after the addition of cryptogein. Niflumic acid was the most potent inhibitor, 50% inhibition of NO₃⁻ efflux being obtained with 38 μM. Potent inhibition also was achieved by the application of ethacrynic acid and glibenclamide, a sulfonylurea known to block ATP binding cassette (ABC) proteins in animal cells. Anthracene-9-carboxylic acid (A9-C) and diphenylamine-2-carboxylic acid (DPC), another ABC protein inhibitor, also displayed an inhibitory effect, but only with concentrations >300 μM.

Table 1.

Effects of Anion Channel Blockers on Cryptogein-Induced NO₃⁻ Efflux, AOS Production, and Alkalinization of the Extracellular Medium

	Inhibitor Concentration Required to Inhibit 50% of the Maximal Response (μM)			Inhibitor Concentration Required for Complete Inhibition of NO3− Efflux, AOS Production, and Extracellular Alkalinization (μM)
Inhibitor	NO3− Efflux	AOS Production	Extracellular Alkalinization
A9-C	350	30% inhibition   at 400 μM	20% inhibition   at 400 μM	ND*
DPC	340	375	410	ND*
Ethacrynic acid	120	105	135	300
Glibenclamide	86	70	91	200
Glibenclamide + KCOs	79	88	93	200
IAA-94	–a	–	–	–
Niflumic acid	38	57	75	200
NPPBb	53	65	58	200
DIDS	ND	–	–	ND
SITS	ND	–	–	ND

All of the inhibitors were added to the suspension cells 10 min before cryptogein (25 nM). The inhibitor concentrations required to inhibit NO₃⁻ efflux by 50% were measured 30 min after cryptogein addition, and those required to inhibit AOS production and extracellular alkalinization by 50% were measured 20 min after cryptogein addition. The KCOs diazoxide or minoxidil sulfate (10 to 2500 μM) were applied to the cell suspensions concomitantly with glibenclamide. ND, not determined; ND

^*, not determined because of insolubility of the inhibitor at concentrations >600 to 800 μM.

^a–, no effect.

^bToxic effects. NPPB alone induced significant cell death.

By contrast, cryptogein-induced NO₃⁻ efflux was insensitive to indanyloxyacetic acid–94 (IAA-94). Phytotoxic effects were observed at concentrations >20 μM with the anion channel antagonists 5-nitro-2,3-phenylpropyl aminobenzoic acid (NPPB), and no discrimination between the general toxicity of the antagonist and the selective effect on NO₃⁻ efflux could be determined. In the same manner, no conclusion could be drawn with the two stilbene derivatives 4,4′-diisothiocyanato-stilbene-2-2′-disulfonic acid (DIDS) and 4-acetamido-4′-isothiocyanato-stilbene-2-2′-disulfonic acid (SITS), because at concentrations as low as 5 μM, both compounds had a dramatic inhibitory effect on the NO₃⁻ reductase–catalyzed reaction used to measure NO₃⁻ concentration (data not shown).

When used at 200 μM, niflumic acid induced almost complete inhibition of cryptogein-induced NO₃⁻ efflux for at least 75 min (Figure 2, Table 1). With longer times, an attenuation of the inhibitory effect was observed, and after 4 h of treatment, niflumic acid was not effective. The addition of niflumic acid (or another active NO₃⁻ efflux antagonist) at concentrations ranging from 100 to 500 μM during the experiment did not restore the inhibition (data not shown). Similar observations were made when experiments were performed with glibenclamide and ethacrynic acid (data not shown), indicating that the inhibition escape was not specific to niflumic acid but rather was a common mechanism.

Figure 2.

Niflumic Acid Is Transiently Effective in Inhibiting Cryptogein-Induced NO₃⁻ Efflux.

Niflumic acid was added to tobacco cell suspensions 10 min before cryptogein. External medium aliquots were taken at the times indicated. NO₃⁻ content was determined using a colorimetric assay. Open circles indicate control cells; open squares indicate control cells treated with 200 μM niflumic acid; closed circles indicate cells treated with 25 nM cryptogein; and closed squares indicate cells treated with 25 nM cryptogein and 200 μM niflumic acid. The data are representative of five experiments. FW, fresh weight.

In animal cells, the inhibitory effect of sulfonylureas on ABC proteins can be reversed by K⁺ channel openers (KCOs) such as diazoxide or minoxidil sulfate (Sheppard and Welsh, 1992). These competitors act through their binding to ABC proteins. As seen in Table 1, concomitant application of diazoxide or minoxidil sulfate with glibenclamide did not reverse the glibenclamide-induced inhibition of the cryptogein-mediated NO₃⁻ efflux, even when millimolar concentrations of KCOs were used in the assays (Table 1). Moreover, when added alone at concentrations ranging from 20 μM to 1 mM, KCOs did not induce any NO₃⁻ efflux (data not shown). Together, these results suggest that the large NO₃⁻ efflux observed in response to cryptogein is mediated by anion channels sensitive to structurally unrelated inhibitors, including sulfonylurea, but is insensitive to KCOs.

Phosphorylation and Calcium Dependence

To assess the involvement of protein kinases and phosphatases in transmitting the cryptogein signal leading to NO₃⁻ efflux, staurosporine, a general protein kinase inhibitor, and calyculin A, a Ser/Thr phosphatase type 1 and 2A inhibitor, were used. These compounds have been shown previously to block or to mimic cryptogein-induced events, respectively (Tavernier et al., 1995; Lecourieux-Ouaked et al., 2000). The addition of 5 μM staurosporine to tobacco cell suspensions suppressed NO₃⁻ efflux (Figure 3A). Moreover, when added in the middle of the response, staurosporine rapidly stopped NO₃⁻ efflux.

Figure 3.

Cryptogein-Induced NO₃⁻ Efflux Acts Downstream of Protein Phosphorylation and Calcium Influx.

(A) Effects of staurosporine. Staurosporine (5 μM) was added to the cell suspensions 5 min before 25 nM cryptogein or at 30 min (arrow). At the indicated times, external medium aliquots were taken, and their NO₃⁻ contents were determined using a colorimetric assay. The data are representative of three experiments.

(B) Effects of calyculin A. Calyculin A (500 nM) and cryptogein (2.5 or 25 nM) were added alone or together at time 0. NO₃⁻ content was measured as in (A). The data are representative of three experiments.

(C) Inhibition of cryptogein-induced NO₃⁻ efflux by a Ca²⁺ surrogate or chelator. Lanthanum (500 μM) was added to the cell suspensions 5 min before 25 nM cryptogein or at 30 min (arrow). EGTA (2 mM) was added 5 min before cryptogein. NO₃⁻ content was measured as in (A). The data are representative of three experiments.

cal, calyculin A; cry, cryptogein; FW, fresh weight; stauro, staurosporine.

In cells treated with calyculin A, a rapid NO₃⁻ efflux occurred within 5 min and attained a maximal value after 15 min (Figure 3B). Calyculin A–mediated NO₃⁻ efflux was approximately twofold lower compared with that in cells treated with a concentration of cryptogein that allows maximal NO₃⁻ efflux (25 nM). When 25 nM cryptogein and calyculin A were used together, NO₃⁻ efflux corresponded to that triggered by 25 nM cryptogein alone. However, the addition of calyculin A together with cryptogein at a lower concentration (2.5 nM) led to an additive NO₃⁻ efflux comparable to those obtained with 25 nM cryptogein alone.

To investigate the possible influence of Ca²⁺ on cryptogein-mediated NO₃⁻ efflux, tobacco suspension cells were treated with the elicitor in the presence of 500 μM lanthanum or 2 mM EGTA. These compounds have been shown repeatedly to block the entrance of calcium and subsequent Ca²⁺-dependent events in cryptogein-incubated tobacco cells (Tavernier et al., 1995; Pugin et al., 1997; Lebrun-Garcia et al., 1998; Binet et al., 2001). As shown in Figure 3C, lanthanum and EGTA dramatically reduced elicitor-induced NO₃⁻ efflux. Moreover, lanthanum had an inhibitory effect when added during cryptogein treatment (at 30 min). By contrast, niflumic acid, ethacrynic acid, glibenclamide, DPC, and A9-C had no inhibitory effects on cryptogein-induced Ca²⁺ entrance into tobacco cells (data not shown). Together, these results indicate that the elicitor-mediated NO₃⁻ efflux acts downstream from Ca²⁺ influx and phosphorylation events.

Cryptogein-Induced NO₃⁻ Efflux Determines AOS Production and Extracellular Alkalinization

AOS production and extracellular medium alkalinization are two early cryptogein-induced events. To determine a possible link between these events and the cryptogein-mediated NO₃⁻ efflux, we examined the effects of anion channel inhibitors on the oxidative burst and external medium alkalinization. As shown in Table 1, DPC, ethacrynic acid, glibenclamide, and niflumic acid inhibited both AOS production and extracellular medium alkalinization with efficiencies comparable to those for NO₃⁻ efflux inhibition: the most potent inhibition was obtained with niflumic acid, followed by glibenclamide, ethacrynic acid, and DPC.

As reported for NO₃⁻ efflux, complete inhibition was obtained only with niflumic acid (200 μM), glibenclamide (200 μM), and ethacrynic acid (300 μM). When applied at concentrations >300 μM, A9-C also reduced AOS production and extracellular medium alkalinization by 30 and 20%, respectively. By contrast, as noted above for NO₃⁻ efflux, AOS production and extracellular medium alkalinization were insensitive to IAA-94. Moreover, neither DIDS nor SITS showed inhibitory effects, even when used at 500 μM. Thus, with the exception of DIDS and SITS, which required further investigation, it is apparent that each inhibitor tested had the same efficiency in inhibiting cryptogein-induced NO₃⁻ efflux, AOS production, and external medium alkalinization.

These data suggest that the NO₃⁻ efflux acted upstream of the pathway(s) leading to the oxidative burst and to extracellular alkalinization. On the other hand, diphenylene iodonium (DPI), a potent inhibitor of neutrophil NADPH oxidase shown previously to inhibit AOS production in cryptogein-treated cells (Simon Plas et al., 1997), did not prevent the activation of elicitor-induced NO₃⁻ efflux (data not shown), confirming that anion efflux was not triggered by the oxidative burst.

A 40-kD Cryptogein-Activated Protein Kinase Is Dependent on Anionic Channel

The treatment of tobacco cell suspensions with cryptogein was shown previously to induce the rapid activation of MAPK homologs with apparent molecular masses of 50 and 46 kD; these were identified recently as salicylic acid–induced protein kinase (SIPK) and wounding-induced protein kinase (WIPK), respectively (Lebrun-Garcia et al., 1998, 2002; Zhang et al., 2000). To determine whether the activation of both kinases would depend on cryptogein-induced NO₃⁻ fluxes, protein extracts from cells cotreated with 25 nM cryptogein and niflumic acid at the concentration required for the complete inhibition of NO₃⁻ efflux (200 μM) (Figure 2, Table 1) were analyzed for kinase activities by an in-gel kinase assay.

As shown in Figure 4A, niflumic acid did not counteract the activation of SIPK and WIPK, suggesting that both kinases acted independently (or upstream) of NO₃⁻ efflux. To further determine the possible role of NO₃⁻ efflux on cryptogein-induced kinases, the same protein extracts were subjected to an in-gel kinase assay using histone III-SS (HIIISS) as a substrate. In this condition, in addition to SIPK and WIPK (data not shown), a protein kinase with a molecular mass of 40 kD was detected transiently in cryptogein-treated cells (Figure 4B). The activity of the 40-kD kinase peaked at 90 min and returned to the basal level within 3 h. In the presence of niflumic acid, the activation of the 40-kD kinase was delayed strongly and became fully active 4.5 h after the addition of the elicitor before returning to the basal level.

Figure 4.

A 40-kD Protein Kinase, but Not SIPK and WIPK, Acts Downstream of Cryptogein-Induced Anion Efflux.

(A) Activation of both SIPK and WIPK is anion channel independent. Tobacco cells were pretreated for 10 min with niflumic acid (200 μM) before cryptogein (25 nM) addition. Samples were taken at the times indicated. Protein extracts were analyzed for kinase activity by an in-gel kinase assay using myelin basic protein as a substrate.

(B) Niflumic acid delayed the activation of a 40-kD protein kinase by cryptogein. Tobacco cells were pretreated with niflumic acid (200 μM) 10 min before cryptogein (25 nM) addition, and samples were taken at the times indicated. Kinase activity was determined with an in-gel kinase assay using HIIISS as the substrate.

(C) The anion channel antagonists affected the cryptogein-induced 40-kD protein kinase differently. Tobacco cells were preincubated for 10 min with ethacrynic acid (300 μM), glibenclamide (200 μM), IAA-94 (400 μM), or niflumic acid (200 μM) before cryptogein (25 nM) addition. After 90 min of elicitor treatment, samples were taken and kinase activities were determined by an in-gel kinase assay using HIIISS as a substrate.

cry, cryptogein; Et-Ac, ethacrynic acid; Gli, glibenclamide; Nif, niflumic acid.

To confirm that the inhibition of the NO₃⁻ efflux does affect the activity of the cryptogein-induced 40-kD protein kinase, we tested the effect of glibenclamide and ethacrynic acid at concentrations that allow a complete inhibition of NO₃⁻ efflux (Table 1). As shown in Figure 4C, like niflumic acid, glibenclamide and ethacrynic acid strongly reduced the activity of the 40-kD protein kinase monitored after 90 min of cryptogein treatment, whereas the inefficient antagonist IAA-94 had no inhibitory effects. These results indicate that the activation of the 40-kD protein kinase depends, at least partially, on NO₃⁻ efflux in response to cryptogein.

Inhibitors of Anionic Channels Delayed the Cryptogein-Induced HR

Cell shrinkage is a major hallmark of the hypersensitive reaction. Based on animal studies, this process may be mediated by a net efflux of water caused by the release of anions and K⁺ (Maeno et al., 2000). Therefore, we tested the effects of the most efficient NO₃⁻ efflux antagonists on the induction of cell death by cryptogein. When applied to tobacco cell suspensions, 25 nM cryptogein induced almost 65% cell death within 24 h (Figure 5). Although IAA-94 had no effect on cryptogein-induced cell death, niflumic acid and glibenclamide caused a significant inhibition (analysis of variance; P < 0.002) and reduced elicitor-mediated cell death by ∼25 to 35% at 24 h. Ethacrynic acid also reduced cell death, but to a lower extent than niflumic acid and glibenclamide (20% reduction). Collectively, these results showed that only the antagonists of cryptogein-mediated NO₃⁻ efflux were efficient at reducing cell death. This finding suggests that NO₃⁻ efflux is an upstream step in the cryptogein transduction pathway leading to cell death.

Figure 5.

Cryptogein-Induced Cell Death Is Reduced by NO₃⁻ Efflux Antagonists.

Tobacco cells were pretreated for 10 min with ethacrynic acid (300 μM), glibenclamide (200 μM), IAA-94 (400 μM), or niflumic acid (200 μM) before addition of cryptogein (25 nM). The percentage of dead cells was estimated after 24 h of treatment by staining with neutral red. Bars represent means ± se of six independent experiments. cry, cryptogein; Et-Ac, ethacrynic acid; Gli, glibenclamide; Nif, niflumic acid.

To gain further insight into the involvement of anion channels in cryptogein-mediated tobacco cell death, we examined whether niflumic acid had an inhibitory effect on the elicitor-induced HR in tobacco plants. Local infiltration of 100 nM cryptogein into upper tobacco leaves resulted in the development of dry lesions within 16 h (Figure 6A). After 20 to 24 h, the entire infiltrated zone was completely necrotized (Figures 6B and 6C). When tobacco leaves were preinfiltrated with 500 μM niflumic acid at 30 min before the injection of cryptogein into the same area, necrotic symptoms appeared more slowly than in leaves treated with cryptogein alone and were clearly apparent only 20 h after elicitor injection (Figures 6D and 6E). Moreover, necrotic symptoms were less severe and were restricted to a small area in the infiltrated zone (Figure 6F).

Figure 6.

Antagonists of the Cryptogein-Induced Anion Channel Delayed the Development of the HR.

(A) to (C) A tobacco leaf was infiltrated with DMSO buffer at 30 min before cryptogein (100 nM) injection in the same area, and HR development was monitored at 16 h (A), 20 h (B), and 24 h (C) after elicitor infiltration.

(D) to (F) A tobacco leaf was infiltrated with niflumic acid (500 μM) at 30 min before cryptogein (100 nM) injection in the same area, and HR development was monitored at 16 h (D), 20 h (E), and 24 h (F) after elicitor infiltration.

(G) A control leaf 24 h after a first infiltration with DMSO buffer followed 30 min later by a second infiltration in the same area with water.

(H) A control leaf 24 h after a first infiltration with niflumic acid (500 μM) followed 30 min later by a second infiltration in the same area with water.

(I) A tobacco leaf was infiltrated with IAA-94 (500 μM) at 30 min before cryptogein (100 nM) injection in the same area, and HR development was monitored at 24 h after elicitor infiltration.

cry, cryptogein; Nif, niflumic acid.

Similar effects, but less extensive, were observed when cryptogein-treated tobacco leaves were preinfiltrated with 200 μM niflumic acid, the efficient concentration used on tobacco suspension cells (data not shown). This discrepancy could result from differences in sensitivity between cell suspensions and whole plants. In contrast to niflumic acid, IAA-94 had no inhibitory effect on the development of HR symptoms induced by cryptogein (Figure 6I).

To investigate the effects of niflumic acid on cryptogein-induced HR at the molecular level, we analyzed by RNA gel blot hybridization the kinetics of accumulation of several gene transcripts in leaf tissues infiltrated with cryptogein in the presence or absence of niflumic acid. Most of these transcripts accumulated during HR and correspond to different classes of so-called defense genes: genes encoding proteins with antioxidant properties (GST1a and GST1b) (Levine et al., 1994); hypersensitive-related (hsr) genes, including hsr515 and hsr203J, which encode proteins showing homology with cytochrome P450 and esterase, respectively (Marco et al., 1990; Pontier et al., 1994); sensitivity-related (str) genes, including str319, which belongs to the sesquiterpene cyclase gene family (Keller et al., 1998); PAL, which encodes Phe ammonia-lyase, a key enzyme of the phenylpropanoid biosynthetic pathway (Dixon, 2001); PR-3 (pathogenesis-related), which encodes a basic chitinase (Fritig et al., 1998); and SAR8.2, which belongs to a 12-member gene family encoding small highly basic proteins of unknown function that are expressed during the HR and SAR (Alexander et al., 1992). We also investigated NpCaM-2, a calmodulin gene that remains expressed constitutively in response to wind and cold shock (Van Der Luit et al., 1999).

With the exception of NpCaM-2, for which no increased mRNA levels were detected, infiltration of cryptogein into the leaves led to the accumulation of all of the transcripts in the infiltrated area (Figure 7). GST1a, GST1b, hsr515, str319, PAL, hsr203J, and SAR8.2 showed rapid induction (within 4 h), with increased expression at 8 h after infiltration of the elicitor. This expression was transient, occurred earlier than HR symptom appearance, and was not detectable after 16 h. The mRNA for PR-3 began to accumulate within 8 h and reached a maximum abundance at 24 h. By contrast, no induction of any of the genes analyzed was detected in control plants.

Figure 7.

Niflumic Acid Affected the Cryptogein-Induced Accumulation of Defense-Related Gene mRNA.

Tobacco leaves were treated as indicated in the legend to Figure 6. Leaf discs were harvested at the indicated times, and total RNA was extracted from these tissues. Total RNA then was subjected to gel blot analysis using GST1a, GST1b, hsr515, str319, PAL, PR-3, hsr203J, SAR8.2, and NpCaM-2 cDNA clones as probes. Ethidium bromide–stained 28S rRNA is shown as a control for gel loading. cry, cryptogein; Nif, niflumic acid.

When leaves were treated with niflumic acid and cryptogein, transcripts of hsr515, str319, PAL, hsr203J, and SAR8.2 were detectable 8 to 16 h after cryptogein infiltration, accumulated transiently, and still were detected after 24 h. Thus, niflumic acid delayed the expression kinetics of these genes. Interestingly, in the case of GST1a and GST1b, niflumic acid completely suppressed the accumulation of both transcripts induced by cryptogein. The slight accumulation detected at 4 h also was observed in control plants infiltrated with niflumic acid alone, suggesting that this effect was not caused primarily by cryptogein. By contrast, PR-3 mRNA that accumulated in response to cryptogein was not decreased significantly by niflumic acid. This observation and the fact that niflumic acid treatment did not affect NpCaM-2 accumulation excluded the possibility that the antagonist delayed gene expression nonspecifically.

DISCUSSION

In the present study, we investigated cryptogein-induced NO₃⁻ efflux in tobacco cell suspensions and analyzed its involvement in the elicitor signaling pathway. Using mainly pharmacological approaches, we provide evidence that this anion efflux, which may be channel mediated, plays a critical role in the mediation of cryptogein-induced events, including the oxidative burst, the activation of a 40-kD protein kinase, and the induction of several defense-related genes and hypersensitive cell death.

Pharmacological Properties of Cryptogein-Induced NO₃⁻ Efflux

Ion channel antagonists constitute powerful tools to investigate the contribution of channels in biological responses. In plants, they have been used widely to reveal the involvement of anion channels in the responses to hormones, elicitors, hypoosmotic stress, and light (Ward et al., 1995; Barbier-Brygoo et al., 1999). In this report, we show that anion channel blockers that belong to distinct chemical families attenuate cryptogein-induced NO₃⁻ efflux with the following sequence of efficiency: niflumic acid > glibenclamide > ethacrynic acid > DPC > A9-C. The active concentrations all were within the concentration ranges reported previously to block anion channels in animal and plants cells. By contrast, IAA-94, one of the most common anion channel antagonists, had no significant effects on elicitor-mediated NO₃⁻ efflux. Moreover, because of their apparent toxicity and/or nonspecificity, no conclusive answer can be given regarding NPPB, DIDS, and SITS effects. These data indicate that cryptogein-induced NO₃⁻ efflux may be channel mediated. This channel also may mediate the cryptogein-induced Cl⁻ efflux that was reported recently (Pugin et al., 1997). Indeed, NO₃⁻ and Cl⁻ permeability seems to be a general feature of plant plasma membrane anion channels (Barbier-Brygoo et al., 2000). In addition, both cryptogein-induced NO₃⁻ and Cl⁻ effluxes were dependent on an influx of calcium and occurred after a similar lag period of ∼5 min.

Interestingly, cryptogein-induced NO₃⁻ efflux shows a distinct pharmacological profile compared with the other elicitor-triggered anion effluxes described to date (Ebel et al., 1995; Jabs et al., 1997; Zimmermann et al., 1998). Moreover, the pharmacological profile of the cryptogein-induced anion channel seems to be different from those of the patch clamp–characterized anion channels involved in hormone signaling, blue light signaling, and cell osmoregulation (for review, see Barbier-Brygoo et al., 2000). However, because sensitivity to anion channel antagonists could differ in patch clamp and in vivo conditions and also between species, further studies are needed to determine whether the anion channel involved in cryptogein signaling belongs to a new class of anion channels.

All of the efficient antagonists had a transient inhibitory effect on cryptogein-induced NO₃⁻ efflux, and the addition of antagonists during the treatment did not sustain inhibition. This latter observation excludes a rapid turnover of the NO₃⁻ efflux–mediating channel, whose density on the plasma membrane may increase consequently in parallel to a decrease of inhibitor availability. A possible explanation for this finding might be that several anion channels or transporters with distinct pharmacological profiles participate in cryptogein-induced NO₃⁻ efflux. The transient effect also could be the result of a metabolization or detoxification process triggered by cells. This hypothesis is supported by the findings that (1) niflumic acid (and other anion channel antagonists; data not shown) induces a rapid accumulation of GST transcripts in tobacco leaves (Figure 7), and (2) the efficiency of niflumic acid was reduced strongly when the inhibitor was added to the suspension cells at 2 h instead of 10 min before cryptogein (data not shown). The fact that anion channel inhibitors are effective transiently may partly explain why most of the cryptogein-induced events that depend on NO₃⁻ efflux are delayed.

Connections between Anion Channel Activation and Other Early Events

Protein phosphorylation followed by Ca²⁺ influx are key upstream steps of cryptogein signal transduction (Viard et al., 1994; Tavernier et al., 1995; Lecourieux-Ouaked et al., 2000; Binet et al., 2001). The present study indicates that activation of the NO₃⁻ efflux depends on protein phosphorylation. Moreover, a prolonged activation of a kinase (or several kinases) is required to maintain the NO₃⁻ efflux because the addition of staurosporine in the middle of the cryptogein response prevented any further NO₃⁻ efflux.

Consistent with these results, calyculin A, a potent inhibitor of Ser/Thr phosphatase type 1 (PP1) and Ser/Thr phosphatase type 2A (PP2A) that was shown to mimic cryptogein-induced events, including polypeptide phosphorylation, Ca²⁺ influx, AOS production, and extracellular medium alkalinization (Lecourieux-Ouaked et al., 2000), stimulated NO₃⁻ efflux in the absence of the elicitor. NO₃⁻ efflux monitored in the presence of calyculin A plus a saturating concentration of cryptogein was comparable to that obtained with cryptogein alone, but an additive effect of both compounds was observed when a nonsaturating elicitor concentration was used.

Although calyculin A should influence multiple processes in tobacco cells, these data confirm previous results indicating that phosphatases negatively control the cryptogein cascade and that some proteins involved in elicitor signal transduction should be targets of calyculin A (Lecourieux-Ouaked et al., 2000). Such antagonistic modulation of anion channels by kinase and phosphatase has been described previously in abscisic acid signal transduction (Schmidt et al., 1995). Here, too, kinases would act as positive regulators, whereas phosphatases would act as negative regulators in the chain of events leading to anion channel activity. The question of whether a cryptogein-activated anion channel is (de)phosphorylated directly will require further investigation.

Experiments with Ca²⁺ channel antagonists indicate that Ca²⁺ influx from the extracellular space is required for the initiation and maintenance of anion channel activation in cryptogein-treated cells. Similarly, Ca²⁺ influx was shown to be a prerequisite for the activation of plasma membrane anion channels in response to several signals, including abscisic acid (Ward et al., 1995), elicitors of the plant defense reaction (Jabs et al., 1997), and cold shock (Lewis et al., 1997). The nature of the link between Ca²⁺ influx and anion efflux is unresolved at present, but it probably involves a complex network of signals. Indeed, plant anion channels are tightly regulated by many factors, including nucleotides, phosphorylation/dephosphorylation events, cytoplasmic free Ca²⁺, voltage, and cytoplasmic pH.

The AOS-producing system activated in response to cryptogein was shown to be DPI sensitive, NADPH dependent, and involved in pH alterations (extracellular medium alkalinization and cytosol acidification), sharing properties with NADPH oxidase from neutrophils (Pugin et al., 1997). Here, we provide evidence that NO₃⁻ efflux is required as a step in the cryptogein pathway leading to the oxidative burst and extracellular medium alkalinization.

Because cryptogein-induced NADPH oxidase seems partly responsible for the alkalinization of the extracellular medium (Pugin et al., 1997), inhibition of the extracellular alkalinization may be, at least in part, a direct consequence of the inhibition of the AOS-generating enzyme. Moreover, the fact that AOS production was not delayed but suppressed completely by NO₃⁻ efflux inhibitors suggests that the activation of cryptogein-induced NADPH oxidase results from a specific temporal combination of several essential events (including NO₃⁻ efflux activation), some of which are activated transiently and independently of anion efflux. Thus, NO₃⁻ efflux activation seems essential but not sufficient to induce NADPH oxidase.

Anion efflux also was shown to be necessary for the induction of the oxidative burst in elicitor-treated parsley or soybean cells (Ebel et al., 1995; Jabs et al., 1997), in soybean cells challenged with avirulent Pseudomonas syringae pv glycinea (Rajasekhar et al., 1999), and in tobacco cells exposed to hypoosmotic stress (Cazalé et al., 1998). Cazalé et al. (1998) reported that DPI slightly inhibits Cl⁻ efflux induced by hypoosmotic shock, suggesting that AOS production also may play a role in the modulation of anion channel activity. Such feedback regulation of the oxidative burst upon anion efflux is unlikely in our system because DPI had no effects on cryptogein-induced NO₃⁻ efflux (data not shown).

Recently, Long and Iino (2001) found a link between channel-mediated anion efflux and alkalinization of the extracellular medium. Indeed, alkalinization could occur if an efflux of anions resulting from channel activation provided substrate for a H⁺/anion symporter at the plasma membrane. In our system, we obtained no evidence for the contribution of H⁺/NO₃⁻ symporters in elicitor-induced extracellular medium alkalinization (data not shown). Indeed, we observed no changes in NO₃⁻ efflux (kinetics and intensity) in assays at pH 5.75 or 7.25, the latter value being similar to the cytosolic pH (Pugin et al., 1997; Lebrun-Garcia et al., 2002). Thus, in the presence or absence of a proton motive force, NO₃⁻ efflux is identical, which excludes a significant NO₃⁻ influx and linked pH change in cryptogein-treated cells.

We showed that the kinetics of activation of a 40-kD protein kinase that preferentially phosphorylates HIIISS was clearly affected by the NO₃⁻ efflux antagonist niflumic acid (and other efficient inhibitors). The antagonist delayed its activation by 3 h, a delay that corresponds to the time required for the cells to overcome the inhibitor effects (Figure 2). These results suggest that the 40-kD protein kinase may be positioned downstream of anion channels in the cryptogein signaling cascade. Moreover, this finding indicates that the activation of the 40-kD kinase is dependent on the duration and intensity of the NO₃⁻ efflux, suggesting that the 40-kD kinase could be activated in response to a hypoosmotic shock as a result of the anion loss. Consistent with this hypothesis, kinases have been shown to be activated in response to osmotic shock. In particular, Hoyos and Zhang (2000) recently proposed that the cryptogein-induced 40-kD kinase is likely to be the same kinase as HOSAK (high osmotic stress–activated kinase), a tobacco kinase activated by hyperosmotic stress caused by the exposure of cells to high concentrations of NaCl, Pro, or sorbitol. A tobacco 40- to 42-kD kinase showing identical characteristics (substrate preference and hyperosmotic stress sensitivity) and identified as a Ser/Thr kinase also was described by Mikolajczyk et al. (2000). Additional experiments will be necessary to determine the identity of the cryptogein-induced 40-kD kinase.

In contrast to this kinase, SIPK and WIPK seem to be activated independently (or upstream) of anion channel opening, as observed in tobacco cells exposed to hypoosmotic stress (Cazalé et al., 1999). This result differs from that of Ligterink et al. (1997), who reported that the parsley elicitor Pep-13–induced MAPK acted downstream of Cl⁻-permeable anion channel. Thus, although anion efflux and MAPK activation occurred in diverse signaling pathways, the causal links between the events should be different.

The Cryptogein-Induced Anion Channel Participates in Pathways Leading to Hypersensitive Cell Death

In animals, cell volume loss is a characteristic feature of cells undergoing PCD. This process (defined as apoptotic volume decrease) is promoted by the plasma membrane channel–mediated loss of Cl⁻, organic anions, K⁺, and therefore water. Recently, the relationship between apoptotic volume decrease and apoptosis was questioned, and data now indicate that that apoptotic volume decrease is not a passive secondary feature of PCD but a driver of this process (Maeno et al., 2000; Ping Yu and Choi, 2000). Consistent with these findings, pharmacological block of the plasma membrane Cl⁻ channel was shown to prevent or delay apoptosis in numerous cell types (Szabo et al., 1998; Maeno et al., 2000; Nietsch et al., 2000).

In plants, it is becoming apparent that cell death induced in the HR shows features that are similar to a certain extent to those in animal cells undergoing PCD (Lam et al., 2001). In particular, plant cells challenged by pathogens or elicitors display profound changes in cellular morphology, including blebbing of the plasma membrane, cell shrinkage, condensation of the cytoplasm, and condensation of the nucleus. Similar features of PCD were observed in tobacco cells treated with cryptogein (Levine et al., 1996).

Based on the assumption that cryptogein-induced cell shrinkage may be driven, at least in part, by anion channels, we investigated whether the cryptogein-induced NO₃⁻ channel participates in HR. Our results show that transient inhibition of cryptogein-induced NO₃⁻ efflux by the anion channel antagonists correlated with a 16 to 30% reduction in elicitor-triggered cell death. These data were extended to plant leaves, in which the infiltration of niflumic acid before cryptogein significantly delayed HR and resulted in the development of mild lesions rather than dry lesions. This was accompanied by a delay or suppression of the cryptogein-induced transcript accumulation of early expressed defense genes. Thus, anion channel activity should be involved in cryptogein-induced pathway(s) that lead to HR cell death. This finding is consistent with those of Levine et al. (1996), who demonstrated that anion channel blockers caused a marked inhibition of cell death and HR development in soybean suspension cells and leaves inoculated with P. syringae pv glycinea harboring the avrA gene.

Niflumic acid treatment differentially affected the accumulation of the transcripts analyzed in the present study. Cryptogein-induced hsr515, str319, PAL, hsr203J, and SAR8.2 transcript accumulation was delayed substantially by niflumic acid (4 to 8 h, depending on the transcript). As reported for the 40-kD kinase activation, this delay may be explained partly by the transient inhibitory effect of niflumic acid on NO₃⁻ efflux. By contrast, GST1a and GST1b transcript accumulation was blocked in the presence of niflumic acid. This complete inhibition led us to hypothesize that the expression of GST1a and GST1b is mediated by NO₃⁻ efflux via the oxidative burst. Indeed, when used at 200 μM, niflumic acid completely blocked and did not delay the activation of cryptogein-induced AOS production, and DPI completely suppressed GST1a and GST1b expression in response to the elicitor (data not shown).

Thus, the elicitor-activated anion channel would initiate at least two pathways: a first pathway mediated by the oxidative burst and leading to GST1a and GST1b expression, and a second pathway independent of AOS production and leading to the accumulation of transcripts str319, PAL, SAR8.2, and the HR-related hsr515 and hsr203J genes. This assumption is in agreement with data indicating the following: (1) phytoalexin synthesis, which may be associated with increased PAL and str319 transcript accumulation (Starks et al., 1997; Keller et al., 1998; Dixon, 2001), was unaffected by DPI in cryptogein-treated cells (Rustérucci et al., 1996); (2) elicitor-induced AOS production was shown to be a primary signal for the activation of GST transcription (Levine et al., 1994); (3) the pathway(s) leading to hsr203J activation does not involve AOS (Pontier et al., 1998); and (4) AOS are not linked to cryptogein-induced hypersensitive cell death (Dorey et al., 1999; Rustérucci et al., 1999; Binet et al., 2001).

Concerning PR-3, its induction by cryptogein was unaffected by niflumic acid treatment. Thus, PR-3 expression seems to be independent of anion channel activity. The fact that its expression was not affected under the conditions in which HR development was delayed by niflumic acid confirms previous reports indicating that PR genes and HR cell death may be regulated by different pathways activated through the same pathogen or elicitor (Mittler et al., 1997).

Among the defense genes investigated here, only hsr203J has been shown to be correlated tightly with PCD during the HR induced by bacterial and viral pathogens and by inducers of HR-like responses such as harpin, PopA1 proteins, and cryptogein (Pontier et al., 1998). HSR203 protein, which has been reported to be a Ser hydrolase showing esterase activity (Baudouin et al., 1997), was postulated to regulate either the establishment or the limitation of cell death (Pontier et al., 1998). Until now, the pathway(s) leading to hsr203J activation has remained enigmatic, and among the effectors of HR, methyl jasmonate, salicylic acid, and H₂O₂ were shown not to be involved in this process (Pontier et al., 1998). Our data clearly indicate that hsr203J induction and related PCD are mediated by anion channel activation. Interestingly, del Pozo and Lam (1998) recently demonstrated that caspase-specific inhibitors suppressed hsr203J transcript accumulation in tobacco leaves infiltrated with an HR-inducing bacteria strain. With regard to this finding, an important goal of future studies should include verification that caspase-like activity and associated cell death might be controlled by anion channel activity, as observed in mammals by Maeno et al. (2000).

In summary, this work greatly extends previous studies describing the functions of anion channels in plant defense. Cryptogein-induced anion channel activity modulates the threshold of activation of multiple defense responses, including the AOS-generating enzyme, several defense genes, and the HR.

METHODS

Plant and Cell Culture Conditions

Tobacco plants (Nicotiana tabacum cv Xanthi nc) were grown in a growth chamber at 24°C with a 14-h light period at ∼15,000 lux. All treatments were performed 6 weeks after germination. Tobacco cell suspensions were maintained as described by Pugin et al. (1997) except that cells were used for elicitor treatment 24 h after subculturing.

Plant and Cell Culture Treatments

With the exception of indanyloxyacetic acid–94 (IAA-94; Alexis Biochemicals, Laüfelfingen, Switzerland), all of the channel blockers were purchased from Sigma. Glibenclamide and diphenylamine-2-carboxylic acid were made up in DMSO as 200 mM stock solutions. The other channel antagonists were dissolved in 4% DMSO solution alkalinized by KOH (DMSO buffer) and prepared as 20 mM stock solutions. Because of insolubility, most of the channel blockers could not be assayed when prepared at concentrations >600 to 800 μM.

Diazoxide, minoxidil sulfate, calyculin A, staurosporine, EGTA, and lanthanum chloride were purchased from Sigma. EGTA and lanthanum chloride were prepared in water, and the other compounds were prepared in DMSO. When these molecules were tested, appropriate controls were included to ensure that DMSO did not interfere with the experiments. The final concentration of DMSO never exceeded 0.25%. Cryptogein was purified according to Bonnet et al. (1996) and prepared as a 10 μM stock solution in water.

For experiments with plants, 1 mL of a 500 μM anion channel antagonist (niflumic acid or IAA-94) solution was injected into the intracellular space of upper tobacco leaves. After a 30-min incubation, 1 mL of a 100 nM cryptogein solution was infiltrated directly into the same leaf area, and plants were exposed to continuous light at 24°C. Control leaves were infiltrated first with 1 mL of DMSO buffer (0.1% DMSO final concentration) or 1 mL of a 500 μM anion channel antagonist (niflumic acid or IAA-94) solution, and 30 min later, they were injected for a second time with 1 mL of ultrapure water in the same area. At various times, leaf discs corresponding to the treated area were harvested, quickly frozen, and stored at −80°C until they were analyzed for defense-related gene transcript accumulation.

Treatment of cell suspensions was performed as described previously (Binet et al., 2001). Briefly, before use, cells were washed and resuspended at 0.1 g fresh weight/mL in a suspension buffer containing 175 mM mannitol, 0.5 mM CaCl₂, 0.5 mM K₂SO₄, and 2 mM Hepes, pH 5.75, and equilibrated for 2 h at 24°C on a rotary shaker (150 rpm). Tobacco cells then were treated with cryptogein (5 to 250 nM). The pharmacological compounds were added 10 min before cryptogein, except lanthanum, EGTA, and staurosporine, which were added 5 min before the elicitor.

Nitrate Content Analysis

After various periods of treatment, aliquots of 2 mL of cells (0.2 g) were filtered on a GF-A glass filter (Whatman, Maidstone, UK). The extracellular medium obtained after filtration was collected and used for the measurement of nitrate (NO₃⁻) content. For the measurement of NO₃⁻ intracellular concentrations, cells (0.2 g) separated from the extracellular medium by filtration were harvested, ground in liquid nitrogen, and suspended in 2 mL of ultrapure water. After centrifugation for 5 min at 4000g, the supernatant was collected and stored before use.

The NO₃⁻ concentration was determined using a colorimetric assay kit (Alexis Biochemicals) according to the procedure recommended by the supplier. In a 96-well microtiter plate, 20 μL of each sample was added to 60 μL of assay buffer, 10 μL of NO₃⁻ reductase, and 10 μL of enzyme cofactor mixture. The microtiter plate was covered and incubated at room temperature for 1 h. During the incubation time, NO₃⁻ was reduced to nitrite by NO₃⁻ reductase. Nitrite then was converted to a deep purple azo compound by adding 50 μL of Griess reagent R1 (sulfanilamide) followed by 50 μL of Griess reagent R2 [N-(1-naphthyl)ethylenediamine]. After a 10-min incubation required to obtain optimal color development, absorbance was read at 540 nm using a plate reader (Molecular Dynamics, Sunnyvale, CA).

Quantification of NO₃⁻ concentrations was performed using a NO₃⁻ calibration curve (0 to 50 μM). To ensure that the pharmacological compounds did not interfere with the NO₃⁻ reductase–catalyzed reaction and the chemistry of the Griess reagents, these molecules were included at concentrations appropriate to the NO₃⁻ calibration curve. The yield of the NO₃⁻ reductase–catalyzed reaction was determined by comparing the NO₃⁻ calibration curve with a nitrite calibration curve (0 to 50 μM). Depending on the experiment, the yield was estimated as 85 to 90%.

Extracellular Alkalinization and Active Oxygen Species

Aliquots (0.25 mL) of cell samples were analyzed for active oxygen species production by chemiluminescence (Lumat 9501 luminometer; Berthold, Bad Wilbad, Germany) as described previously (Pugin et al., 1997). Extracellular pH was measured directly in the medium of tobacco cells.

Cell Viability Assay

Cell viability was assayed using the vital dye neutral red as described by Binet et al. (2001). Cells (1 mL) were taken, washed with 1 mL of a solution containing 175 mM mannitol, 0.5 mM CaCl₂, 0.5 mM K₂SO₄, and 2 mM Hepes, pH 7.0, and incubated for 5 min in the same solution supplemented with neutral red to a final concentration of 0.01%. Cells that did not accumulate neutral red were considered dying. At least 500 cells were counted for each treatment. The experiment was repeated six times.

Protein Kinase Activity Assay

After treatment, cells (0.25 g) were harvested as described above, frozen in liquid nitrogen, and ground in a mortar. Preparation of protein extracts was performed as reported previously (Lebrun-Garcia et al., 1998). The concentration of protein extracts was determined according to Bradford (1976). Protein extracts (20 μg) were electrophoresed on 10% SDS–polyacrylamide gels embedded with 0.25 mg/mL myelin basic protein (MBP) or 0.25 mg/mL histone IIISS (HIIISS). After electrophoresis, SDS was removed by washing the gels for 1 h with the washing buffer (50 mM Tris-HCl, pH 8.0, and 20% 2-propanol). The gels then were equilibrated for 1 h in buffer B (50 mM Tris-HCl, pH 8.0, and 5 mM β-mercaptoethanol).

Subsequently, proteins were denatured for 1 h with 6 M guanidine-HCl in buffer B and allowed to renature overnight at 4°C in buffer B containing 0.04% Tween 40 (five changes). The gels were equilibrated for 30 min at room temperature in 10 mL of the reaction buffer (40 mM Hepes, pH 7.5, 0.1 mM EGTA, 20 mM MgCl₂, and 2 mM DTT when MBP was used as the substrate; 40 mM Hepes, pH 7.5, 0.5 mM CaCl₂, 20 mM MgCl₂, and 2 mM DTT when HIIISS was used as the substrate) and then for 1.5 h in the reaction buffer supplemented with 25 μM ATP and 10 or 15 μCi of γ-³²P-ATP (Amersham) when MBP or HIIISS was used as the substrate, respectively.

The reaction was stopped by transferring the gels to 5% (w/v) trichloroacetic acid and 1% (w/v) sodium phosphate. The unincorporated γ-³²P-ATP was removed by extensive gel washing in the same solution for at least 2.5 h. Finally, the gels were dried onto 3MM paper (Whatman) and exposed to X-Omat AR film (Kodak). Prestained molecular mass markers (Bio-Rad) were used to calculate the apparent sizes of the kinases.

RNA Gel Blot Analysis

Total RNA was extracted from leaf discs using Trizol reagent (Gibco BRL) according to the supplier's instructions. Ten micrograms of total RNA per lane was separated on 1.2% agarose gels containing 1.1% formaldehyde, blotted to a nylon membrane (Hybond-XL; Amersham), and cross-linked by UV light. The cDNA clones used as probes for hybridization were labeled by random priming (Prime-a-Gene Labeling System; Promega).

Membrane hybridization was performed at 65°C as described by Church and Gilbert (1984). The membrane was washed with 2 × SSC (1× SSC is 0.15 M NaCl and 0.015 M sodium citrate) twice for 5 min at room temperature, with 0.5% SDS and 2 × SSC twice for 30 min at 65°C, and subsequently with 0.1 × SSC twice for 30 min at room temperature. The membrane then was exposed to X-Omat AR film (Kodak).