A transient outward-rectifying K⁺ channel current down-regulated by cytosolic Ca²⁺ in Arabidopsis thaliana guard cells

Abstract

Sustained (noninactivating) outward-rectifying K⁺ channel currents have been identified in a variety of plant cell types and species. Here, in Arabidopsis thaliana guard cells, in addition to these sustained K⁺ currents, an inactivating outward-rectifying K⁺ current was characterized (plant A-type current: I_AP). I_AP activated rapidly with a time constant of 165 ms and inactivated slowly with a time constant of 7.2 sec at +40 mV. I_AP was enhanced by increasing the duration (from 0 to 20 sec) and degree (from +20 to −100 mV) of prepulse hyperpolarization. Ionic substitution and relaxation (tail) current recordings showed that outward I_AP was mainly carried by K⁺ ions. In contrast to the sustained outward-rectifying K⁺ currents, cytosolic alkaline pH was found to inhibit I_AP and extracellular K⁺ was required for I_AP activity. Furthermore, increasing cytosolic free Ca²⁺ in the physiological range strongly inhibited I_AP activity with a half inhibitory concentration of ≈ 94 nM. We present a detailed characterization of an inactivating K⁺ current in a higher plant cell. Regulation of I_AP by diverse factors including membrane potential, cytosolic Ca²⁺ and pH, and extracellular K⁺ and Ca²⁺ implies that the inactivating I_AP described here may have important functions during transient depolarizations found in guard cells, and in integrated signal transduction processes during stomatal movements.

At least two general classes of voltage-dependent K⁺ channels have been characterized in the plasma membrane of plant cells: hyperpolarization-activated inward-rectifying K⁺ channels (K⁺_in), which mediate K⁺ influx (for review see refs. 1 and 2), and depolarization-activated outward-rectifying K⁺ channels (K⁺_out), which mediate K⁺ efflux. K⁺_out currents have been identified in a variety of cell types and plant species, such as guard cells of Vicia faba, maize, tobacco, and Arabidopsis (3–7); trap-lobe cells of Dionaea muscipula (8); Samanea and Mimosa pulvinar motor cells (9, 10); mesophyll cells of Arabidopsis and tobacco (11, 12), and root cells of various species (13–15). The most common property of these plant K⁺_out currents is time- and voltage-dependent activation and lack of inactivation, with currents remaining stable for many minutes during continuous depolarization (16). These sustained K⁺_out currents have been proposed to function as important pathways for prolonged K⁺ efflux during, for example, stomatal closing (3), leaf movements (9), and repolarization processes after depolarizations (for review see ref. 17).

Interestingly, an outward-rectifying K⁺ channel (KCO1) recently was cloned from Arabidopsis, which is activated by cytosolic Ca²⁺ when expressed in insect cells (18). The primary structure of KCO1 contains two pore-forming domains within one subunit, showing structural similarities to two pore-domain K⁺ channels isolated recently in yeast, Drosophila, and human (19–21). In maize suspension cells and Mimosa pulvinar cells K⁺_out have been demonstrated to be activated by cytosolic Ca²⁺ (10, 22).

Depolarization-activated K⁺ channels have been found in many types of animal cells. These channels have diverse functions in these cell types, such as controlling neuronal excitability and hormone release (23). In animal cells, subsequent to channel activation by depolarization there is a process of channel inactivation with time courses ranging from ms to sec (A-type K⁺ channel; refs. 23–26). Inactivation is brought about by the channel entering in nonconducting (inactivated) states that are distinct from the closed states (23, 27). One prominent difference in the physiological properties of these K⁺ channels lies in their inactivation, which can be extremely diverse with respect to time course and can occur by multiple mechanisms (23). Differences in inactivation properties in many cases have a strong influence on the physiological response of the expressing cells (e.g., refs. 28 and 29).

In plant cells, K⁺ channels showing inactivation as described for A-type channels in animal cells have not been characterized in detail, although a transient outward-rectifying current component recently was observed in some tobacco guard cells, in addition to the sustained K⁺_out currents (5). In the present study, we describe the detailed characterization of an outward-rectifying K⁺ current with inactivation and unexpected regulation mechanisms in Arabidopsis guard cells, which may have important functions in stomatal movements, such as in repolarization processes after transient depolarizations reported in guard cells (30, 31) and in short-term K⁺ efflux during regulation of stomatal apertures.

MATERIALS AND METHODS

Guard Cell Isolation.

A. thaliana plants were grown in a controlled environment growth chamber for 4–6 weeks, and guard cell protoplasts were isolated as described previously (7). In brief, two Arabidopsis leaves were blended. The epidermal strips were collected on a 292-μm mesh and then incubated in 10 ml of medium containing 1.3% cellulase R10 and 0.7% macerozyme R10 (Yakult Honsha, Tokyo), 0.1 mM KCl, 0.1 mM CaCl₂, 500 mM d-mannitol, 0.5% BSA, 0.1% kanamycin sulfate, and 10 mM ascorbic acid-Tris (pH 5.5) for 16 hr at 22 ± 2°C on an orbital shaker. Isolated guard cell protoplasts were then collected, washed twice, and stored on ice for use.

Data Acquisition and Analysis.

Patch clamp electrodes were prepared from soft glass capillaries (Kimax 51, Kimble, Toledo, OH), and pulled on a multistage puller (model P-87, Sutter Instruments, Novato, CA). Conventional whole-cell voltage-clamp was performed (32). Giga-ohm seals between electrode and plasma membrane were obtained by suction and usually appeared within 2–3 min. Cells were pulled up to the bath solution surface to reduce stray capacitance. Whole-cell configurations were established by applying increased continuous suction to the interior of the pipette. Guard cell protoplasts were voltage-clamped by using Axopatch 200 and Axopatch 1D amplifiers (Axon Instruments) (7). Data were analyzed by using axograph (version 3.5, Axon Instruments). Statistical analyses were performed by using excel (version 5.0; Microsoft). Data are the mean ± SEM.

Solutions.

The standard pipette solution contained 30 mM KCl, 70 mM K-glutamate, 2 mM MgCl₂, 6.7 mM EGTA, 3.35 mM CaCl₂, 5 mM Mg-ATP, and 10 mM Hepes-Tris, pH 7.1. The standard bath solution contained 30 mM KCl, 1 mM CaCl₂, 2 mM MgCl₂, and 10 mM Mes-Tris, pH 5.6. Total CaCl₂ concentrations in pipette solutions were changed to give indicated cytosolic free Ca²⁺ with 6.7 mM EGTA in all solutions. Free Ca²⁺ concentrations were calculated after accounting for ionic strength, temperature, and other parameters with Chelator software (Theo J.M. Schoenmakers, University of Nijmegen, The Netherlands). For pH experiments 6.7 mM EGTA and no CaCl₂ were added to pipette solutions. For ion substitution experiments, CsCl and K-gluconate were used. Other changes of components in solutions are described in detail in the text and figure legends. Osmolalities of all solutions were adjusted to 485 mmol⋅kg⁻¹ for bath solutions and 500 mmol⋅kg⁻¹ for pipette solutions by addition of d-sorbitol. Ionic activities and liquid junction potentials were measured and corrected (33).

RESULTS

In Arabidopsis guard cell protoplasts, when the membrane potential was held at −40 mV and then pulsed to positive potentials, small outward whole-cell currents were observed (Fig. 1A). At the beginning of whole-cell recordings (before “rundown”) we found large noninactivating K⁺_out currents in Arabidopsis guard cells (data not shown) that were analogous to recently reported K⁺_out currents (6). These noninactivating K⁺_out currents in V. faba guard cells have been reported to also show rundown (3). Because of this rundown of K⁺_out currents, we were able to discern an additional type of outward current in the present study.

Figure 1.

Hyperpolarization before subsequent depolarization gives rise to transient outward currents in Arabidopsis guard cells. (A) Whole-cell outward currents recorded at positive potentials after a holding potential of −40 mV. (B) Whole-cell outward currents recorded in the same guard cell shown in A after a holding potential of −180 mV. Results similar to those in A and B were found in >100 guard cells. Membrane potentials were stepped from −180 mV to potentials ranging from +100 mV to 0 mV in −20 mV increments with an interval time between pulses of 30 sec in A and B. (C) Hyperpolarization recovered currents calculated by subtraction of current recorded in A from that in B at +100 mV. (D) Normalized transient outward peak conductance (G/Gmax) from experiments performed as in B plotted as a function of the applied membrane potentials. (E) Time course analysis of the transient outward current activation (see text). (F) Times to reach peak currents (t_p) and activation time constants (τ_a) of transient outward currents plotted as a function of the applied membrane potentials. (G) Single exponential decay functions fitted to the inactivation of the transient outward currents. (H) Time constants of the decay (inactivation; τ_i) of transient outward currents plotted against the applied membrane potentials. The standard pipette solution and a bath solution with 30 mM KCl, 40 mM CaCl₂, 2 mM MgCl₂ were used in A–D (see Materials and Methods). Pipette solutions with nominally zero CaCl₂ and a bath solution with 1 mM CaCl₂ were used in E–H (see Materials and Methods). Arrows and dashed line show zero current levels.

Enhancement of a Transient Outward Current by Prehyperpolarization.

Because the resting potential in guard cells is negative of −100 mV (30, 31), in subsequent experiments the membrane potential was clamped to more hyperpolarized potentials before depolarizing voltage pulses. When the membrane potential was initially held at −180 mV, depolarizations caused transient outward currents (Fig. 1B). An example of difference currents between Fig. 1 A and B at +100 mV clearly shows that prehyperpolarization to −180 mV enhanced a transient outward current (Fig. 1C). The currents were activated at depolarizing potentials, and subsequently showed a decline during prolonged depolarization. As described later, this decline in currents corresponds to inactivation similar to transient voltage-dependent outward-rectifying K⁺ currents in animal cells (A-type current, I_A; refs. 24–26). We therefore named the transient outward currents I_AP (I_A in planta).

The peak conductance-voltage relationship of I_AP showed that the currents were activated at membrane potentials positive of ≈−20 mV under the imposed conditions (Fig. 1D). The time- and voltage dependence of activation and inactivation of I_AP were analyzed. A sigmoidal rise of I_AP was resolved after activation (Fig. 1E). The activation time course of currents at membrane potentials could be described by a simple equation:

where I_L is an instantaneous current component; I_∝ is the maximum current after activation, τ_a is the activation time constant, and P is modeled as the number of gating particles. The currents were well fitted by the equation described above when P = 4 (n = 5 cells). τ_a and the time for the currents to reach I_∝(t_p) at +40 mV were 165 ± 7 ms and 1.15 ± 0.16 sec, respectively (Fig. 1F; n = 5). τ_a and t_p showed a voltage dependence, with increased depolarization resulting in more rapid activation. A single exponential decay function could be fitted to the inactivation of I_AP (Fig. 1G). Inactivation times of I_AP were significantly longer (τ_i = 7.16 ± 1.37 sec at +40 mV) than activation times and did not show a strong voltage-dependence (Fig. 1H; n = 5).

I_AP Is Carried Mainly by K⁺.

Experiments were performed to test which ions carried I_AP by ion substitution and reversal potential determination. Cytosolic ion substitution of K⁺ for Cs⁺, which is largely impermeable to most plant K⁺ channels (16, 34), completely abolished I_AP (Fig. 2A). Hyperpolarization-activated K⁺_in currents activated at potentials ranging from −120 to −180 mV were not abolished as expected for cytosolic K⁺ substitution by Cs⁺. Note that the lack of outward tail currents for K⁺_in currents at depolarizing potentials (Fig. 2A) is consistent with strong unidirectional Cs⁺ block of K⁺_in channels reported previously in V. faba guard cells and mesophyll cells of Avena sativa (16, 34). The lack of I_AP with Cs⁺ in the cytosol suggested that I_AP may be carried by K⁺ ions.

Figure 2.

Transient outward currents are carried mainly by K⁺. (A) Whole-cell current recordings under conditions with 100 mM Cs⁺ in the pipette solution. The voltage protocol was similar to that given in Fig. 1B (note large K⁺_in currents at potentials negative of −100 mV). Arrow shows zero current level. (B) Tail current recordings of the transient outward currents. From a holding potential of −180 mV, transient outward currents were activated by a voltage step to +100 mV. Tail currents of the transient outward currents were recorded at the indicated membrane potentials (I_AP Tail). (C) Tail currents of K⁺_in currents recorded in the same cell as in B. After a prepulse potential of −180 mV, where K⁺_in currents were activated, tail currents were recorded at the indicated membrane potentials ranging from −140 mV to +80 mV with +20 mV increments (K⁺_in Tail). At increasingly depolarized tail potentials, the decay times of K⁺_in current deactivation became shorter and therefore could be clearly distinguished from the transient outward currents. (D) Tail currents from recordings in B and C were plotted as a function of applied membrane potentials. ru, relative units. I_AP tail currents and K⁺_in tail currents were normalized by the currents at +40 mV and −60 mV, respectively. Pipette solutions with zero [Ca²⁺]_i and the standard bath solution were used.

To further test this hypothesis, reversal potentials of these currents were determined by measuring tail currents. Fig. 2B shows typical tail currents recorded for I_AP. From the holding potential of −180 mV, the membrane potential was stepped to +100 mV and I_AP was activated. Before the currents were inactivated, the membrane potential was stepped back to more negative values. This protocol allowed the deactivation of I_AP to be monitored. Fig. 2C shows typical tail currents recorded for K⁺_in from the same cell. Tail current amplitudes of both I_AP and K⁺_in currents were quantitatively analyzed and plotted against corresponding membrane potentials (Fig. 2D). The reversal potentials for I_AP and K⁺_in currents were −18 ± 2.2 mV (n = 4) and −24 ± 1.6 mV (n = 6), respectively. Potassium ions were the only ions under the imposed conditions that had a significantly negative equilibrium potential (E_K⁺ = −28.7 mV). The reversal potential therefore indicates that the outward I_AP is carried mainly by K⁺. The 10.7 mV more positive reversal potential for I_AP when compared with E_K⁺ could be caused by Cl⁻, Ca²⁺, or Mg²⁺ permeation because E_Cl⁻ ≈ 0 mV, E_Ca²⁺ > +120 mV and E_Mg²⁺ = 0 mV, which are all positive to E_K⁺. After substitution of KCl by K-gluconate to reduce background anion currents, K⁺_in reversed at −31.8 ± 2.0 mV (n = 8; E_K⁺ = −30 mV), whereas the reversal potential of I_AP was virtually unaffected (E_IAP = −16.5 ± 2.6 mV; n = 3), indicating that I_AP is not significantly anion permeable. Therefore, both ion substitution and reversal potential measurements show that the outward I_AP is carried mainly by K⁺ with negligible anion permeability (35), and that K⁺_in was more K⁺ selective than I_AP. In additional experiments, increasing the extracellular Ca²⁺ concentration from 1 mM to 40 mM with either Cl⁻ or gluconate as the anions caused a positive shift in the reversal potential of I_AP (n = 12), showing that I_AP channels have a permeability to Ca²⁺.

Effect of Extracellular Ca²⁺ and K⁺ on I_AP.

When K⁺ ions in the bath solution ([K⁺]_o) were substituted by Ca²⁺, I_AP was abolished (Fig. 3A). For these cells when the bath was then perfused with a solution containing 30 mM KCl and 1 mM CaCl₂, I_AP was restored (Fig. 3B). These data suggest that Ca²⁺ may inhibit I_AP and/or [K⁺]_o may be required for I_AP activation. Further experiments were designed to test these possibilities by simultaneously adding both 30 mM K⁺ and 40 mM Ca²⁺ to the bath solution. Fig. 3C shows that increasing extracellular Ca²⁺ concentration from 1 mM (Center) to 40 mM (Right) decreased I_AP amplitudes, suggesting an inhibitory effect of extracellular Ca²⁺ on I_AP. Increasing [K⁺]_o from 0 mM (Fig. 3C, Left) to 30 mM (Fig. 3C, Right) dramatically enhanced I_AP amplitudes, indicating that [K⁺]_o is required for I_AP activation. Varying [K⁺]_o (0, 1, 10, and 40 mM) confirmed that [K⁺]_o activates I_AP, whereas I_AP activation potentials were not significantly shifted by varying [K⁺]_o (n = 23; data not shown). This increase in outward K⁺ current, on increasing [K⁺]_o is reminiscent of depolarization-activated K⁺ channels in animal cells (26). These characteristics of I_AP are opposite to those described for the sustained K⁺_out currents found in V. faba and Arabidopsis guard cells, where increasing [K⁺]_o decreases K⁺_out currents, via a positive shift in the activation potential (6, 16, 36).

Figure 3.

Extracellular Ca²⁺ inhibits I_AP and extracellular K⁺ is required for I_AP. (A) Whole-cell currents recorded in a bath solution with zero mM KCl and 40 mM CaCl₂ show inhibition of I_AP. (B) I_AP was restored when the bath was perfused with a standard bath solution containing 30 mM KCl, 1 mM CaCl₂. I_AP was recorded in the same cell as in A. Similar results were found in eight cells. Arrows show zero current levels in A and B. (C) Average effects of extracellular Ca²⁺ and K⁺ concentrations on I_AP (n = 5–8). A pipette solution with nominally zero Ca²⁺ and bath solutions with varying Ca²⁺ and K⁺ concentrations as indicated (in mM) were used.

I_AP Recovery Is Dependent on the Degree and Duration of Prehyperpolarization.

To quantitatively analyze the effect of negative prepulse holding potentials on recovery of I_AP, experiments were designed to compare I_AP peak amplitudes after various prepulse holding potentials ranging from +40 to −160 mV (Fig. 4A). The potential was held for 60 sec at the indicated holding potentials before depolarization to saturate I_AP recovery. More negative holding potentials increased I_AP recovery, demonstrating that a hyperpolarized membrane potential was required for I_AP recovery (Fig. 4 A and B). I_AP magnitudes reached a plateau for prepulse holding potentials more negative than ≈−100 mV (Fig. 4B). The midpoint of membrane potentials required for I_AP recovery was ≈−20 mV under the imposed conditions. K⁺_in currents were not required for I_AP because I_AP magnitudes were large at prepulse membrane potentials ranging from −40 to −100 mV, whereas at these potentials K⁺_in currents were not activated. Thus these data demonstrate that physiological resting potentials negative of −100 mV favor the full recovery of I_AP in Arabidopsis guard cells.

Figure 4.

I_AP magnitudes are enhanced by more negative prepulse holding potentials and by longer hyperpolarization times. (A) I_AP magnitudes were enhanced when prehyperpolarization potentials were more negative. Left arrow shows the changes of holding potentials, and right arrow indicates the associated increasing I_AP amplitudes. Prehyperpolarizations ranged from +40 mV to −160 mV in increments of −20 mV. The interval time between pulses was 60 sec at holding potential (HP). (B) Normalized I_AP peak amplitudes and K⁺_in currents plotted against respective prepulse holding potentials for the data in A. Similar results were analyzed in three representative cells. (C) Increasing durations of prehyperpolarizations enhanced the magnitude of I_AP. Data from a representative cell show increasing prehyperpolarization times at −180 mV. (Upper) Voltage protocol is shown. In the voltage protocol, voltage ramps were used after depolarization to stabilize the cells because large prolonged and repetitive depolarizing voltage pules often led to loss of whole-cell recordings. (D) I_AP peak amplitudes plotted against respective prepulse hyperpolarization times from experiments performed as in C. Similar results were obtained in four cells. A pipette solution with nominally zero Ca²⁺ concentration and a standard bath solution were used.

The duration of prehyperpolarization required for I_AP recovery was analyzed. Voltage protocols were used with increasingly prolonged prehyperpolarization times between pulses from zero to 10.5 sec (Fig. 4C, Upper). I_AP peak amplitudes increased with increasing hyperpolarization times (Fig. 4 C and D). I_AP peak amplitudes did not reach maximum levels even with 10.5 sec of prehyperpolarization. Maximum current levels could be reached only with approximately 20 sec of prehyperpolarization at −180 mV. The half prehyperpolarization time at −180 mV required for I_AP recovery was 5.2 ± 1.7 sec (n = 4). The finding that I_AP activation can be recovered by both the degree and duration of prehyperpolarization demonstrated that the decay of I_AP during depolarization is analogous to inactivation found for A-type K⁺ channels in animal cells (23–26), as well as inactivation found for rapid anion channels (37) and voltage-dependent Ca²⁺ channels (38) in plant cells.

Cytosolic Ca²⁺ Inhibition of I_AP.

Stimulus-induced increases in cytosolic Ca²⁺ play an important role in many plant signal transduction processes, for example in abscisic acid (ABA)-induced stomatal closing (39–41). The sustained K⁺_out currents appear not to be modulated by physiological changes in cytosolic Ca²⁺ in guard cells (35, 42, 43). Unexpectedly, I_AP was strongly inhibited by increases in cytosolic free Ca²⁺ concentrations ([Ca²⁺]_i) in the physiological range (Fig. 5). I_AP peak amplitudes were larger at nominally zero [Ca²⁺]_i than those at 2 μM [Ca²⁺]_i (Fig. 5A). This inhibition by [Ca²⁺]_i was further tested over a range of [Ca²⁺]_i from nominally zero to 2 mM and confirmed the strong down-regulation of I_AP by [Ca²⁺]_i (Fig. 5 B and C). The average effect of cytosolic Ca²⁺ shows an 8.9-fold ± 2.5-fold decrease of I_AP by increasing [Ca²⁺]_i from nominally zero to 2 μM (Fig. 5C). A Hill curve could be fitted to the data showing a K_d of ≈94 nM for a Hill coefficient of 1 (26). These data indicate that even small physiological changes in [Ca²⁺]_i could efficiently up- and down-regulate I_AP, which may imply an important physiological role of I_AP in connection to Ca²⁺-dependent signal transduction pathways in Arabidopsis guard cells.

Figure 5.

Cytosolic free Ca²⁺ elevation inhibits I_AP. (A) Representative whole-cell I_AP recorded at [Ca²⁺]_i of 2 μM (Upper) and nominally zero (Lower). The voltage protocol was as given in Fig. 1B. (B) I_AP measured under various [Ca²⁺]_i in the recording pipettes ranging from zero to 2,000 μM. Voltage protocol was the same as in Fig. 1B. Only current traces at +100 mV are shown. Dashed line shows zero current level. (C) Effect of [Ca²⁺]_i on I_AP peak amplitudes at +100 mV from experiments as performed in B (n = 5–12). Pipette solutions varied only in the free Ca²⁺ concentrations. The bath solution was the standard bath solution.

Inhibition of I_AP by Cytosolic Alkaline pH.

ABA induces cytosolic alkalization (41, 44). Alkalization of cytosolic pH (pH_i) has been found to up-regulate the sustained K⁺_out currents in the plasma membrane of V. faba guard cells (44, 45). We therefore determined whether cytosolic alkalization affects I_AP found in Arabidopsis guard cells.

Surprisingly, when pH_i was shifted to more alkaline values from 7.1 to 8.3, I_AP was dramatically decreased (Fig. 6). Examples of currents recorded at two different pH_i values of 7.1 and 8.3 illustrate that pH_i 8.3 inhibited I_AP (Fig. 6A). Fig. 6B shows in more detail that I_AP was decreased by cytosolic alkalization. Fig. 6C shows the decrease of I_AP as pH_i increased from 7.1 to 8.3. Besides the decreased current amplitudes with cytosolic alkalization, the half-activation time at +100 mV for I_AP appeared to be dramatically increased from τ_a = 86 msec at pH_i 7.1 to τ_a = 820 msec at pH_i 8.3 (Fig. 6B). It is possible that I_AP is completely inhibited by alkaline pH_i and that the small currents recorded at pH_i 8.3 (Fig. 6 A and B) correspond to a small residual sustained K⁺_out currents (6) rather than I_AP.

Figure 6.

Cytosolic alkaline pH inhibits I_AP. (A) Representative I_AP recorded under pH_i 8.3 (Upper) and pH_i 7.1 (Lower). (B) I_AP at +100 mV measured under various pH_i from 7.1 to 8.3. The voltage protocol was the same in A and B as given in Fig. 1B, but only current traces at +100 mV are shown in B. (C) Effect of pH_i on I_AP peak amplitudes at +100 mV from experiments as performed in B (n = 4–12). Pipette solutions with nominally zero free Ca²⁺ concentrations and varied pH_i, and standard bath solution were used.

DISCUSSION

I_AP shows functional characteristics different from the sustained plant K⁺_out channels characterized in many cell types and the cloned Arabidopsis KCO1 channel. Several major differences are apparent between I_AP and K⁺_out currents. First, I_AP is a higher plant K⁺ current that shows a pronounced inactivation with time (Fig. 1), a characteristic not seen for the noninactivating K⁺_out currents (3–16). Second, the pH sensitivity of I_AP is opposite to that of the K⁺_out currents in V. faba guard cells: I_AP is inhibited by alkaline cytosolic pH (Fig. 6), whereas K⁺_out currents are enhanced by alkaline pH shifts (44, 45). Third, the regulation of I_AP by [K⁺]_o is also opposite to that of K⁺_out and KCO1. Decreasing [K⁺]_o shifts the activation potential of K⁺_out to more negative values and enhances current amplitudes (6, 14, 16, 36). Similar results to K⁺_out were observed for KCO1 expressed in insect cells (18). However, decreasing [K⁺]_o did not greatly affect the activation potential of I_AP and in addition decreased the amplitudes of I_AP. In a recent study, two components of outward K⁺ currents, one showing instantaneous and another slow time-dependent activation with slight inactivation, have been characterized in Arabidopsis guard cells (6). Both the instantaneous and time-dependent currents were enhanced by decreasing [K⁺]_o via a negative shift of activation potentials (6), which is opposite and therefore distinct from the time-dependent I_AP. Finally, cytosolic Ca²⁺ in general does not regulate K⁺_out currents in guard cells (35, 42, 43), and in some cells [Ca²⁺]_i up-regulates K⁺_out currents (10, 18, 22). I_AP showed the unique characteristics of being down-regulated by physiological [Ca²⁺]_i increases (Fig. 5).

I_AP Has Properties Similar to A-Type K⁺ Channels in Animal Cells.

First, I_AP inactivates with prolonged depolarization time (Fig. 1). Animal A-type K⁺ channels show variable inactivation with two distinct types of inactivation mechanisms, N-type and C-type (23, 46–48). N-type inactivation involves the binding of a tethered blocking particle, containing positively charged N-terminal amino acids to the cytoplasmic face of the channel (46, 47). The exact mechanisms of C-type inactivation (containing C-terminal amino acids) are unknown, but may involve a constriction of the outer mouth of the pore, preventing conduction (48, 49). The mechanisms of inactivation for I_AP remain to be determined. Second, I_AP was dependent on [K⁺]_o (Fig. 3). Depolarization-activated K⁺ channel currents in animal cells are enhanced by increasing [K⁺]_o (26). Furthermore, the recovery of A-type currents from both N-type and C-type inactivation is dependent on [K⁺]_o (50, 51). Third, activation of I_AP is voltage-dependent and the inactivation is voltage-independent at positive potentials (Fig. 1 F and H), which also has been reported for A-type channels (23). Fourth, the recovery of I_AP activity from inactivated states is time- and voltage-dependent at negative potentials (Fig. 4), which resembles A-type channel recovery from both C-type and N-type inactivation (50, 51). Finally, I_AP was reduced by increasing [Ca²⁺]_i (Fig. 5). This property may be similar to A-type channels where after inactivation, high [Ca²⁺]_i prevents K⁺ current reactivation (52). Whether I_AP channels are structurally related to animal A-type K⁺ channels, however, remains unknown.

Possible Physiological Functions of I_AP.

K⁺_out channels in guard cells are proposed to serve as major sustained K⁺ efflux pathways, leading to full stomatal closure in response to strong stimuli such as long periods of dark and drought stress (1, 3). The prominent inactivation of I_AP indicates that this current is unlikely to be involved in sustained K⁺ efflux lasting many minutes, but may be involved in short-term and transient K⁺ efflux as a signal leading to adjustments in stomatal apertures. For example, during ABA-induced stomatal closure, an initial transient efflux of K⁺ (Rb⁺) and Cl⁻ is observed followed by a sustained efflux phase (53). It is possible that I_AP contributes a component to this transient K⁺ efflux, which acts as a signal to initiate stomatal closing. A transient component of total outward current also was observed in some tobacco guard cells (5), suggesting that I_AP may occur and function in other plant species. Furthermore, [K⁺]_o was required for I_AP, also implicating a physiological role for autoregulation of K⁺ currents by [K⁺]_o as feed forward mechanism to enhance transient K⁺ efflux in Arabidopsis guard cells.

Potassium channels play a central role in controlling cell membrane potential (1, 2, 17, 23, 26). Repetitive transient depolarizations or oscillations have been observed in V. faba and Arabidopsis guard cells (30, 31, 54). These membrane potential oscillations have been proposed to function as a feedback mechanism for fine tuning of stomatal apertures (30, 31, 54). Rapid depolarization has been proposed to be initiated by anion channel activation, mediating anion efflux, in particular rapid anion channel activation (55). Interestingly, both rapid anion channel currents (56) and I_AP (Fig. 6) show enhancement at more acidic pH_i. Consistent with these findings, a recent study in Arabidopsis guard cells has demonstrated that buffering pH_i to ≤pH 7.0 triggers repetitive rapid depolarizations (31). Furthermore, the repolarization time constants measured in guard cells (≈150 ms; refs. 30, 31, and 54) are similar to the activation time constants of I_AP (τ_a ≈ 80–220 ms; Fig. 1), indicating that I_AP could contribute to the repolarization process. Therefore, physiological membrane potential depolarizations may activate I_AP in vivo, subsequently causing repolarization, suggesting that I_AP may be an integral component of guard cell firing patterns as described for A-type channels in animal cells (26, 28, 29).

Both [Ca²⁺]_i and pH_i are important second messengers in guard cells (39–45). For instance, during ABA-induced stomatal closing increases of [Ca²⁺]_i have been reported in guard cells (35, 39–41). However, it appears unlikely that I_AP functions downstream of the Ca²⁺-dependent ABA signaling pathway, because ABA-induced increases of [Ca²⁺]_i may inhibit I_AP in Arabidopsis guard cells. Increases of [Ca²⁺]_i in guard cells also have been observed during light-induced stomatal opening (41). [Ca²⁺]_i has been suggested to function as a versatile signal transducer in guard cells that can affect both stomatal opening and closing (41). For example, Ca²⁺-dependent protein kinases activate vacuolar Cl⁻ channel and malate uptake currents, which were suggested to function in Ca²⁺-dependent-stomatal opening (57). Therefore, Ca²⁺-permeable I_AP may function in other Ca²⁺-dependent signal transduction pathways.

In conclusion, we present a characterization of an inactivating K⁺ channel current in a higher plant cell. The unique characteristics of I_AP inactivation and regulation by diverse factors including membrane potential, cytosolic Ca²⁺, cytosolic pH, and extracellular K⁺ imply that I_AP may have important roles in the adjustment of stomatal apertures during repetitive transient depolarizations, during early rapid phases of K⁺ efflux for stomatal closing, and/or during integrated signal transduction processes.