Dec 7, 1991 - dation grant DMB-84 19759 to M.H.M.G., by Department of Energy grant 85ER 13359 to C.L.S., and by a Volkswagen-Stiftung Akademie-.
Plant Physiol. (1992) 99, 96-102
Received for publication September 16, 1991 Accepted December 7, 1991
Channels in Arabidopsis Plasma Membrane1
Transport Characteristics and Involvement in Light-induced Voltage Changes Edgar P. Spalding*, Clifford L. Slayman, Mary Helen M. Goldsmith, Dietrich Gradmann2, and Adam Berti Department of Biology, Kline Biology Tower, Yale University, New Haven, Connecticut 06511 (E.P.S., M.H.M.G.); and Department of Cellular and Molecular Physiology, Yale School of Medicine, New Haven, Connecticut 06510 (C.L.S., D.G., A.B.) ABSTRACT White light (25 watts per square meter) induced an increase in plasma membrane K+-channel activity and a 30- to 70-millivolt transient membrane depolarization (completed in 2-3 minutes) in Arabidopsis thaliana leaf mesophyll cells. Transport characteristics of three types of ion channels in the plasma membrane were determined using inside-out patches. With 220 millimolar K+ on the cytoplasmic side of the patch and 50 millimolar K+ in the pipette, (220/50 K), the open-channel current-voltage curves of these channels were sigmoidal and consistent with an enzyme kinetic model. Two channel types were selective for K+ over Na+ and CI. One (named PKC1) had a maximum conductance (Gmax) of 44 picosiemens at a membrane voltage (Vm) of -65 mV in (220/ 50 K) and is stimulated by light. The other (PKC2) had Gmax = 66 picosiemens at Vm = 60 millivolts in (220/50 K). The third channel type (PCC1) transported K and Na+ about equally well but not Cl. It had Gmax = 109 picosiemens at Vm = 55 millivolts in (250/ 50 K) with 10 millimolar Ca2+ on the cytoplasmic side. Reducing Ca2+ to 0.1 millimolar increased PCC1 open-channel currents by approximately 50% in a voltage-independent manner. Averaged over time, PKC2 and PCCI currents strongly outward rectified and PKC1 currents did so weakly. Reductants (1 millimolar dithiothreitol or 10 millimolar O-mercaptoethanol) added to the cytoplasmic side of an excised patch increased the open probability of all three channel types.
This study focused on plasma membrane ion channels (reviewed in ref. 27) in Arabidopsis thaliana mesophyll cells and demonstrates the involvement of one channel type in a light-induced voltage change. Mutants that are defective in photoperception and/or signal transduction will be useful in future investigations of the response of Vm to light. Thus, combinations of genetic and patch-clamp techniques, both having resolution at the molecular level, are anticipated to advance the study of membrane transport and its control.
MATERIALS AND METHODS Plant Material
Light can rapidly modulate electrogenic transport systems in plant cells, resulting in large changes in the voltage across the plasma membrane (10, 20, 24). Changes in activity of the H+-ATPase and voltage-gated ion channels are probably components of the mechanism (1, 25, 28). Elucidation of the details would yield results relevant to numerous physiological processes that depend on Vm3.
Arabidopsis thaliana (L.) Heynh. (Landsberg genotype) plants were grown from unsterilized seeds sown on wet vermiculite:sand (10:1 mixture) in plastic containers (100 x 100 x 85 mm) placed on an irrigation table flooded twice daily with a nutrient solution (Hyponex Co., Copley, OH). Fluorescent and incandescent bulbs delivered light (20 W-mm2) on a 16-/8-h light/dark cycle. Rosette leaves (approximately 25 x 10 mm) from 3-week-old plants were excised, and portions of the abaxial epidermis were peeled offwith forceps. The peeled areas were floated peeled side down on 2 mL of protoplast isolation medium (0.6 M sorbitol, 10 mM KCl, 5 mM Mes, 1 mm CaCl2, adjusted to pH 5.8 with BTP) containing 5 mg Cellulysin (Calbiochem, San Diego, CA), 2 mg Pectinase (Sigma, St. Louis, MO), and 5 mg BSA (Sigma) in a small Petri dish (40 x 10 mm). The dish was gently shaken at 60 rpm for 1 h at 28°C. Undigested material was removed, and the protoplast suspension was centrifuged at 730g for 2 min. The pellet was resuspended in fresh medium, centrifuged, resuspended again, and stored on ice for up to 8 h.
'This research was supported, in part, by National Science Foundation grant DMB-84 19759 to M.H.M.G., by Department of Energy grant 85ER 13359 to C.L.S., and by a Volkswagen-Stiftung AkademieStipendium 11/66 647 to D.G. 2Permanent address: Pflanzenphysiologisches Institut der Universitat Gottingen, D-3400 Gottingen, Federal Republic of Germany. 3Abbreviations: Vm, membrane voltage; BTP, 1,3-bis[tris-(hydroxymethyl)-methylamino] propane; IO, current through an open channel; I-V curve, a plot of I, versus Vm; WL, white light; PO, open probability; PKC, plasma membrane K+ channel; PCC, plasma membrane cation channel; EK (EcI), equilibrium voltage for K+ (Cl-); Gmax, maximum slope conductance.
Single-Channel Recording Apparatus Patch-clamp experiments were performed generally as described in ref. 12. Micropipettes were pulled from Kimax-5 1 capillary glass (Kimble Products, Vineland, NJ) to a tip diameter of 1 to 2 ,tm and heat polished. The resulting electrodes, when filled as described below, had resistances of approximately 10 MQ. A Ag/AgCl reference electrode contacted the bath by way of a salt bridge. Electrode offsets were nulled before each experiment with the correction circuitry of the amplifier (Axopatch-lA with CV-3 headstage; Axon 96
ION CHANNELS IN ARABIDOPSIS PLASMA MEMBRANE
Instruments, Burlingame, CA). Amplifier output was continuously recorded on a Toshiba M-120 videotape recorder via a pulse-code modulator (model VR-IOA; Instrutech Corp., Elmont, NY). Recorded data were replayed, filtered (lowpass, 8-pole Bessel filter; model 902LPF, Frequency Devices, Haverhill, MA) and sampled at a rate at least twice the filter frequency using pCLAMP 5.51 software and a TL-1 data interface (both from Axon Instruments). Data were also obtained with the apparatus described in ref. 4. Gaussian distributions were fitted to all-points amplitude histograms. The mean current corresponding to the closed state was subtracted from that of the open state to give the open-channel currents plotted in the I-V curves. Sealing and Experimental Conditions
Protoplasts (25-50 Aim diameter) were transferred to the recording chamber (volume = 0.5 mL) containing bathing solution (220 or 250 mm KCI, 10 mM CaCl2, 5 mm Hepes, adjusted to pH 7.0 with BTP), and adhered to a piece of acidwashed glass coverslip on the bottom of the chamber. With 50 mM KCI, 1 mm CaCl2, 250 mm sorbitol, 5 mM Hepes (adjusted to pH 7.0 with BTP) in the pipette, seals of >I0 GQ formed in >50% of the attempts. Similar results were obtained using 5 mM Tris-Mes as buffer in both pipette and bath. Seal formation was aided by millimolar concentrations of CaC12, adherence of the protoplasts, and proper osmolality of the bathing solution (protoplasts swollen but safely short of their bursting point). Inside-out patches were produced after seal formation by withdrawing the pipette and exposing it very briefly to the air. Although a few patches were maintained for 2 h after excision, many did not last long enough to obtain useful data. High [Ca2+] on the cytoplasmic side produced more stable seals and higher channel activity compared with patches bathed in 1 uM free Ca2+ (10 mM EGTA + 7 mM CaC12, pH 7.0). Use of unphysiologically high [Ca2] did not seem to alter channel selectivities or general I-V relationships, the features on which we focused. The same characteristics obtained in studies with 1 uM cytoplasmic free Ca2+ and cell-attached patches. Excised patches may not retain the molecular machinery for sensing physiological [Ca2+] (21). The chamber was perfused throughout each experiment at rates up to 7 mL * min-' with a solution selected by means of a manifold. At the fastest perfusion rates, the time between switching and delivery of the new solution was about 3 s. Measurement of Light Effects
Light-related effects on Vm were measured in excised rosette leaves held by Plasticine at the tip of the blade and petiole and superfused with 0.1 mM KCI and 0.1 mm CaCl2. A cell two to three cell layers from the adaxial surface was impaled with a conventional glass microelectrode filled with 1 M KCI. The potential difference between the microelectrode and a bath electrode was led to a high-impedence electrometer and then to a chart recorder. The preparation was darkened for at least 10 min before WL (25 W_m-2) from a 150-W projector bulb (model EKE, General Electric, Cleveland, OH) was
directed onto the leaf with a fiberoptic. (The same light treatment was used in the channel-activation experiments.) The change in Vm followed the onset of illumination by less than the approximately 1 s of resolution obtainable with manual switching of the light. Irradiating submerged electrodes alone had no effect, ruling out possible artifacts due to light sensitivity of Ag/AgCl electrodes. Notation and Sign Convention Vm is defined as the electric potential of the cytoplasm minus that of the extracellular space. Currents through an open channel, IO, are taken to be positive when they carry positive charge out of the cytoplasm (outward current). Positive currents and voltages are plotted above or to the right of a reference. In this report, the term "I-V curve" refers to a plot of I,, versus Vm. Important ion concentrations at the two sides of an inside-out patch are given in millimolars in the form jcytoplasmic/extracellularl. For example, (250/50 K) indicates that there was 250 mM K+ at the cytoplasmic side (bath) and 50 mM K+ at the extracellular side (pipette). (200 Na + 50 K/50 K) means the bath contained 200 mM Na+ + 50 mM K+ and the pipette contained 50 mM K+.
Kinetic Modeling A cyclic two-state model with four apparent rate constants (units of s-') was used to describe the I-V curves (1 1, 13). One voltage-dependent pair of rate constants (kio, and koi) stands for the reversible transition of the charge between the cytoplasm (inside, i) and the outside (o): kio,= k%if and koi= k=,il f The superscript 0 denotes the value at zero voltage andf = exp(VVmze/2kT). z, e, k, and T have their usual thermodynamic meanings. The factor 2 reflects the assumption that the Eyring (energy) barrier is symmetric. The second pair of rate constants (Koi and Ki,) stands for the remaining, voltageindependent part of the reaction cycle. The current-voltage relationship of this model reads:
koi + Kio + Koi)
A three-state model is appropriate to describe two I-V curves obtained with different inside substrate concentrations. The three states correspond to the (sole) binding site loaded with the charge z facing inside (1) or outside (2) and the empty, neutral binding site inside (3). The voltage-dependent step is now k12 = k012f and k21 = k°2/f Substrate binding to, and debinding from, the inward-oriented binding site is described by k3 = k%3I[S]i (c03 being the rate constant with 1 M inside substrate concentration) and k13, respectively. A pair of formal rate constants (k23 and k32) stands for the unaffected part of the reaction cycle. The I-V relationship of the three-state model reads:
Io(Vm) = ze
2k23k3I - k2lkl3k32)
+ k32k21 + k23k3l +
k,2k32 + k,3k32
k3ikl2 + kl3k23 + kl2k23 + k2lkl3) In the I-V diagrams presented below, the solid curves are least-squares fits of Equations 1 or 2 to the plotted data. +
Plant Physiol. Vol. 99, 1992
SPALDING ET AL.
cation channel named PCC 1. The different channel types were encountered at the following approximate frequencies: PKC1, 60%; PKC2, 20%; and PCC1, 20%.
E .0 CZ
Plasma Membrane K+ Channel, PKC1
5 -200 5 min
light on 5
pA| 30 s
5 500 ms
Figure 2 (top) shows examples of channel currents through PKC 1 recorded from an inside-out patch at different membrane voltages. The two-state reaction-kinetic model was fitted to data obtained from four patches with 1220/50 K}, and the result is displayed in Figure 2 (bottom). The sigmoidal curve intersects the voltage axis at V1=0 = -40 mV. This is close to EK (-38 mV) and far from ECI (39 mV), indicating a high selectivity for K+ over CI-. Gmax was 44 pS at -65 mV. This demonstrates inward rectification of the open-channel currents, despite the outward K+ concentration gradient. On
Figure 1. Vm, ion channels, and the effects of light. A, WL (25 W.
m-2) induced a transient membrane depolarization in mesophyll cells of an intact Arabidopsis leaf. Hyperpolarization followed light off. B, WL-stimulated K+ channel activity in a plasma membrane cell-attached patch. K+ concentrations were 250 mm in the bath and 10 mm in the pipette. Pipette-holding voltage was -50 mV, and filter frequency was 50 Hz. C, Cell-attached patch containing three different types of channels. Differences between open-channel currents (amplitudes of individual pulses) distinguish the different channel types. K+ concentrations were 250 mm in the bath and 50 mm in the pipette. Pipette-holding voltage was -80 mV, and filter frequency was 400 Hz.
Effects of Light on Vm and Channel Activity Figure IA displays transient shifts in Vm that began upon the onset of WL (25 W. m-2). The peak change in Vm ranged from 30 to 70 mV. The shoulder during the depolarization was a usual feature. The repolarization usually proceeded beyond the initial Vm and then settled at a slightly more negative voltage 5 to 10 min after the "light-on" condition. Transient hyperpolarizations (approximately 20 mV) were induced without a detectable lag by the "light-off' condition. The return to a new steady Vm following the hyperpolarization also included a transient overadjustment. Figure lB shows a recording of currents from a cell-attached patch containing at least two channels. Within the first 1 s of WL, the activity of the channel increased. Analysis of this experiment revealed an approximately fourfold increase in the P. of the channels after the onset of WL. The lag between light-on and the onset of channel activation ranged between