Is nitric oxide a critical key factor in ABA-induced stomatal closure?

The role of nitric oxide (NO) in abscisic acid (ABA)-induced stomatal closure is a matter of debate. We conducted experiments in Vicia faba leaves using NO gas and sodium nitroprusside (SNP), a NO-donor compound, and compared their effects to those of ABA. In epidermal strips, stomatal closure was induced by ABA but not by NO, casting doubt on the role of NO in ABA-mediated stomatal closure. Leaf discs and intact leaves showed a dual dose response to NO: stomatal aperture widened at low dosage and narrowed at high dosage. Overcoming stomatal resistance by means of high CO2 concentration ([CO2]) restored photosynthesis in ABA-treated leaf discs but not in those exposed to NO. NO inhibited photosynthesis immediately, causing an instantaneous increase in intercellular [CO2] (Ci), followed by stomatal closure. However, lowering Ci by using low ambient [CO2] showed that it was not the main factor in NO-induced stomatal closure. In intact leaves, the rate of stomatal closure in response to NO was about one order of magnitude less than after ABA application. Because of the different kinetics of photosynthesis and stomatal closure that were observed, we conclude that NO is not likely to be the key factor in ABA-induced rapid stomatal closure, but that it fine-tunes stomatal aperture via different pathways.

Introduction

Stomata are the gateways for gas exchange between the inside of the leaf and the external atmosphere. In terrestrial plants, fine regulation of stomatal aperture is critical for balancing CO2 uptake for photosynthesis whilst preventing excessive transpirational water loss. Evolution has resulted in a complex signalling network of pathways that trigger opening or closing of stomata in response to environmental cues. Abscisic acid (ABA) is considered to be the main phytohormone that induces stomatal closure, although an ABA-independent pathway for closure of stomata has also been proposed (Yoshida ; Huang ; Montillet ; Roychoudhury ).

Nitric oxide (NO) acts as important signalling molecule in diverse physiological processes, including adaptations to abiotic and biotic stresses (Wendehenne ; Courtois ; Neill ; Jeandroz ; Fancy ) and it interacts with multiple plant hormones, including salicylic acid, ABA, and cytokinin (Freschi, 2013; Prakash ). Its bioactivity is realized through redox-based post-translational protein modifications such as S-nitrosylation, which involves the addition of a NO group to a protein cysteine thiol to generate an S-nitrosothiol (Astier ). Application of sodium nitroprusside (SNP), a NO donor, to detached leaves of Triticum aestivum, Tradescantia virginiana, and Poncirus trifoliata results in reduced water loss (García-Mata and Lamattina, 2001; Fan and Liu, 2012), whilst treatment of epidermal peels of Tradescantia species, Vicia faba, and Pisum sativum results in significant reductions in stomatal aperture (García-Mata and Lamattina, 2001; Neill ). ABA promotes NO synthesis in guard cells, while the application of NO-scavengers prevents ABA-induced stomatal closure (García-Mata and Lamattina, 2001; Neill ). Consequently, NO has repeatedly been suggested to be a key component of the ABA signalling pathway for stomatal closure, acting downstream of ABA (Desikan ; Neill ; Garcia-Mata ; Bright ; Murata ).

However, other reports have raised doubts about this proposed role of NO. While ABA treatment reduces the water loss of detached leaves of Arabidopsis, application of SNP to leaves undergoing rapid dehydration does not result in pronounced stomatal closure compared with dehydrating control leaves, and scavenging of NO prevents neither the dehydration-induced nor the ABA-induced stomatal closure of the dehydrating leaves (Ribeiro ). These findings suggest that in Arabidopsis, NO signalling is redundant in the ABA-induced dehydration response. Indeed, dehydrating leaves of both the Arabidopsis wild-type and the nia1 nia2 mutant with reduced NO biosynthesis show similar rates of stomatal closure following ABA treatment (Ribeiro ). In addition, a critical role for NO downstream of ABA can also be questioned in non-dehydrating leaves, as stomata of the nia1 nia2 noa1-2 triple-mutant with impaired NO biosynthesis are hypersensitive to ABA and the plants show a strong resistance to drought (Lozano-Juste and León, 2010). Overall, this suggests that NO is not an indispensable component in the ABA signalling pathway.

Furthermore, stomata of the NO-overproducing Arabidopsis mutant gsnor1-3 are less sensitive to ABA (Wang ), and NO may function in a negative feedback mechanism to prevent over-activation of ABA signalling in guard cells by S-nitrosylation of Open Stomata 1 (OST1), a kinase that functions as a positive regulator for ABA-mediated stomatal closure. Laxalt suggested that NO acts both in promoting and arresting the ABA-induced/phospholipid-mediated signals that trigger stomatal closure.

Thus, the role of NO in ABA-induced stomatal closure is far from clearly defined.

Epidermal peels are often used to study the interactions between NO and ABA; however, stomatal responses in peels can differ from those in intact leaf tissues, as shown for calcium (Atkinson ), CO2, and light (Mott ; Mott, 2009; Fujita ). Lee and Bowling (1992) found a stomatal response to light when an isolated epidermis was incubated in the presence of mesophyll cells or chloroplasts isolated from an illuminated leaf, but there was no response without the mesophyll or in the presence of chloroplasts isolated from dark-adapted leaves. They suggested that guard-cell responses are linked to a product of photosynthetic activity in the mesophyll via a diffusible factor. Lawson suggested that stomatal movements are coordinated with mesophyll carbon assimilation via a complex interaction between ion channels, sugars, malate, and photosynthesis.

NO-donor compounds such as sodium nitroprusside (SNP) are often used to investigate stomatal responses to NO. This has several disadvantages. NO donors produce decomposition products in addition to NO that can themselves affect cells (Feelisch, 1998). For example, SNP can release iron complexes, disulphides, and cyanide (CN–) together with NO (Feelisch, 1998; Wang ). Interactions between CN–, NO, and ABA are known in seed dormancy (Bethke ) and CN– will affect photosynthesis (Wodala ). Because of these possible side-effects, the results of experiments that use SNP and other NO donors to study stomatal responses should be interpreted carefully. Moreover, precise control of NO dosage (time × concentration) is difficult when NO donors are used.

In light of this, in addition to using SNP (the most common NO donor) in the current study, we also used NO in its pure gas form when investigating stomatal responses to NO in epidermal peels, leaf discs, and intact leaves of Vicia faba. Adding gaseous NO to flow-through gas-mixing systems enables instantaneous administration and removal of NO in order to observe the dynamics of stomatal closure as well as photosynthesis responses. Despite these advantages, to our knowledge there are no studies of NO as an ABA-signalling intermediate in stomatal closure in which it has been supplied as a gas rather than via a donor. Our results showed that in intact leaves the rate of stomatal closure after NO application was about one order of magnitude less than after ABA application. Experiments with epidermal peels were largely affected by factors other than NO, most likely CO2.

Materials and methods

Plant material and growth conditions

Fava bean (Vicia faba L. cv Longpod) plants were cultivated in a growth chamber at 20±1 °C and 60±5% relative humidity (RH) with a 12-h photoperiod (300 μmol m–2 s–1). Plants were grown in soil in pots and were regularly watered with nutrient solution (for details see Aliniaeifard ). Fully developed leaves (fourth and fifth in acropetal order) were used for measurements.

Stomatal responses to NO gas, SNP, and ABA in epidermis peels

Epidermal strips were isolated from the abaxial surfaces of fully developed leaves in order to investigate stomatal responses to NO gas (5000 µmol mol –1 in N2; Linde Gas Benelux BV, Schiedam, The Netherlands), sodium nitroprusside (SNP, Na2[Fe(CN)5NO].2H2O; Sigma), and ABA (Sigma). The strips were pre-incubated for 1 h in a stomatal-opening medium (50 mM KCl, 10 mM MES-KOH, pH 6.15, 50 µM CaCl2 in de-gassed distilled water; Aliniaeifard and van Meeteren, 2014) in Petri dishes placed in a test chamber at 20–25 °C and 30 µmol m–2 s–1 photosynthetically active radiation (PAR, 400–700 nm; spectrum given in Supplementary Fig S1A at JXB online). The strips were then transferred to freshly de-gassed stomatal-opening solutions with different concentrations of SNP (50, 200, and 1000 µM) and ABA (100 µM), or to fresh medium alone (control) and kept for 1 h in the test chamber. Images were taken for the measurement of stomatal apertures.

To study the effects of NO gas, a self-built gas-mixing system was connected to a desiccator, into which the Petri dishes with the epidermal strips were placed (12 µmol m–2 s–1 PAR inside the desiccator; spectrum as above). The strips were pre-incubated for 1 h in stomatal-opening medium and kept in an atmosphere with a lowered O2 concentration (20 mmol mol–1 O2, 380 µmol mol–1 CO2, remainder N2) in order to diminish the oxidation of NO gas in the next step. Before exposure to NO gas, some strips were taken out of the desiccator as controls and images were taken immediately for determination of stomatal apertures. The gas-mixing system was adjusted to obtain different concentrations of NO in the same O2 and CO2 concentrations as during pre-incubation, and the strips were treated for 1 h.

In each experiment, 10–12 strips were taken from two leaves of the same plant (fourth and fifth: see above). For each treatment, three strips were randomly selected and an image was taken of each, from which the pore widths of 10 randomly selected stomata were measured. Various concentrations of NO gas and SNP were tested in separate experiments, each of which included a control (for NO) or control and ABA (for SNP). All experiments were repeated three times.

Images (500 × magnification) were taken using a DXM-1200 digital camera (Nikon) attached to an Aristoplan microscope (Leica). Pore widths were measured using the ImageJ software.

Stomatal responses to NO gas, SNP, and ABA in leaf discs

Leaf discs were cut with a 1.5-cm diameter cork-borer, using the middle of a leaf between the main vein and leaf margin. In each experiment three discs per treatment from leaves of one plant were used. The discs were incubated with their adaxial side upwards for 1 h in Petri dishes filled with the stomata-opening medium and then transferred to dishes with either opening medium alone (control and NO gas treatments) or with different concentrations of SNP (50, 200, and 1000 µM) or ABA (100 µM). To achieve fast, uniform uptake of the solutions, the dishes were twice subjected to a vacuum for 30 s (Filippou ). After 90 min, the dishes were placed in a flow-through cuvette under a system for chlorophyll fluorescence imaging (see below) and different concentrations of NO gas were applied. Different concentrations of SNP were also tested in separate experiments. Each experiment included a control (for NO) or control and ABA (for SNP). All experiments were repeated three times.

Confocal laser-scanning microscopy

To verify its presence in guard cells, a fluorescent NO-indicator dye, 4,5-diaminofluorescein diacetate (DAF-2DA; Sigma), was used (Bright ). The DAF-2DA solution was prepared in darkness in DMSO on the day of the experiment. Epidermal strips were pre-incubated for 1 h in the stomatal opening medium and 15 µM DAF-2DA was added for 20 min. The epidermal strips were then rinsed with opening medium for 20 min followed by incubation in opening medium alone (control) or with NO gas, SNP, or ABA before imaging using a Zeiss LSM 510-META confocal laser-scanning microscope (excitation 488 nm, emission 515–560 nm). The fluorescence intensity was assessed using ImageJ and for each stoma it was corrected as described by Gavet and Pines (2010):

Corrected fluorescence = Guard cells fluorescence – (Guard cells area × Mean fluorescence background)/1000

where the background reading corresponds to the pixel intensity of areas outside the stoma.

Mapping of PSII photochemical efficiency using chlorophyll fluorescence

Imaging of chlorophyll fluorescence and calculation of PSII efficiency (Φ PSII) was done as described by Rezaei Nejad , with slight modifications. A gas-mixing system was attached to a flow-through cuvette that was placed in the imaging area of a chlorophyll fluorescence imaging system (FluorCam 700MF, Photon System Instruments, Brno, Czech Republic) and Petri dishes containing leaf discs were placed in the cuvette. Light intensity in the cuvette was 75 µmol m–2 s–1 PAR (see Supplementary Fig S1B for spectrum), the temperature was 22±1 °C and RH was 40±3%. Measurements in the cuvette were conducted under 20 mmol mol–1 O2, 380 µmol mol–1 CO2, and the remainder N2; these non-photorespiratory conditions were deemed appropriate because, by means of Φ PSII, they allowed areas where the stomata might be closed to be identified. NO gas was applied as described above. Preliminary experiments showed that 10 min light exposure was sufficient to reach the steady state of Φ PSII. Hence, images were taken after 10 min and values for Ft and Fm´ were determined and then averaged over all pixels per leaf disc. Φ PSII was then calculated as (Fm´–Ft)/Fm´ (Rezaei Nejad ).

To determine whether changes in Φ PSII were caused by a decrease in stomatal conductance or by a direct effect of the treatments on photosynthetic activity, images of chlorophyll fluorescence were taken after 5 min exposure to a high CO2 concentration (50 000 µmol mol –1 CO2, other gases unchanged) following the measurements at 380 µmol mol–1 CO2.

Stomatal impressions

After chlorophyll fluorescence imaging, the stomatal apertures of the leaf discs were determined using a silicon rubber impression technique. Silicone (elite HD+ light body; Zhermack, Badia Polesine, Italy) was applied on the abaxial leaf surface and removed after hardening. For positive replicas of the impressions, the surface of the silicone was painted with a clear nail polish, which was removed after hardening. The stomatal apertures of the positive replicas were determined using light microscopy (Rezaei Nejad ) for 10 randomly selected stomata per disc.

Gas exchange measurements

Net photosynthetic rate (P), leaf intercellular CO2 concentration (Ci), and stomatal conductance (gs)were measured on one leaf per plant using a LI-6400 photosynthesis system (LI-COR). Light intensity in the cuvette was 300 μmol m–2 s–1 with a 90/10% red/blue mixture (see Supplementary Fig S1C for spectrum), except when testing for the effects of the PAR spectrum on NO-induced stomatal closure, in which case 100% red light was used. RH was set to 60%, the block temperature was 23 °C, and the CO2 concentration was set at either 50 μmol mol –1 or 380 μmol mol–1. A gas-mixing system was used for the NO treatment. Leaf samples were initially exposed to an atmosphere containing 20 mmol mol–1 O2, 380 µmol mol–1 CO2, and remainder N2. The NO treatment was started after the leaves had reached a steady state for gs. For the ABA treatment, the leaf petioles were cut under water and placed in 10-ml vials containing water before the leaves were positioned in the cuvette of the gas analyser. After reaching a steady state for gs, ABA was added to the vial to a final concentration of 100 µM.

Statistical analysis

Data were subjected to ANOVA, except for Φ PSII. Data from one plant were treated as a block in a randomized block design. Treatment means were compared using Fischer’s least-significant difference (LSD) post hoc test. Genstat 17th Edition (VSN International, Hemel Hempstead, UK) was used for statistical analysis. Because the Φ PSII data showed non-homogeneous variances, they were analysed as a complete randomized design using Welch’s ANOVA followed by the Games–Howel post hoc test, using IBM-SPSS v. 23. In all cases, P<0.05 was considered statistically significant.

Results

Stomata in epidermal peels do not close in response to NO

In a first series of experiments, stomatal responses to NO gas, SNP, and ABA were investigated in epidermal peels. Fumigation with different concentrations of NO (1–100 µmol mol–1) did not close the stomata (Table 1), and neither did exposure to different SNP concentrations (50–1000 µM) (Table 2). When the epidermal peels were placed in an ABA solution the stomatal apertures were reduced by ~50% (Table 2), indicating that the stomata had not lost their ability to close.

Table 1.

Stomatal aperture (µm) in epidermal strips of Vicia faba after 1 h fumigation with NO gas

Control	NO gas
	1 µmol mo –1	10 µmol mol–1	100 µmol mol–1
9.39 a	10. 06 a
8.64 a		8.45 a
8.54 a			8.15 a

Different concentrations of NO gas were tested in separate experiments (each including a control treatment). Each experiment was repeated three times and in each one the apertures of 30 stomata from three epidermal strips (10 per strip) were measured for each treatment. All strips in each experiment were obtained from the same plant. Means followed by different letters within a row differ significantly (P<0.05) as determined by Fisher’s LSD test.

Table 2.

Stomatal aperture (µm) in epidermal strips of Vicia faba in response to the NO-donor SNP (sodium nitroprusside) and ABA

Control	SNP			ABA (100 µM)
	50 µM	200 µM	1000 µM
7.37 b	7.60 b			4.09 a
7.56 b		7.65 b		4.01 a
7.19 b			7.78 b	3.25 a

Different concentrations of SNP were tested in separate experiments (each including a control and ABA treatment). All other details are as described in Table 1. Means followed by different letters within a row differ significantly (P<0.05) as determined by Fisher’s LSD test.

In light of the unexpected absence of SNP- and NO-induced stomatal closure, a cell-permeable fluorescent detector of NO (DAF-2DA) was applied to the epidermal strips together with the treatments. The resulting fluorescence indicated that application of NO gas, SNP, and ABA all resulted in significant accumulation of NO in guard cells (Fig. 1), while only stomata in the ABA treatment were (partly) closed. Exposure to 10 µmol mol–1 NO gas resulted in an increase in fluorescence that was comparable to application of 100 µM ABA. Exposure to 100 µmol mol–1 NO gas and treatment of the strips with 1 mM SNP solution resulted in fluorescence levels that were at least four times higher than those in the ABA treatment.

Fig. 1.

NO is accumulated in guard cells of Vicia faba after exposure of epidermal strips to ABA, NO gas, or the NO-donor sodium nitroprusside (SNP). Epidermal strips were loaded with the fluorescent NO detector DAF-2DA and treated with ABA (100 µM), nitric oxide in its pure gas form (NO; 10 µmol mol–1 or 100 µmol mol-1), or SNP (1000 µM). The presence of NO in the guard cells is measured as fluorescence. Data are means (±SE) of 10–15 replicates. Different letters indicate significant differences as determined by Fisher’s LSD test (P<0.05).

These results indicated that even when NO was clearly present, it was not a critical component in ABA-induced stomatal closure in isolated epidermal tissues of Vicia faba.

Stomata in leaf discs respond differently to NO gas, SNP, and ABA

To test whether the lack of stomatal response to NO in epidermal peels was due to the absence of mesophyll tissue, the efficiency of photosystem II (Φ PSII) and stomatal pore width were examined in leaf discs. Under the non-photorespiratory conditions at which it was measured (20 mmol mol–1 O2), Φ PSII relates to stomatal conductance (gs) if CO2 availability is the only factor limiting photosynthesis. To determine the contribution of gs to the changes in Φ PSII under the different treatments, measurements of Φ PSII were taken at physiological (380 µmol mol–1) and at extremely high (50 000 µmol mol–1) CO2 concentrations: at the latter concentration, gs does not limit photosynthesis.

NO gas (10 µmol mol–1 and 100 µmol mol–1), SNP, and ABA all caused significant reductions in Φ PSII in leaf discs (Fig. 2), and the higher SNP concentrations (200 µM and 1000 µM) resulted in similar or larger decreases in Φ PSII as observed with ABA (Fig. 2B, C). Exposure to high CO2 recovered Φ PSII in ABA-treated discs (91–93% of control) but did not recover it under NO- or SNP-treatments. This indicated that ABA decreased Φ PSII mainly because of stomatal closure, while in the case of NO gas and SNP linear photosynthetic electron transport was inhibited by non-stomatal factors. Exposure to control air (0 µmol mol–1 NO and 380 µmol mol–1 CO2) for 10–15 min after NO gas treatment recovered Φ PSII to ~90% of its initial value (‘R’ in Fig. 2A).

Fig. 2.

Efficiency of photosystem II (Φ PSII) of Vicia faba leaf discs in response to treatment with NO gas, sodium nitroprusside (SNP), or ABA. Measurements were made with 75 µmol m–2 s–1 PAR, and under 20 mmol mol–1 O2 plus the gases indicated and the remainder was N2. Data are means (±SE). Different NO and different SNP concentrations were tested in separate experiments. Each experiment was repeated three times with three replicates per treatment in each one. (A) Φ PSII after exposure to various concentrations NO gas for 10–15 min. Measurements were made under 380 µmol mol–1 CO2 except for the NO+CO2 treatment (high CO2) where 50 000 µmol mol–1 CO2 was used. Following the NO+CO2 treatment, the discs were exposed to 380 µmol mol–1 CO2 without NO for a 10–15-min recovery period, R. Different letters indicate significant differences as determined by Fisher’s LSD test (P<0.05). (B) Representative images of Φ PSII in the leaf discs as represented by false colours. (C) Φ PSII after treatment with different concentrations of SNP or 100 µM ABA. Measurements were made under 380 µmol mol–1 CO2 (open columns) or under 50 000 µmol mol–1 CO2 (filled columns). Different letters indicate significant differences as determined by Welch’s ANOVA followed by the Games–Howell post hoc test (P<0.05).

Given that the effects of NO and SNP on Φ PSII were not related to the stomata, impressions of the leaf discs were made to measure stomatal apertures directly. Some small, concentration-dependent effects on stomatal aperture were observed as a result of exposure to the three concentrations of NO gas (Table 3). At 1 µmol mol–1, the apertures were significantly wider compared with the control, at 10 µmol mol–1 there was a small, but significant reduction in aperture, and at 100 µmol mol–1 there was no significant effect of NO gas. In contrast with epidermal strips, in leaf discs the stomatal aperture was decreased by SNP by ~50% for all three concentrations tested (Table 3). ABA treatment resulted in a reduction in aperture of ~90% compared with the control, which was much more than the ~50% reduction seen in epidermal strips (Table 2) and a significantly larger reduction compared to the SNP treatments of the discs (Table 3). In addition to the effect on mean aperture, the frequency distributions of pore sizes were also different between ABA and SNP concentrations (Supplementary Fig. S2). ABA treatment resulted in a skewed distribution towards smaller apertures with a median pore size (0.64 µm) of about half of the mean (1.20 µm), while the three SNP treatments had distributions that were about normal (median values 6.3, 4.9, 5.3 and means 6.3, 4.9, 5.8 µm for 50, 200, and 1000 µM, respectively).

Table 3.

Stomatal aperture (µm) in leaf discs of Vicia faba in response to NO gas (10 min fumigation), SNP, and ABA

Control	NO gas			SNP			ABA (100 µM)
	1 µmol mol–1	10 µmol mol–1	100 µmol mol–1	50 µM	200 µM	1000 µM
11.71 c	12.69 d			6.27 b			1.52 a
11.53 d		10.87 c			4.91 b		1.32 a
10.84 c			11.05 c			5.84 b	0.76 a

Different concentrations of NO and SNP were tested in separate experiments (each including a control and ABA treatment). All other details are as described in Table 1. Means followed by different letters within a row differ significantly (P<0.05) as determined by Fisher’s LSD test.

These results indicated that stomata in leaf discs responded differently to the various treatments than stomata in epidermal peels. The effects between NO gas, SNP, and ABA were also different.

Is photosynthesis involved in the stomatal response to NO?

To examine whether the mesophyll tissue in the leaf discs influenced the stomatal response to NO through photosynthetic activity, chlorophyll fluorescence and gas exchange were measured simultaneously in intact leaves subjected to NO or ABA treatment. During 1 h of exposure to 10 µmol mol–1 NO, stomatal conductance to water vapour (gs) gradually decreased by 30%, and then slowly recovered after the NO was flushed out (Fig. 3A). With ABA feeding (100 µM), gs had decreased by 37% within 15 min, and it further decreased by 67% after 60 min. The dynamic behaviour of net photosynthetic rate (Pn) and intercellular CO2 concentration (Ci) differed from that of gs when the leaves were exposed to NO (Fig. 3B, C). Pn dropped sharply to a very low level within minutes of exposure and then remained low for the rest of the exposure time. With ABA feeding, Pn showed a logistic pattern of decrease: it declined slowly during the first 15 min, followed by a faster decrease in the next 15 min. When NO was flushed out, Pn recovered to 65, 87, and 93% of its original level after 3, 30, and 60 min of exposure to NO-free air, respectively. Ci increased sharply within 5 min of exposure to NO gas, remained at the same high level during the rest of the exposure period, and decreased quickly to its original level after removal of the gas (Fig. 3C). The time course of Ci in response to ABA treatment was similar to that of Pn.

Fig. 3.

Effects of NO gas and ABA on gas-exchange parameters in leaves of Vicia faba. Time-courses of (A) stomatal conductance for water vapour (gs), (B) net photosynthetic rate (Pn), and (C) intercellular CO2 concentration (Ci) of leaves in response to 10 µmol mol–1 NO gas or 100 µM ABA After 60 min, the NO gas was flushed out with NO-free air (arrows). Measurements were performed in an atmosphere containing 20 mmol mol–1 O2, 380 µmol mol–1 CO2, and the remainder N2. Data are means (±SE), n=14.

These results indicated that ABA induced closure of stomata that resulted in a decrease in photosynthesis due to restricted influx of CO2, whilst NO directly inhibited photosynthesis (confirming the Φ PSII measurements in leaf discs), which resulted in an instantaneous increase in Ci that was followed by stomatal closure.

To investigate a possible effect of NO on gs via Ci, we decreased Ci experimentally by lowering the CO2 concentration (in combination with NO) in the ambient gas flow. At a [CO2] of 380 µmol mol–1, gs increased during the first 9–15 min of NO application (Fig. 4A, C, Phase 1a), whilst photosynthesis sharply declined immediately after NO application (Fig. 4B). Ci initially increased, as would be expected from the inhibition of photosynthesis. After about 15 min of NO application, the stomata started to close (Fig. 4A, C, Phase 1b). The rate of closure had the form of an exponential time curve; gs changed at a mean rate of –0.10 mmol m–2 s–2 during the period 15–30 min (Fig. 4A, C, Phase 1b), followed by a rate of –0.02 mmol m–2 s–2 during the following 2-h period (Phase 1c). The external [CO2] was then lowered (20 µmol mol–1, Phase 2), and in response Ci decreased sharply to 23–25 µmol mol–1 CO2 (Fig. 4B) whilst stomatal aperture increased very slowly (at a rate of 0.005 mmol m–2 s–2). The NO gas was then removed (Phase 3), and in response the rate of stomatal opening increased (Fig. 4A, C) while Pn remained at a low level due to the low [CO2] (Fig. 4B). These results indicated that high Ci had a significant but only small negative effect on stomatal aperture, while removal of NO resulted in much faster stomatal opening compared to lowering Ci alone. For comparison, gs decreased from ~5–30 min after the start of ABA application (Fig. 4D) at a rate of about –0.30 mmol m–2 s–2 during the linear part of the logistic time curve, and thus stomata closed much faster after ABA treatment than after NO treatment (compare Fig. 4A and 4D).

Fig. 4.

(A) Time-courses of stomatal conductance (gs) for three individual Vicia faba leaves from different plants (shown in different colours). After gs reached a steady-state under 380 µmol mol–1 CO2 (time 0), NO was applied at 10 µmol mol–1 (+NO +380 CO2; Phase 1a–c). After ~150 min (indicated by a solid arrow for each leaf), the CO2 concentration was lowered to 20 µmol mol–1 (+NO +20 CO2; Phase 2). At ~190–210 min (indicated by an open arrow for each leaf), NO was flushed out (–NO +20 CO2; Phase 3). (B) Typical time-courses of net photosynthetic rate (Pn; solid line) and intercellular CO2 concentration (Ci; dashed line) for one of the leaves in (A). All measurements were performed in an atmosphere with 20 mmol mol–1 O2 and the remainder N2. (C) Mean changes in gs during the phases indicated in (A). The changes were calculated as the slope of linear regression for gs versus time during each phase. Because gs initially increased, followed by an exponential decrease during NO exposure under normal CO2, Phase 1 was sub-divided into 1a, 1b, and 1c. Data are the means (±SE) of the three leaves. (D) Time-courses of g for three individual leaves (shown in different colours) after application of 100 µM ABA. ABA application to the cut petioles started at time 0.

Confirming that stomatal closure during NO treatment was not caused by increased Ci, photosynthesis was reduced by low CO2 before the NO treatment began. At an ambient [CO2] of 50 µmol mol–1, Ci was ~40 µmol mol–1 and photosynthesis was decreased to ~2 µmol m-2 s-1 (Fig. 5). Stomata were opened wide (gs=0.46 mol m–2 s–1), probably because of the low Ci. When NO was added, photosynthesis dropped to ~0 µmol m–2 s–1, and Ci increased to 50 µmol mol–1. About 10 min from the start of the NO treatment, gs started to decrease quickly and continued to do so at the same rate as seen previously (about –0.13 mmol m–2 s–2; Phase 1b in Fig. 4A, C) for the next ~60 min.

Fig. 5.

Time-courses of stomatal conductance for water vapour (gs), net photosynthetic rate (Pn), and intercellular CO2 concentration (Ci) of a Vicia faba leaf kept under 50 µmol mol–1 CO2, 20 mmol mol–1 O2, and the remainder N2. Logging was started when gs reached a steady state (after ~90 min in the apparatus), and 10 µmol mol–1 NO was added after 2 min.

These results indicated that NO by itself induced a decrease in stomatal aperture, while an increase of [CO2] in the sub-stomatal cavities had only a limited effect on aperture.

Discussion

NO and mesophyll-driven signals for stomatal closure: sub-stomatal [CO2] is not the main factor

In our experiments, application of ABA to epidermal peels induced stomatal closure, whereas NO (gaseous or via the NO-donor SNP) did not (Tables 1, 2). However, exposing intact leaves to NO resulted in a gradual decrease of stomatal conductance over a period of 1–2 h (Figs 3–5), suggesting that NO interacted with mesophyll-driven signals for stomatal closure. Feedback of photosynthetic activity on stomatal aperture is of central importance for plants. Light acts on stomatal opening through two parallel signalling pathways: blue light specifically acts via phototropins in the guard cells, whereas red light induces opening via photosynthesis (Shimazaki ). Coordination between mesophyll photosynthesis and stomatal conductance is a well-known phenomenon (Lawson and Blatt, 2014). When leaves (intact or as discs) were exposed to NO, photosynthesis dropped instantaneously (Figs. 2A, 3B, 4B) and was not recovered by adding a very high concentration of CO2. Thus, the reduction was not due to stomatal closure (as in the case of ABA treatment). After removal of the NO gas, photosynthesis quickly reverted to its original rates (Figs. 2A, 3B). Decreases in Φ PSII and photosynthesis as a result of fumigation with NO have been reported previously (Clyde Hill and Bennett, 1970; Procházková ), and it reversibly inhibits photophosphorylation and ATP synthesis in thylakoid membranes of spinach chloroplasts (Takahashi and Yamasaki, 2002). Concurrently with the sharp decline in Pn in response to NO, there was a steep increase in Ci (Figs. 3, 4B), probably because CO2 was no longer consumed by photosynthesis. In contrast to the instantaneous changes in Pn and Ci, gs gradually decreased over 1–2 h of NO exposure. When the NO gas was removed, Pn and Ci quickly recovered to their original levels, and this was followed by a gradual increase of gs (Figs. 3, 4). These results at first suggested that stomatal closure by NO may have been induced by increased sub-stomatal [CO2], as closing of stomata due to high Ci is a well-known phenomenon (Mansfield, 1965; Allaway and Mansfield, 1967).

However, when Ci was additionally lowered by a low ambient [CO2] during the NO treatment, only minor effects on stomata were observed (Fig. 4), indicating that an increase in Ci caused by suppression of photosynthesis via NO was not the main factor. This was confirmed by decreases of gs after NO treatment when Ci was already low before the treatment started due to low ambient [CO2] (Fig. 5). Therefore, a factor other than high sub-stomatal [CO2] must be involved in mesophyll-driven signals of stomatal closure that interact with NO. This could be further examined using transformants with reduced Rubisco (Quick ): because such plants have low Pn whilst gs is still at normal levels, NO would not cause strong increases in Ci. Stomatal closure induced by elevated CO2 in tomato has been shown to be dependent on NO production in guard cells in an ABA-independent manner (Shi ).

NO- and mesophyll-driven signals for stomatal closure: no relationship with CO2 assimilation

Lawson argued that mesophyll-driven signals coordinate stomatal movements with mesophyll carbon assimilation via a complex interaction between ion channels and metabolites derived from photosynthesis. However, our results showed that NO also affected stomatal aperture when CO2 assimilation was around zero due to low ambient CO2 (Figs. 4, 5). A current hypothesis is that, rather than Ci, it is the redox state of the chloroplastic plastoquinone (QA) pool that is the photosynthetic signal that interacts with the stomatal guard cells to balance gs with Pn (Busch, 2014; Głowacka ; Kromdijk ). Limitation of CO2 supply to the Calvin–Benson cycle affects the photosynthetic electron sink capacity in the chloroplasts and the QA pool may become more reduced (Quick ), resulting in an increase of gs. NO reversibly blocks linear electron flow between PSII and PSI (Takahashi and Yamasaki, 2002), which means that there is no plastoquinone reduction; this will result in a more oxidized QA pool (Busch, 2014) and probably in a decrease of gs. Transgenic plants with differing QA redox states resulting from differing expression levels of Photosystem II Subunit S (Głowacka ) could be used to provide insights into the role of mesophyll-driven signals for NO signalling in stomatal closure and the inter-relationship of ABA and NO.

Dual dose–response relationship between NO and stomatal aperture

NO gas showed a dual dose–response effect on stomatal aperture. After 10 min exposure of leaf discs to 1 µmol mol–1 NO the stomatal aperture was increased, whilst exposure to 10 µmol mol–1 resulted in smaller apertures than the control (Table 3). An increase in gs was observed during the first 9–15 min of exposure of leaves to 10 µmol mol–1 NO, followed by a decrease thereafter (Fig. 4). This dual dose-dependent effect may be the result of two opposite effects of NO on gs: an increase due to S-nitrosylation of OST1, as reported by Wang ), and a decrease because of a more oxidized QA pool, as discussed above. It would therefore be expected that the positive effect of NO on gs at low doses would be absent in ost1 loss-of-function mutants (Mustilli ).

The effects of NO on stomatal closure in epidermal peels are controversial

The lack of stomatal closure that we observed when exposing epidermal peels to SNP or NO gas (Tables 1, 2) contradicts previous results for epidermal fragments of Pisum sativum, Salpichroa organifolia, Tradescantia species, and Vicia faba (García-Mata and Lamattina, 2001; Desikan ; Neill ; She ), whilst the response to ABA in our peels was similar or greater to what has previously been reported in V. faba (Iwai ; Yan ; Arve ). Hettenhausen also observed no reduction in stomatal aperture when applying 100 μM SNP in Nicotiana attenuate, and Sakihama even found an increase in aperture in V. faba leaf peels treated with the NO-donor S-Nitroso-N-Acetyl-D,L-Penicillamine (SNAP). These contradictions could be related to variations in the procedures of the different experiments: together with some other studies, we used a degassed (CO2-free) MES/KCl stomata-opening medium as the incubation solution, while other studies did not degas the medium, or subjected it to shaking. Furthermore, the exposure time to SNP/SNAP before the measurement of stomatal aperture varied from 20–60 min in our experiments and those of Sakihama and Hettenhausen , and from 1–3 h in the experiments of García-Mata and Lamattina (2001, Desikan , and She . Wang ) showed that elevated CO2 concentrations can induce stomatal closure in epidermal strips within about 2 h. It is likely that removing dissolved CO2 from the opening medium by degassing and/or the length of the period in which CO2 can be produced by epidermal tissue or re-dissolve from the surrounding air into the medium are factors that will influence stomatal aperture. To determine the importance of the time between degassing the bathing solution and the observations of the epidermal peels, we performed an additional experiment in which stomatal apertures were measured 1 h and 3 h after degassing, with ABA (100 µM) and SNP (200 µM) being applied to the strips immediately after degassing (Supplementary Fig. S3A). Over the 2 h-period between the measurements, the apertures decreased in the control (12%) and the SNP (19%) treatments. In the ABA treatment, the mean aperture was already small after 1 h (4.0 µm) and did not decrease over the next 2 h. In addition, the shape of the frequency distributions of stomatal apertures differed between the ABA and SNP treatments (Supplementary Fig S3C). When the bathing solutions were not degassed there was no decrease in stomatal aperture for the control, and the SNP treatment actually showed a small but significant opening of the stomata between 1–3 h (Supplementary Fig S3B).These results demonstrated the importance of taking precautions to prevent differences and/or changes in [CO2] in the bathing solution of epidermal peels.

Different outcomes between NO gas and SNP treatments of leaf discs

When leaf discs were used, NO gas showed a concentration-dependent effect on stomatal aperture, while the three SNP concentrations all resulted in a decrease in stomatal aperture of ~50% compared to the control discs (Table 3). There are several possible explanations for these different outcomes between treatment with NO gas and SNP, as follows.

First, SNP-induced stomatal closure is the result of other products in addition to the NO that is released, including CN−, as has been demonstrated for some other physiological processes. For example, gases generated by SNP have been shown to increase Arabidopsis seed germination from 0% in controls (exposed to water vapour) to 90%, while germination after exposure to NO gas is ~30% (Libourel ). Wodala demonstrated that application of an NO-scavenger (2-phenyl-4,4,5,5-tetramethylimidazoline-1-oxyl 3-oxide) only partially restored SNP-induced changes in Φ PSII, which suggests that CN− and NO are both responsible for the effect of SNP on Φ PSII. It has previously been reported that CN− affects Φ PSII (Jones ), and Zelitch (1965) mentions that CN− inhibits stomatal opening in the light and prevents closure in the dark. These results indicate that SNP is an unsuitable NO donor for studies on stomatal closure.

Second, in our study NO treatment was performed for 10 min while SNP was applied for 90 min before the stomatal apertures were measured. The strong inhibition of Φ PSII after 10 min of NO treatment demonstrated that penetration of the gas into the leaf discs was not a limiting factor during this time. However, as shown in Fig. 4, gs increased during the first 9–15 min of NO application to leaves, and stomata only started to close around the end of this period. So, the fact that stomatal closure was not observed in the leaf discs after treatment with NO was very likely due to the short exposure time.

Third, during the treatment with NO the leaf discs in the flow-through cuvette of the chlorophyll fluorescence imaging system were exposed to red light, while during the SNP treatments the discs were exposed to light from white fluorescent tubes (spectra in Supplementary Fig. S1B and S1A, respectively). It is known that stomata open in response to weak blue light under a background of strong red light and that ABA inhibits blue light-dependent opening (Shimazaki ). It is also known that NO inhibits stomatal opening specific to blue light (Zhang ). The gs time-courses shown in Figs 3 and 4 were obtained under a spectrum of 90% red plus 10% blue light. The possibility cannot be excluded that NO may be less effective in stomatal closure under 100% red light.

To get a better understanding of the limited effect of NO in stomatal closure that was found in leaf discs (Table 3), additional measurements were done in which the discs were exposed to NO gas for 2 h under an irradiance of 90% red plus 10% blue. At the end of this treatment, mean aperture was decreased by 50% compared with the control (Supplementary Fig. S4A; the bathing solution was not degassed at start to prevent an increase of [CO2] by ambient CO2). Without NO, the aperture increased under the same light and CO2 conditions over the same time period (Supplementary Fig. S4B). The effect of NO after 2 h resembled the effect of SNP on leaf discs (Table 3). In addition, the dynamics of gs of intact leaves exposed to NO were recorded under 100% red or 90% red plus 10% blue light. This demonstrated that the spectral composition interacted with the NO effect on gs (Supplementary Fig. S5): when blue light was added to the red light, the start of stomatal closing was delayed by 6 min, but the rate of closure was about twice as fast.

Taken together, it is likely that the main reason for the limited and dual effects of NO on stomatal aperture (Table 3) were due to the relatively short exposure time of the leaf discs.

Is NO a critical key factor in ABA-induced stomatal closure?

A critical role for NO in ABA-induced stomatal closure has been proposed downstream of ABA (García-Mata and Lamattina, 2001; Neill ); however, multiple conflicting results for this role have been published (see Introduction). Most data on NO production and function rely on experimental approaches based on the application of NO donors and scavengers. Inhibition of ABA-induced stomatal closure by treatment with 2-4-carboxyphenyl-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide (c-PTIO) as a NO scavenger has often used as an argument for a key role of NO in ABA-induced stomatal closure. Although c-PTIO reacts rapidly with NO, its effects are diverse and its specificity has been questioned (Pfeiffer ). Bethke demonstrated that c-PTIO prevents both cyanide- and NO-stimulated seed germination in Arabidopsis. In our experiments, the effects on stomatal closure as well as their dynamics differed between ABA and NO gas treatments. The effect of ABA was fast and was seen in epidermal peels as well as in intact leaves, while the effect of NO gas was slow (delayed and with a low rate) and was affected by the type of tissue sampled (epidermis peel or with mesophyll present) and by exposure time. In epidermis peels, none of the concentrations of NO gas that we used (1, 10, and 100 µmol mol–1) resulted in stomatal closure, while floating the peels in an ABA solution decreased the aperture to ~50% of that in the controls. As discussed above, the effects of SNP in epidermal peels in previous studies could be (partly) related to an increase of [CO2] in the bathing solution during long exposure times. When mesophyll was present (leaf discs or intact leaves), NO had a dual dose–response effect with low concentrations or short exposures increasing stomatal aperture. Upon exposure to 10 µmol mol–1 NO, stomatal conductance started to decrease after 10–15 min (depending on the spectral composition of the light; Supplementary Fig. S5), with a rate of ~20% of that of ABA-induced stomatal closure. On the other hand, after application of ABA via cut leaf petioles, stomatal conductance decreased after ~5 min and reached its minimal value after ~30 min. These results indicate that NO affects stomatal aperture but that it is not the critical key factor in rapid responses of stomata to ABA.

Conclusions

Despite NO being ubiquitously produced in plants, many details regarding its generation and signalling pathways are still unclear (Domingos ). As a reactive oxygen species, NO is an important signalling molecule under various abiotic stresses. Our results suggest that NO is not the critical key factor in ABA-induced rapid stomatal closure but that it acts independently of ABA, thus implicating NO as an effector of stomatal closure in its own right, especially in the presence of mesophyll tissue. This points to an alternative mechanism that plants use to reduce stomatal aperture to restrict water loss under stressful conditions.

Supplementary data

Supplementary data are available at JXB online.

Fig. S1. Spectra of the various light sources used in the different experiments.

Fig. S2. Frequency distribution of stomatal apertures in leaf discs as affected by NO gas, SNP, and ABA.

Fig. S3. Interactions between degassing the bathing solution, SNP treatment, and time since degassing on stomatal apertures in epidermal peels.

Fig. S4. Stomatal apertures in leaf discs after 2 h treatment with 10 µmol mol–1 NO gas.

Fig. S5. Interaction of spectral composition and NO on stomatal conductance and photosynthesis.

Click here for additional data file.