Modulation of Guard Cell Plasma Membrane Potassium Currents by Methyl Jasmonate

Previous studies have suggested a role for jasmonates in the promotion of stomatal closure (Raghavendra and Reddy, 1987; Gehring et al., 1997). This report describes whole-cell patch-clamp experiments that demonstrate that methyl jasmonate (Me-JA) has concentration-dependent effects on guard cell K+channels and promotes stomatal closure by biasing potassium transport at the plasma membrane for K+ efflux. Stomatal closure is brought about by a reduction in guard cell solutes. Potassium salts are abundant in guard cells (Outlaw, 1983) and many regulators of stomatal movements, such as ABA, exert their activity, at least in part, through the modulation of potassium fluxes (MacRobbie, 1981; Blatt, 1990). Concentrations of jasmonates increase in response to stresses that alter water status and necessitate a change in stomatal aperture (Creelman and Mullet, 1995; Lehmann et al., 1995). In the majority of species investigated, jasmonates have been implicated in the control of stomatal movements by their effects on transpiration (Leshem et al., 1994;Wang, 1999), gas exchange (Lee et al., 1996; Herde et al., 1997), and stomatal aperture (Raghavendra and Reddy, 1987; Gehring et al., 1997). However, little is known about the mechanisms by which jasmonates affect stomatal behavior. To investigate the effect of Me-JA on guard cell plasma membrane K+ channels in patch-clamp experiments,Vicia faba guard cell protoplasts were prepared by enzymatic digestion (Lemtiri-Chlieh, 1996). Protoplasts were perfused with 10 mm KCl, 1 mm CaCl2, 2 mm MgCl2, 5 mm MES, and sometimes 20 mmpotassium Glu (pH 5.5, 480 or 520 mOsmol kg−1). Patch pipettes prepared from Kimax-51 glass (Kimble/Kontes, Vineland, NJ) contained 200 mm potassium Glu, 2 mmK2ATP, 2 mmMgCl2, and 10 mm HEPES buffer (pH 7.6–8, 520 mOsmol kg−1). Once the whole-cell configuration was obtained, the membrane was clamped close to K+ equilibrium potential (EK) and no current measurements made for 2 min. Recordings were only made when the seal resistance exceeded 1 GΩ. Currents were monitored using a Dagan 3900A or Axopatch 200A Integrating Patch Clamp Amplifier (Dagan Corporation, Minneapolis; Axon Instruments Inc., Union City, CA). Voltage commands were given and signal recordings assessed by microcomputer using pClamp 6.0 software (Axon Instruments Inc.). The microcomputer and amplifier were connected via an input/output device and current traces low pass filtered at 5 kHz before analog to digital conversion. Liquid junction potentials were corrected for as described by Neher (1992). Me-JA (Duchefa Biochemie, Haarlem, The Netherlands) was a racemic mixture and solubilized in ethanol. Student's t tests and analysis of variance were performed to compare current characteristics and channel conductance before and after Me-JA application. Effect of 0.1 μm Me-JA on IKout and IKin. Data from a typical guard cell protoplast. a, Electrical recordings made in response to a series of 600-ms voltage pulses (above) ranging in 20 mV steps from 64 to −196 mV (holding potential −76 mV). Recordings were made before (left) and at the times indicated after Me-JA had been added to the bath. Scale: vertical, 250 pA; horizontal, 200 ms. b, IKout and IKin (the time-dependent component of the current traces in a) plotted as a function of membrane potential. Cross-referenced to part a by symbol. Significant changes in the conductance of Kout and Kin evoked by Me-JA are concentration dependent and voltage independent1-a Conductance values were calculated for protoplasts with currents >± 50 pA prior to Me-JA application. Time-dependent currents were recorded after 600 ms, at which point they were approaching steady state. Channel conductance post-Me-JA application was normalized to its pre-application value and relative conductance values are shown in the table ± se. Student'st tests indicated the effect of 0.1 μm Me-JA on the conductance of K out andK in was significant (see main text). Analysis of variance indicated the relative change in steady-state conductance ofK outand K in was voltage independent in all cases (P ≥ 0.6). Significant changes in the conductance of Kout and Kin evoked by Me-JA are concentration dependent and voltage independent1-a Conductance values were calculated for protoplasts with currents >± 50 pA prior to Me-JA application. Time-dependent currents were recorded after 600 ms, at which point they were approaching steady state. Channel conductance post-Me-JA application was normalized to its pre-application value and relative conductance values are shown in the table ± se. Student'st tests indicated the effect of 0.1 μm Me-JA on the conductance of K out andK in was significant (see main text). Analysis of variance indicated the relative change in steady-state conductance ofK outand K in was voltage independent in all cases (P ≥ 0.6). In the majority of experiments (16 of 20), the effect of 0.1 μm Me-JA on IKout and IKin was detected within 15 min of exposure. This is illustrated in Figure 1 where electrical recordings made from a typical guard cell protoplast are shown. Ten minutes after the addition of 0.1 μm Me-JA, IKout has clearly already increased. In this example, Kin is slower to respond than Kout, but a decrease in inward current at −176 and −196 mV can be detected 20 min after the addition of the growth regulator. The activation and deactivation kinetics of both currents were examined after the addition of 0.1 μm Me-JA (Tables II andIII). Although the activation time constant for IKout was voltage dependent (P = <0.001), no effect of 0.1 μm Me-JA on the activation time constant for either current could be detected (Kout:P = 0.786; Kin: P= 0.572). Deactivation kinetics were examined after the return of the membrane potential to the holding potential from a series of activating voltage pulses. There was no significant effect of Me-JA on any of the deactivation time constants (P ≥ 0.1), regardless of the magnitude of the conditioning voltage step (data not shown, P ≥ 0.26). Me-JA does not affect the activation time constant of IKout or IKin a Outward currents were fitted to the equation IKout = Il + (IKss(1 − exp(−t/τ))2), where Il is the leak and IKss is the steady-state component of the current. Analysis of variance-indicated τ was dependent on voltage (P = 0.001) but independent of Me-JA (P= 0.786). Inward currents were fitted to the equation IKin = Il + (IKss (1 − exp(−t/τ))). Analysis of variance-indicated τ was independent of voltage (P = 0.121) and Me-JA (P = 0.572). All values are means ± se. Me-JA does not affect the activation time constant of IKout or IKin a Outward currents were fitted to the equation IKout = Il + (IKss(1 − exp(−t/τ))2), where Il is the leak and IKss is the steady-state component of the current. Analysis of variance-indicated τ was dependent on voltage (P = 0.001) but independent of Me-JA (P= 0.786). Inward currents were fitted to the equation IKin = Il + (IKss (1 − exp(−t/τ))). Analysis of variance-indicated τ was independent of voltage (P = 0.121) and Me-JA (P = 0.572). All values are means ± se. Me-JA does not affect the deactivation kinetics of IKout or IKin a Deactivating currents were recorded at −76 mV following a 600-ms conditioning voltage step at 64 mV (IKout) or −196 mV (IKin). Outward currents were fitted to double exponential functions, deactivating inward currents to single exponentials. Analysis of variance indicated all time constants were independent of voltage (P ≥ 0.3) and Me-JA treatment (P ≥ 0.1). All mean values are ± se. Me-JA does not affect the deactivation kinetics of IKout or IKin a Deactivating currents were recorded at −76 mV following a 600-ms conditioning voltage step at 64 mV (IKout) or −196 mV (IKin). Outward currents were fitted to double exponential functions, deactivating inward currents to single exponentials. Analysis of variance indicated all time constants were independent of voltage (P ≥ 0.3) and Me-JA treatment (P ≥ 0.1). All mean values are ± se. The only other concentration of Me-JA tested to have a significant effect was 50 μm Me-JA (Table I). This concentration of Me-JA led to a significant reduction in the steady-state conductance of Kout (Table I;P = 0.01). Considering the effect of 0.1 μm Me-JA on this conductance, the data suggest Me-JA may have a bimodal effect on Kout. However, jasmonates have not been reported to have a corresponding dual effect on stomatal movements and unlike the other concentrations tested, 50 μm Me-JA is vastly in excess of estimates of jasmonate concentrations in planta (Creelman and Mullet, 1997). Pollen and seed germination, root growth, and cell division have all been reported to show a bimodal response to jasmonates when the concentrations tested have included those in excess of the physiological and the significance of the generally inhibitory effect of such supraoptimal concentrations is unclear. The changes in the chord conductance of Kout and Kin evoked by 0.1 μm Me-JA would increase the capacity for K+ efflux across the guard cell plasma membrane and decrease the capacity for K+ influx. This suggests that Me-JA is capable of promoting a net loss of potassium from the guard cell cytoplasm and by this mechanism, stomatal closure. This is consistent with the findings of Raghavendra and Bhaskar Reddy (1987), who reported a significant reduction in the potassium content of isolated Commelina benghalensis guard cells after a 4-h incubation in 1 μm Me-JA. It has also been suggested that the ability of jasmonic acid to accelerate pulvinar movements in Cassia fasciculata (Bourbouloux et al., 1994) and inhibit light-induced turgor increases inMimosa pudica pulvini (Tsurumi and Asahi, 1985) may result from comparable effects on K+ channels in these cells. Single-channel studies are now required to determine if Me-JA directly affects the conductance of Kout and Kin or whether additional signal transduction components are involved. Cytosolic alkalinization precedes jasmonate-induced stomatal closure in Paphiopedilumspp. and acid loading of the guard cells eliminated this response (Gehring et al., 1997). This would suggest some overlap between jasmonate and abscisic acid signal transduction pathways (Blatt and Armstrong, 1993). The likely involvement of cytosolic calcium is entirely speculative in the absence of any measurements of jasmonate-induced increases in cytosolic calcium in guard cells. However, calcium has been implicated in other jasmonate-dependent processes (Leon et al., 1998;Kenton et al., 1999) and jasmonates are known to alter the activity of at least one calcium transport protein in mammalian systems (Starling et al., 1994). The ability of Me-JA to alter the properties of potassium conductances in the guard cell plasma membrane has now been demonstrated. Aims for the future must be to investigate how these effects of Me-JA are mediated and under what conditions such control is important. I would like to thank Prof. Enid MacRobbie, Dr. Stephen Roberts, and Dr. Martin McAinsh for comments on the manuscript.