Potassium Channels in Motor Cells of Samanea saman1

Transcription

1 Plant Physiol. (1988) 88, /88/88/0643/06/$0 1.00/0 Potassium Channels in Motor Cells of Samanea saman1 A PATCH-CLAMP STUDY Received for publication February 16, 1988 and in revised form May 26, 1988 NAVA MORAN2*, GERALD EHRENSTEIN, KUNIHIKO IWASA, CHARLES MISCHKE, CHARLES BARE, AND RUTH L. SATTER Department ofneurobiology, Weizmann Institute ofscience, Rehovot 76100, Israel (N.M.); Biophysics Laboratory, National Institute ofneurological and Communicative Disorders and Stroke, National Institutes ofhealth, Bethesda, Maryland (G.E., K.I.); Weed Science Laboratory, U.S. Department ofagriculture, Agricultural Research Service, Beltsville, Maryland (C.M., C.B.); and Department of Molecular and Cell Biology, U-42, University ofconnecticut, Storrs, Connecticut (R.L.S.) ABSTRACI Leaflet movements in Samanea saman are driven by the shrinking and swelling of cells in opposing (extensor and flexor) regions of the motor organ (pulvinus). Changes in cell volume, in turn, depend upon large changes in motor cell content of K+, Cl and other ions. We performed patch-clamp experiments on extensor and flexor protoplasts, to determine whether their plasma membranes contain channels capable of carrying the large K currents that flow during leaflet movement. Recordings in the "whole-cell" mode reveal depolarization-activated K+ currents in extensor and flexor cells that increase slowly (t½ = ca. 2 seconds) and remain active for minutes. Recordings from excised patches reveal a single channel conductance of ca. 20 picosiemens in both cell types. The magnitude of the K+ currents is adequate to account quantitatively for K+ loss, previously measured in vivo during cell shrinkage. The K+ channel blockers tetraethylammonium (5 millimolar) or quinine (1 millimolar) blocked channel opening and decreased light- and dark-promoted movements of excised leaflets. These results provide evidence for the role of potassium channels in leaflet movement. Leaflet movements in nyctinastic (night closure) plants often involve significant changes in the volume and up to severalfold variation in the ionic content of motor cells in the pulvinus (reviewed in Ref. 21). These variations may occur in response to light, darkness, and an endogenous biological clock. Cells in the extensor region of the pulvinus take up K+ and Cl- as they swell during leaflet opening, and lose both ions as they shrink during leaflet closure, while cells in the opposing (flexor) region behave in the reverse manner (12, 22, 23, 25, 30, 32). We undertook this study to examine a possible role for K+ channels in leaflet movement-related K+ fluxes and changes in cell volume in the nyctinastic legume Samanea saman. K+ channels have already been described in giant algae (1, 5) and in protoplasts isolated from wheat mesophyll cells (13, 14), Vicia faba guard cells (27, 29), carrot callus cells (14), and Asclepias 1 Supported by grant US from the United States-Israel Binational Agricultural Research and Development Fund to N. M., R. L. S., and C. M., by grant from the United States-Israel Binational Science Fundation, Jerusalem to N. M., and by grant DMB from the National Science Foundation to R. L. S. 2 Reprint requests may be sent to either N. M. or R. L. S. tuberosa cultured cells (28). Preliminary accounts of K+ channels in protoplasts isolated from Samanea pulvinar cells have appeared in abstract form (15, 16). MATERIALS AND METHODS Plant Material. Samanea saman trees were grown (a) in controlled chambers, with a 16 h light:8 h dark cycle (cool-white fluorescent light, 200,umol m-2 s-1 at 27 +1± C) or (b) in a greenhouse, with daily temperature variations between 22 and 36 C in the summer and between 20 and 27 C in winter, and with illumination as high as 700 umol m-2 s-' at noon on a bright day. The photoperiod in the greenhouse was extended to 16 h by irradiation with cool-white fluorescent bulbs (5,umol m-2 s-1). Protoplasts isolated from plants grown in controlled growth chambers and in a greenhouse gave similar results in patch-clamp experiments. Protoplast Isolation. Terminal secondary pulvini from the fourth to ninth leaf (counting from the apex) were harvested 2 to 3 h after the beginning of the light period in the growth chamber, or 2 to 3 h after sunrise in the greenhouse. Protoplasts were prepared by enzymic digestion of slices of extensor or flexor tissue (pooled separately), as described in (7) but with the following modifications. (a) The osmotic pressure of the pre-digestion solution was raised in two steps to 680 mosmol to ensure plasmolysis. (b) Pectolyase Y-23 (Seishin Pharmaceutical, Tokyo, Japan) was added to the digestion solution (which contained cellulase, pectinase and Driselase) to a final concentration of 0.2% (w/v). (c) A second purification step was added, as follows: the cells collected from the Ficoll interface were layered again on a sucrose gradient (2), spun at 60 to IOOg for 5 min, and collected and kept on ice for up to 24 h for patch-clamp experiments. Forty to fifty protoplasts of each type, flexor and extensor, were used for our experiments. Patch-Clamp Experiments: Methodology. A detailed description of the patch-clamp technique can be found in (8), while a brief tutorial description of patch-clamp methodology as applied to plant cells can be found in (24). Briefly, a drop ofthe protoplast suspension was added to 1.5 ml of the recording solution and the protoplasts were allowed to settle and stick to the bottom of the experimental chamber (Falcon Primaria [Becton Dickinson Labware, Oxnard, CA], or Nunc [Roskilde, Denmark] tissue culture dish). The patch pipette was then brought into contact with the protoplast. Upon the formation of a tight seal between the patch electrode and the cell membrane, the patch was either (a) broken to form a "whole-cell" configuration or (b) excised, by withdrawing the patch-pipette from the cell, to form an -643

2 644 MORAN ET AL. "outside-out" patch for single-channel recording. The recording setup and procedures were described previously (13). All experiments were performed in a voltage clamp mode (using the EPC- 7 amplifier, List-electronic, Darmstadt/Eberstadt, West Germany) and were under computer control (PDP-l 1/23, Plessey, Irvine, CA, or Digital Equipment Corp, Maynard, MA). Membrane potential was varied according to a pre-programmed schedule, and the resulting membrane current was filtered at 0.3 to 1 khz (-3 db), digitized at 0.5 to 5 khz, and stored for further analysis. The error in voltage clamping of the whole-cell membrane, largely due to the access resistance of the patch pipette, was compensated at 60 to 80% by analog circuitry of the EPC-7 amplifier (3 1). Recording Solutions for Patch-Clamp Experiments. The bath solution contained 2 to 50 mm KC1, 5 to 20 mm Mes at ph 6, and 1 mm CaCl2. The internal (pipette) solution contained 5 to 10 mm Hepes at ph 7 to 7.2, 125 mm KCl, and Ca-EGTA buffer to yield free Ca concentrations between and M (3). In some experiments 2 to 5 mm Mg-ATP (Sigma) was added to the internal solution. The osmolarity of all solutions was adjusted to 530 mosmol with sorbitol. The experiments were conducted at room temperature, 21 to 24'C, but temperature did not vary more than 1 C during any single experiment. White illumination from the microscope lamp was passed through a green filter, since protoplast longevity is decreased by high intensity irradiation with white light. However, low intensity white light was present at all times in the room in which the experiments were conducted. Measurements of Electrical Properties of the Protoplast. All measurements except for those of unitary conductance were performed in the whole-cell configuration. Membrane capacitance measurements involved nulling of the capacitative transients of current accompanying a repeatedly applied step change in membrane potential. We used the negative capacitance compensation circuitry of the amplifier to null these transients (31). The specific capacitance was calculated as the ratio of the whole cell capacitance to the cell surface area, which, in turn, was calculated from the diameter of the spherical protoplast, measured with a microscope graticule. The "resting" membrane potential was measured shortly after gaining access to the cytoplasm, by determining the value of the holding potential required to null the holding current. The "resting" membrane conductance was measured from the current-voltage relation ofthe whole cell at membrane potentials at which the known channels were closed (between -60 and -100 mv). This conductance is equivalent to the "leakage" conductance. At potentials that activate channels, the leakage current can be measured from the current jump at the very beginning of a step from a nonactivating to an activating potential, before any channels have had the chance to open (leakage currents are designated by short horizontal lines at the left of Fig. 1A). The whole cell conductance activated by depolarization was calculated from the net current remaining after subtracting the leakage current from the total current measured at the end of a voltage pulse. For the calculation of gk, the conductance of all the open channels in the whole cell, the following ohmic relation was used (9) IK = gk (VM-Vrev), where IK is the current through all the open K+ channels, and (V4- Vrev) is the driving force for the current, i.e. the difference between VM, the membrane potential, and Vr,e, the reversal potential for this current. The specific membrane conductance was calculated as the ratio of the whole cell conductance to the cell surface area. An identical calculation to that described for whole cell conductance was used to determine the single channel conductance, Plant Physiol. Vol. 88, 1988 except the amplitude of the unitary (single channel) current was used instead of the sum of the currents through all of the open channels. Thus, the number of open channels can be obtained from the ratio between the whole cell current and the unitary current at the same membrane potential, or from the ratio between the whole cell conductance and the unitary conductance. Leaflet Movement. Leaflet (pinnule) pairs subtended by tertiary pulvini were used for experiments testing the effects of channel blockers on light-promoted leaflet opening and darkpromoted leaflet closure. Tertiary pulvini are anatomically and physiologically similar to the secondary pulvini that provided protoplasts for the patch clamp experiments (21). Leaflet pairs were excised in the closed condition toward the end of the 16 h photoperiod (since they are least sensitive to manipulation at this time) and were trimmed. A dim green safelight provided the only illumination for manipulations and measurements during otherwise dark periods. The angle between paired leaflets was measured with a protractor. To determine the effect of channel blockers on leaflet movement, leaflet pairs were floated on experimental and control solutions immediately prior to the end of the photoperiod, darkened for 5 h, and then irradiated with white light (120,umol m-2 s-') for 150 min. The angle between paired leaflets was measured before and after the light treatment. Alternatively, leaflet pairs were floated on water, darkened for 8 h (as during the usual light/dark cycle), and then irradiated with white light for 3 h to promote opening. The water was then replaced by the control or experimental solution. After an additional 5 h of light (required for adequate uptake of the experimental solution), leaflets were darkened for 40 to 60 min and angles were measured before and after the dark treatment. Ten replicate leaflet pairs were used for each experiment. Each experiment was repeated three or more times, with similar results. RESULTS Properties of the "Resting" Protoplast. Protoplast diameters ranged from 24 to 44,m (mean value of 32 ± 4 Mm [SD]). There was no significant difference in size between the two types of cells (extensors: 31.6 ± 4.0 sm, n = 33, and flexors: 31.7 ± 4.6,um, n = 49). The estimated specific membrane capacitance was 0.72 ± 0.02 uf cm 2 (SEM) for the extensors (n = 26) and 0.67 ± 0.02 MAF cm-2 for the flexors (n = 18); this difference is practically insignificant (level of significance of only 0.10). The "resting" specific membrane conductance, measured at membrane potentials between -100 mv to -60 mv (i.e. when channels were not activated), ranged between 10 and 50 MuS cm-2, and the resting potentials of protoplasts in the whole-cell configuration was between -80 and -20 mv, with external [K+] varying between 1 and 50 mm. Whole-Cell K+ Currents in Flexor Cells. Figure la depicts whole-cell membrane currents measured from a flexor protoplast upon membrane activation by step depolarizations. The membrane potential was held at -70 mv between the depolarizing pulses. The outward currents increased gradually with time during each 7 s voltage pulse, indicating a gradual increase in conductance in response to depolarization. Note the slow time course of activation (in the range of seconds). At moderate depolarizations, these currents did not inactivate for several minutes (data not shown). The steady-state membrane current also increased with the degree of depolarization (Fig. 1B). Note that depolarizations above -40 mv elicited distinguishable membrane currents; these voltages are within the range of membrane potentials measured in Samanea flexor cells in vivo (19). Also note the sudden decrease in current upon step return to the holding potential of -70 mv and the slow decay of the (still outward) tail-currents (Fig. 1A). The direction ofthe tail-currents

3 K+ CHANNELS IN SAMANEA MOTOR CELLS: PATCH-CLAMP STUDY l(na) 20 ; ~~~-20D --20 V(mV) (C) (D) I(PA) 1200 pa -=o -90 / so ~~~~ close to the Nernst potential of -77 mv for K+ in this experiment. The Nernst potentials of all other permeating ions in the solutions bathing the membrane were positive. Thus, the voltagesensitive, time-dependent membrane current appears to be carried largely by K+. Single K' Channels in a Flexor Membrane Patch. To gain more information regarding the pathway for the membrane current, an outside-out patch was excised from the same flexor cell that had been used for the whole-cell recording shown in Figure 1. This patch was also depolarized by a series of square voltage pulses, lasting 2 s each. Figure 2A depicts typical single channel activity. While a small depolarizing pulse (e.g. depolarization to -30 mv) resulted in a quiescent record, larger depolarizations evoked channel opening (seen as upward square "jumps" in the current record, during a constant membrane potential); the higher the depolarization, the more fluctuations among different current levels. Jumps to multiple current levels (marked by [*] in Fig. 2A, middle trace) indicate simultaneous openings of an increasing number of channels. The single-channel conductance was 22 ± 2 ps3 and the reversal potential for the single-channel current was -52 ± 2 mv (Fig. 2B), if the current-voltage curve is linearly extrapolated. The only ion that I ~~~~~~~-91.,O/ ~~-40 FIG. 1. Membrane currents from a flexor cell, recorded in a wholecell recording configuration (see inset). A, Superimposed records ofwhole cell membrane currents (top panel) elicited by a series of 7 s step depolarizations (superimposed, bottom panel) to potentials (mv) indicated at the right, applied at 20 s intervals. The membrane potential between the pulses (the holding potential) was -70 mv. Upward (positive) direction indicates current (cations) flowing out of the cell. B, The steady-state current-voltage (I- V) relationship of the membrane currents. The values of membrane current at the end ofthe pulse, after subtraction of "leakage" current (the current at the start of the pulse, marked by horizontal dashes) are plotted versus membrane potential during the pulse. C, Superimposed records (top) of membrane currents elicited by three pairs of voltage pulses (superimposed, bottom), applied at 20 s intervals. The first pulse of each pair (an identical, depolarizing, conditioning pulse to +60 mv), was followed by a 6 s step to one of the three indicated membrane potentials (test pulses). Following each pair of pulses, the holding potential was restored. D, Instantaneous I-Vrelationship of tail-currents. The value of the tail-current immediately after the step to a test pulse (after subtraction of leakage current at that potential) was plotted versus membrane potential of the test pulse. Solutions (in mm): bath: 500 sorbitol, 10 Mes, 5 KC1, 1 CaCl2, ph 6.1; pipette: 250 sorbitol, 125 KCI, 10 Hepes, 4 NaCl, 3 EGTA, 2 Mg-ATP, 1.5 CaCl2, ph suggests that the reversal potential for the membrane current is below the holding potential. The slow decay indicates a gradual decrease in membrane conductance at the holding potential. The actual value of the reversal potential was determined by the "tail-current" method (9). Following a conditioning pulse which elicited an outward membrane current, the membrane potential was stepped to a test pulse, which elicited a tail-current (Fig. IC). Although the outward currents during the conditioning pulse superimposed, the tail-current was, at the first instant after the step, either outward (positive) or inward (negative), depending upon whether the membrane potential during the test pulse was above or below the reversal potential, respectively. As in Figure IA, the tail-currents decayed to absolute lower values. The instantaneous current-voltage (I- V) relationship of the tail-currents (Fig. ID) yields a reversal potential of -75 mv, (A) (C) 30 _30 r_. -30f U 0 J5PA 0.31 FIG. 2. Unitary currents from a membrane patch (outside-out configuration, see inset) excised from the flexor cell of Figure 1. A, Single channel current (top three traces) during step depolarizations (bottom trace) from a holding potential of-70 mv to the values (in mv) indicated above each current trace. B, I-V relationship of the open channel. Amplitudes of the fluctuations in current between adjacent levels during various pulses are plotted versus membrane potential during the pulse. The scatter in amplitudes at any given potential (SD shown) is probably due, at least in part, to errors in measurement. C, Averaged single channel currents. The recordings show current responses to voltage pulses, as in Figure 1A, but five current records were averaged for each voltage pulse and the data are superimposed. Disregard the very brief capacitative current transients accompanying step changes in potential. Solutions: same as in Figure 1, except KC1 in the bath was increased to 25 mm (final concentration), by an addition of 40 AL of 500 mm KCI to 1 ml of external solution. 'Abbreviations: ps, ns, and,s, pico-, nano-, and microsiemens; TEA, tetraethylammonium; TP3, inositol trisphosphate.

4 646 MORAN ET AL. had a Nernst potential near this value was K+, with a Nernst potential of -38 mv. We do not have a definitive explanation for the difference between these two values. However, following a subsequent increase of external [K+] to 50 mm, the reversal potential of the single channel current changed to -25 mv, in close correspondence with the new Nernst potential of -23 mv (data not shown). These results indicate that K+ is the major current carrier through these channels. To test whether the observed single channels are the pathway for K+ currents through the whole cell membrane, five records of single channel currents were averaged for each voltage pulse. The averaged currents through the channels in this patch (Fig. 2C), as in the whole cell (Fig. IA), increased with time during a sufficiently large depolarizing pulse, confirming the trend apparent in Figure 2A of gradual recruitment of single open channels during a pulse. The time course of this recruitment (Fig. 2C) resembled that of the gradual increase in current seen during the first 2 s of each pulse in the whole cell recording. The average number of channels open at the end of a pulse increased as membrane depolarization increased, i.e. the voltage dependence of channel opening was similar to the voltage dependence of the whole cell current, although the depolarization needed to activate channels in the excised patch (Fig. 2) was ca. 20 mv higher than that required for currents in the whole cell (Fig. 1). A shift in the range of membrane potentials required to activate channels after the isolation of a patch has already been observed in some cells (4), but not in others (18). Our results might be explained by a difference in external [K+] in experiments shown in Figures 1 and 2, since the range of activation of channels in Samanea (whole cell configuration) varies with external [K+] (16). The whole membrane conductance of this cell, activated by a depolarization to 20 mv (Fig. 1B), reached about 8 ns. Since the conductance of a single channel is about 20 ps (Fig. 2B), we estimate that about 400 channels were open simultaneously at 20 mv depolarization. In this particular cell, with a diameter of 31,um, the calculated average channel density was at least 1 channel for every 7 /Am2. The specific membrane conductance of this cell at 20 mv (Fig. 1B) reached about 280,S cm-2 (compared to a resting value of about 10,uScm-2). Whole-Cell Currents in Extensor Cells. Membrane currents were also recorded from extensor protoplasts (Fig. 3A). As with flexor protoplasts, channel opening was promoted by depolarization and channels activated slowly, in the range of seconds. Note the absence of tail currents, suggesting that the holding potential of -80 mv was the reversal potential for these currents. Since the Nernst potential was -116 mv for K+ and above +100 mv for both C1- and Ca2+ (see legend), the reversal potential of -80 mv implies that the currents were carried primarily by K+ (this reversal potential corresponds to a membrane permeability ratio of about 100:3 for K+ relative to Cl-). Increasing external [K+] by superfusing the cell from an adjacent pipette with an external solution containing 83 mm KCI (Fig. 3B), resulted in appreciable inward tail-currents following the same sequence of pulses as in Figure 3A, indicating a shift of the reversal potential to a value much more depolarized than the holding potential. This supports the notion of K+ as the major membrane current carrier, since increase in external [K+] would make the Nernst potential for K+ less negative. Figure 3C illustrates the effect of TEA, a commonly used K+ channel blocker. Addition of 6 mm TEA to the bath abolished the whole-cell currents within a few seconds. Partial removal of TEA by a puff of the external solution containing 83 mm KC1 reduced temporarily (for the duration of the puff) the blocking effect of TEA (Fig. 3D) and partially restored the outward currents and the inward tails. TEA was similarly effective in reversibly blocking both whole-cell and single-channel currents in flexorprotoplasts (data not shown). In addition, quinine, (A) (B) Plant Physiol. Vol. 88, TEA FIG. 3. The effect of voltage, KCl, and TEA on membrane currents from an extensor protoplast. A, Superimposed records of whole-cell currents elicited by 10 s voltage pulses from a holding potential of -80 mv to the following test potentials, applied at 20 s intervals: -50, -30, -10, 10, 30, 50, and 70 mv, The lowermost trace corresponds to the lowest membrane potential. Positive (upward) current indicates outward current. The current level at the holding potential is indicated by a horizontal dash at the left. The major component of the large current jumps at the beginning of depolarizing pulses consists of K+ current via K+ channels that did not have sufficient time to close during the interpulse interval. The same sequence of pulses (including an additional one, to -70 mv) was applied to the membrane in B, C, and D. B, Membrane current recorded during superfusion of the cell with a 83 mm KCl-extended solution, by a pipette positioned 30 to 50 Am away from the cell (see inset). C, Currents recorded several seconds after the addition of 6 mm TEA to the bathing medium, followed by D, superfusion from an adjacent pipette, as in B. Solutions (in mm): Bath: 500 sorbitol, 20 Mes, KCI, 1 1 CaC12, ph 6.0; in addition, 6 TEA present in the bath in C and D. Patch pipette: 250 sorbitol, 125 KC1, 20 Hepes, 3 EGTA, 0.5 CaCl2, ph 7.0; superfusion pipette: 333 sorbitol, 83 KCI, 13 Mes, 1 CaCl2, ph 6.0. another K+ channel blocker (6), abolished single channel and whole-cell currents in both types of cells (at mm; data not shown). The pathway for depolarization-activated K+ membrane currents in extensors, as in flexors, are K+ channels, based on the reversal potentials for single channel currents, the voltage dependence of their gating (opening promoted by depolarization) and their susceptibility to blockage by externally applied TEA or quinine (data not shown). In the experiment shown in Figure 3, the TEA-sensitive conductance of the whole cell was about 8 ns at 70 mv (Fig. 3B). The unitary conductance of extensor K+ channels (data not shown) and flexor K+ channels (Fig. 2B) are similar (about 20 ps) under comparable conditions. Given the TEA-sensitive whole-cell current flows through the 20 ps channels, the estimated number of channels in this cell was at least 400. In this particular cell, 40,um in diameter, the calculated average channel density was at least1 channel/12,um2. Effect of Channel Blockers on Leaflet Movement. IfK+ chan- (C) (D) 11OOpA Is +TEA

5 K+ CHANNELS IN SAMANEA MOTOR CELLS: PATCH-CLAMP STUDY clomnm 5 OmM 5 IOmM 5 10l O Jl winine ceh QUININE TEA FIG. 4. Effects of quinine and TEA on (A) light-promoted leaflet opening and (B) dark-promoted leaflet closure. See "Materials and Methods" for experimental details. The bar histograms indicate white lightpromoted leaflet opening and dark-promoted closure of the experimentally treated leaflets as compared to their controls. In A, the control leaflets opened 109 ± 7' (mean ±SD, n = 12) during 150 min irradiation. In B, the controls closed 108 ± 4' during the 40 to 60 min dark period. The controls contained mannitol of osmolarity equal to that of the experimental solution. (*) and (**) indicate that differences from control are significant at the P < 0.05 and P < 0.01 levels, respectively. nels are the pathways for the flow of K+ ions that drive leaflet movement, application of channel blocking agents should reduce the extent of such movement. This indeed is the case; TEA (25 mm) and quinine (> 1 mm) reduced white-light-promoted leaflet opening and dark-induced leaflet closure (Fig. 4) of isolated Samanea leaflets floating on test solutions. It was not possible to test reversibility of these drug effects, since excised leaflets were not viable for a sufficiently long period. Although neither TEA nor quinine prevented movement completely, this might be due to limited uptake of the drugs. Alternatively, the drugs might have affected K+ movement out of shrinking cells but not into swelling cells, and therefore impeded the movement of the pulvinus only incompletely. We are currently testing these possibilities, by determining the effect of K+ channel blockers on K+ fluxes measured in situ in whole pulvini. DISCUSSION Data from small membrane patches and from whole motor cells demonstrate channels that open by depolarization and that are permeable to K+. Their identification as K+ channels is further supported by their blockage by TEA and by quinine, both commonly used to identify K+ channels in animal cells (6). TEA has also been used recently to identify K+ channels in giant algal cells (1, 5, 11), wheat cells (14) and Asclepias tuberosa cells (28). Since TEA and quinine also reduced leaflet movement (at concentrations similar to those that blocked K+ channels in motor cell protoplasts), the K+ channels we describe could provide pathways for K+ fluxes in vivo. It is noteworthy that K+ channels in extensor and flexor cells exhibit similar gating properties, i.e. they are opened by depolarization to membrane potentials within similar ranges, and they have similar time constants for opening. In addition, the unitary conductance of K+ channels in the two cell types is similar. Quantitative Relationship between K+ Currents through Whole-Cell Membranes and K' Fluxes in Vivo. If leaflet movement depends upon K+ diffusion through the channels we observed, the total amount of K+ that can diffuse through these channels during an appropriate period must be adequate to account for changes in K+ that occur in situ. In Albizzia, another nyctinastic legume with leaflets movements and K+ values similar to those of Samanea (21), shrinking cells lose 15 to 32% of their K+ during the first 30 min of dark-promoted leaflet closure (extensor cells shrink) or light-promoted leaflet opening (flexor 647 cells shrink) (20, 23). The density of K+ channels observed in both types of Samanea motor cells appears to be more than sufficient to account for similar losses of K+. Assuming the average motor cell internal [K+] = 500 mm (22), an average cell measuring 32 gm in diameter (see Results section) would contain 8 pmol K+. In only 6 min, the measured whole-cell current of 0.5 na (Fig. lb, flexor data at 0 mv) can transport about 2 pmol of K+, corresponding to about 1/4 of the cell's K+ content, providing the current does not inactivate. As indicated above, these channels remain active for several minutes during moderate depolarizations. K+ channels that can transport sufficient K+ to account for the volume changes occurring in Vicia guard cells during stomatal movement have also been described recently (29). In the experiments reported here, the channels that can provide a pathway for the passive (electrochemical potential driven) efflux of K+ ions from motor cells are opened by depolarization. If the same mechanism occurs in vivo, flexor cells should depolarize as they shrink during light-promoted leaflet opening, while extensor cells should depolarize as they shrink during darkpromoted leaflet closure. In fact, the potential of flexor cells in a whole pulvinus depolarized from -50 to -20 mv and remained at -20 mv for several minutes during light-promoted leaflet opening (19), thereby supporting our hypothesis. The effect of a light/dark transition on the membrane potential of extensor cells has not yet been reported. The data described in this paper permit us to suggest the final steps in a mechanism for cell shrinking, namely, membrane depolarization -- K+ channel opening -* K+ efflux -. cell shrinkage. We have also observed channels that open upon hyperpolarization of the plasma membrane, and they could be relevant to cellular swelling. These are currently under investigation. Changes in the volume of Samanea motor cells depend upon large fluxes of Cl- as well as K+ (25), and we are currently using patch clamp methods to determine whether motor cell plasma membranes contain Cl- channels. Cl- accumulates in motor cells against its electrochemical gradient; data obtained with excised strips of motor tissue suggest that it is taken up via co-transport with H+ (26). Since Cl- uptake is a secondarily active process, passive efflux through a channel would appear to be a logical mechanism. Coupling between Changes in Illumination and Changes in Cell Volume. In intact plant motor cells, the movement of leaflets occurs in response to an endogenous clock and to changes in illumination which modulate it (21). Thus, the model for cell shrinkage proposed above implies that the motor cells also contain mechanisms to transduce a light stimulus to membrane potential, as already demonstrated for flexor cells (19). Extensor cells contain an outwardly directed H+ pump that is activated by white light and inactivated by darkness, while flexor cells contain an outwardly directed H+ pump with the reverse behavior, i.e. it is activated by darkness and inactivated by white light (10). H+ secretion through these pumps has been postulated to play an important role in regulating the membrane potential in response to changes in illumination (19, 21). These changes in membrane potential would change the electrochemical gradient for K+ to favor K+ uptake when the H+ pump is activated and K+ efflux when the H+ pump is inactivated. The membrane potential, and thus the electrochemical gradient for K+, would of course be affected also by the transport of other ions in addition to H+. We do not yet have data on the roles of these other ions, nor have we addressed steps that precede light-regulated changes in H+ pump activity. It is possible that a second messenger system involving IP3 is part ofthe transduction mechanism, since a brief (15-30 s) white light pulse increases the IP3 level in Samanea pulvini (17).

6 648 MORAN ET AL. Acknowledgments-We thank Youngsook Lee and Jerry Moylan for help with the experiment in Figure 4, and Dr. Ezra Galun and the Dobrinin Nutrition Center at the Weizmann Institute of Science for providing financial support and greenhouse space. LITERATURE CITED 1. BEILBY MJ 1986 Potassium channels at Chara plasmalemma. J Membr Biol 89: EDWARDS GE, SP ROBINSON, NJC TYLER, DA WALKER 1978 Photosynthesis by isolated protoplasts, protoplast extracts, and chloroplasts of wheat. Plant Physiol 62: FABIATO A, G FABIATO 1979 Calculator programs for computing the compositions of the solutions containing multiple metals and ligands used for experiments in skinned muscle cells. J Physiol (Paris) 75: FENWICK EM, A MARTY, E NEHER 1982 Sodium and calcium channels in bovine chromaffin cells. J Physiol (Lond) 331: FINDLAY GP, HA COLEMAN 1983 Potassium channels in the membrane of Hydrodictyon africanum. J Membr Biol 75: FISHMAN MC, I SPECTOR 1981 Potassium current suppression by quinidine reveals additional calcium currents in neuroblastoma cells. Proc Natl Acad Sci USA 78: GORTON HL, RL SATTER 1984 Extensor and flexor protoplasts from Samanea pulvini. I. Isolation and characterization. Plant Physiol 76: HAMILL PO, A MARTY, E NEHER, B SAKMANN, FJ SIGWORTH 1981 Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pfluegers Arch Eur J Physiol 391: HODGKIN AL, AF HUXLEY 1952 A quantitative description of membrane current and its application to conduction and excitation in nerve. J Physiol (Lond) 117: IGLESIAs A, RL SATTER 1983 H+ fluxes in excised Samanea motor tissue. I. Promotion by light. Plant Physiol 72: KEIFER DW, WJ LUCAS 1982 Potassium channels in Chara corallina: control and interaction with the electrogenic H+ pump. Plant Physiol 69: MAYER WE 1977 Kalium und Chloridverteilung im Laminargelenk von Phaseolus coccineus L. waehrend der circadianen Blattbewegung im tagesperiodischen Licht-Dunkelwechsel. Z Pflanzenphysiol 83: MORAN N, G EHRENSTEIN, K IWASA, C BARE, C MISCHKE 1984 Ion channels in plasmalemma of wheat protoplasts. Science 226: MORAN N, G ERHENSTEIN, K IWASA, C BARE, C MISCHKE 1986 Ionic channels in plant protoplasts. In R Latorre, ed, Ionic Channels in Cells and Model Systems. Plenum Press, New York, pp MORAN N, G EHRENSTEIN, K IWASA, C MISCHKE, C BARE, RL SATTER 1986 Potassium channels in protoplasts from Samanea saman (abstract). Biophys J 49: 165 Plant Physiol. Vol. 88, MORAN N, K IWASA, G EHRENSTEIN, C MISCHKE, C BARE, RL SATTER 1987 Effects of external K+ on K+ channels in Samanea protoplasts. Plant Physiol 83: S MORSE MJ, RC CRAIN, RL SATTER 1987 Light-stimulated phosphatidylinositol turnover in Samanea saman leaf pulvini. Proc Natl Acad Sci USA 84: PALLOTA BS, JR HEPLER, SA OGLESBY, TK HARDEN 1987 A comparison of calcium-activated potassium channels in cell-attached and excised patches. J Gen Physiol 89: RACUSEN RH, RL SATTER 1975 Rhythmic and phytochrome-regulated changes in transmembrane potential in Samanea pulvini. Nature 255: SATTER RL, AW GALSTON 1971 Phytochrome-controlled nyctinasty in Albizzia julibrissin. III. Interaction between an endogenous rhythm and phytochrome in control of potassium flux and leaflet movement. Plant Physiol 48: SATTER RL, AW GALSTON 1981 Mechanisms of control of leaf movements. Annu Rev Plant Physiol 32: SATrER RL, GT GEBALLE, PB APPELWHITE, AW GALSTON 1974 Potassium flux and leaf movement in Samanea saman. I. Rhythmic movement. J Gen Physiol 64: SATrER RL, P MARINOFF, AW GALSTON 1970 Phytochrome controlled nyctinasty in Albizzia julibrissin. II. Potassium flux as a basis for leaflet movement. Am J Bot 57: SATTER RL, N MORAN 1988 Ionic channels in plant cell membranes. Physiol Plant 72: SATTER RL, M SCHREMPF, J CHAUDRI, RL SATrER 1977 Phytochrome and circadian clocks in Samanea: rhythmic redistribution of potassium and chloride within the pulvinus during long dark periods. Plant Physiol 59: SArTER RL, Y Xu, A DEPASS 1987 Effects of temperature on H+ secretion and uptake by excised flexor cells during dark-induced closure of Samanea leaflets. Plant Physiol 85: SCHAUF CL, KJ WILSON 1987 Effects of abscisic acid on K+ channels in Vicia faba guard cell protoplasts. Biochem Biophys Res Commun 145: SCHAUF CL, KJ WILSON 1987 Properties of single K+ and Cl- channels in Asclepias tuberosa. Plant Physiol 85: SCHROEDER JI, K RASCHKE, E NEHER 1987 Voltage dependence ofk+ channels in guard cell protoplasts. Proc Natl Acad Sci USA 84: ScoTT BIH, HF GULLINE 1975 Membrane changes in a circadian system. Nature 254: SIGWORTH, FJ 1985 EPC-7 Users Manual. List-electronics, Darmstadt/Eberstadt, W Germany 32. TORIYAMA H 1955 Observational and experimental studies of sensitive plants. VI. The migration of potassium in the primary pulvinus. Cytologia 20: