Guard cells are electrically isolated from other plant cells and therefore offer the unique possibility to conduct current- and voltage-clamp recordings on single cells in an intact plant. Guard cells in their natural environment were impaled with double-barreled electrodes and found to exhibit three physiological states. A minority of cells were classified as far-depolarized cells. These cells exhibited positive membrane potentials and were dominated by the activity of voltage-dependent anion channels. All other cells displayed both outward and inward rectifying K+-channel activity. These cells were either depolarized or hyperpolarized, with average membrane potentials of −41 mV (SD 16) and −112 mV (SD 19), respectively. Depolarized guard cells extrude K+ through outward rectifying channels, while K+ is taken up via inward rectifying channels in hyperpolarized cells. Upon a light/dark transition, guard cells that were hyperpolarized in the light switched to the depolarized state. The depolarization was accompanied by a 35 pA decrease in pump current and an increase in the conductance of inward rectifying channels. Both an increase in pump current and a decrease in the conductance of the inward rectifier were triggered by blue light, while red light was ineffective. From these studies we conclude that light modulates plasma membrane transport through large membrane potential changes, reversing the K+-efflux via outward rectifying channels to a K+-influx via inward rectifying channels.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Guard cell ion-transport has been studied with single cells, using either the patch-clamp technique on protoplasts or microelectrodes on cells in epidermal strips (Hedrich and Roelfsema, 1999). The development of these techniques broke new grounds in the analysis of the biophysical properties of ion-transport proteins (Hedrich and Schroeder, 1989). New insights were obtained into the regulation of ion-transport by physiological factors, such as hormones (Assmann, 1993). Guard cell responses to changes in the light conditions, however, received little attention, although light is a potent promoter for stomatal opening.
Opening of stomata is induced by blue as well as red light, but both light qualities act through different signalling pathways (Assmann and Shimazaki, 1999). The red light effect can be inhibited by 3-(3, 4-dichlorophenyl)-1-1-dimethyl-urea (DCMU) (Schwartz and Zeiger, 1984) and therefore is likely to be mediated by guard cell photosynthesis. In the presence of red light, blue light further stimulates stomatal opening. The blue light receptors that have been found in Arabidopsis thaliana appear not to be involved in this guard cell response (Lascève et al., 1999), while the involvement of zeaxanthin is currently under debate (Eckert and Kaldenhoff, 2000). Guard cells of the zeaxanthin free mutant, npq1, were reported to lack a blue light-specific response (Frechilla et al., 1999). However, a blue light-specific stimulation of the transpiration was still observed with npq1 plants (Eckert and Kaldenhoff, 2000).
Stomatal opening is the result of osmotic swelling of two guard cells that surround the stomatal pore. The osmotic swelling is accomplished by the accumulation of K+-salts within the cell. During stomatal opening, guard cells of Vicia faba increase their K+-concentration from values close to 0.1 m to values exceeding 0.5 m (MacRobbie, 1987). These changes in the K+-concentration are most likely mediated by K+-selective channels in the guard cell plasma membrane (Schroeder et al., 1984).
Two types of K+-channels have been identified in the guard cell plasma membrane. Inward rectifying K+-channels activating at hyperpolarized membrane potentials, and outward rectifying channels activating at more depolarized potentials (Blatt, 1988; Blatt, 1992; Schroeder et al., 1987). The inward channels are suitable for the uptake of K+, since the negative membrane potential at which the channel opens can drive an influx of K+. Likewise, outward channels enable K+-extrusion since these channels are active at potentials favouring K+-efflux (Ache et al., 2000; Blatt, 1991; Schroeder, 1988). Although the K+-uptake may be facilitated by inward K+-channels, the uptake is energized by H+-ATPases in the plasma membrane. The H+-ATPases extrude H+, thereby creating a H+-gradient and an electrical potential difference across the plasma membrane (Lohse and Hedrich, 1992).
The H+-ATPase is activated by blue light pulses superimposed on a background of red light (Shimazaki et al., 1986). This activation is mediated via a blue light-dependent phosphorylation of the C-terminus of H+-ATPases (Kinoshita and Shimazaki, 1999). Blue light-induced stimulation of the H+-ATPase can be measured in the whole-cell patch-clamp mode, as a change of 3 pA in plasma membrane current (Assmann et al., 1985; Taylor and Assmann, 2000). Larger blue light-induced currents up to 20 pA were found, using a slow-whole-cell mode (Assmann, 1993; Schroeder, 1988). This indicates that the amplitude of the blue light response depends on diffusible cytoplasmic components. Although most reports have focused on the blue light-specific stimulation of H+-ATPases, red light-induced outward currents have been measured in guard cells too (Serrano et al., 1988). In these experiments, a high intensity red light was found to trigger an outward current of 2 pA.
Guard cells in epidermal strips, impaled with microelectrodes, were found in two physiological states. Cells in the depolarized state have membrane potentials close to the K+-Nernst potential and extrude K+, while hyperpolarized cells have more negative potentials driving K+-uptake (Roelfsema and Prins, 1998; Thiel et al., 1992). Guard cells can shift from the depolarized to the hyperpolarized state and vice versa. These membrane potential changes were autonomous, rapid and often repetitive and therefore classified as oscillations (Gradmann et al., 1993). Recently, these membrane potential changes were linked to cytoplasmic Ca2+ concentration waves (Grabov and Blatt, 1998). The phytohormone ABA promotes guard cells to enter the depolarized state (Blatt and Armstrong, 1993), while IAA has the opposite effect (Blatt and Thiel, 1994). Based on these results, light-induced changes from one state to the other are to be expected. Although light has been shown to affect the membrane potential (Assmann et al., 1985; Moody and Zeiger, 1978), light-induced changes from one to the other state have not been described so far.
Here, we present the first study in which guard cells were impaled in their natural environment, the intact plant. Consequently, the impaled guard cells are still neighboured by viable and turgerescent epidermal and mesophyll cells. Furthermore, the extracellular ion concentration will closely resemble that of intact leaves. In this configuration, we could measure large light-induced membrane potential changes and determine their consequence for K+-transport across the plasma membrane.
Current- and voltage clamp of guard cells in intact plants
Mature guard cells do not possess functional plasmodesmata (Wille and Lucas, 1984) and are electrically isolated from other cells. Due to this unique property, guard cells within the intact plant are suitable for current- and voltage clamp studies. However, before these techniques can be applied, a proper connection between a reference electrode and the guard cell wall has to be established.
In the present experiments, guard cells in the intact leave were visualized with a water immersion objective. The reference electrode was placed in the solution between the objective and the epidermis. Using blunt electrodes that are in contact with the guard cell wall via an open stoma, a potential difference between the guard cell wall and reference electrode could be measured. This surface potential was sensitive to changes in the light intensity and could reach values up to 50 mV in older leaves. Consequently, we used the youngest fully unfolded leaves in our experiments, which had an average surface potential of −0.6 mV (SD 7, n = 7). Switching the microscope lamp on or off induced transient changes in the surface potential, with an average of 8 mV (SD 4, n = 7) for light off and 8 mV (SD 6, n = 7) for light on. Voltage-clamp measurements may be hampered further by the resistance between the guard cell wall and reference electrode. To get an estimate of this resistance, blunt double-barreled electrodes were brought into contact with the guard cell wall, via open stomata. The series resistance measured under current clamp was 4.9 MΩ (SE 0.6, n = 7) in the light and 4.3 MΩ (SE 0.7, n = 7) after 10 min in the dark. With membrane currents up to 2 nA, the applied voltage was thus maximally overestimated by 10 mV. This overestimation of the clamped voltage, however, does not alter with changes in the light intensity. When Vicia faba leaves were mounted on the microscope table, stomata on the abaxial leaf side were initially open, but closed within the following 15 min. This response may well be due to an increase in the cytoplasmic Ca2+ concentration, known to occur in plant cells during mechanical manipulation (Haley et al., 1995). Thereafter, stomata opened again allowing the impalement of single guard cells with microelectrodes. In plant cells, the tip of a microelectrode is normally located in the cytoplasm rather than in a vacuole (Felle, 1988), a feature of plant cells that also holds for guard cells in epidermal strips (Blatt et al., 1990; McAinsh et al., 1990). This subcellular localization was confirmed for guard cells in intact plants by pressure injection of lucifer-yellow (Figure 1). After dye injection, fluorescent light was emitted from the body of the cell, rather than from one of the vacuoles.
Three states of V. faba guard cells in intact plants
Based on their free-running membrane potential and plasma membrane conductance, individual guard cells could be divided into three groups (Figure 2). The plasma membrane of cells in all three groups depolarized immediately after impalement, while they differed in the magnitude and velocity of the subsequent repolarization (Figure 2a). Membrane potentials of cells in the first group slowly repolarized but remained positive (Figure 2a, upper graph). When the plasma membrane of these cells was clamped from a holding potential of −140 mV to more positive values, activation of inward conducting channels was found (Figure 2b, upper graph). The inward currents could be separated into an instantaneous and a slowly activating component. The currents activated at potentials positive of −120 mV and reversed at positive potentials (Figure 2c, upper graph). This voltage dependence is similar to that of anion channels previously identified in guard cells (Hedrich et al., 1990; Keller et al., 1989; Linder and Raschke, 1992; Schroeder and Keller, 1992). Out of 117 cells impaled, 10 had positive membrane potentials and conductance properties similar to those in the upper graphs of Figure 2. These cells were classified as ‘far-depolarized’ cells.
The majority of cells (66 of 117 cells measured) were characterized by free-running membrane potentials that repolarized to moderately negative values (Figure 2a, middle graph); the mean membrane potential after 5 min was −41 mV (SD 16). Clamping the plasma membrane of these cells from −100 mV to values positive of −60 mV resulted in the activation of outward rectifying channels, while activation of inward rectifying channels was found at potentials negative of −120 mV (Figure 1b, middle graph). The activation kinetics were similar to those of outward- and inward-rectifying K+-channels, previously identified in guard cells (Blatt, 1988; Schroeder et al., 1987). The reversal potential, determined by a tail current analysis (data not shown), was −56 mV for outward channels (SE 5, n = 11) and −73 mV (SE 2, n = 25) for inward channels. A difference in the reversal potential of both channels was not expected, since both channels were reported to be highly K+-selective (Blatt and Gradmann, 1997; Blatt, 1988; Blatt, 1992). However, the selectivity of outward rectifying channels in A. thaliana guard cells was pH dependent (Roelfsema, 1997). In intact V. faba leaves the extracellular pH ranges from 5.3 to 5.7 (Felle et al., 2000), while the selectivity of outward K+ channels has been determined at pH 7.4 (Blatt and Gradmann, 1997; Blatt, 1988). In intact leaves, the outward rectifying channels therefore may be less K+-selective than determined previously. The Nernst potential of K+ is most likely close to the reversal potential of inward rectifying channels, which do not have a pH dependent selectivity (Blatt, 1992).
The activation of outward and inward channels was also apparent in the current-voltage plot (Figure 2c, middle graph). Note that only small current changes were found between −120 and −60 mV. The current-voltage curve intersected the zero current axis at −50 mV, close to the free-running membrane potential of this cell. In accordance with previous reports (Roelfsema and Prins, 1997) these cells were classified as ‘depolarized cells’.
A third group of cells (41 of 117 cells measured) reached more negative membrane potentials; their average membrane potential was −112 mV (SD 19) (Figure 2a, lower graph). The plasma membrane conductance of these cells was similar to that of depolarized cells (Figure 2b, lower graph). However, the current-voltage curve was shifted to more positive current values (Figure 2c, lower graph). As a result, the current-voltage curve intersected the zero current axis at −110 mV, a value slightly negative of the threshold potential of inward rectifying channels. These cells were therefore classified as ‘hyperpolarized cells’.
Light-induced membrane potential changes
After the characterization of the variations in membrane properties, guard cells were exposed to changes in light intensity by switching the microscope lamp on and off. Cells that were depolarized in the light mostly responded with a transient depolarization of 8 mV (SD 8, n = 16), both to light on or off signals (Figure 3a). These cells repolarized to the original membrane potential after 10 min. The membrane potential change in these cells was probably related to changes in the surface potential.
Guard cells that were hyperpolarized in the light were much more light responsive. Switching off the microscope lamp resulted in a large depolarization in 9 out of the 11 cells tested, the average membrane potential change was 64 mV (SD 12). These cells could repeatedly switch from the hyperpolarized state in the light to the depolarized state in the dark and vice versa (Figure 3b). As a result of these membrane potential changes the electrochemical gradient of K+ reverses, causing a dramatic change in the K+-flux across the plasma membrane. In the light, K+ is taken up through inward rectifying channels, while K+ is released through outward rectifying channels in the dark. Out of the 11 cells tested, 2 cells did not depolarize, but remained hyperpolarized in the dark. This indicates that factors other than light can also induce hyperpolarized membrane potentials.
Conductance changes at the plasma membrane
Light-induced conductance changes were measured by clamping the plasma membrane from a holding potential of −100 mV to test potentials ranging from −200 to 20 mV. The conductance was measured before, during and after 15 min exposure to darkness. The mean current-voltage relation in the light was compared to that in the dark. Figure 4 shows a compilation of current-voltage curves of 5 cells, 4 of which remained depolarized throughout the experiment. Currents were normalized to those measured in the light, at −200 mV for inward currents and 20 mV for outward currents. Currents mediated by inward rectifying channels increased in the dark, while no conductance changes were apparent at potentials more positive than −120 mV.
Figure 5(a) shows the free-running membrane potential of a cell that was hyperpolarized in the light, but slowly depolarized from −117 mV to −55 mV, after a transition to the dark. Before and during the depolarization, the plasma membrane was clamped from a holding potential of −100 mV to test potentials ranging from −180 to 0 mV (Figure 5b). A small increase in inward currents was found after 14 min in the dark, while outward currents remained unchanged. The same response is displayed in the current-voltage curves of Figure 5(c) (left graph). When current-voltage curves were scaled to a smaller current range (Figure 5c, right graph) an additional change in currents from −100 to −60 mV was apparent. Currents at these potentials all decreased by 35 pA after the light was switched off. The free-running membrane potential should be equal to the zero current potentials of the current-voltage curve, provided that time-dependent currents saturate during test pulses of 2 sec. Note that the zero current potential indeed correlates with the free-running membrane potential. Similar changes in current-voltage curves were found for all cells that altered their membrane potential in response to light/dark transitions. On average, currents at −80 mV increased by 35 pA (SE 14, n = 5) in the light.
The change of steady state currents in Figure 5(c) was further analysed, extracting time activated currents from those measured at the end of the capacity compensation peak. In the dark, currents measured early after the onset of the clamp (interval a, Figure 6) changed towards more inward directed values (Figure 6a). In contrast, only small changes occurred in time-activated currents (interval difference b–a, Figure 6b). This indicates that changes in the membrane potential are due to a light-dependent response of fast activating ion-transporters.
Guard cell responses to red and blue light
Changes in the electrical properties of the plasma membrane in response to light/dark transitions could have been evoked by either blue or red light, since guard cells are sensitive to both light qualities. The sensitivity to blue and red light was tested with guard cells that were impaled at a low intensity red light background, provided by the microscope lamp. All cells were depolarized after impalement and remained depolarized after the onset of additional red light, introduced from the abaxial side at a photon flux density of 450 µmol m−2 s−1. The plasma membrane conductance was measured every 3 min and guard cells were exposed to additional blue light, only when inward and outward conductance had remained stable over 12 min. The conductance was regarded as stable when the changes in current during subsequent pulses were less than 10%.
An additional beam of blue light (photon flux of 200 µmol m−2 s−1) applied from the abaxial side induced a transient hyperpolarization, in 4 out of 6 cells (Figure 7a). The plasma membrane was clamped from a holding potential of −100 mV to potentials ranging from −160 to 0 mV (Figure 7b). Blue light had little effect on time-dependent outward currents, but inhibited time-dependent inward currents (Figure 7d). In addition, blue light induced a change of currents measured early after the onset of the voltage clamp, relative to the zero current level (Figure 7c). The change in current was identical for potentials ranging from −100 to −60 mV, indicating that a voltage-independent conductance change underlies this response. As expected, the potentials at which the steady state current-voltage curves intersect with the zero current axis (Figure 7e) are the same as found for the free-running membrane potential, as seen in Figure 7(a).
The blue light effect on the plasma membrane conductance was averaged for 6 cells that were exposed on 9 occasions to blue light. On average, the outward rectifying channels remained unaffected by blue light (Figure 8a). The shift in currents measured early after onset of the voltage clamp (as shown in Figure 7c), measured at −80 mV, increased in amplitude by 60 pA (Figure 8b). The blue light effect was transient; the currents returned to their pre-stimulus value 7.5 min after switching on the blue light. In contrast, a sustained inhibition of inward rectifying channels was observed (Figure 8c).
To test whether or not the light responses can be triggered by red light, guard cells were exposed to increased intensities of red light. A small beam of red light, at a photon flux of 450 µmol m−2 s−1, was given on top of background red light (1.5 µmol m−2 s−1). Under these conditions, red light could not induce a hyperpolarization in any of the cells measured (n = 5, data not shown). In agreement with these results, additional red light neither had an effect on the currents measured at −80 mV (Figure 8b), nor did it affect the outward or inward rectifying channels (Figure 8a,c).
In the present experiments, guard cells were exposed for a prolonged period to blue light instead of blue light pulses that have been used in most other reports. Nevertheless, transient changes in the membrane potential and plasma membrane current were observed. To compare our data with blue light-induced proton extrusion, the same light conditions were applied to guard cells in an epidermal strip of V. faba. The blue light-induced acidification was measured with a flat-tipped pH-electrode placed in contact with the epidermal strip (Figure 9, inset). An exponential function was fitted to the time-dependent changes in proton concentration, and from its derivative the rate of proton extrusion in time was calculated (Figure 9). After a lag period of 60 sec, blue light induces a rapid rise in H+ extrusion, reaching its maximum after 120 sec and decaying slowly afterwards.
The impalement of guard cells within the intact plant provides a refined tool to study guard cells within their natural environment. The use of this method, however, is complicated by the presence of a potential difference between the guard cell wall and external solution. We found that this problem can be minimized using young V. faba leaves, where the surface potential was relatively small and only small light-induced transients were measured. In our opinion, the advantage of measuring guard cells within the intact plant justifies the complications that arise from the surface potential and the resistance between the guard cell wall and reference electrode.
Injection of guard cells with fluorescent dye revealed that the tip of the electrode is, in most cases, located in the cytoplasm. However, the possibility cannot be excluded that on occasion the electrodes may also have penetrated the vacuolar membrane. Nevertheless, the conductance properties of cells in all three physiological states resemble those of ion-channels known to be present in the plasma membrane only (Figure 2). Vacuolar membranes of guard cells are dominated by cation channels, with properties quite different from ion-channels in the plasma membrane (Allen et al., 1998; Hedrich and Neher, 1987).
Most guard cells were depolarized and displayed the activity of ion-channels with properties similar to those of previously identified K+-selective outward- and inward rectifying channels. Some of these cells displayed light-induced membrane potential changes that brought about a shift from the depolarized to the hyperpolarized state. A schematic presentation of the transporters involved in this response and their current-voltage relation is shown in Figure 10. In the dark, cells are in the depolarized state and extrude K+ through outward rectifying channels. This may be accompanied by the extrusion of Cl– via anion channels, since anion channels are voltage activated at depolarized potentials. However, in guard cells that display light-induced membrane potential changes, the activity of anion channels was small. Current-voltage relations of these cells (Figures 5c and 7c,d) lacked a negative slope from −80 to −100 mV, typical for cells with active anion channels (Figure 2c). This indicates that conductances other than anion channels may charge balance the K+-efflux. These ion-transporters, yet to be identified, could either facilitate an efflux of anions or an influx of cations. In the light, H+-ATPases become active and hyperpolarize the plasma membrane. Inward rectifying K+-channels are activated, while outward K+ and anion channels are deactivated. The change in membrane potential reverses the driving force for K+, and in turn K+ is taken up via inward rectifying channels.
A typical voltage dependence of the K+-channels in guard cells is presented in Figure 10(b). The current-voltage relation of the K+-channels was calculated from Boltzmann equations, which were fitted to data depicted in Figure 3. Transporters other than K+-channels are represented as a linear conductance with the same slope as the current voltage relation of the cell in Figure 5. In the light, the I-V relation of these transporters is shifted over the whole voltage range by 35 pA. Such a change in the I-V curve is typical for the activation of H+-ATPases, since their activity is virtually constant over the voltage range of −180 to 50 mV (Lohse and Hedrich, 1992; Taylor and Assmann, 2000). As a result of the change in the background I-V-curve, the cells shift from depolarized to hyperpolarized potentials (Figure 10c).
Light also reduces the activity of inward rectifying K+-channels, which is often interpreted as a reduction of the guard cells capacity to take up K+ (Assmann and Shimazaki, 1999). However, a decrease in maximum conductance will cause a small hyperpolarization of the plasma membrane (Figure 10c), which will in turn activate inward channels. Overall, the actual conductance of inward rectifying K+-channels will be little affected by a change in maximum conductance (Roelfsema and Prins, 1998).
Provided the 35 pA increase of currents is due to the activation of H+-ATPases, it would drive a H+-extrusion of 1.3 pmol h−1. Given a stochiometry of 1K+ taken up per H+, this drives an K+-influx of 1.3 pmol h−1. Guard cells of V. faba take up 1.8–2.5 pmol K+ during stomatal opening (Allaway and Hsiao, 1973; Humble and Raschke, 1971; Outlaw and Lowry, 1977), which will thus take about 2 h. Previously, a K+-uptake rate of 0.7 pmol h−1 was estimated to occur during stomatal opening in V. faba (Raschke et al., 1988).
In guard cells depolarized in red light, a beam of blue light induced an outward current with an average of 60 pA after 90 sec. In contrast to the effect of white light, the blue light-induced current change was transient. The same was found for proton extrusion rates measured in epidermal strips, indicating that indeed the activation of H+-ATPases underlies the current change. Stimulation of H+-ATPases most likely occurs via phosphorylation of the C-terminus of this protein (Kinoshita and Shimazaki, 1999). Phosphorylation of the C-terminus, as well as ATP-hydrolysis by ATPases, were at a maximum 2.5 min after a blue light pulse. In the present experiments blue light was applied for a prolonged period, but the change in currents at −80 mV peaks between 1.5 and 4.5 min to decrease afterwards. This may relate to the nature of blue light receptors in guard cells. After being excited by blue light, these receptors recover only slowly (half time of 12 min, Iino et al., 1985). With a red light background, blue light will simultaneously stimulate all receptors, giving rise to a maximum response. Subsequently, no receptors are available any longer and the response will be quenched until functional receptors reappear.
Blue light also decreased the conductance of the inward channel, but this response remained stable after 4 min. This indicates that the light-dependent regulation of the proton pump and the inward rectifier occur via different signalling pathways. Further evidence for independent signalling pathways comes from cells that remain depolarized in white light. These cells did not display a change in currents measured at −80 mV, but the conductance of inward rectifying channels was still inhibited in the light.
A beam of red light, superimposed on a low intensity red light background, failed to affect the H+-ATPase or the conductance of the inward rectifier. Under the same conditions, blue light triggered changes in the activity of both transporters. This suggests that the response to white light is due to irradiation in the blue spectrum. However, the response of the plasma membrane differs for both light qualities. White light induces a steady hyperpolarization, while blue light causes a response that is maximal after 2 min and decays afterwards. The origin of this difference is at present not fully understood. In the present experiments the CO2 levels were kept low by a steam of CO2 free air directed on the drop of solution between the leaf and objective. Low CO2 concentrations are known to enhance the blue-light sensitivity of stomata (Assmann, 1988; Karlsson, 1986), but a red light response, depending on photosynthesis, may be inhibited by low CO2. Future experiments will therefore be directed towards the exploration of interactions between CO2, red-and blue light, in guard cells.
Plant growth and experimental setup
Broad bean (Vicia faba L. cv. Grünkernige Hangdown, Gebag, Hannover, Germany) plants were grown in a green house. Four- to 6-week-old plants were placed next to an upright microscope (Axioskop 2FS, Carl Zeiss, Göttingen, Germany) The adaxial side of a leaf was mounted on a plexiglas holder in the focal plane using double-sided adhesive tape. In this experimental configuration, the abaxial side of the leaf was accessible for impalement with microelectrodes. For this purpose, a water immersion objective (Achroplan 40x/0.80 W, Carl Zeiss) was submerged into a drop of the following solution: 50 mm KCl, 5 mm 2-(N-Morpholino) ethane sulphonic acid/Bis-Tris propane pH 6.5 and 0.1 mm CaCl2, which was placed on the leaf surface. Stomata were shielded for the variable CO2 concentration in the laboratory by a flow of CO2-free air directed at the drop of solution. The reference electrode, a 1 m KCl and 0.1% agarose salt bridge, connected to an AgCl/Cl half cell, was placed in the solution between the cuticula and objective. Electrodes were moved towards an open stomate with a micro manipulator (type 5171, Eppendorf, Hamburg, Germany). A piezo translator (P-280.30, Physik Instrumente, Waldbronn, Germany) was used to impale the electrode into a guard cell. Guard cells under investigation were 41 μm (SD = 3, n = 60) in length and 14 μm (SD = 1, n = 60) in diameter.
Electrodes and electrical configuration
Double-barreled electrodes were pulled from two glass capillaries (GC100F-10, Clark Electromedical Instruments, Pangbourne, Reading, UK) that were aligned, heated and twisted 360° on a customized vertical electrode puller (L/M-3P-A, List Medical Electronic, Darmstadt, Germany). A first pull was executed on the vertical electrode puller, while the final pull was carried out on a horizontal laser puller (P-2000, Sutter Instumant Co., Novato, CA, USA). The electrodes were filled with 300 mm KCl and had resistances ranging from 60 to 110 MΩ. Blunt double-barreled electrodes, used to measure the surface potential and series resistance, had a tip resistance ranging from 3 to 14 MΩ. The electrodes were connected via a 1 m KCl bridge and AgCl/Cl half cells to a double microelectrode amplifier (VF-102, Bio-Logic, Claix, France) equipped with headstages of 1011Ω input impedance.
Voltage step protocols were applied via an ITC−16 interface (Instrutech, Corp., Elmont, NY, USA) under control of Pulse software (Heka, Lambrecht, Germany). The test voltages were fed into a differential amplifier (CA-100, Bio-Logic, Claix, France) connected to the VF-102 amplifier. The data were low pass filtered at 300 Hz with 8-pole Bessel (type 902, Frequency Devices, Haverhill Ma, USA) and sampled at 1 kHz.
For pressure injection, the same double-barreled microelectrodes were used as for voltage clamp experiments. The tip of one of the barrels was filled with 25 µm Lucifer-yellow and this barrel was glued to a micropipette holder. The dye was injected with a pressure of 15 bar applied after impalement of the guard cell. Fluorescence was detected on a laser scanning microscope (LSM 410, Carl Zeiss. Göttingen, Germany) using an excitation wavelength of 488 nm and an emission filter at 515–565 nm.
Leaves were illuminated with the microscope lamp (HAL 12 V/100 W, Carl Zeiss), focused on the adaxial side at an area of 2 cm in diameter. The photon flux densities were 98 and 4.4 µmol m−2 s−1 at the adaxial and abaxial side, respectively. Switching the microscope on caused an increase in temperature of less than 0.5°C, as measured with a temperature sensor in the medium above the abaxial epidermis. Light passing a longpass (edge wavelength 610 nm) glass filter (RD 610, Schott, Mainz, Germany) was defined as red light. Red light photon flux densities were 55 and 1.5 µmol m−2 s−1 at the adaxial and abaxial side, respectively. Additional red light, provided by a second light source (KL 1500, Schott), was projected onto the abaxial side of the leaf, at an area 0.3 mm in diameter and at a photon flux density of 450 µmol m−2 s−1. Blue light with a central wavelength of 470 nm (bandwidth 12 nm) was provided by a monochromator (Polychrome 1, T.I.L.L. Photonics, Martinsried, Germany) and projected at the abaxial side, at an area of 0.2 mm in diameter and at a photon flux density of 200 µmol m−2 s−1. Photosynthetic active radiation was measured with a LI-COR 250 light meter (Quantum sensor, LI-190, LI-COR, Lincoln NE, USA).
Epidermal strips were peeled from the abaxial side of leaves and mounted on a microscope slide using Medical Adhesive (VM 335, Ulrich AG, St. Gallen, Switzerland). The epidermal strip was placed in a measuring chamber (Roelfsema et al., 1998) filled with 50 mm KCl, 0.1 mm CaCl2 and 0.5 mm 2-(N-Morpholino sulphonic acid/Bis-Tris propane pH 6.5. A combined pH-reference electrode with a flat tip (Ingold Electrodes, Wilmington, MA, USA) was placed in contact with the epidermal strip. The strip was illuminated at a 45° angle, with red light at a photon flux density of 450 µmol m−2 s−1 and additional blue light at a flux density of 200 µmol m−2 s−1. Light sources and filters were the same as described previously (Roelfsema et al., 1998).
We thank M. Knoblauch and A.J.E. van Bel (University of Giessen, Giessen, Germany) for their useful suggestions concerning microscopy and the leaf holder; H.B.A. Prins (University of Groningen, Haren, The Netherlands) for technical support; D. Carden (University Giessen, Giessen, Germany) and P. Dietrich (University of Würzburg) for help with the preparation of the manuscript. This work was funded by Deutsche Forschungsgemeinschaft grants to R.H.