Tomato–Pseudomonas syringae interactions under elevated CO2 concentration: the role of stomata

Summary Elevated CO2 enhanced defence against Pseudomonas syringae in tomato plants, not only by reducing stomata-mediated entry of P. syringae but also by invoking a stomata-independent pathway to counteract P. syringae.


Introduction
In modern greenhouse cultivation, CO 2 enrichment is considered as an important means of increasing the yield and quality of agricultural produce, particularly in C 3 horticultural crops (Uprety, 1998;Dion et al., 2013). A steady increase in atmospheric CO 2 levels has been observed over the past 150 years, and this increase is projected to continue so that CO 2 levels will be doubled over the next 50 years (Levine et al., 2008). The rising CO 2 concentration ([CO 2 ]) has influenced recent research interests, and a strong emphasis has been placed on the impact of elevated [CO 2 ] on photosynthesis and plant growth in the era of climate change (Long et al., 2004;Levine et al., 2008). Despite this trend, the effects of rising atmospheric [CO 2 ] on plant-pathogen interactions have received little attention. Studies conducted in free-air CO 2 enrichment sites, open-top chambers, and growth chambers involving plant diseases have shown that responses to elevated [CO 2 ] may vary with the host-pathogen system, and thus the severity and/or incidence of disease may either increase, decrease, or remain unaffected (Eastburn et al., 2010;Melloy et al., 2010;West et al., 2012). As plant-pathogen interactions under increasing [CO 2 ] have the potential to interrupt both agricultural and natural systems significantly, additional work is required to understand the extent and mechanisms through which elevated [CO 2 ] affects plant diseases. Such studies will be essential for making accurate predictions regarding future plant disease dynamics and proper management of agricultural and natural ecosystems under changing climatic conditions.
Plants have a complex array of defence mechanisms. For example, the cell wall is covered with a waxy cuticle that serves as a potent physical barrier. Although some pathogenic fungi infect plants by penetrating the cell wall, many foliar bacterial plant pathogens invade plants primarily through natural surface openings, namely, through the stomata (Kumar et al., 2012). Stomata are the small pores in the leaf epidermis, formed by a pair of guard cells that have developed mechanisms to sense and respond to various endogenous and environmental stimuli. By changing the size of the stomatal pore, stomata can regulate gas exchange between the plant and environment, as well as control water loss (Melotto et al., 2008). Manning and Vontiedemann (1995) speculated that reducing the stomatal conductance and the size of the stomatal aperture could inhibit the entry of bacterial pathogens through the stomata. Notably, reduced stomatal conductance is one of the primary effects of rising atmospheric [CO 2 ] on plants (Drake et al., 1997;Long et al., 2004); however, it is not clear whether the CO 2 -induced alterations in the stomatal characteristics are associated with foliar bacterial pathogenic infections. In a recent study involving oilseed Brassica juncea, the decrease in the disease index of downy mildew caused by the stomata-invading pathogen Hyaloperonospora brassicae was suggested to be associated with a decrease in stomatal density, pore size, and stomatal conductance in response to elevated [CO 2 ] (Mathur et al., 2013). Similarly, in Populus, the clones that opened their stomata late in the morning were more resistant to Melampsora larici-populina Kleb., compared with those that opened their stomata earlier (Siwecki and Przybyl, 1981). By contrast, there are also reports suggesting that stomatal conductance and stomatal density are not correlated with pathogenic infection (Riikonen et al., 2008;O'Keefe et al., 2013). Furthermore, elevated [CO 2 ] not only reduces stomatal conductance but also stimulates higher rates of photosynthesis (Drake et al., 1997;Ainsworth and Rogers, 2007). The altered leaf chemistry, including elevated levels of carbon gain or secondary compounds, also has the potential to affect disease severity (McElrone et al., 2005;Ghasemzadeh and Jaafar, 2011;Huang et al., 2012). The dependence of foliar bacterial disease on stomatal conductance or the leaf chemistry under elevated [CO 2 ] has yet to be elucidated.
Pseudomonas syringae pv. tomato DC3000 (P. syringae) is a Gram-negative bacterium capable of infecting tomato and Arabidopsis. This bacterium is a model pathogen used to investigate the molecular mechanisms underlying plant-pathogen interactions (Whalen et al., 1991;Preston, 2000;Zeng et al., 2011). Furthermore, P. syringae is also considered as one of the stomata-invading pathogens (Melotto et al., 2008). Elucidating how elevated [CO 2 ] affects host symptoms will advance our current understanding of the dynamics between plants and pathogens in future natural ecosystems. Therefore, the present study was conducted with the following objectives: (i) to examine whether elevated [CO 2 ] has positive effects on plant defence against P. syringae infection, and if so, (ii) to identify the mechanism involved. In the present study, tomato seedlings were grown under ambient and elevated [CO 2 ] and inoculated with P. syringae pv. tomato DC3000. Using pharmacological and gene silencing approaches, we obtained evidence that elevated [CO 2 ] induces resistance to P. syringae in tomato plants. In addition to reducing the stomata-mediated pathogen entry of P. syringae, elevated [CO 2 ] also invokes stomata-independent pathways to counteract P. syringae.

Pathogen inoculation and bacterial growth analysis
The bacteria P. syringae pv. tomato DC3000 was cultured in King's B medium containing rifampicin (25 mg ml -1 ) at 28 °C. A single bacterial colony was cultured at 28 °C with shaking until log phase. Then, the bacterial cells were collected by centrifugation at 4000 g for 10 min and resuspended in 10 mM MgCl 2 until an OD 600 of 0.2 was reached, which corresponded to approximately 10 8 colony-forming units (cfu) ml −1 . The inoculation was carried out by spraying the bacterial suspension on the whole plant at a final concentration of 10 6 cfu ml -1 , which was obtained by serial dilution, according to Katagiri et al. (2002). The MgCl 2 buffer was applied as a mock inoculation. The bacterial populations were measured from four leaves per plant at 48 hours post-inoculation (hpi), according to the method described previously (Wolfe et al., 2000).
To exclude the factor for P. syringae entry through stomata, P. syringae was artificially infiltrated directly into the leaf apoplast using 1 ml needleless syringe at a final concentration of 10 5 cfu ml -1 . At 4 days after inoculation, disease symptoms were assessed using a chlorophyll fluorescence imaging system (Liao et al., 2013). Generally, in actinic light (300 µmol m −2 s −1 ), maximal fluorescence (F m ′) and steady-state fluorescence prior to the flash (F) were measured while saturated light flashes were applied every 20 s, and the photochemical quantum yield at photosystem II (ΦPSII) of lightadapted leaves was calculated as F m ′-F/F m ′.

Measurement of stomatal conductance and aperture
Stomatal conductance was measured using an infrared gas analyserbased portable photosynthesis system (LI-6400; LI-COR, Lincoln, NE, USA). The air temperature, relative humidity, and light intensity were maintained at 25 °C, 85%, and 1000 µmol m -2 s -1 , respectively. The CO 2 concentration was set to match the growth CO 2 concentration, i.e., 380 and 800 µmol mol -1 for ambient and elevated [CO 2 ]-treated plants, respectively.
To view the stomatal aperture by light microscopy, abaxial epidermises were peeled with forceps and floated on buffer (10 mM MES, 30 mM KCl, 0.5 mM Ca 2+ , pH 6.2). The temperature during the experiment was set at 25 °C, and the stomatal apertures were measured using a light microscopy equipped with a digital camera (Leica Microsystems, Wetzlar, Germany), as well as the image analysis software Adobe Photoshop CS5 (Adobe, San Jose, CA, USA).
Samples for scanning electron microscopy (SEM) were prepared following standard procedures (Singh et al., 2000). Leaf samples were immersed in 5% glutaraldehyde overnight, washed in 0.1 M phosphate buffer (pH 7.0), then treated in 1% osmium tetroxide for 2 h. The samples were dehydrated using serially diluted ethanol (70% to 100%) and iso-amyl acetate. Specimens were dried in a Hitachi HCP-2 critical point dryer (Tokyo, Japan) with liquid CO 2 and mounted onto double-coated carbon conductive tapes that were attached onto specimen holders. Samples were sputter-coated with gold-platinum and examined in a Hitachi TM-1000 TSEM (Tokyo, Japan).

Quantification of NO content and trypan blue staining
The NO accumulation was assayed and calculated by following the conversion of oxyhaemoglobin (HbO 2 ) to methaemoglobin (MetHb) spectrophotometrically at 401 nm and 421 nm as described by Pasqualini et al. (2009). Briefly, samples corresponding to 0.3 g of fresh leaf tissue were ground in a mortar and homogenized in 100 mM phosphate buffer (pH 7.0) and 0.6% (w:v) insoluble polyvinylpolypyrrolidone. The extracts were clarified by adding powdered activated carbon and centrifuged at 10 000 g for 10 min at 4 °C. The supernatant was filtered through a PTFE Millipore membrane (0.45 µm) and immediately assayed for NO. Five minutes before HbO 2 addition, samples were pre-treated with catalase (100 U) and superoxide dismutase (100 U) to remove reactive oxygen species.
Trypan blue staining was performed as described by Bai et al. (2012). Tomato leaves were boiled for 10 min in the lactophenoltrypan blue solution (10 ml of lactic acid, 10 ml of glycerol, 9.3 g of phenol, 10 mg of trypan blue, dissolved in 10 ml of distilled water) and then decolourized in chloral hydrate (25 g chloral hydrate dissolved in 10 ml distilled water) for 12 h. The dark-brown spots were photographed to evaluate cell death.

Virus-induced gene silencing
The tobacco rattle virus (TRV) virus-induced gene silencing (VIGS) constructs used for the silencing of the tomato nitrate reductase (NR) and guard cell slow-type anion channel 1 (SLAC1) gene were generated by cloning a 379-bp NR cDNA fragment and 430-bp SLAC1 cDNA fragment, which were PCR-amplified using the following primers: forward primer 5ʹ CGgaattcAGACCCTCAACACTCAACCC 3ʹ and reverse primer 5ʹ CGggatccCAGAATGCTGTCAACACCAC 3ʹ (with EcoRI and BamHI restriction sites) for NR; forward primer 5ʹ CGCggatccTGATGAATTTGATGGCTTGG 3ʹ and reverse primer 5ʹ TGCtctagaGGCGCACTCTTTGTCTCC 3ʹ (with BamHI and XbaI restriction sites) for SLAC1. The amplified fragments were digested with restriction enzymes and ligated into the corresponding restriction sites of pTRV2. An empty pTRV2 vector was used as a control. The resulting plasmids were transformed into Agrobacterium tumefaciens strain GV3101 and were grown for 16-18 h in LB broth supplemented with the appropriate antibiotics (Liu et al., 2002). The cells were pelleted, washed, and resuspended in infiltration buffer [10 mM MgCl 2 , 10 mM 2-(N-morpholine)-ethane-sulphonic acid (MES), pH 5.5, 150 mM acetosyringone). Three hours after induction at room temperature, A. tumefaciens containing pTRV1 and pTRV2 (empty vector or with target gene insert) were mixed in a 1:1 ratio to a final OD 600 of 0.6. Leaves and cotyledons of 2-week-old tomato seedlings were infiltrated using a blunt syringe. Plants were then kept at 21 °C for 4 weeks before they were used for the experiments. The silencing efficiency was assessed by quantitative real-time PCR as described in previous studies (Livak and Schmittgen, 2001).

Statistical analysis
At least four independent replicates were conducted for each determination. The data were subjected to analysis of variance using SAS 8.0 software package (SAS Institute, Cary, NC), and the means were compared using Tukey's test at the P<0.05 level.

Response of stomatal movement to P. syringae in ambient and elevated [CO 2 ]
After 48 h of elevated [CO 2 ] pre-treatment (800 µmol mol -1 ), the average stomatal aperture of the tomato leaves was much smaller than that of their ambient counterparts (Fig. 1A, B). Upon P. syringae spray inoculation, the stomatal openings of ambient [CO 2 ]-treated leaves decreased rapidly in epidermal peels. At 4 hpi, the average stomatal aperture of the leaves decreased to the lowest level, which was only 34.3% of the average stomatal aperture of the mock-inoculated plants grown under ambient [CO 2 ]. The P. syringae-induced stomatal closure in ambient [CO 2 ]-treated leaves was only transient because the stomatal aperture recovered to mock levels at 6 hpi and remained constant thereafter. By contrast, under elevated [CO 2 ], P. syringae spray inoculation did not lead to evident changes in stomatal movement, and the stomatal aperture of the inoculated plants was much smaller than that of the ambient counterpart but approximately the same as that of the mock-inoculated and elevated [CO 2 ]-treated plants (Fig. 1A, B). The distribution of bacteria on the surface of tomato epidermal cells in response to [CO 2 ] elevation and P. syringae spray inoculation was observed by SEM (Fig. 1C). Twenty five stomata from at least 4 plant leaves for each treatment were observed, and there were general observations that elevated [CO 2 ]-treated plants had a smaller average stomatal aperture than that of their ambient [CO 2 ] counterparts and showed a transient closing of their stomata at 4 hpi, which is in accordance with the results obtained by light microscopy. The bacterial inoculum flocked together near the guard cells at 4 hpi, irrespective of CO 2 concentration treatment. Thereafter, following the reopening of the stomata, the bacteria remained distributed around the open stomata in ambient [CO 2 ] but not in elevated [CO 2 ]-treated plants. By contrast, the bacteria remained scattered on the surface of tomato epidermal cells of leaves under the elevated [CO 2 ] treatment.
As P. syringae can produce polyketide toxin COR to reopen stomata, we examined whether elevated CO 2 was preventing the reopening of stomata by disrupting the function of COR (Fig. 1D). Indeed, after 4 h of COR treatment, COR slightly but significantly increased stomatal aperture sizes in ambient [CO 2 ]-treated plants, whilst such effects were not observed in elevated [CO 2 ]-treated plants.

Involvement of NO in elevated [CO 2 ]-induced stomatal movement
Compared with ambient [CO 2 ], plants treated with elevated [CO 2 ] showed a rapid and sharp decrease in stomatal conductance, which was accompanied by a simultaneous increase in endogenous NO content ( Fig. 2A). At 12 h after treatment with elevated [CO 2 ], the stomatal conductance had decreased by 37.7%, whereas the NO content increased by 59.2%. After 48 h of ambient or elevated [CO 2 ] treatment, exogenous pharmacological SNP (NO donor) and cPTIO (NO scavenger) were applied to the tomato leaves to study the role of NO in [CO 2 ]-regulated stomatal movements (Fig. 2B, C). As expected, the endogenous NO content was either increased by exogenous SNP or decreased by cPTIO (Fig. 2C). SNP further decreased the stomatal aperture of plants grown in both ambient and elevated [CO 2 ] conditions, whereas cPTIO application fully compromised the elevated [CO 2 ]-induced stomatal closure, suggesting that NO plays a role in the elevated [CO 2 ]-induced stomatal closure (Fig. 2B, C).

Effect of exogenous NO modulator on P. syringae infection in ambient and elevated [CO 2 ] conditions
As NO plays a role in the elevated [CO 2 ]-induced stomatal closure, we studied the effects of NO on plant innate immunity against P. syringae under conditions of both ambient and elevated [CO 2 ] (Fig. 3). We pre-treated the tomato leaves of both ambient and elevated [CO 2 ] plants with SNP and cPTIO before spray inoculating them with P. syringae. We found that elevated [CO 2 ] effectively alleviated the bacterial infection, and the bacterial population growth on plants under elevated [CO 2 ] was reduced to 56.0% compared with those plants grown in the ambient [CO 2 ] at 48 hpi. The NO scavenger cPTIO fully blocked elevated [CO 2 ]-induced alleviation in P. syringae infection, whereas SNP application on the plants significantly reduced the bacterial infection in both the ambient and elevated [CO 2 ] plants. Under SNP-pre-treated condition, the bacterial population of P. syringae was reduced to a similar level in both the ambient and elevated [CO 2 ] plants.

Effects of NR and SLAC1 gene silencing on NO content, stomatal aperture, and P. syringae infection
To verify the role of elevated [CO 2 ]-induced stomatal closure and NO generation in basal defence against P. syringae spray inoculation, the key genes involved in stomatal closure and NO generation, i.e. SLAC1 and NR, respectively, were silenced via VIGS in tomato plants. Under ambient [CO 2 ], neither stomatal aperture nor endogenous NO content was affected by NR silencing (Fig. 4). Under elevated [CO 2 ], NR silencing significantly reduced elevated [CO 2 ]-induced NO accumulation and subsequent stomatal closure; both were reversed by exogenous SNP application. By contrast, SLAC1 gene silencing significantly increased the stomatal aperture in both ambient and elevated [CO 2 ], although it had no evident effect on endogenous NO content. Exogenous SNP did not seem to affect the stomatal aperture in SLAC1-silenced plants, even though the endogenous NO concentration was significantly elevated in these plants (Fig. 4).
The elevated [CO 2 ]-induced P. syringae defence response also changed in conjunction with the SLAC1-and NRsilencing effects on the stomatal aperture and NO content (Fig. 5). In ambient [CO 2 ], we observed that both the NRand SLAC1-silenced plants, especially the pTRV-SLAC1 plants, were more sensitive to the infection of bacteria and exhibited significantly higher bacterial colony counts and cell death at 48 hpi. The elevated [CO 2 ]-induced P. syringae defence response was blocked by NR-silencing. However, in SLAC1-silenced but not NR-silenced plants, elevated [CO 2 ] still had significantly lower bacterial growth population compared with the ambient [CO 2 ] plants. Under both ambient and elevated [CO 2 ], exogenous SNP significantly reduced the bacterial population of all three plant materials, although the alleviation effect was smaller in pTRV-SLAC1 compared with pTRV and pTRV-NR plants.

Effect of elevated [CO 2 ] on P. syringae infection under syringe-inoculation condition
To exclude the stomata-limiting factor for entry of bacterial inoculum, P. syringae was also inoculated into leaf interior by needleless syringe infiltration method after 48 h of ambient or elevated [CO 2 ] pre-treatment. As pathogen infection often results in a reduction in the operating efficiency of PSII, the chlorophyll fluorescence imaging method was used to analyse the response of ΦPSII to P. syringae infection (Fig. 6). In absence of restriction in stomatal movement, elevated [CO 2 ]-treated leaves still exhibited remarkably higher defence against P. syringae than ambient [CO 2 ]-treated leaves which was evident by minor spread of P. syringae lesions in elevated [CO 2 ]-treated leaves at 4 days post inoculation (Fig. 6A). Furthermore, elevated [CO 2 ]-treated plants also had higher ΦPSII than that of ambient [CO 2 ] counterpart (Fig. 6B).

Discussion
Increasing atmospheric [CO 2 ] is creating novel environment for plants and is likely to have significant consequence with regard to the relationship between plant pathogens and their hosts. Previous studies on the effects of elevated [CO 2 ] on plant-pathogen interactions have produced conflicting results (Lake and Wade, 2009;Newton et al., 2011). For example, the plant fungal pathogen Colletotrichum gloeosporiodes exhibited increased fecundity and aggressiveness over 25 infection cycles in the host Stylosanthes scabra under elevated [CO 2 ] (Chakraborty and Datta, 2003). However, investigations into the systemic responses of tomato to Tomato yellow leaf curl virus (TYLCV) and of tobacco to Potato virus Y found that elevated [CO 2 ] decreased disease incidence and severity (Matros et al., 2006;Huang et al., 2012). As a model pathogen, P. syringae has been well studied; however, the impacts of elevated [CO 2 ] on P. syringae diseases are largely unknown. The present study is among the first to examine the effects of elevated [CO 2 ] on host symptoms associated with P. syringae infection. Generally, elevated [CO 2 ] enhanced plant resistance to infection with P. syringae via a dual mechanism. This information is helpful for predicting the responses of bacterial pathogens to elevated atmospheric [CO 2 ].
Pathogen entry into hosts is a critical step before the onset of the infection and disease progression. There is evidence indicating that P. syringae bacteria use the leaf surface stomata for entry into the host tissue (Boureau et al., 2002). Reduced stomatal opening in response to elevated atmospheric [CO 2 ] has previously been well documented (Drake et al., 1997). In this study, we investigated whether stomatal closure mediated the defence effects against P. syringae in response to elevated [CO 2 ]. Based on the light microscopy and SEM observations, we noticed that the ambient [CO 2 ]-treated leaves exhibited a transient stomatal closing in response to P. syringae spray inoculation at approximately 4 hpi (Fig. 1). Similarly, A. thaliana stomata close in response to live bacteria and purified pathogen/microbe-associated molecular patterns (PAMPs/ MAMPs) in a solution system; thus, Melotto et al. (2006) suggested that stomata played an active role in restricting bacterial invasion as part of the innate immune response, but not only as a passive port of entry. It has also been reported that bacterium-induced stomatal closure requires PAMPs signalling and homeostasis of the defence hormone salicylic acid, and is upstream of signalling regulated by abscisic acid in the guard cell (Melotto et al., 2006;Melotto et al., 2008). In addition, it seems that NO is also required for PAMPs and bacteria to close stomata, as NO production in guard cells could also be rapidly induced by the known MAMPs, flagellin (flg22), or lipopolysaccharide (LPS), and inhibitor of NO synthase could effectively block flg22-, LPS-, and E. coliinduced stomatal closure (Melotto et al., 2006). To counter host defences during infection and in the apoplast, P. syringae and other plant pathogenic bacteria have evolved a variety of virulence factors to subvert host defences or to obtain nutrients (Nomura and He, 2005). P. syringae pv. tomato DC3000, used in this study, can produce polyketide toxin COR to promote stomatal opening and disrupts plant defence responses (Melotto et al., 2006;Zeng et al., 2011). The stomatal movement in the P. syringae-inoculated ambient [CO 2 ] plants in this study (Fig. 1) was in accordance with the model proposed in previous studies (Melotto et al., 2006;Melotto et al., 2008). By contrast, under elevated [CO 2 ], the stomatal aperture was constantly smaller than the ambient counterpart and did not show any evident transient changes in response to P. syringae inoculation ( Fig. 1A-C). Furthermore, COR-induced stomatal opening was also effectively counteracted by elevated [CO 2 ] compared with an ambient counterpart (Fig. 1D), which might partly contribute to tomato defence against P. syringae infection under elevated [CO 2 ]. These differences in P. syringae-and COR-induced stomatal movement between ambient and elevated [CO 2 ] treatment may contribute to the different behaviour of P. syringae bacteria in these plants; namely, the concentrated distribution of P. syringae around the reopened stomata in ambient [CO 2 ]-treated plants might allow pathogens to enter the intercellular space more successfully compared with the elevated [CO 2 ] plants, in which bacteria were dispersed on the surface of tomato epidermal cells among the closed stomata (Fig. 1).
In the present study, the elevated [CO 2 ]-induced rapid decrease in stomatal conductance was accompanied by a simultaneous increase of endogenous NO content (Fig. 2). Furthermore, exogenous application of the NO donor SNP further decreased stomatal aperture, whereas its scavenger cPTIO fully abolished elevated [CO 2 ]-induced stomatal closure (Fig. 2). It seems that NO plays a pivotal role in elevated [CO 2 ]-induced guard cell movement and stomatal closure. In previous studies, an NO-induced Ca 2+ increase and subsequent stomatal closure have been observed in the guard cells of Arabidopsis and Vicia faba (Neill et al., 2002;Garcia-Mata and Lamattina, 2003). NO has also been demonstrated to be a signalling intermediate during bicarbonate-induced (mimic of elevated [CO 2 ]) stomatal closure in Pisum sativum (Kolla and Raghavendra, 2007). The elevated [CO 2 ]-induced and NO-mediated stomatal closure might in turn participate in the defence against P. syringae infection. To verify and extend this knowledge in planta, pharmacological and gene silencing approaches were used to control stomatal opening, and the mechanisms of elevated [CO 2 ]-induced defence against P. syringae were investigated accordingly. In correspondence with the involvement of NO in stomatal closing, elevated [CO 2 ]-induced P. syringae defence was fully abolished by cPTIO but enhanced by SNP (Fig. 3). Previous studies have demonstrated that NR is the key enzyme in the reduction of nitrite to NO in several physiological contexts, including stomatal closure (Rockel et al., 2002;Bright et al., 2006;Bellin et al., 2013). The guard cell gene SLAC1 has been identified as an important player in rapid stomatal responses to environmental factors (Negi et al., 2008;Vahisalu et al., 2008;Merilo et al., 2013). Elevated bicarbonate, more so than elevated [CO 2 ], activates the intracellular free calcium ion [Ca 2+ ] i sensitivity of SLAC1-mediated anion channels, thus reducing stomatal opening (Xue et al., 2011). In this study, NR silencing significantly reduced elevated [CO 2 ]-induced NO accumulation and stomatal closure, which was reversed by SNP application, whereas SLAC1 silencing also blocked stomatal closure, but could not be reversed by exogenous SNP (Fig. 4). These results indicate that both NR and SLAC1 play crucial roles in elevated [CO 2 ]-induced stomatal closure (Bright et al., 2006;Negi et al., 2008;Xue et al., 2011), but NR acts in an NO-dependent manner, whereas SLAC1 acts in an NO-independent manner. The increased stomatal opening in NR-and SLAC1-silenced plants resulted in significant increases in P. syringae infection and disease expression, and also blocked elevated [CO 2 ]-induced P. syringae defence to varying extents (Fig. 5), confirming the role for stomata in restricting bacterial entry into plant tissue. In this study, elevated [CO 2 ] also inhibited COR-induced stomatal opening (Fig. 1D). How NO is involved in this process remains unclear. In a previous study on epidermal peels of Arabidopsis plants, even though NO is suggested to be required for PAMPs and bacteria to close stomata, COR did not block synthesis of NO induced by PAMPs perception; the researchers thus speculated that COR acts downstream or independently of NO synthesis to block stomatal closure (Melotto et al., 2006). In a recent study on Arabidopsis, COR-induced decrease in photosynthetic efficiency at ambient CO 2 was found to be eliminated by supplementation of plants with high CO 2 for only a 2-h period, suggesting atmospheric [CO 2 ] had fundamental effects on COR-mediated function (Attaran et al., 2014). Thus, further studies will be necessary to determine the precise effect mode of elevated [CO 2 ] on COR-mediated stomatal movement.
SLAC1-silenced plants, however, grown under elevated [CO 2 ] still had significantly higher resistance compared with those grown under ambient [CO 2 ] (Fig. 5), implying that factors other than the stomata also play a role in elevated [CO 2 ]-induced P. syringae resistance. In accordance, we also subjected tomato plants to P. syringae inoculation by syringe infiltration method. In such condition, entry of P. syringae through stomata is not a restricting factor, and we found that elevated [CO 2 ] plants still had higher defence level than that of ambient [CO 2 ] counterpart (Fig. 6), verifying the results obtained on NR-and SLAC1-silencing plants. In addition, as the NO content was higher in the elevated [CO 2 ]-treated SLAC1-silenced plants when compared with the ambient counterpart (Fig. 4), we speculate that NO-related physiological events other than stomatal closure also contribute to the elevated [CO 2 ]-induced resistance. NO is now recognised as a crucial player in plant defence against pathogens. Many proteins that are specifically regulated by NO and that participate in signalling during the plant defence response have been identified, highlighting the physiological relevance of these modifications in plant immunity (Bellin et al., 2013). Alternatively, other unknown factors may also be involved in elevated [CO 2 ]-induced P. syringae resistance, thus underscoring the need for further investigation.
In conclusion, we found that the susceptibility of tomato plants to P. syringae is reduced under elevated [CO 2 ] conditions. Here, we propose a dual mechanism involving elevated [CO 2 ]-induced reductions in stomata-dependent and -independent pathways underlying this elevated [CO 2 ]-induced defence response. This information is important for making proper predictions with regard to disease pressure and for designing strategies to improve plant pathogen resistance, especially to foliar bacterial pathogens, under changing agricultural and natural ecosystems.