SEARCH

SEARCH BY CITATION

Keywords:

  • apoplast;
  • CO2 response;
  • Commelina communis L. ;
  • epidermis;
  • light response;
  • mesophyll;
  • photosynthesis;
  • stomata

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Previous studies have suggested that the mesophyll contributes to stomatal CO2 responses.
  • The effects of changes in CO2 concentration (100 or 700 ppm) on stomatal responses in red or white light were examined microscopically in a leaf segment, an epidermal strip and an epidermal strip placed on a mesophyll segment of Commelina communis, all mounted on a buffer-containing gel.
  • In both red and white light, stomata of the leaf segment opened/closed rapidly at low/high CO2. In red light, epidermal strip stomata barely responded to CO2. In white light, they opened at low CO2, but hardly closed at high CO2. Stomata of the epidermal strip placed on the mesophyll responded in the same manner as those on the leaf segment. Insertion of a doughnut-shaped cellophane spacer (but not polyethylene spacer) between the epidermal strip and the mesophyll hardly altered these responses. Stomata in leaf segments treated with 3-(3,4-dichlorophenyl)-1,1-dimethylurea (DCMU), a photosynthesis inhibitor, did not open in red light, but opened/closed at low/high CO2 in white light.
  • These results indicate that the apoplast transfer of ‘mesophyll signals’ and the stomatal opening at low CO2 are dependent on photosynthesis, whereas the stomatal closure at high CO2 is independent of photosynthesis.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Stomata regulate the CO2 supply for photosynthesis and transpiration in the leaf (Zeiger, 1983; Willmer & Fricker, 1996). Stomatal responses (i.e. opening and closure) are controlled by various environmental parameters, including light and CO2 concentration; however, the mechanisms of stomatal responses to these factors have not been elucidated completely.

Stomata exhibit at least two types of response to light (Kuiper, 1964; Sharkey & Raschke, 1981; Zeiger, 1983). One type is termed the blue-light response, for which the action spectrum peaks at c. 450 nm. Blue light is efficient in stomatal opening (Hsiao & Allaway, 1973; Iino et al., 1985). Blue light induces H+ extrusion from the guard cells, and thereby the membrane potential of a guard cell becomes hyperpolarized (Assmann et al., 1985; Shimazaki et al., 1986; Roelfsema et al., 2001). It is well established that phototropins, the blue-light receptors in guard cells, absorb blue light and induce stomatal opening by sequentially causing the following events: activation of the plasma membrane H+-ATPases, hyperpolarization of the plasma membrane and activation of K+ uptake channels (Kinoshita & Shimazaki, 1999; Kinoshita et al., 2001; Shimazaki et al., 2007). In addition to the regulation of H+-ATPases, blue light inhibits S-type anion channels (Marten et al., 2007). Stomata have also been shown to open in green light, even when leaf photosynthesis is inhibited (Wang et al., 2011). Because the green light used in the study by Wang et al. (2011) would not be absorbed by phototropins, certain light receptors other than phototropins might be responsible for the stomatal opening in green light. The second type of stomatal response is termed the red-light response, in which the action spectrum resembles that of photosynthesis. The red-light response is strongly inhibited by 3-(3,4-dichlorophenyl)-1,1-dimethylurea (DCMU), which is a potent inhibitor of photosynthesis. In addition, there is a strong correlation between stomatal conductance and the photosynthetic rate in red light. Accordingly, it is widely believed that photosynthesis is involved in the red-light response (Messinger et al., 2006; Wang et al., 2011). Red light also inhibits S-type anion channels when it is projected on a large area of a leaf (Roelfsema et al., 2002). It has also been shown that illumination of a single stoma and its surroundings with a red beam is not sufficient to induce stomatal opening (Mott et al., 2008).

CO2 is another environmental factor that regulates stomatal responses (Willmer, 1988). It has been shown that Nicotiana tabacum mitogen-activated protein kinase 4 (NtMPK4)-silenced plants lacking stomatal sensitivity to CO2 are less responsive to red light, indicating that both responses are interconnected (Marten et al., 2008). Guard cells in the epidermal strip respond directly to CO2 (Schwartz et al., 1988; Webb et al., 1996; Webb & Hetherington, 1997). Molecular studies using Arabidopsis thaliana have revealed that stomatal closure at high CO2 is mediated by carbonic anhydrases (CAs) localized in the plasma membrane of guard cells (Fabre et al., 2007; Hu et al., 2010; Kim et al., 2010). By contrast, several lines of evidence have indicated that the mesophyll is involved in stomatal responses to CO2.

Many studies use epidermal strips to evaluate stomatal responses to light and/or CO2. However, some studies have indicated that stomata in epidermal strips respond to light and/or CO2 much less than those in intact leaves (Lee & Bowling, 1992, 1995; Mott et al., 2008). Moreover, several studies have indicated that the role of photosynthesis in guard cells in controlling stomatal responses to red light is minor (Schwartz & Zeiger, 1984; Tominaga et al., 2001). For example, the stomatal guard cells of Paphiopedilum leeanum leaves have no chloroplasts; however, the stomata in the intact leaves open in response to red light (Nelson & Mayo, 1975). Furthermore, in Chlorophytum comosum, red-light-induced stomatal opening requires the mesophyll with active chloroplasts. The stomata over the chloroplast-less mesophyll do not respond to red light (Roelfsema et al., 2006). In a study by Mott et al. (2008), the stomata in epidermal strips exhibited a limited response to light or CO2, whereas those in epidermal strips placed on a mesophyll layer responded to light and CO2 in a manner similar to those in leaf segments (Mott et al., 2008). On the basis of these experiments, Mott et al. (2008) suggested that signals produced in mesophyll control the stomatal response. Their work is pioneering and highly suggestive in that they clearly showed the importance of the mesophyll in the stomatal response by a very straightforward method. However, in the experiment by Mott et al. (2008), the stomata in the epidermal strips might have opened widely in a hydropassive manner, as the stomata in the epidermal strips did not respond to environmental factors, such as light and CO2, but remained open. Therefore, it remains unclear whether the stomatal response in epidermal strips can be compared with that in leaves.

Sibbernsen & Mott (2010) found that stomatal opening decreased when various liquids were injected into the intercellular spaces of leaves, suggesting that mesophyll signals are gaseous. By contrast, Lee & Bowling (1993, 1995) showed that stomata responded to light when epidermal strips were floated on a solution containing illuminated mesophyll cells or chloroplasts. When the epidermis was floated on the same buffer without mesophyll or chloroplasts, the stomata did not respond to light (Lee & Bowling, 1992, 1995). Stomatal opening was also observed when epidermal strips were floated on the supernatant of a solution containing illuminated mesophyll cells (Lee & Bowling, 1992). The authors also observed that guard cell protoplasts swelled when suspended in the supernatant (Lee & Bowling, 1993). These studies indicate that mesophyll signals are aqueous.

In this study, referring to Mott et al. (2008), we devised a novel method to observe microscopically stomatal responses under more physiological conditions. By placing epidermal strips on a buffer-containing gel, rather than on a solution, we prevented the epidermal strips from being subject to extreme desiccation or hydration for up to 8 h. With this new system, we aimed to clarify whether the mesophyll plays an important role in stomatal responses to CO2 by comparing the stomatal responses of the leaf segments, epidermal strips and epidermal strips restored onto mesophyll segments. We used red light, in addition to white light containing a blue-light component. We also investigated how photosynthesis regulates stomatal responses using DCMU. Whether mesophyll signals move to the epidermis via the aqueous phase in the apoplast was further examined by inserting doughnut-shaped spacers made of polyethylene film or cellophane between the epidermal strip and the mesophyll segment.

Materials and Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Plant materials

Commelina communis L. plants were grown from seeds in a mixture of vermiculite (Nittai, Osaka, Japan) and culture soil (Metro-Mix 350; Sun Gro Horticulture, Bellevue, WA, USA) in pots (diameter, 7 cm; height, 8 cm; one seedling per pot), and watered every other day with 10−3 strength nutrient solution (Hyponex 6-10-5; Hyponex Japan, Tokyo, Japan). The plants were placed in an environmentally controlled room with a 14-h light and 10-h dark cycle at 23°C in c. 60% relative humidity. A bank of fluorescent lamps served as the light source. The photosynthetically active photon flux density (PPFD) at plant height was c. 350 μmol m−2 s−1. Fully expanded mature leaves (showing no signs of senescence) of plants grown for at least 5 wk were used for the experiments. Before preparing the samples, light-grown plants with open stomata were kept in the dark for 1 h to close stomata.

Sample preparations

Four types of sample were used. These samples were placed on a gel containing 30 mM KCl, 1 mM CaCl2, 10 mM Mes-KOH (pH 6.15) and 1% (w/v) gellangum (Wako, Osaka, Japan). When gellangum solidifies, Ca2+ acts as a cross-linker. The concentration of free Ca2+ in the gel, measured with a calcium assay kit (Metalloassay Calcium assay kit; AKJ Global Technology, Chiba, Japan), ranged from 0.4 to 0.6 mM.

Leaf segment

The mesophyll with the adaxial epidermis on the gel was enclosed by the abaxial epidermis to prevent desiccation from the cut surface of the mesophyll (Fig. 1). Therefore, the area of the abaxial epidermis in the leaf segment was greater than that of the mesophyll with the adaxial epidermis. The leaf segment (15-mm square) was prepared by cutting a leaf with a razor blade. Mesophyll with the adaxial epidermis was carefully removed from the four edges to create a square-shaped mesophyll with the adaxial epidermis (10-mm square) on the abaxial epidermis (15-mm square). The leaf segment was placed on a buffer-containing gel with the abaxial side upwards, and the free abaxial epidermis was attached to the gel. The air between the abaxial epidermis and the lateral sides of the mesophyll was removed by gently pushing the abaxial epidermis along the lateral edges of the mesophyll.

image

Figure 1. Preparation of the leaf segment sample from Commelina communis. (a) Procedures for the preparation of a leaf segment. From the adaxial side, along the dashed lines, a leaf segment (15-mm square) was shallowly cut with a razor blade. Mesophyll with the adaxial epidermis was carefully removed from the four edges to prepare a square-shaped mesophyll with the adaxial epidermis (10-mm square) on the abaxial epidermis (15-mm square). (b) The leaf segment was placed on a buffer-containing gel with the abaxial side upwards. The free abaxial epidermis was attached to the gel with a pair of forceps. The light was illuminated from the adaxial side. Bars, 10 mm.

Download figure to PowerPoint

Epidermal strip

The abaxial epidermal strips (10-mm square) were prepared using a pair of forceps (Fig. 2). The epidermal strip was first floated on a buffer containing 30 mM KCl, 1 mM CaCl2 and 10 mM Mes-KOH (pH 6.15), with the outer part of the epidermis facing upwards. The epidermal strip was gently slid onto the gel to remove the buffer occupying the substomatal cavities. When the buffer remained in the substomatal cavities, the epidermal strip appeared transparent, whereas the epidermal strip with air-filled substomatal cavities appeared whitish. After removing excess buffer, the epidermal strip was placed on a fresh gel.

image

Figure 2. Preparation of the epidermal strip sample from Commelina communis. An abaxial epidermal strip (10-mm square) was placed on a buffer-containing gel. The gel was placed on a leaf segment (20 mm square). Light was illuminated through the leaf segment and the gel. Bar, 10 mm.

Download figure to PowerPoint

Epidermal strip placed directly on the mesophyll segment

The mesophyll with the adaxial epidermis (10-mm square) was prepared by peeling off the abaxial epidermis, and was placed on the buffer-containing gel with the abaxial side upwards (Fig. 3). We placed the abaxial epidermal strip (15-mm square) on the mesophyll with the adaxial epidermis. The edges of the abaxial epidermis were attached to the gel. The epidermal strip was gently pressed against the mesophyll segment using a pair of forceps to obtain close contact.

image

Figure 3. Preparation of the epidermal strip placed on a mesophyll sample from Commelina communis. (a) Procedures for the preparation of an epidermal strip placed on a mesophyll. The abaxial epidermal strip was first floated on the buffer containing 30 mM KCl, 1 mM CaCl2 and 10 mM Mes-KOH (pH 6.15). After removing the buffer from the substomatal cavities, the epidermal strip was placed on the mesophyll with the adaxial epidermis. The epidermal strip was gently pressed against the mesophyll segment using a pair of forceps. (b) An epidermal strip placed on a mesophyll with the adaxial epidermis. Bar, 10 mm. (c) Light was illuminated from the gel side.

Download figure to PowerPoint

Epidermal strip and the mesophyll segment sandwiching a spacer

The mesophyll segment (10-mm square) was prepared by peeling off the abaxial epidermis and was placed on the buffer-containing gel (Fig. 4). A polyethylene (L-LDPE; Okura, Marugame, Japan) or cellophane (PT#300; Heiko, Haga, Japan) spacer (20-mm square; thickness, 50 μm) with a square hole (4-mm square) in the centre of the spacer was placed on the mesophyll segment. The epidermal strip was placed on the spacer. A piece of filter paper (Whatman No. 5; Whatman International, Maidstone, Kent, UK) with a circular hole (diameter, 5 mm), which had been dipped in a buffer containing 30 mM KCl, 1 mM CaCl2 and 10 mM Mes-KOH (pH 6.15), was placed on the sample to prevent desiccation. A plastic sheet with a hole (diameter, 5 mm) was placed on the filter paper to prevent excess evaporation from the filter paper.

image

Figure 4. Preparation of the spacer-inserted sample from Commelina communis. Components of the spacer-inserted sample are shown. The components were placed on the gel (left). Side view of a spacer-inserted sample (right). Light was from the gel side. A doughnut-shaped spacer was inserted between an abaxial epidermal strip and a mesophyll segment. A piece of buffer-containing filter paper and a plastic sheet were placed on this sample to prevent desiccation. The filter paper and the plastic sheet had a common hole (diameter, 5 mm). The holes were arranged to form one common hole.

Download figure to PowerPoint

DCMU treatment

The distal half of the leaf lamina was cut off and the resultant half leaf with the petiole was immersed in a solution containing 0.2 mM DCMU (Sigma-Aldrich Japan, Tokyo, Japan) and 0.1% (v/v) dimethyl sulfoxide (DMSO; Wako). The pressure was reduced to 0.08 MPa for 5 min to infiltrate the solution into the leaf; 0.1% (v/v) DMSO was used as the control. The infiltrated leaf was maintained in the dark for c. 1.5 h to allow the solution in the intercellular spaces to evaporate. The leaf section was then prepared as described above. The leaf section was placed in the chamber and maintained in the dark for > 30 min before measurements were taken.

Measurement of chlorophyll fluorescence

Chlorophyll fluorescence was measured using a pulse amplitude-modulated fluorometer (PAM-2500; Waltz, Effeltrich, Germany). The measuring light (ML; PPFD = 0.03 μmol m−2 s−1) was switched on and the minimum fluorescence (Fo) was measured. Next, a saturation pulse (SP; PPFD = 4000 μmol m−2 s−1) was provided for 0.8 s to obtain the maximum fluorescence (Fm). Actinic light (AL; PPFD = 200 μmol m−2 s−1) was then provided for 2 min to observe the Kautsky transient.

Microscopic observation system

We constructed a system to control the environment of the leaf segment or the isolated epidermis, and to observe the stomata microscopically (Fig. 5). The sample chamber consisted of two brass blocks (45 × 55 × 10 mm3). Both half-chambers had glass windows (20 × 20 × 10 mm3). By circulating water from a temperature-controlled bath, the chamber temperature was maintained at 23°C. We placed the gel with a sample in the lower chamber. The chamber with the sample was mounted on a microscopic stage, and the sample was observed under a microscope (BH2; Olympus, Tokyo, Japan) with a long focal objective lens (SLMPN × 20; working distance, 25 mm; Olympus). Digital images were obtained using a digital camera (D5100; Nikon, Tokyo, Japan) and analysed using digital image analysis software (Macromax GOKO Measure; GOKO Camera, Kawasaki, Japan).

image

Figure 5. Microscopic observation system. The chamber was mounted on the microscope stage. The sample placed on the gel was set in the chamber. The sample was illuminated with the light source from the microscope. Digital images of stomata were obtained using a digital camera attached to the microscope. Water from a temperature-controlled bath was circulated through the chamber and the reservoir. The mixed gas consisted of humidified N2 and O2, and the dew point of this mixture was controlled with a condenser. This mixture was mixed with 1% CO2 before entering the chamber. The mixed gas that passed through the chamber entered the infrared gas analyser (IRGA) for the measurement of the concentrations of CO2 and H2O.

Download figure to PowerPoint

N2 and O2 from cylinders were mixed at a ratio of 80 : 20 using mass flow controllers (Horiba Stec, Kyoto, Japan). The mixture was humidified by bubbling it in water at 23°C. The dew point of the mixture was controlled using a condenser chilled with a Peltier element; 1% CO2 in N2 was then mixed using another mass flow controller to produce air containing 100 ppm or 700 ppm CO2. The air was divided into two lines, and each was introduced to the half-chamber at a flow rate of 50 ml min−1. The sample temperature was measured using a copper-constantan thermocouple. CO2 and H2O concentrations were monitored using an infrared gas analyser (LI-840; LI-COR, Lincoln, NE, USA).

The sample was illuminated with a light source attached to the microscope. To illuminate a large area, we removed the condenser unit below the stage and fully opened the diaphragm of the lighting unit. The light from the lighting unit of the microscope, passing through a reflective filter (03SWP614; Melles Griot, Tokyo, Japan) which removes the infrared component, was used as the white light. Red light was obtained using a red-light filter (03LWP610; Melles Griot) in addition to the reflective filter. The spectra of these lights are shown in Fig. 6. In the present study, light was illuminated from the adaxial side of the sample. Therefore, the light that reached the abaxial epidermis of the leaf segment was transmitted through the leaf. To examine the stomata in the epidermal strip under similar light conditions, a leaf segment was placed under the gel (Fig. 2). We measured the apertures of the stomata whose substomatal cavities were filled with air. The stomata responded minimally when the substomatal cavities were filled with liquid.

image

Figure 6. Spectra of the red light (a) and white light (c) that illuminated the samples. The spectra in (b) and (d) are the red and white light, respectively, transmitted through the leaf of Commelina communis.

Download figure to PowerPoint

Statistical analysis

The data are shown as means ± SEM of at least three independent experiments. Differences between the mean values of the data were analysed using analysis of variance (ANOVA), Welch's t-test or Student's t-test. All statistical analyses were conducted using the R statistical software package (ver. 2.15.1.; R Development Core Team, 2003).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Novel method for observing stomatal responses in a stable physiological state

A previous study (Mott et al., 2008) has demonstrated that stomata in epidermal strips respond minimally to light or CO2. However, these data indicated that stomata in the epidermal strip were fully open from the onset of the experiment, and that the aperture size did not change throughout the experiment. We repeated this experiment, to check its reproducibility, by assessing the stomatal responses of C. communis epidermal strips. Namely, the epidermal strip was placed on a segment of filter paper containing an incubation buffer (50 mM KCl, 1 mM CaCl2 and 10 mM Mes-KOH (pH 6.15)). Because we maintained the epidermal strip in the dark for at least 1 h, stomata were closed at the onset of illumination. Stomata opened fully within 3 h of being placed in the light and at 100 ppm CO2, but gradually became insensitive to high CO2 or darkness (data not shown). Even with very high humidity in the chamber, it was difficult to keep the epidermis on the wet filter paper turgid for a long time. However, numerous studies have revealed that stomata in epidermal strips floating on aqueous buffers are able to respond to environmental changes.

We also conducted a preliminary examination of stomatal responses in an epidermal strip floated on aqueous buffer containing various concentrations of CaCl2. The epidermal strip, which had been kept in the dark for 1 h, was illuminated with white light (PPFD = 350 μmol m−2 s−1) in a growth chamber. We observed that stomata with air-filled cavities opened and that stomatal opening was most rapid in Ca2+-free buffer and slowed down with an increase in the Ca2+ concentration in the buffer (data not shown). In addition, we noticed that stomata with their substomatal cavities filled with the buffer showed a much weaker response to light than those with air-filled cavities, and that the proportion of the stomata with buffer-filled cavities increased with time. Because we wanted to examine the role of mesophyll, it was necessary to keep the substomatal cavities in the air-filled state for a longer time. Therefore, to adequately supply water to the sample, we devised a method by which we can easily keep the substomatal cavities filled with air, namely the ‘gel method’. We devised the gellangum gel containing 30 mM KCl, 1 mM CaCl2, 10 mM Mes-KOH (pH 6.15) and 1% (w/v) gellangum. When we placed the epidermal strip on the gel, most of the substomatal cavities retained air. Using this system, stomatal opening and closure in the epidermal strips placed on the gel were observed in white light and in the dark, respectively (Fig. 7e,g; Supporting Information Fig. S1). The epidermal strips did not show any symptoms of desiccation for > 8 h from the onset of the experiments. This indicates that stomata in the epidermal strip placed on the gel were able to both open and close when light and CO2 conditions were manipulated appropriately. In other words, placing the sample on the gel allowed us to observe stomatal responses in the epidermal strip in a more physiological state. In addition, using this system, we were able to observe the stomatal responses in the leaf segment for > 3 h.

image

Figure 7. Changes in stomatal aperture of the leaf segment (a, d), the epidermal strip (b, e, g) and the epidermal strip placed on a mesophyll segment in Commelina communis (c, f). The sample was illuminated with red light (RL; photosynthetically active photon flux density (PPFD) = 550 μmol m−2 s−1) or white light (WL; PPFD = 700 μmol m−2 s−1). Data are the mean ± SEM of at least 51 stomata obtained from three independent measurements. Before measurements, the samples were maintained in the dark for 30 min to close the stomata. In (a–f), the CO2 concentration was changed from 100 to 700 ppm at 2 h after the onset of illumination (dashed lines). In (g), the epidermal strip was placed in the dark at 2 h after the onset of illumination (dashed line).

Download figure to PowerPoint

CO2 responses under different light conditions

We checked whether the mesophyll regulates stomatal responses to CO2. We compared the stomatal responses in a leaf segment, an epidermal strip and an epidermal strip placed on a mesophyll segment at low CO2 (100 ppm) or high CO2 (700 ppm). We supplied red light (PPFD = 550 μmol m−2 s−1) or white light (PPFD = 700 μmol m−2 s−1) to the samples. In the experiments shown in Figs 7, 8 and 10, the CO2 concentration was changed from 100 to 700 ppm at 2 h after the onset of illumination. Before starting illumination, the samples were maintained in the dark in air containing CO2 at 100 ppm for 30 min.

image

Figure 8. Changes in the stomatal aperture of the epidermal strip placed directly on the mesophyll segment (a, d), the epidermal strip placed on the mesophyll segment with a polyethylene spacer inserted between them (b, e) and the epidermal strip placed on the mesophyll segment with a cellophane spacer inserted between them (c, f) in Commelina communis. The sample was illuminated with red light (RL; photosynthetically active photon flux density (PPFD) = 550 μmol m−2 s−1) or white light (WL; PPFD = 700 μmol m−2 s−1). Data are the mean ± SEM of at least 41 stomata obtained from three independent measurements. Before measurements, the samples were maintained in the dark for 30 min to close the stomata. The CO2 concentration was changed from 100 to 700 ppm at 2 h after the onset of illumination (dashed lines).

Download figure to PowerPoint

Figure 7 and Table 1 show changes in the stomatal aperture of the leaf segment, the epidermal strip and the epidermal strip placed on the mesophyll segment in red or white light. In red light, stomata in the leaf segment and the epidermal strip placed on the mesophyll segment opened when CO2 levels were low (Fig. 7a,c). The stomatal aperture in the leaf segment 2 h after the onset of illumination was 1.8 μm greater than that of the epidermal strip placed on the mesophyll segment. By contrast, stomata in the epidermal strip barely opened (Fig. 7b and Table 1). Stomata in the leaf segment and the epidermal strip placed on the mesophyll segment closed rapidly at high CO2 (Fig. 7a,c). The rate of stomatal closure was more rapid in the leaf segment than in the epidermal strip placed on the mesophyll segment (Table 1). In the leaf segment, rapid stomatal closure was observed within 30 min of the onset of high CO2 treatment (4.1 μm in 30 min). Because stomata in the epidermal strip only opened slightly at low CO2, we were unable to evaluate the rate of stomatal closure in the epidermal strip (Fig. 7b).

Table 1. Extent of stomatal opening and closure for the leaf segment, the epidermal strip and the epidermal strip placed on the mesophyll segment in Commelina communis (shown in Fig. 7)
 Leaf segment (μm)Epidermal strip (μm)Epidermal strip placed on mesophyll (μm)Epidermal strip Dark treatment (μm)P value
  1. RL, red light (photosynthetically active photon flux density (PPFD) = 550 μmol m−2 s−1); WL, white light (PPFD = 700 μmol m−2 s−1).

  2. 1Difference between the stomatal aperture at 2 h and the stomatal aperture at 0 h.

  3. 2Difference between the stomatal aperture at 3 h and the stomatal aperture at 2 h.

  4. Data are shown as the mean ± SEM (n ≥ 50). P values were calculated using analysis of variance (ANOVA). When ANOVA was significant at < 0.05, Tukey's multiple comparison test among samples was conducted at a significance level of < 0.05. Different lower case letters denote significant differences in Tukey's test for the data shown in each row.

RL opening15.39 ± 0.215a0.702 ± 0.0938b3.60 ± 0.199cNo data< 0.001
RL closure2−4.66 ± 0.199a−0.282 ± 0.0874b−2.08 ± 0.172cNo data< 0.001
WL opening15.59 ± 0.247a3.08 ± 0.199b4.27 ± 0.299c3.50 ± 0.229bc< 0.001
WL closure2−3.98 ± 0.181a−0.354 ± 0.0802b−2.62 ± 0.245c−2.182 ± 0.122c< 0.001

In white light, stomata in the leaf segment opened widely at low CO2. Stomata in the epidermal strip also opened at low CO2; however, the stomatal aperture at 2 h after the onset of illumination was 2.5 μm smaller than that in the leaf segment (Fig. 7d,e and Table 1). Stomata in the epidermal strip placed on the mesophyll segment opened widely at low CO2. The aperture was 1.2 μm wider than that of the epidermal strip and 1.3 μm narrower than that of the leaf segment (Fig. 7d–f and Table 1). Stomata closed rapidly in the leaf segment and the epidermal strip placed on the mesophyll segment when the CO2 concentration was increased (Fig. 7d,f). The rate of stomatal closure was more rapid in the leaf segment than in the epidermal strip placed on the mesophyll segment (Table 1). As observed in red light, rapid stomatal closure was observed within 30 min from the onset of high CO2 treatment in the leaf segment (Fig. 7a,d). By contrast, stomata in the epidermal strip barely closed within 1 h at high CO2 (Fig. 7e). To check whether stomata in the epidermal strip placed on the gel lost the ability to close, we maintained the epidermal strip in the dark. In this case, the stomata closed rapidly within 1 h (Fig. 7e,g and Table 1).

Effect of the epidermis being in contact with the mesophyll

We investigated the importance of physical contact between the epidermis and mesophyll in stomatal responses. Small molecules in the liquid may diffuse through a cellophane film, but not through polyethylene. We inserted a polyethylene or a cellophane spacer (50 μm thick) between the epidermal strip and the mesophyll segment. Aqueous substances released from the mesophyll segment would reach the epidermal strip in the cellophane-inserted sample, but not in the polyethylene-inserted sample. In these spacer-inserted samples, the stomata positioned in the centre of the spacer hole were observed.

The data in Fig. 8 and Table 2 show changes in the stomatal aperture in the epidermal strip placed directly on the mesophyll segment, the epidermal strip placed on the mesophyll segment with a polyethylene spacer inserted between them (the polyethylene spacer-inserted sample) and the epidermal strip placed on the mesophyll segment with a cellophane spacer inserted between them (the cellophane spacer-inserted sample) in red or white light. In red light, stomata opened widely in the epidermal strip placed directly on the mesophyll segment and in the cellophane spacer-inserted sample at low CO2 (Fig. 8a,c). Stomata opened 1.5 μm more widely in the cellophane spacer-inserted sample than in the epidermal strip placed directly on the mesophyll segment. By contrast, stomata opened only slightly in the polyethylene spacer-inserted sample at low CO2 (Fig. 8b and Table 2). Stomata closed rapidly in the epidermal strip placed directly on the mesophyll segment and in the cellophane spacer-inserted sample at high CO2 (Fig. 8a,c). Stomata barely closed within 1 h in the polyethylene spacer-inserted sample at high CO2 (Fig. 8b and Table 2).

Table 2. Extent of stomatal opening and closure for the epidermal strip placed directly on the mesophyll segment, the polyethylene spacer-inserted sample and the cellophane spacer-inserted sample in Commelina communis (shown in Fig. 8)
 Epidermal strip placed on mesophyll (μm)Polyethylene spacer (μm)Cellophane spacer (μm)P value
  1. RL, red light (photosynthetically active photon flux density (PPFD) = 550 μmol m−2 s−1); WL, white light (PPFD = 700 μmol m−2 s−1).

  2. 1Difference between the stomatal aperture at 2 h and the stomatal aperture at 0 h.

  3. 2Difference between the stomatal aperture at 3 h and the stomatal aperture at 2 h.

  4. Data are shown as the mean ± SEM (n ≥ 40). P values were calculated using analysis of variance (ANOVA). When ANOVA was significant at < 0.05, Tukey's multiple comparison test among samples was conducted at the significance level of < 0.05. Different lower case letters denote significant differences in Tukey's test for the data shown in each row.

RL opening13.18 ± 0.145a0.733 ± 0.108b4.72 ± 0.335c< 0.001
RL closure2−1.55 ± 0.111a−0.0291 ± 0.0766b−2.28 ± 0.189c< 0.001
WL opening13.61 ± 0.262a3.56 ± 0.171a5.61 ± 0.230b< 0.001
WL closure2−2.29 ± 0.185a−0.126 ± 0.0913b−2.70 ± 0.120a< 0.001

In white light, the stomata opened widely in all samples (Fig. 8d–f ). The stomata opened most widely in the cellophane spacer-inserted sample (5.61 μm at 2 h after the onset of illumination). Stomata closed rapidly in the epidermal strip placed directly onto the mesophyll segment and in the cellophane spacer-inserted sample at high CO2 (Fig. 8d,f). Stomata barely closed within 1 h in the polyethylene spacer-inserted sample at high CO2 (Fig. 8e and Table 2).

Stomata in the polyethylene spacer-inserted sample responded to CO2 in a manner similar to stomata in the epidermal strip, regardless of light colour (Fig. 7b vs Fig. 8b, Fig. 7e vs Fig. 8e).

Effects of photosynthesis on stomatal responses

To investigate the role of photosynthesis on the stomatal responses, we used leaves treated with 0.2 mM DCMU in an aqueous solution of 0.1% (v/v) DMSO. The control leaves were treated with 0.1% (v/v) DMSO.

Chlorophyll fluorescence transients in these leaves are shown in Fig. 9. Chlorophyll fluorescence in the control leaf increased sharply in actinic light and then decreased gradually, exhibiting the typical Kautzky transient (Fig. 9a). Chlorophyll fluorescence in the DCMU-treated leaf increased to the level attained by the saturating flash, and remained at that level for 2 min, indicating the complete inhibition of photosynthetic electron transport (Fig. 9b).

image

Figure 9. Changes in chlorophyll fluorescence from the control leaf (a) and the 3-(3,4-dichlorophenyl)-1,1-dimethylurea (DCMU)-treated leaf of Commelina communis (b). Measuring light (ML; photosynthetically active photon flux density (PPFD) = 0.03 μmol m−2 s−1) was switched on and the minimum fluorescence (Fo) was measured. Next, the sample was illuminated with the saturation pulse (SP; PPFD = 4000 μmol m−2 s−1) for 0.8 s and the maximum fluorescence (Fm) was measured. Finally, the sample was illuminated with actinic light (AL; PPFD = 200 μmol m−2 s−1) for 2 min to observe the Kautsky transient.

Download figure to PowerPoint

Changes in the stomatal aperture in the control leaf segment and the DCMU-treated leaf segment in red and white light are shown in Fig. 10 and Table 3. In red light, stomata in the control leaf segment opened rapidly at low CO2 and closed rapidly at high CO2 (Fig. 10a). In comparison, stomata in the DCMU-treated leaf segment barely opened at low CO2 (Fig. 10b and Table 3). As the stomata barely opened in the DCMU-treated leaf segment at the onset of high CO2 (shown in Fig. 10b), it was difficult to determine whether the stomata in the DCMU-treated leaf segment closed.

Table 3. Extent of stomatal opening and closure between the control leaf segment and the 3-(3,4-dichlorophenyl)-1,1-dimethylurea (DCMU)-treated leaf segment in Commelina communis (shown in Fig. 10)
 Control leaf segment (μm)DCMU-treated leaf segment (μm)P value
  1. RL, red light (photosynthetically active photon flux density (PPFD) = 550 μmol m−2 s−1); WL, white light (PPFD = 700 μmol m−2 s−1).

  2. 1Difference between the stomatal aperture at 2 h and the stomatal aperture at 0 h.

  3. 2Difference between the stomatal aperture at 3 h and the stomatal aperture at 2 h.

  4. Data are shown as the mean ± SEM (n ≥ 54). P values were calculated using Welch's t-test (RL and WL open) and Student's t-test (WL closed).

RL opening13.03 ± 0.1300.175 ± 0.0643< 0.001
RL closure2−2.32 ± 0.134−0.223 ± 0.0578< 0.001
WL opening14.69 ± 0.2102.61 ± 0.166< 0.001
WL closure2−3.04 ± 0.161−2.33 ± 0.2000.00638
image

Figure 10. Changes in the stomatal aperture in the control leaf segment (a, c) and the 3-(3,4-dichlorophenyl)-1,1-dimethylurea (DCMU)-treated leaf segment (b, d, e) in Commelina communis. The sample was illuminated with red light (RL; photosynthetically active photon flux density (PPFD) = 550 μmol m−2 s−1) or white light (WL; PPFD = 700 μmol m−2 s−1). Data are the mean ± SEM of at least 55 stomata obtained from three independent measurements. Before measurements, the samples were maintained in the dark for 30 min to close the stomata. In (a–d), the CO2 concentration was changed from 100 to 700 ppm at 2 h after the onset of illumination (dashed lines). In (e), the CO2 concentration was maintained at 100 ppm throughout the measurements.

Download figure to PowerPoint

In white light, stomata in the control leaf segment opened and closed rapidly (Fig. 10c). Stomata in the DCMU-treated leaf segment opened at low CO2 (Fig. 10d,e); however, the aperture was smaller than that in the control leaf segment (Table 3). When the CO2 concentration was maintained at 100 ppm, the DCMU-treated leaf segment stomata remained open for at least 3 h (Fig. 10e). Stomata in the DCMU-treated leaf segment closed rapidly at high CO2 (Fig. 10d).

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

In this study, we constructed a system to control the environment of leaf segments or isolated epidermis and to observe stomatal responses to CO2 (100 ppm or 700 ppm) on these materials microscopically. By using the buffer-containing gel rather than aqueous buffers, we were able to observe the stomatal response in a reasonably physiological state for a long time. We compared stomatal responses in a leaf segment, epidermal strip and epidermal strip placed on a mesophyll segment in red light or white light. The present results clearly indicate that the mesophyll is important for both stomatal opening and closure, supporting the existence of ‘mesophyll signals’, signals which have been proposed in previous studies (Lee & Bowling, 1992, 1993, 1995; Mott et al., 2008). We also investigated whether mesophyll signals are gaseous by inserting a polyethylene or cellophane spacer between the epidermal strip and the mesophyll segment. By comparing the stomatal responses between the polyethylene-inserted sample and the cellophane-inserted sample, we concluded that mesophyll signals are aqueous. We then treated the leaf with DCMU to inhibit photosynthesis throughout the whole leaf. It has been suggested that mesophyll signals inducing stomatal opening are dependent on photosynthesis at low CO2, whereas those inducing stomatal closure are independent of photosynthesis at high CO2.

Conditions for observing stomatal responses

Previous studies have reported that 1 mM CaCl2 in the incubation buffer tends to induce stomatal closure (De Silva et al., 1985; Schwartz, 1985; McAinsh et al., 1995). When the epidermal strip was placed on the gel, the epidermal cells were closely associated with the gel, whereas the guard cells located above the air-filled substomatal cavities were not directly associated with the gel. Epidermal cells are closely associated with the mesophyll in intact leaves, with mesophyll apoplastic Ca2+ concentrations probably ranging from 0.1 to 1 mM (Sattelmacher, 2001). The apoplastic Ca2+ concentrations in epidermal cell walls that are closely associated with the mesophyll are expected to be in the same range as that of the mesophyll apoplast. Therefore, in our system, 0.4–0.6 mM free Ca2+ in the gel would not severely inhibit stomatal opening. Certainly, stomatal opening was severely inhibited in white light when the substomatal cavities were filled with buffer from the gel and the guard cells were directly associated with the buffer (data not shown). We also investigated stomatal responses in the samples placed on a Ca2+-free gel (Fig. S2). In this gel, Mg2+ was used as cross-linker for the solidification of gellangum. With this gel, the stomatal opening was moderately speeded up, especially in white light (compare Fig. 7 and Fig. S2). In the epidermal strip, the induction of stomatal closure at high CO2 was slower under the Ca2+-free condition than under the 1 mM Ca2+ condition (compare Fig. 7e and Fig. S2e). Although the absolute value of the stomatal aperture and the speeds of opening and closure differed depending on the presence of Ca2+ in the gel, both stomatal opening and closure were greatly accelerated when the epidermal strip was placed on the mesophyll. Thus, we concluded that the extracellular Ca2+ concentration did not influence markedly the tendencies of stomatal responses, and the role of mesophyll in stomatal responses could be assessed adequately with 1 mM CaCl2 buffer-containing gel.

Ethylene is expected to arise from the cut surface of samples (Boller & Kende, 1980). Although ethylene is known to affect the regulation of the stomatal aperture, its effect remains unclear. In some species, ethylene induces stomatal closure (Desikan et al., 2006), whereas, in others, it mediates auxin-induced stomatal opening (Merritt et al., 2001). In our system, when compared with other environmental factors, the effects of ethylene on stomatal responses are considered to be small, because stomata in each sample responded to light and/or CO2 as in intact leaves (data not shown). If the effects of ethylene were stronger than those of other environmental factors, the effects of ethylene would be superimposed on the physiological responses to light and/or CO2.

Stomatal opening

Previous studies have supported the contention that photosynthesis in both guard cells and mesophyll is involved in the regulation of stomatal responses (Sharkey & Raschke, 1981; Messinger et al., 2006; Wang et al., 2011). It is also known that chloroplasts in guard cells exhibit high photosynthetic electron transport activity (Lawson et al., 2002, 2003; Lawson, 2009). When a leaf is illuminated with red light, photosynthesis occurs in both the guard cells and the mesophyll. The results of the current study showed that stomata in the epidermal strip of C. communis barely opened in red light at low CO2. By contrast, stomata in the leaf segment opened widely (Fig. 7a,b and Table 1). However, when the epidermal strip was placed on the mesophyll segment, stomata opened widely (Fig. 7c). Moreover, stomata in the leaf segment that had been treated with DCMU, which is a potent inhibitor of photosynthesis, barely opened in red light at low CO2 (Fig. 10b and Table 3). These findings indicate that stomatal opening in red light is more strongly dependent on photosynthesis in the mesophyll than in guard cells.

Unlike the stomatal response to red light, stomata in the epidermal strip opened widely when illuminated with white light (compare Fig. 7b and e). The blue-light receptors, phototropins, are localized in the guard cells of A. thaliana, and it has been established that blue light excites these phototropins and leads to stomatal opening (Kinoshita et al., 2001; Shimazaki et al., 2007). It is highly likely that the guard cells of C. communis also contain phototropins (Iino et al., 1985); hence, when the leaf was illuminated with white light, the blue component in white light strongly induced stomatal opening (Fig. 6d).

Stomata in the leaf segment treated with DCMU opened when illuminated with white light (Fig. 10d,e). When stomata open, H+-ATPase in guard cell plasma membranes consumes cytosolic ATP (Shimazaki et al., 2007). As photosynthesis was inhibited in the DCMU-treated leaf segment, the ATP for H+-ATPases would be produced by respiration, with carbohydrates stored in the guard cell being consumed (Mawson, 1993). However, stomata in the leaf segment treated with DCMU did not open in red light (Fig. 8b), even though carbohydrates would be available. Although blue light activates H+-ATPases in guard cell protoplasts (without mesophyll cells), red light has no such function in guard cell protoplasts (Taylor & Assmann, 2001). Hence, it is probable that mesophyll signals induced stomatal opening via the activation of H+-ATPase in guard cells.

In addition to the regulation of H+-ATPases, the regulation of S-type anion channels in guard cells would be important for fast stomatal opening. In intact plants, the inhibition of S-type anion channels in guard cells occurs in red light or at low CO2 (Roelfsema et al., 2002; Marten et al., 2008). Loss of SLAC1, a major S-type anion channel in guard cells, leads to slow stomatal closure in the dark (Negi et al., 2008; Vahisalu et al., 2008). In other words, the deactivation of S-type anion channels would lead to fast stomatal opening in the light. Mesophyll signals at low CO2 in the light may deactivate S-type anion channels, and speed up stomatal opening.

Stomatal closure

Although stomata in the leaf segment closed rapidly in white light and at high CO2, stomata in the epidermal strip barely closed within 1 h under these conditions (Fig. 7d,e). Previous studies have indicated clearly that stomata in the epidermal strip close at high CO2 (Schwartz et al., 1988; Webb et al., 1996; Webb & Hetherington, 1997; Hu et al., 2010). To check whether this phenomenon occurred in our system, the epidermal strip was maintained at high CO2 for > 1 h. The stomata on the epidermal strip closed slowly from an aperture of 4.6 to 2.8 μm over a 3-h period (Fig. S1). The degree of stomatal closure was comparable with that observed in previous studies. In addition, when the epidermal strip was maintained in the dark, stomata closed rapidly within 1 h (Figs 7g, S1, Table 1). Therefore, the stomata in the epidermal strip had the potential to close; the closure under high CO2 conditions was not an experimental artefact. By contrast, when the epidermal strip was placed on the mesophyll segment, stomata closed rapidly (Fig. 7f ). These data indicate that the mesophyll contributes to the rapid induction of stomatal closure at high CO2. Stomata in the leaf segment treated with DCMU closed rapidly in white light at high CO2 in a manner similar to that of the control (Fig. 10c,d). Hence, we questioned what was responsible for the rapid stomatal closure in the DCMU-treated leaf segment. As photophosphorylation was inhibited in the DCMU-treated leaf segment, the ATP required for stomatal opening would gradually decline. However, stomata in the DCMU-treated leaf segment continued to open gradually for at least 3 h when the CO2 concentration was maintained at 100 ppm (Fig. 10e). On the basis of this information, it is unlikely that stomatal closure in the DCMU-treated leaf segment at high CO2 is caused by a shortage of ATP. It is also unlikely that stomatal closure at high CO2 is dependent on mesophyll photosynthesis.

In contrast with the effects of low CO2, S-type anion channels were activated at high CO2 (Roelfsema et al., 2002). Mesophyll signals at high CO2 may cause the activation of S-type anion channels for fast stomatal closure.

Abscisic acid (ABA) is one of the candidates of the mesophyll signal that might induce stomatal closure. CA catalyses the reversible reaction of CO2 + H2[LEFT RIGHT ARROW] H+ + inline image, with this enzyme also being present in mesophyll cells (Badger & Price, 1994). When ambient CO2 concentrations are high, CO2 concentrations in mesophyll cells inevitably increase. Then, the equilibrium favours the production of H+ and inline image. Thus, the pH inside the mesophyll cells becomes more acidic. Under these acidic conditions, ABA would exist in its uncharged form (ABAH), which is able to diffuse across the cell membrane (Slovik et al., 1992). Therefore, when CO2 concentrations are high, ABA would be released from the mesophyll. Thus, ABA moves from the mesophyll to the guard cells, in turn inducing stomatal closure. Further studies are required to test this hypothesis.

Mesophyll signals move from the mesophyll to the epidermis via the apoplast

Our data demonstrate that the mesophyll plays a critical role in controlling stomatal aperture. Like previous studies, we also propose the existence of ‘mesophyll signals’ (substances controlling stomatal aperture), which are released from the mesophyll and move towards the epidermis (Lee & Bowling, 1992, 1993, 1995; Mott et al., 2008). Sibbernsen & Mott (2010) found that stomata close rapidly when liquids are microinjected into the intercellular spaces of leaves. As the transfer of gaseous substances from the mesophyll to the epidermis was blocked by microinjection, the authors suggested that mesophyll signals are gaseous. However, the results of Sibbernsen & Mott (2010) would not eliminate the involvement of aqueous mesophyll signals. Our study showed that mesophyll signals inducing stomatal opening are probably produced during the course of photosynthesis and are released from the mesophyll. For photosynthesis, a CO2 supply to the mesophyll is needed (Fig. 10a,b). When the intercellular space floods, the CO2 supply to the mesophyll is blocked, which would consequently suppress the production of both gaseous and aqueous mesophyll signals. It is also possible that the liquid occupying the substomatal cavities suppresses respiration in guard cells, inhibiting the O2 supply from the intercellular space to the guard cell.

We examined whether the mesophyll signals are gaseous. Small molecules in the liquid may diffuse across a cellophane film, but not across a polyethylene one. We inserted spacers made of these films between the epidermis and the mesophyll, and investigated whether the mesophyll signals regulating stomatal responses were aqueous. In our system, the mesophyll segment was sandwiched by the gel and the doughnut-shaped spacer (Fig. 4, right). As described in the section on ‘Stomatal opening’, stomatal opening was strongly dependent on photosynthesis. With the spacer having no holes, CO2 for photosynthesis cannot be supplied to the mesophyll, and thereby the release of the mesophyll signals would be inhibited. Indeed, when we inserted the spacer without a hole, we could not observe stomatal opening in red light (data not shown). Therefore, we used the ‘doughnut-shaped’ spacers to deliver CO2 through the stomata in the epidermis to the mesophyll (Fig. 4).

When the doughnut-shaped polyethylene spacer (50 μm thick) was inserted between the epidermis and the mesophyll, the stomata could not respond to CO2 (Fig. 8b,e). However, when the doughnut-shaped cellophane spacer (50 μm thick) was inserted between the epidermis and the mesophyll segment, the stomata opened at low CO2 (Fig. 8c) and closed at high CO2 (Fig. 8f). This indicates that the stomata in the cellophane spacer-inserted samples respond rapidly to CO2 in a manner similar to that of the stomata in the leaf segment (Fig. 7a,d). On the basis of these results, the mesophyll signals inducing both stomatal opening and closure appear to move from the mesophyll to the epidermis via the aqueous phase in the apoplast. If gaseous mesophyll signals were present, the effects on the regulation of stomatal aperture would be less important relative to those of the aqueous mesophyll signals.

In summary, the data presented in this study indicate that both stomatal opening and closure are strongly regulated by aqueous signals from the mesophyll. Stomatal opening is dependent on mesophyll photosynthesis, whereas stomatal closure is less dependent on mesophyll photosynthesis. In an effort to identify such mesophyll signals, we are currently devising a method to collect the mesophyll apoplastic solution.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

We are grateful to Mr S. Otsuka (Workshop, Department of Physics, University of Tokyo, Tokyo, Japan) for manufacturing the condenser in the microscopic observation system. This work was supported by a Sasagawa Scientific Research Grant from The Japan Science Society, a Grant-in-Aid for Scientific Research on Innovative Areas (no. 21114007) and a Grant-in-Aid for JSPS Fellows (no. 12J08951).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office.

FilenameFormatSizeDescription
nph12261-sup-0001-FigS1-S2.pdfapplication/PDF82K

Fig. S1. Stomatal responses in the epidermal strip of Commelina communis.

Fig. S2. Stomatal responses with the Ca2+-free gel.