Salicylic acid (SA), a ubiquitous phenolic phytohormone, is involved in many plant physiological processes including stomatal movement. We analysed SA-induced stomatal closure, production of reactive oxygen species (ROS) and nitric oxide (NO), cytosolic calcium ion ([Ca2+]cyt) oscillations and inward-rectifying potassium (K+in) channel activity in Arabidopsis. SA-induced stomatal closure was inhibited by pre-treatment with catalase (CAT) and superoxide dismutase (SOD), suggesting the involvement of extracellular ROS. A peroxidase inhibitor, SHAM (salicylhydroxamic acid) completely abolished SA-induced stomatal closure whereas neither an inhibitor of NADPH oxidase (DPI) nor atrbohD atrbohF mutation impairs SA-induced stomatal closures. 3,3′-Diaminobenzidine (DAB) and nitroblue tetrazolium (NBT) stainings demonstrated that SA induced H2O2 and O2– production. Guard cell ROS accumulation was significantly increased by SA, but that ROS was suppressed by exogenous CAT, SOD and SHAM. NO scavenger 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (cPTIO) suppressed the SA-induced stomatal closure but did not suppress guard cell ROS accumulation whereas SHAM suppressed SA-induced NO production. SA failed to induce [Ca2+]cyt oscillations in guard cells whereas K+in channel activity was suppressed by SA. These results indicate that SA induces stomatal closure accompanied with extracellular ROS production mediated by SHAM-sensitive peroxidase, intracellular ROS accumulation and K+in channel inactivation.
Plants have developed adaptive mechanisms like regulation of stomatal guard cells to withstand against pests and extreme environmental conditions. In order to regulate gas exchange and water loss via transpiration and defend invasion of microorganisms, plants control volume of guard cells resulting in opening and closure of stomata in response to phytohormones and various environmental signals (Schroeder, Kwak & Allen 2001). Salicylic acid (SA) is a phenolic plant hormone and regulates various physiological processes like defence against harmful microorganisms, growth and ions uptake (Hayat, Ali & Ahmad 2007). SA-induced stomatal closures in Commelina communis and Vicia faba have been reported (Manthe, Schulz & Schnabl 1992; Lee 1998; Mori et al. 2001) but SA-induced stomatal closure in Arabidopsis remains to be investigated.
It is well known that ABA-induced stomatal closure involves Ca2+-dependent and Ca2+-independent signalling pathway. ABA activates anion channels and induces stomatal closure via this parallel pathway (Allan et al. 1994; Grabov & Blatt 1997). In Vicia faba, ABA-induced Ca2+-independent stomatal closure has been reported (Levchenko et al. 2005), on the other hand, ABA-induced Ca2+-dependent and Ca2+-independent activation of anion channels has been reported in guard cells of Nicotiana tabacum (Marten et al. 2007). ABA- and MeJA-induced oscillation of [Ca2+]cyt is modulated by Ca2+ influx from extracellular space mediated by Ca2+-permeable non-selective cation channels, which are activated by ROS production mediated by NADP(H) oxidase (Pei et al. 2000; Murata et al. 2001; Kwak et al. 2003; Munemasa et al. 2007; Islam et al. 2010). ABA and MeJA lead [Ca2+]cyt elevation to activate S-type anion channels, which causes in plasma membrane depolarization, leading activation of outward-rectifying K+ (K+out)channels, resulting in stomatal closure (Allen et al. 1999). In addition, inward-rectifying K+ (K+in) channels in guard cells have been suggested to provide a pathway for K+ uptake into guard cells during stomatal opening. Suppression of K+in channel is favourable to stomatal closure (Kwak et al. 2001; Saito et al. 2008; Siegel et al. 2009). K+in channels are inhibited by increases in [Ca2+]cyt (Schroeder & Hagiwara 1989; Lemtiri-Chlieh & MacRobbie 1994) but [Ca2+]cyt-independent K+in channel activation has also been reported (Grabov & Blatt 1997).
In this study, to elucidate SA signalling in Arabidopsis, we examined stomatal closure, ROS production, NO production, [Ca2+]cyt oscillations, and K+in channel activity.
MATERIALS AND METHODS
Plant materials and growth
Arabidopsis (Arabidopsis thaliana) L. ecotype Columbia was used as wild-type plant in this study. Columbia and atrbohD atrbohF mutant plants were grown in growth chambers (22 °C, 80 µmol m−2 s−1 under a 16 h light/8 h dark regime). Arabidopsis genome initiative numbers for AtRBOH genes are AtRBOHD (At5g47910) and AtRBOHF (At1g64060).
Measurement of stomatal aperture
Stomatal apertures were measured as described previously with modification (Murata et al. 2001; Munemasa et al. 2007; Jahan et al. 2008; Islam et al. 2009, 2010). Excised rosette leaves of 4–6 weeks old Arabidopsis plants were blended in water for 30 s in blender and epidermal tissues were collected with 48 µm nylon sieves. The epidermal tissues were dipped into stomatal assay solution containing 50 mM KCl, 50 µM CaCl2 and 10 mM MES-Tris, pH 6.15, for 2 h in the light (80 µmol m−2 s−1). Then, SA was added to the stomatal assay solution and stomatal apertures were measured 2 h after incubation. Twenty stomatal apertures were measured on each epidermal peel. Three replications were maintained for each stomatal assay experiment. Student's t-test was used to determine statistical significance of the data. We regarded difference at the level of P < 0.05 as significant.
Measurement of ROS and NO in guard cells
To analyse ROS and NO production in guard cell, 2′, 7′-dichlorodihydrofluorescein diacetate (H2DCF-DA) (Sigma, St. Louis, MO, USA) (Murata et al. 2001; Suhita et al. 2004; Munemasa et al. 2007; Saito et al. 2008; Islam et al. 2010) and 4,5-diaminofluorescein-2 diacetate (DAF-2DA) (Sigma) (Foissner et al. 2000; Neill et al. 2002; Huang et al. 2004; Munemasa et al. 2007; Saito et al. 2008; Islam et al. 2010) were used, respectively. In case of ROS detection, epidermal peels were incubated for 3 h in the medium containing 50 mM KCl, 50 µM CaCl2, 10 mM MES-Tris, pH 6.15 and then 50 µM H2DCF-DA was added to this medium. The epidermal peels were incubated for 40 min at room temperature, and then the excess dye was washed out with distilled water. The dye-loaded tissues were treated with 500 µM SA, and then fluorescence of guard cells were imaged using Bio-zero software. The fluorescence was then quantified with Adobe IMAGE J software (Adobe, Bethesda, MD, USA). For NO detection 20 µM DAF-2DA was added instead of 50 µM H2DCF-DA.
Visualization of H2O2 and O2– in whole leaf
The rosette leaves of Arabidopsis plants were excised, floated on Stomatal Assay Solution containing 0.1% Tween 20 and incubated for 2 h under light (80 µmol m−2 s−1). Then the leaves were transferred in 1 mg/mL 3, 3′-diaminobenzidine (DAB)-HCl (Sigma) and gently infiltrated in a vacuum for 2 h. Then, SA was added and infiltrated for 3 h. Inhibitors were applied 30 min before SA application. After incubation, the leaves were cleared in boiling ethanol (99%) for 10 min. Localization of H2O2 is visualized as a reddish-brown coloration. For detection of O2–, nitroblue tetrazolium (1 mg/mL) (Sigma) was used instead of DAB. Localization of O2– was visualized as blue coloration. In both cases, after clearing the leaves in ethanol, the leaves were mounted on cover glass and pictures were taken. The intensity of coloration was quantified using Adobe Photoshop CS (Adobe Systems Inc.; San Jose, CA, USA) software.
Measurement of [Ca2+]cyt oscillations in guard cells
Oscillations of [Ca2+]cyt in guard cells of Arabidopsis expressing YC3.6 were measured according to Islam et al. (2010). The abaxial side of an excised leaf was gently mounted on a glass slide by using a medical adhesive, followed by removal of the adaxial epidermis and the mesophyll tissue with a razor blade in order to keep intact the lower epidermis on the slide. The mounted abaxial epidermal peel was kept in a solution containing stomatal assay solution to promote stomatal opening. The turgid guard cells were considered for ratiometric [Ca2+]cyt measurement. Then guard cells were treated with 500 µM SA in the incubation buffer by a peristatic pump after 5 min from the start of measurement. The fluorescence intensity of guard cells was imaged and analysed using AQUA COSMOS software (Hamamatsu Photonics K.K., Hamamatsu, Japan).
Measurement of K+in currents in guard cell protoplasts
We used a whole-cell patch-clamp technique to measure K+in currents of guard cell protoplasts (GCPs). Arabidopsis GCPs were prepared from rosette leaves of 4- to 6-week-old plants with the digestion solution as previously described (Munemasa et al. 2007). Whole-cell currents were recorded by a CEZ-2200 patch clamp amplifier (Nihon Kohden, Tokyo, Japan). The resulting values were corrected for liquid junction potential and leak currents were not subtracted. For data analysis, pCLAMP 8.2 software (Axon Instruments, Foster City, CA, USA) was used. For whole-cell current recordings, the pipette solution contained 30 mM KCl, 70 mM K-Glu, 2 mM MgCl2, 3.35 mM CaCl2, 6.7 mM O,O'-Bis(2-aminoethyl)ethyleneglycol-N,N,N',N'-tetraacetic acid (EGTA), 10 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) adjusted to pH 7.1 with Tris, and bath solution was composed 30 mM KCl, 2 mM MgCl2, 40 mM CaCl2, 10 mM MES titrated to pH 5.5 with Tris (Saito et al. 2008). To investigate the effects of SA, GCPs were treated with 500 µM SA for 2 h under dark condition before measurements.
SA induces stomatal closure
We investigated SA-induced stomatal closure in wild-type (Col-0) plants. Application of 200 µM, 500 µM, and 1000 µM SA reduced stomatal apertures by 17% (P < 0.003), 21% (P < 0.0002), and 23% (P < 0.0003), respectively, in Col-0 plants and showed a dose-dependent manner (Fig. 1a). Solvent control (0.01% ethanol) did not cause significant closure of stomata (data not shown). These observations are similar to the findings that SA induced stomatal closure in V. faba (Manthe et al. 1992; Mori et al. 2001) and Commelina communis (Lee 1998).
To investigate whether SA-induced stomatal closure requires ROS as second messenger, we examined effects of catalase (CAT) and superoxide dismutase (SOD) on SA-induced stomatal closure. SA-induced stomatal closures were significantly inhibited by exogenous application of 100 units mL−1 of CAT (P < 0.0009) and 330 units mL−1 of SOD (P < 0.0002) (Fig. 1b), where neither CAT nor SOD by itself did not induce stomatal movement (data not shown). Since both CAT and SOD are cell impermeable chemicals, it is predicted that these enzymes scavenged extracellular ROS. These results suggest that SA induced stomatal closure via production of extracellular ROS as second messenger.
In order to clarify that cell wall peroxidases are involved in SA-induced ROS production, we examined effects of SHAM on SA-induced stomatal closure. The SA-induced stomatal closure was completely suppressed by SHAM (P < 0.02) (Fig. 1c), which suggest that cell wall peroxidases are involved to mediate extracellular ROS production during SA-induced stomatal closure. Note that SHAM by itself did not change stomatal aperture (data not shown). In Vicia faba, SA-induced stomatal closures were also suppressed by SHAM (Mori et al. 2001). To investigate involvement of NADPH oxidases in SA-induced stomatal closure, we examined the effects DPI (NADPH oxidase inhibitor) and atrbohD atrbohF mutation on SA-induced stomatal closure. Exogenous application of DPI did not inhibit the SA-induced stomatal closures (P = 0.06) (Fig. 1c) and the atrbohD atrbohF mutation did not affect SA-induced stomatal closure (P < 0.002) (Fig. 1d). Note that DPI by itself did not affect stomatal aperture (data not shown). These results indicate that SA induces stomatal closure via production of extracellular ROS mediated by cell wall peroxidases.
ROS production in whole leaves
We used two histochemical staining approaches to investigate production of ROS in Arabidopsis whole leaves. One is that in vivo formation of H2O2 was visualized using DAB (Thordal-Christensen et al. 1997) and another is that accumulation of O2– was examined by staining with NBT (Doke 1983). DAB staining demonstrated that SA induced H2O2 production in whole leaves (Fig. 2a) and that SHAM inhibited SA-induced H2O2 production (P < 0.009) (Fig. 2a). These results indicate that SA-induced H2O2 production involves SHAM-sensitive peroxidases. NBT staining illustrated that SA elicited O2– production (Fig. 2b). SHAM significantly suppressed SA-induced O2– production (P < 0.002) (Fig. 2b). These results indicate that O2– production is attributed to activation of SHAM-sensitive peroxidases.
Accumulation of ROS in guard cells
To clarify that ROS acts as second messenger in SA signalling, we examined SA-induced ROS accumulation in guard cells using 2′,7′-dichlorodihydrofluorescein diacetate (H2DCF-DA). SA increased ROS levels in guard cells compared with control (P < 0.001) (Fig. 3a). SA-induced ROS accumulation was significantly impaired by pre-treatment with CAT, SOD and SHAM (P < 0.006 for CAT; P < 0.003 for SOD; P < 0.002 for SHAM) (Fig. 3a). In addition, SA induced ROS accumulation in the atrbohD atrbohF mutants (P < 0.006) (Fig. 3b). These results suggest that SA induces ROS production in extracellular spaces mediated by SHAM-sensitive peroxidases, and that ROS is then accumulated in guard cells by diffusion.
NO production in guard cells induced by SA
NO has been reported to function as second messenger in ABA- and MeJA-induced stomatal closure in Arabidopsis (Munemasa et al. 2007). To investigate involvement of NO in SA signalling, we examined the effect of a NO specific scavenger [2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (cPTIO)] on SA-induced stomatal closure. SA-induced stomatal closures were significantly suppressed by pre-treatment with cPTIO (P < 0.006) (Fig. 4a), whereas SA-induced ROS accumulation in guard cells was not suppressed by pre-treatment with cPTIO (P = 0.99) (Fig. 4b). In addition, in SHAM (P < 0.005) pre-treated cells, NO production induced by SA was completely suppressed (Fig. 4c). These results indicate that NO production occurs during SA-induced stomatal closure and NO production proceeds after ROS production in SA signalling.
Involvement of extracellular free Ca2+ in SA-induced stomatal closure
A large proportion of Ca2+ is localized in extracellular spaces. External application of Ca2+ has been shown to promote stomatal closure and induces [Ca2+]cyt oscillation in guard cells (McAinsh et al. 1995; Allen et al. 2001). In order to clarify involvement of extracellular Ca2+ in SA-induced stomatal closure, we investigated the effect of Ca2+ chelator (EGTA) and calcium ion channel blocker (LaCl3), in SA-induced stomatal closure. Both EGTA and LaCl3 significantly suppressed SA-induced stomatal closure (P < 0.0002 for EGTA; P < 0.004 for LaCl3) (Fig. 5), where neither EGTA nor LaCl3 by itself did not affect stomatal aperture (data not shown). These results suggest that extracellular free Ca2+ might be involved in modulation of SA-induced stomatal closure.
SA-induced [Ca2+]cyt oscillations in guard cells
A recent report shows that Ca2+-dependent and Ca2+-independent pathways may exist in parallel in plants (Roelfsema & Hedrich 2010). [Ca2+]cyt oscillations in guard cells play vital role in stomatal closure (Siegel et al. 2009; Islam et al. 2010).To analyse SA-induced [Ca2+]cyt oscillations, we used Yellow Cameleon 3.6 (YC3.6) to measure change in [Ca2+]cyt (Allen et al. 1999). When treated with 500 µM SA, [Ca2+]cyt oscillations were recorded in 18.75 % guard cells and 81.25 % guard cells did not show detectable [Ca2+]cyt elevations (Fig. 6). It is speculated that SA uses cytosolic Ca2+-independent pathway to induce stomatal closure.
Suppression of Kin channels by SA in guard cells
Several reports suggest that suppression of K+in currents is favourable to ABA-induced stomatal closure but is not favourable to light-induced stomatal opening (Schroeder, Ward & Gassmann 1994; Kwak et al. 2001). We measured K+in channel currents in SA-treated guard cell protoplasts using the whole-cell patch clamp technique. The amplitudes of K+in currents were significantly decreased by treatment with 500 µM SA (P < 0.03) (Fig. 7). These results suggest that SA suppresses K+in channel during SA-induced stomatal closure.
Extracellular ROS are involved in SA-induced stomatal closure
ABA- and MeJA-induced stomatal closure require ROS production in guard cells and the ROS production is mediated by NAD(P)H oxidases (Pei et al. 2000; Murata et al. 2001; Munemasa et al. 2007; Jahan et al. 2008; Saito et al. 2008; Islam et al. 2009, 2010). This study shows that SA-induced stomatal closure also requires ROS production (Fig. 1b). However, the ROS production is not mediated by NAD(P)H oxidases but by SHAM-sensitive peroxidases (Fig. 1c), which is consistent with the results that SA induced ROS production in the atrbohD atrbohF mutants (Fig. 3b). In tobacco suspension cultured cells, SA-induced ROS production is inhibited by peroxidase inhibitors (Kawano & Muto 2000). In addition, SA treatment induced pathogenesis related gene (PR1) in both wild-type and atrbohD atrbohF mutant plants (data not shown) and ROS is responsible for inducing pathogenesis related genes in plants (Neuenschwander et al. 1995). Taken together, it is concluded that SA induces ROS production mediated by peroxidases to lead stomatal closure and to express pathogenesis related genes.
Moreover, this study has elucidated that SA-induced stomatal closure involves extracellular oxidative burst (Figs 2 & 3) and that the extracellular oxidative burst is mediated with SHAM-sensitive peroxidases (Fig. 2) but not NAD(P)H oxidases (Figs 1d & 2). Mori et al. (2001) reported that SA induces stomatal closure via SHAM-sensitive peroxidase dependent oxidative burst in Vicia faba. SA induces productions of O2– and H2O2 in extracellular spaces in tobacco suspension cultured cells (Kawano et al. 1998; Kawano & Muto 2000). This study further indicates that SA induces peroxidase-mediated ROS production in extracellular spaces and then this ROS moved into guard cells (Figs 2 & 3). Allan & Fluhr (1997) have demonstrated that ROS migrates from one cell to another cell via apoplastic route.
NO acts as second messenger in SA-induced stomatal closure
We further demonstrated that pre-treatment with cPTIO did not inhibit SA-induced ROS accumulation in guard cells (Fig. 4b), whereas pre-treatment with SHAM suppressed NO production in guard cells (Fig. 4c). These results suggest that NO functions downstream of ROS production in SA signalling in Arabidopsis. Several studies have demonstrated that involvement of NO in ABA signalling and NO production occurs following production of ROS (Dong et al. 2005; Bright et al. 2006; Li et al. 2009).
Extracellular calcium act as priming in modulation of stomatal closure
La3+ completely inhibited Ca2+ channel currents in Arabidopsis guard cells to suppress [Ca2+]cyt oscillations (Allen et al. 2001). In this study, treatments with LaCl3 partially suppressed SA-induced stomatal closure (Fig. 5). On the other hands, EGTA treatment completely inhibited SA-induced stomatal closure (Fig. 5). Furthermore, SA did not elicit [Ca2+]cyt oscillation (Fig. 6). These results suggest that inhibition of SA-induced stomatal closure by LaCl3 and EGTA is not attributed to that of Ca2+ influx to mediate [Ca2+]cyt oscillation.
Separately, property of Ca2+-binding is closely related with salt stress tolerance via regulation of K+ and Na+ transport activities (Murata et al. 1998, 2000; Kasukabe et al. 2006). EGTA can remove Ca2+ bound to plasma membrane through chelation and LaCl3 can compete with Ca2+ at binding sites on plasma membrane. These results suggest that the inhibitory effects of EGTA and LaCl3 on SA-induced stomatal closure are caused by the reduction of Ca2+-binding on plasma membrane.
In addition, extracellular Ca2+ acts as priming for modulation of stomatal closure (Klüsener et al. 2002; Kim et al. 2010). It is suggest that extracellular Ca2+ could take part in SA-induced stomatal closure to prime modulation of some signal component(s). However, the priming mechanism remains to be clarified.
ABA-induced H2O2 production inhibits K+in channels in Vicia guard cells (Xiao et al. 2001) and SA induced extracellular and intracellular ROS production (Fig. 3). Moreover, extracellular Ca2+ inhibits inward K+in channel currents and activated K+out channel currents in GCPs of Vicia faba (Xiang et al. 2008) whereas SA-induced stomatal closure involved extracellular Ca2+ (Fig. 5) but not [Ca2+]cyt oscillation (Fig. 6). Taken together, SA induces stomatal closure is accompanied by K+in channel inactivation possibly by peroxidase-mediated ROS production and extracellular Ca2+ fluctuation.
SA signalling in Arabidopsis guard cells
Based on our results, we propose a simple model of SA signalling in Arabidopsis guard cells in Fig. 8. SA elicits extracellular ROS production mediated by SHAM-sensitive peroxidase, followed by intracellular ROS accumulation and NO production, leading to stomata closure and SA inactivates K+in channels, which is favourable to stomatal closure.
We thank Taniya Rahman for critical reading this paper. This research was funded in part by a Grant-in-Aid for Young Scientists and Grants for Scientific Research on Priority Areas from the Ministry of Education, Culture, Sports, Science and Technology of Japan.