Ionic movements in a single neuron.

Intracellular ionic activities of K and C1 were measured in the giant cell of Aplysia californica before and after the inhibition of cellular metabolism. Ion-specific liquid ion-exchanger microelectrodes were used to make these measurements. The equilibrium potential for K, calculated from directly measured intracellular and extracellular ionic activities was -80 mV, while the resting membrane potential was -45 to -55 mV. This indicates the presence of an active transport system which pumps K into the cell. Cooling below 3°C, ouabain (2 x 10[-4]M), 2,4-dinitrophenol (0.2mM), and cyanide (5-10mM) all caused a net efflux of K. The efflux, which was generally mono-exponential, had a rate constant which varied between 4.2 x 10[-5]sec[-1] and 5 x 10[-5]sec[-1] from cell to cell, with no significant difference among the fluxes consequent to the various treatments. Therefore, it was concluded that each of these treatments completely inhibited all active K influx. The permeability coefficient for K, calculated on the basis of the net efflux of intracellular K, was 1.5 x 10[-8]cm/sec. The calculated ionic conductance (chord conductance) was 1.5 x 10[-5]mho/cm[2]. These values are in good agreement with values obtained using electrical measurements. The equilibrium potential for CI was from 2 to 18 mV more negative than Em, which indicated the presence of an outwardly directed C1 pump. Cooling the cell below 4°C caused a passive influx of C1 and EC1 equaled Em within 20-80 minutes. The C1 influx was mono-exponential and had a rate constant of -1.4 x 10[-4]sec[-1]. Although ouabain increased C1 influx, its effect was much smaller than that of cooling. It is possible that the fall in Em caused by ouabain was responsible for the somewhat parallel decrease in EC1. Studies on the effects of DNP and cyanide on intracellular chloride activity were inconclusive. The passive net influx of C1 yielded a calculated permeability coefficient and calculated ionic conductances which were almost two orders of magnitude greater than the values calculated from independent electrical measurements. It is suggested the C1 crosses the membrane primarily as a neutral complex with some carrier cation.

IONIC MOVEMENTS IN A SINGLE NEURON by John McCandless Russell A thesis submitted to the faculty of the University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Pharmacology University of Utah August 1971 This Thesis for the Doctor of Philosophy Degree by John McCandless Russell has been approved June 1971 Super'v; sory Committee ------------ Reader, Supervisory Commlttee Reader, Supervlsory Committee , . . ------- Head, Major Department ACKNOWLEDGEMENTS I am grateful to Drs. E. A. Swinyard, A. M. Brown, J. W. Gibb, S. C. Harvey and D. M. Woodbury for their helpful advice to me while serving as members of my thesis committee as well as throughout my graduate education. I am especially indebted to Dr. A. M. Brown for his encouragement and personal interest in my education. I am also indebted to Drs. H. M. Brown and D. L. Kunze with whom I have had numerous helpful discussions during the course of my thesis research. Special thanks go to Dr. D. N. Franz for his thoughtful criticism of this manuscript. Thanks for technical assistance are extended to Sally Combes, Ton; Gillett, and Pablo Berman. Thanks also to Nancy Wilcox for typing this manuscript through all its rough drafts and in its present form. Finally, I wish to extend my appreciation to my wife, Donna, and my son, Joshua, without whom the present effort would not have been worthwhile. This investigation was supported by U. S. Public Health Service Grants (5-F01-GM-32, 600-04 and NS 09545-01). i ;i TABLE OF CONTENTS Page ACKNOWLEDGEMENTS iii ABSTRACT vi INTRODUCTION METHODS 4 Materials and experimental procedure 4 Measurement of membrane potential 7 Measurement of intracellular potassium and chloride activities 8 Correction factors used for long-term measurements 13 RESULTS 15 Baseline values for Em. aiK and aiCl in the giant ce 11. R2 i i 15 Effects of cooling on a K and a Cl 17 Effects of ouabain on aiK and aiCl 20 Effects of metabolic inhibitors 22 2,4-Dinitrophenol (DNP) 22 Potassium cyanide (KCN) 23 Further analysis of the effects of cooling on Em and Rm 24 Cellular volume changes 28 Calculation of PK, 9K' PCl and gCl 29 iv Page DISCUSSION 35 Intracellular activity coefficients 35 Evidence for active transport of K and Cl 35 Mechanisms of passive permeation of K and Cl across the membrane of R2 39 REFERENCES 41 FIGURES AND TABLES 45 VITA 76 v ABSTRACT Intracellular ionic activities of K and Cl were measured in the giant cell of Aplysia californica before and after the inhibition of cellular metabolism. Ion-specific liquid ion-exchanger microelec- trodes were used to make these measurements. The equilibrium potential for K, calculated from directly measured intracellular and extracellular ionic activities was -80 mV, while the resting membrane potential was -45 to -55 mY. This indicates the presence of an active transport system which pumps K into the cell. Cooling below 30 C, ouabain (2 x 10-4M), 2,4-dinitro- phenol (0.2 mM), and cyanide (5-10 mM) all caused a net efflux of K. The efflux, which was generally monoexponential, had a rate constant -5 -1 -5 -1 which varied between 4.2 x 10 sec and 5.9 x 10 sec from cell to cell, with no significant difference among the fluKes consequent to the various treatments. Therefore, it was concluded that each of these treatments completely inhibited all active K influx. The permeability coefficient for K, calculated on the basis of the net efflux of intracellular K, was 1.5 x 10-8cm/sec, The cal- culated ionic conductance (chord conductance) was 1.5 x 10-5mho/cm2. These values are in good agreement with values obtained using electrical measurements. The equilibrium potential for Cl was from 2 to 18 mV more negative than Em' which indicated the presence of an outwardly directed o Cl pump. Cooling the cell below 4 C caused a passive influx of Cl vi and ECl equaled Em within 20-80 minutes. The Cl influx was mono-4 exponential and had a rate constant of -1.4 x 10 sec -1 • Although ouabain increased C1 influx, its effect was much smaller than that of cooling. It is possible that the fall in Em caused by ouabain was responsible for the somewhat parallel decrease in EC1 • Studies on the effects of DNP and cyanide on intracellular chloride activity were inconclusive. The passive net influx of C1 yielded a calculated permeability coefficient and a calculated ionic conductance which were almost two orders of magnitude greater than the values calculated from independent electrical measurements. It is suggested that C1 crosses the membrane primarily as a neutral complex with some carrier cation. vii INTRODUCTION According to a current theory (Hodgkin, 1958; Katz, 1965) the electrical properties of cells having excitable membranes are based on two distinct principles: 1) active secretory processes which build up unequal distributions of inorganic ions (mainly potassium and sodium) across cell membranes, 2) passive systems in parallel with the active systems, which permit ions to move at differing rates down their electrochemical gradients. Thus, in relation to the extracellular fluid, living cells usually have a high concentration of potassium and a low concentration of sodium. The unequal distribution is thought to result from an energy-requiring active transport process located in the cell membrane (Glynn, 1964). This process is responsible for extruding sodium from the cell against its electrochemical gradient. Potassium moves into the cell presumably against its electrochemical gradient, although such movement could be passive (Cross, Keynes & Rybov/, 1965). The uneven transmembrane distributions of ions result in the establishment of an electrical potential difference (Em) across the membrane. When an ion, X, is in electrochemical equilibrium across a cell membrane, the equilibrium potential of the ion, EX' should be equal to Em' EX EX can be calculated from the Nernst equation: = 58 log (X)o/(X); at 20 0 C where {x)o is the external ionic concentration, and (X); is the internal ionic concentration. 2 In frog skeletal muscle ECl = Em' and chloride is therefore thought to be passively distributed across the membrane. However, evi- dence for a non-equilibrium distribution of Cl in a number of other tissues has been presented (see Discussion). Deviation of transmembrane ionic distribution from electrochemical equilibrium has been interpreted to be the consequence of an active transport system in the cell membrane. If cells are deprived of their immediate energy sources, usually ATP, by means metabolic poisons or anoxia, the downhill movement of ions occurs. This pro- cess can be reversed in squid axon by the intracellular injection of high energy phosphate compounds (Caldwell, Hodgkin, Keynes &Shaw, 1960). The development of electrodes suitable for intracellular recordings of the ionic activities of K and Cl permits recording of the electrochemical potentials in single neurons. The cell chosen for the present study was one in which the transmembrane distributions K and Cl are thought not to be in electrochemical equilibrium (Brown, Walker &Sutton, 1970). In order to cause net ionic move- ments, the processes responsible for the normal distribution of K and Cl had to be inhibited. By directly measuring the rate of net movements of these ions and the electrochemical driving forces involved, insight into the way by which these ions passively cross the membrane should be gained. In addition, it might be possible to elucidate the processes responsible for the active transport of these ions by the use of agents with differing mechanisms of action. The cell chosen for study was the giant cell of the abdominal 3 ganglion of the marine mollusc Aplysia californica (cell R2' according to the nomenclature of Frazier, Kandel, Kupferman, Waziri & Coggeshall, 1967). This neuron was selected because it is large, electrically quiescent and present in every ganglion. It has been characterized functionally, at least in part, in a preparation whose use for neurophysiological research has become widespread (Aplysia abdominal ganglion). Finally, while the cell does have glia cells associated with it, the glia probably do not represent as great a diffusion barrier as that presented by the Schwann cell layer that surrounds the squid giant axon. METHODS Materials and experimental procedure Experiments were conducted between July, 1970 and April, 1971 on the abdominal ganglion of Ap1ysia ca1ifornica, (Pacific Bio-Marine Supply Company). The animals were kept at 14 0 C in a seawater aquar- ium (Instant Ocean, Inc.). The abdominal ganglion was excised and pinned to a Sy1garJED resin in the bottom of an acrylic-plastic chamber having a volume of 2.0-2.5 m1. In most cases, the connective tissue capsule of the ganglion above the cell of interest was removed by use of a sliver of a razor blade to cut the capsule and a pair of fine forceps to pull it away from the cell. The chamber containing the ganglion formed the roof of a constant-temperature bath through which flowed a 40% V/V mixture of • methanol and water. The temperature of this mixture was maintained at the desired level by means of a refrigeration unit (PolyScience Corporation, Model KR-30) and a combination heater-pump unit with an internal thermostat (PolyScience-Haake, Inc., Model FJ). The fluid bathing the ganglion passed through a series of coils immersed in this constant-temperature bath and emerged at the rate of 1-3 ml/min into the chamber containing the ganglion (Fig. n. The temperature of the fluid bathing the ganglion was measured with a thermistor placed near the gi an t ce 11 . Experiments were begun 30 to 60 minutes after the capsule was 5 removed, during which time ASW (20 + 2oC) had been continually flowing over the ganglion. This procedure allowed the cell to recover from the minor trauma caused by removal of the capsule. In addition, the viscous material often found over the cell after removal of the capsule was washed away. This material may come from nearby cells which were destroyed during the dissection. If not washed away, the mate- rial tended to plug the recording electrodes and give large tip potentials in the Em electrodes and spurious ASW readings for the ionselective electrodes. After the 30-60 minute washing period, the cell was punctured first with an Em electrode. Five to 10 minutes were allowed for the electrode to be sealed into the cell before checking membrane resistance. If the resistance was suitable (about 1 was considered to be satisfactory. M~), the impalement When K- and Cl-sensitive elec- trodes were used simultaneously, they were mounted in a dual-head Zeiss micro-manipulator for impalement of the cell. brought close together (10-20 ~) The tips were and adjusted to exactly the same plane just above the cell with the aid of a dissecting microscope. They were then advanced while the voltmeter reading for either electrode was observed; it was possible to switch recording from the Ksensitive to the Cl-sensitive electrode by means of a Cary pH switch box. Impalement was signaled by a large DC shift in the voltmeter reading as well as some decrease in the measured Em' A good check for successful impalement by the ion-sensitive electrodes was to pass a 2-4 mV hyperpolarizing pulse through the Em recording electrode to see whether both ion electrodes also changed their reading by -2 to -4 mV. 6 The ion electrodes were usually within 50-150 p of the Em electrode, care being taken not to place any of the electrodes into the nucleus. The same voltmeter was used to register the output of both ion-exchanger microe1ectrodes. When all electrodes were satisfactorily in the cell, the experimental readings were begun. Readings from the voltmeter and Brush recorder were obta"ined every 10 minutes by taking 3 readings one minute apart, say at 9, 10 and 11 minutes, and averaging the resulting calculated activities and membrane potentials to get an average 10-minute value. The average values so obtained showed less fluctuation than did the individual readings. In the cooling experiments, the temperature of the fluid bathing the ganglion stabilized in 3-5 minutes after the start of cooling. Bathing solutions containing ouabain, 2,4-dinitropheno1, KCN or 10w-Cl solutions were introduced by switching from one reservoir to another. Usually a maximal flow rate (10 m1/min) for one minute was allowed in order to change quickly and completely the solutions bathing the ganglion. The fluid bathing the ganglion was an artificial seawater (ASW) having the same composition as the extracellular fluid of Aplysia (Hayes & Pe 11 uet, 1947; Brown & Berman, 1970). The sol ut ions were made with analytical reagent grade chemicals (Table 1). The pH of the solutions at room temperature (20 0C) was 7.507.60, measured with a Beckman expanded scale pH meter with a combination pH electrode (Beckman No. 39030) which is suitable for use with Tris-buffer. Since Tris-buffer has a significant temperature 7 coefficient (0.03 pH unit/oC; Sigma Technical Bulletin No. 106B, August 1967), it was necessary to lower the pH of those solutions which were to be cooled, to compensate for the alkaline changes in pH due to cooling. The necessary acidification could be accomplished by a slight increase in Tris-maleate and a slight decrease in NaOH concentrations. In the most extreme case, cooling to 0.5 oC, these changes were less than 1 mM, so that no significant change in ASW composition occurred. Substitution for C1 was made from conversion tables (Handbook of Chemistry and Physics, 1967; Robinson &Stokes, 1968) to maintain activities and osmotic pressure as close as possible to those of the control solution. Activity coefficients were calculated by Davies' modification of the expanded Debye-Huckel equation (Robinson & Stokes, 1968). The osmolality of all solutions measured by an osmom- eter (Advanced Instruments, Inc.) was 950 + 5 milliosmols/kg. Measurement of membrane potential Membrane potential (Em) was measured with conventional glass microelectrodes having tip diameters less than 5-12 MQ. l~ and resistances of The electrodes were filled with 3 M KC1, 0.6 MNa2S04 or 2M Na citrate. The electrode was connected to an electrometer with negative capacitance compensatlon, and the output was displayed on both a Tektronix 502 oscilloscope and a Brush Mark 220 penwriter. The reference electrode was a microelectrode filled with the same solution as the Em-senSing electrode. After its tip was broken off to be greater than 30 ]..I in diameter, the reference electrode was connected to ground. Tip potentials were determined by measuring the 8 baseline shift caused by breaking the tip of the Em-sensing electrode after completion of the experiments. Results from electrodes with tip potentials greater than 5 mV were rejected (Adrian, 1956). This retording arrangement yielded steady DC recordings in the temperature o range employed for these experiments (0.5 - 23 C). Membrane resistance was measured either by passing a test pulse through the Em-sensing microelectrode in a Wheatstone bridge circuit (Martin & Pilar, 1963) or by uSing two intracellular microelectrodes, 8 one to pass current across a 5 x 10 resistor and the other to measure Em' In some experiments, the membrane potential was set at certain levels by passing constant current across the 5 x 108n series resistor from one stimulator. This was done so that the test current applied across the 5 x 108n resistor from a second stimulator would result in a membrane voltage change that was well within the linear portion of the current-voltage curve near the control resting potential. Test pulses in both situations (1-10 x 10 -9 amps for 5-8 sec) induced Em changes of 1-10 mV. Measurement of intracellular potassium and chloride activities Micropipettes were pulled in the usual manner by a vertical pipette puller (David Kopf Instruments, Model 700C) from Pyrex (Corning Code 7750) glass with an internal diameter of 1.2-1.4 mm. The glass was first cleaned by placing appropriate lengths of the glass in a loosely covered 600 ml beaker containing about 50 ml of 100% ethanol and allowing the ethanol to evaporate to dryness over low heat. After pulling, the pipettes were si1iconized. For K 9 electrodes the s;liconizing agent was 2-4% (V/V) solution of tri-Nbutylchlorosilane in l-chloronaphthalene. For Cl electrodes, it was a 2-4% (V/V) solution of Siliclad (Clay-Adams, Inc.) in l-chloronaphthalene. In both cases the tip was dipped into the siliconizing solution until it filled to a height of about 200 ~. After dipping, the electrode was inserted, tip up, into a drill 0 hole in a metal block which was placed in an oven at 250 C. The K electrodes were left in the oven for 10 minutes, and the Cl electrodes were left for 1 hour. After removal from the oven, the pipettes were stored tip up, under an inverted beaker. To convert the siliconized pipette into an ion-specific electrode, 1-2 mm of liquid ion exchanger (Corning, Code 477317 potassium exchanger, or Corning Code 477315 chloride exchanger) was injected as far down the tip as possible through a 30-gauge needle. This left an air space between the exchanger and the pipette tip that was displaced simply by inserting a fine cat whisker or pulled glass needle into the exchanger until it broke the meniscus of the exchanger-air interface nearest the pipette tip. The K electrodes were typically stored in air until used, usually within 24 hours. The Cl electrodes were then stored with the rest of the pipette filled with 1.0 M KCl and the tip immersed in 1.0 M KC1. The KCl solution was simply injected through such a 30-gauge needle leaving no air trapped between the exchanger and the KCl solution. The K electrode was filled with 0.5 MKCl just prior to use. Both types of ion-electrodes were equilibrated in ASW for about 2 hours 11 before testing. These electrodes had a resistance of 10 10 _10 ohms, 10 a time constant of 0.5-1.0 sec and a drift of less than 0.5 mV per hour when immersed in any given solution. The potential of an ion-exchanger microelectrode is described by an empiY'ical equation (Nicolsky, 1937; Lev, 1964; Walker, 19l'1): zi/ z , E = Ec = (nRT/ziF) loge (a i + ~jKijaj J) (1) where E is the electric potential (volts); Ec is a constant (volts); 1 R is the gas constant (8.2 joules deg - mole -1 ); T is the tempera- ture (oK); F is the Faraday, (96,500 coulomb equivalent -1); n is an empirical constant chosen so the nRT/zF is the slope of the line when E is plotted as a function of loge a; with ~jKijaj = 0, zi and Zj are the valences of the ith and jth ions, respectively, ai is the activity of the ion the electrode is designed to measure (the principal ion), aj are interfering ions whose valence sign is the same as that of the principal ion, and K.. is the selectivity constant of the jth ion lJ with respect to the ith ion. When Kij<l the electrode has a higher selectivity for the principal ion than for the competing ion. Kij has been measured (Brown, Wa"lker & Sutton, 1970) in mixtures of one interfering ion and the principal ion at constant ionic strength. Equation (1) can be rearranged to the form shown in equation (2) and when the left side was plotted as a function of Z ./z. a. 1 J, the resulting straight line had a slope of K.. : J _ z/z. lJ exp (E-Eo) z.F/nRT - a. - K.. a. J 1 1 lJ J (2) Tab les 2 and 3 present the values of K.. for a number of i nterlJ fering ions for the potassium and chloride electrodes. were made in solutions whose ionic strengths were 1 M (~ 0.1 M (~ = O.l). Measurements = 1.0) and 11 Potential measurements were made with the ion-specific electrode connected through a Cary pH meter switch box to the input of a vibrating reed electrometer (Cary Model 401) with an input impedance of 16 10 ohms. The reference electrode was the same as that used to measure Em' namely, a 3M KCl micropipette with a broken tip. This recording arrangement provided stable reference values and was much more practical than a calomel half-cell. The output of the elec- trometer was displayed on a digital voltmeter (Honeywell Model 333R). The electrodes were calibrated in a series of KCl solutions 2 varying in concentration from 1.0 to 1.0 x 10- M, prior to and after measuring intracellular activities. The calibrating solutions were maintained at the experimental temperatures to be used by placing them in the 0.5-ml chambers on top of the constant temperature bath When cooling the electrodes, it was important that they be (Fig. 1). cooled slowly (over a 1- to 3-minute period); otherwise the exchanger oil rapidly leaked out and the electrode was destroyed. It was also important to immerse the electrode in the cooled fluid to a level above that of the exchanger oil; otherwise the temperature gradient in the oil produced large fluctuations (10-15 mY) in the potential output of the electrode. This made long-term intracellular recording "impossible. By plotting potassium or chloride ion activities against the voltage output of the liquid ion-exchanger microelectrode at each activity, one obtained the slope of the electrode. For potassium electrodes, the slope was 58 mV per lO-fold change at 200 C and 54 mV at lOCo For chloride, the slopes were 56 mV and 53 mV respectively. 12 These slope changes were accounted for in equation (1) by the decrease in temperature. In addition to measuring the electrode voltages in the KCl solutions, their voltage output in ASW was also read. This, then, gave external potassium and chloride activities (aoK and aOCl)' Knowing the slope of the activity-voltage function for the electrode, the external activity of the ion of interest, and the membrane potential one could calculate the internal activity. By subtracting the Em from the intracellular voltage reading of the ion-sensitive electrode, the resulting net voltage reading was used to read the intracellular ion activity of the principal ion directly from the calibration graph. Since the selectivity of the electrodes for the principal ion was in most cases quite large (i.e. 50:1, K:Na), the contribution by interfering ions was ignored. Therefore, the equation for any electrode becomes: Eo = Ec + nRT/zF loge ao where Eo is the potential of the ion electrode in the ASW. (3) Upon insertion of the ion electrode into the cell, we have: (4) where Ei is the potential of the ion electrode inside the cell and Ec ' n, R, T, Z have the same meanings as in equation (1) and ao and ai are the extracellular and intracellular ionic activities respectively. The factor nRT/zF is the electrode slope (/b/) determined graphically as described above. Ec is a constant potential inherent within each recording set-up, and was assumed not to vary when the cell was impaled. Evidence that this assumption was true was provided by the 13 close agreement between intracellular ionic activity values obtained by this method and those obtained by other methods (Cornwall, Peterson, Kunze, Walker & Brown, 1970; Chow, Kunze, Brown &Woodbury, 1970). Solving equation (4) for intracellular potassium activity (a\) yields: E.1 - Ea - Em i 0 a K -- a K . 10 (5) Ibl and for intracellular chloride activity yields: i 0 Ei - Eo - Em (6) Ibl a Cl a Cl I 10 A more complete discussion of liquid ion-exchanger microelec= trodes is found in a recent paper by Walker (197l). Correction factors used for long-term measurements The experiments required that a cell be continuously impaled with the Em recording microelectrode for relatively long periods of time (4-7 hrs), and thus a certain amount of baseline drift occurred. To minimize drift in the recorder, a Brush Mark 220 DC recorder which drifted only ± 0.2 mV/8 hours was used. However, the drift in the cathode-follower and in the electrode itself, might on occasion, have been greater. Therefore, experiments in which the baseline had drifted more than 5 mV were rejected. Changes of 5 mV or less were assumed to have occurred linearally with time, since such changes were linear over several hours when the electrodes were placed in ASW. The voltages of the ion-sensitive microe1ectrodes sometimes also tended to drift with time. Most of the drift occurred during the first two hours; therefore, the electrodes were allowed to equilibrate in ASW for two hours before being calibrated. When during the course 14 of an experiment the change in baseline was greater than 3 mV, the experiment was rejected. Baseline changes of the ion electrodes were also found to occur linearal1y with time. However, changes in the slope of the ion-sensitive electrodes which could not be accounted for by temperature changes were cause for rejection of the experilllent. RESULTS i i Baseline values for Em' a K and a Cl in the giant cell, R2 It was necessary to record continuously for 4-7 hours to obtain accurate rate constants for the net fluxes of potassium and chloride ions in the giant cell. This raised two questions: 1) Was there a significant leak of K and Cl into the impaled cell from microelectrodes filled with 3M KCl and having resistances of 5-10 Mn? 2) How well did a cell, into which as many as three micropipettes had been inserted, tolerate 4-7 hours of ~ vitro conditions? To answer these questions, Em was monitored continously with mi croe 1ectrodes fi 11 ed with either 3M KC1, 2M Na citrate or, in one case,0.6 MNa2S04' Differential recording, to avoid drift due to liquid junction potentials, was used when the reference electrode contained S04 or citrate. Internal potassium activity (a i K) or internal chloride activity (a i Cl ) or both simultaneously were measured with the liquid ion-exchanger microelectrodes. As was the case for all cells reported herein, the cells recorded from were quiescent, but action potentials could be generated by injecting depolarizing current intracellularly. Table 4 shows the results of seven experiments. Note that Em improved after impalement, that is, Em became more negative. Figure 2 shows the increase in Em with time in one experiment. Em usually reached its maximum control value 30 to 80 minutes after the 16 last impalement of the cell. During the period when Em was becoming i more negative, a and aiCl changed very little. The cells in which K Em was measured with 3M KC1-filled microelectrodes showed an average of 5 mV improvement in Em' a slight decrease in a'K and a slight increase in aiCl over a period of 3-6 hours. The group of cells in which Em was measured with microelectrodes containing either 2M Na citrate or 0.6 MNa2S04 also exhibited an i improvement in Em (about 4 mV) while a K increased. The average increase in aiK is heavily biased by one experiment since the other two show relatively small changes in a i K. In the one cell in which ai Cl was measured, it declined slightly. Thus, the solutions used to fill the Em-recording micropipettes had minimal effects on aiK and aiCl and leakage of KCl from the recording pipette did not appear to affect the results obtained in the present experiments. Two important findings in these preliminary experiments and all subsequent experiments were: 1) that EK was always much more negative than Em' and 2) ECl was usually more negative than Em (occasionally it was less negative than Em; ECl was never equal to Em in any of the 36 cells where aiCl was measured). EK and ECl were calculated from direct measurements by means of the Nernst equation. These facts indicate that an active process was involved in maintaining the transmembrane electrochemical potentials for K and possibly Cl. Therefore, the effects of cooling, ouabain, cyanide, and 2,4-dinitrophenol on intracellular potassium and chloride activities were investigated, since these are all treatments thought to interfere with or abolish active transport. 17 i i Effects of cooling on a K and a Cl o As expected, cooling the giant cell below 3-4 C caused K and Cl to approach electrochemical equilibrium across the cell membrane. If enough time was allowed, such an equilibrium was actually achieved for K (3 experiments; Fig. 5) as well as for Cl (Fig. 3; Table 5). response of Em and membrane resistance categories: (~) The to cooling fell into two 1) A rapid, initial decrease in Em followed by a slower decrease; this was associated with an initial decrease in then gradually returned to control values. Rm which 2) No immediate change or a slight hyperpolarization followed by a slow decrease of Em' accompanied by a large increase in Rm, which gradually returned toward control values. 0 i o Seven cells were cooled to between 0.5 and 2.5 C, and log a K was plotted against time for each cell (Figs. 3,4,5). These figures show that during the first 2-4. hours of cooling, aiK declined exponentially with a single rate constant (k). Rate constants (see below) were calculated from regression lines (determined in each experiment i by the method of least squares) from the plot of log a X vs. time, where ai x is the internal activity of ion X. Experiments reported here had correlation coefficients for linear regression greater than 0.95; therefore all points common to a single regression line define the same rate constant. Values for aix were taken directly from a calculated regression line for a 10-minute interval and the rate constant was calculated as follows: k = 2.303 (log a\l - log a\2) / time (sec) lB i ; where a Xl was ionic activity at time 1 and a X2 was the ionic activity ten minutes later. i a Xl could be selected as desired from any point on the regression line. This was necessary since it was important to determine if the membrane permeabilities and conductances to K or Cl changed during the course of the experiment (see Table 9). The average rate constant for the first 2-4 hours for these seven cells was 4.71 ± .76 x 10-5 sec -l (S.E.), which gives a time constant, t hours. l/e , of 5.9 These values can be compared with values for squid giant axon of B.B x 10-5sec -l (Shanes &Berman, 1955) and 5 x 10-6 sec -l (Hodgkin & Keynes, 1955a). Three of the seven cells showed a later, quicker decline in ai K commencing 2-3 hours after the cell had been cooled (Figs. 3 &5). This second component had a rate constant of 17.5 -5 -1 ± .7 x 10 sec (S.E. )(t = 1.6 hrs). In these three cells, a i K and l/e Em became relatively constant about two hours after the onset of the faster rate constant. At this time EK = Em = -10 experiment, the cell was rewarmed to 23 0 C. mV. In one such ai K and Em began to increase, but EK was always more negative than Em during rewarming (Fig. 5). 0 0 Cooling to temperatures of 9 _11 C had little, if any, effect on i a K or aiCl for periods of 2~ hours (Fig. 6). On the other hand, a 0 maximum effect was obtained at 3 C (no further effect was detected at lower temperatures). 0 Thus, all cells cooled below 3 C were con- sidered as a single group. Since influx of K is an active process (E K more negative than Em)' K influx would be expected to show a high Q10 over a range of temperatures including room temperature. The present data, showing both a threshold and a maximum for temperature 19 effects over a narrow range of temperatures, probably result from the fact that net K flux rather than K influx alone was measured. Thus, the net K flux includes passive K efflux as well as K influx, both of which fluKes may be temperature sensitive to varying and perhaps opposite degrees. Intracellular chloride activity was measured in 11 cells cooled o 0 to between 0.5 and 4.0 C, a range of temperatures over which the i changes in aiel were similar (Table 5). was measured simultaneously (Fig. 3). In five of these cells, a K A rapid increase in aiel which appeared to be monoexponential occurred between 5 and 40-80 minutes after cooling. stant of -1.36 ± This inward chloride movement had a rate con-4 .25 x 10 sec -1 (t l / e = 2 hrs). The chloride influx then slowed down as Eel (calculated from the Nernst equation by means o . of the directly measured values of a Cl and alel) approached and became equal to Em' This usually occurred within 40 to 80 minutes after cooling the cell (Table 5). Thereafter, influx slowed even further as ECl remained equal to the slowly falling Em' The results were similar whether or not aiK was being simultaneously nleasured. It should be noted that the effect of cooling to 10C on the Nernst equation is to decrease EC1 by 3 to 4 mV, whereas the average total decrease in EC1 was 10 mV (Table 5). Chloride influx showed a strong temperature dependence. At gOC, i a Cl changed slightly if at all in 3 experiments (Fig. 6), whereas cooling to 40C caused a rapid influ~ with a rate constant of -1.04 x -4 10 sec -1 which is not significantly different (P>0.2) from values for chloride influx at lower temperatures (1 0 - 2.5 0C). 20 i i Effects of ouabain on a K' a Cl and Em Thus, the results of cooling experiments support the view that both K and Cl are actively transported across the cell membrane. To obtain information on the nature of the metabolic processes involved, the effects of ouabain, a well-known inhibitor of the Na-K activated ATPase system (Glynn, 1964) were examined. The concentration of ouabain required to obtain the maximum rate of K efflux was determined by testing three concentrations: 2 x 10-5M, 2 x 10-4M, and 2 x 10-3M. After stable control values were obtained, one of the two lower ouabain concentrations was applied for 1 hour followed by the application of one of the higher concentrations. This avoided the necessity of using large numbers of cells to determine the optimal dose. The rate constants of K efflux for any two concentra- tions of ouabain tested on the same cell were compared, and no significant difference among the three concentrations tested was found in 3 cells. The membrane depolarization produced by 2 x 10- 3Mouabain was no greater than that by 2 x 10 -5 Mouabain. Ouabain (2 x 10-4M) was used since this approximates the concentrations used by workers investigating the electrogenic pump (Kerkut &Thomas, 1965; Carpenter &Alving, 1968; Pinsker & Kandel, 1969; Marmor & Gorman, 1970), and it was important to be sure a maximally effective ouabain concentration was being used. When ouabain was applied, a marked depolarization of the giant cell began within 1-2 minutes and continued until reaching a maximum within 30 to 40 minutes. The depolarizing effect of ouabain was 21 greatest for those cells having the most negative control Em and least for cells having less negative Em' Figure 7 demonstrates the linear relationship between control Em and the change in Em caused by ouabain which occurred before any significant changes in ai K had occurred. This relationship indicates that ouabain caused about 1 mV depolarization for every mV by which control Em exceeds -36 mY. This occurs in i the absence of any significant change in a and hence in EK• K Intracellular potassium activity began to fall within 10-20 minutes of ouabain application, although lag periods up to 30 minutes were occasionally noted (Fig. 8). average rate constant of 5.96 ± In 14 cells, aiK fell with an .58 x 10 -5 sec -1 (t lie = 4.65 hrs). Although this was somewhat faster than the rate constant for K efflux in cells cooled to 0.5-30 C, the difference between the two groups was not significant (P> 0.6). Intracellular chloride activity and aiK were measured simultaneously in 5 ouabain-treated cells (Fig. 8). Although aiCl did increase and the increase was statistically significant (P< 0.02), Eel never equaled Em even after 6 hours of ouabain treatment (Table 6). The difference between Em and ECl was actually greater after the ouabain treatment than before the treatment. Chloride influx after ouabain appeared to be exponential in only one of the five cells. this case, the rate constant was -6.4 x 10-5sec-l which was about half that observed in cooled cells. In 22 Effects of metabolic inhibitors Cooling below 30 C and ouabain had similar, if not identical, effects on net K efflux. This suggested that the active transport of K was mediated solely by some system inhibited completely by ouabain, the most likely system being an Na-K activated ATPase. Complete inhibition of cellular metabolism would be expected to have effects 0 similar to those obtained by cooling to 3 C or less, if the latter procedure, in fact, completely inhibited active transport of K. Therefore, experiments were done with 2,4-dinitrophenol (DNP) and potassium cyanide to inhibit cellular metabolism. In addition, the effects of these agents on aiCl were also studied. 2,4-Dinitropheno1 (DNP) DNP inhibits formation of adenosine triphosphate (ATP) by uncoupling the electron transport of the cytochrome oxidase system from oxidative phosphorylation of adenosine diphosphate (ADP) (White, Handler &Smith, 1964). Thus, it should inhibit 95% of ATP forma- tion, leaving only anaerobic glycolysis as a source of AlP assuming that this cell has a complete metabolic scheme. DNP was applied to two cells in a concentration of 0.2 mM (Hodgkin & Keynes, 1955a). i . Em and a K were measured; however, alCl could not be measured in the presence of DNP, since at pH 7.55 this agent is essentially present entirely in the form of the anion, DNP-. - - The selectivity of the chloride electrode for Clover DNP is small. The mean rate constant for K efflux, which was monoexponential in these cells {Fig. 9~ was 5.13 ± .83 x 10-5 sec -1 • This value does 23 not differ significantly from cooled or ouabain-treated cells (P>0.3). K efflux commenced within 10 minutes of DNP application. The effect of DNP to cause an initial depolarization prior to significant changes in ai K, similar to the effect of ouabain, is included in Fig. 7. Em fell rapidly for the first 30 minutes, and then more slowly for the remaining 2.5 hours (Fig. 9). Potassium Cyanide (KCN) i Since it was not possible to measure a Cl in the presence of DNP, it was necessary to use another metabolic inhibitor of ATP formation. Cyanide (CN) acts by combining with ferric ion in the prosthetic heme group of cytochrome oxidase, preventing its reduction. Thus, it prevents electron transport and the resulting oxidative phosphorylation ceases. The effects of two concentrations of CN (see Methods) on Em' aiK and aiCl were investigated. The CN solutions were made just before testing. Three cells were treated with 5 mM CN, a concentration chosen because it exerted a maximal effect on active cation transport (Hodgkin &Keynes, 1955a) in squid axon. 10 mM CN (Table 7). One cell was treated with Notice that K efflux began within 20-30 minutes of application of the cyanide (Fig. 10). Since the number of experiments was small, it was not possible to determine if the effect of the 5 mM CN on aiK was significantly less than those of cooling, ouabain, or DNP; however, the effect of 24 10 mM CN was well within the range of the other treatments (Table 7). The effects on aiCl were small and inconsistent, aiCl actually decreased in three cells and increased in only one (Table 7). A rapid depolarization of Em was seen with both concentrations of CN used. There was no difference in the degree of depolarization with either concentration. By plotting these changes in Figure 7, one sees that, in this respect, CN and ouabain had similar effects. Further analysis of the effects of cooling on Em and Rm In the giant cell of Aplysia, it appears that virtually all the transmembrane ionic current is carried by K, Na and Cl. Therefore, changes in membrane resistance must reflect changes in the ability of anyone or all of the ions to cross the membrane. Cooling to 0.S-2.SoC caused changes in membrane resistance (see p. 17) and Em' The immediate effects fell into two groups. One group showed an immediate (within 10 minutes) depolarization and mean Em changed from -50.3 ± 1.9 to -45.2 ± 1.6 (±S.E.; n=12) when cooled from room temperature. This change in Em could be predicted from the Goldman equation, taking into account the change in temperature. However, as will be noted below, a similar depolarization occurred when the temperature was lowered to lOoC. This latter response cannot be accounted for merely by the passive effect of decreased temperature since a lOoC fall in temperature would only cause a 2 mV decrease in a potential based solely upon those passive properties accounted for by the Goldman equation. 25 In the second group. individual cells showed either no change in Em or slight hyperpolarization upon cooling. An average Em change from -47.2 ± 1.4 mV to -50.2 ± 1.2 mV (± S.E.; n=9) occurred 10-20 minutes after cooling. It may be of interest that 10 of the 12 cells in the first group were examined between the months of October and January. while 6 of the 9 cells in the second group were studied in July and August. In addition. the cells which hyperpolarized after cooling tended to have slightly lower control resting potentials (see above). Depolarizations of equal magnitude were obtained over a temperao ture range of 0.5 to 10 C. -46.0 ± Thus. Em decreased from -51.5 ± .2 mV to o .5 mV when cells were cooled from room temperature to 10 C. 0 The effects of temperatures between lOoC and 20 C were not investigated. No data were obtained for the cells which hyperpolarized at o 0 temperatures other than 20 C and 0.5-2.5 C. Cells in both groups showed a second slow phase of depolarization which averaged 5 mV over the subsequent 2 hours of cooling. The cells studied between October and January showed characteristic changes in Rm upon cooling (Fig. 11). Cooling to 60 C caused little or no change in Rm. but below SoC Rm fell to an average of about 66% of its control (20 oC) value. Rm then increased and returned to control values 30-60 minutes after the onset of cooling. On several occasions, Rm exceeded control values after 2 or 3 hours of cooling. Cells cooled in July and August usually showed an initial large increase in ~ (2-4 times) followed by a decrease which reached the 26 control value after 1-2 hours of cooling. By contrast, ouabain caused no significant changes in membrane resistance. Membrane resistance was not measured in the presence of either CN or DNP. The foregoing results make it clear that cooling the giant cell o to 0.5-2.5 C can cause rather marked changes in membrane conductance (gm,where gm = l/Rm)· Since gm = gK + gCl + gNa' a conductance change in any of three current-carrying ions could be involved. However, the rate constant for K efflux was the same for cooled cells as for cells treated with ouabain, DNP, or CN. Values of gK and PK in these different situations were also similar (see below). It appears unlikely, therefore, that gK was altered at the lower temperatures. An increase in gC1 was suspected since rapid net transmembrane movements of C1 were seen only in cooled cells. ECl to become equal to Em' Only cooling caused It was important to know whether this effect was due to inhibition of active transport or to an increase in the ability of Cl passively to cross the membrane. Another indepen- dent measure of gCl was clearly necessary. The membrane conductance of any ion is expressed by the equation, gx = TX ~ gm' The total membrane conductance (gm) was measured directly and TX' the transport number, was determined by the change in membrane potential produced by a change in the concentration of the ion being studied. Thus, aEm/alogX o = 58 TX (Brown et !i, 1970). The effect of changes in tl~ on Em was investigated at 20 0 C and at 0 0.5-2.5 C. 0 At 20 C, substitution of S04 for Cl to yield 1/2 and 1/4 27 normal chloride concentrations showed that Em decreased about 18 mV for a lO-fold decrease in (Cl) (n=6). After cooling to 0.S-2.S oC, o mixed results were seen. Three cells showed hyperpolarizing responses and a fall in total membrane conductance when they were cooled. Three others demonstrated a depolarizing response and an increase in total membrane conductance. Despite the opposite effects on Em and Rm, gCl decreased 14-3~~% in both groups of cells when they were cooled (Table 8). Thus, these small decreases in gCl might contribute to the depolarization upon cooling seen in the October-January group of cells, but are in the wrong direction to account for the rapid, net influx of Cl after cooling. The evidence presented above indicates that large changes in gK or gCl are not responsible for the changes in gm observed after cooling. Effects of cooling on gNa were not investigated in the present work, but others have shown temperature-dependent changes in gNa without changes in gK in the Aplysia giant cell (Carpenter, 1970; Marchiafava, 1970) and squid giant axon (Hidalgo &Latorre, 1970). It is concluded that the effects of cooling (at least for the first 2-3 hours) on gK and gC1 are minimal and that the changes in Rm reflect changes primarily in gNa' Therefore, the measured rates of K and C1 movement after cooling accurately reflect the membrane permeability to these ions at normal temperatures. 28 Cellular volume changes One of the functions of active transport is probably regulation of the cellular volume (Woodbury, 1965a). Therefore, it was necessary to estimate any possible increase in cellular volume that might have resulted from cessation of active transport. From direct measurement of aiK and aiCl a volume change can be calculated in the following way. During the first hour a cell cooled to 10C loses, on the average, 33 mM K and gains 10 mM C1. To maintain electroneutrality, some cation, presumably Na, is assumed to enter the ce 11 . So 33 rnM of Na moves into balance the K that is lost and 10 mM Na moves in to balance the gain in Cl. Cellular osmolarity is not disturbed by the Na-K exchange, but 20 mi1liosmols per liter were gained by the Na and Cl influx. Cellular osmolarity during control is assumed to be 950 milliosmo1s per liter, the same as that of extracellular fluid. The increase in osmolarity is 20/950 X 100 during the first hour of cooling. = 2.1% An intracellular volume increase of 2.1% would decrease aiK by only 2-3 mM, less than 1/10 the actual decreas~ seen. After the first hour, the volume changes would occur much more slowly since aiCl increases slowly after the first hour. Cells treated with ouabain or KCN also would experience little increase in volume since aiCl changes are slow and small. Such a calculation assumes that, if there is a fraction of intracellular bound water, it does not change. In addition, it must be assumed that intracellular proteins do not change their osmotic properties while the intracellular ionic composition is changing. 29 Calculation of PK, gK' PCl and gel The rate at which an ion passively crosses a biological membrane is a function of its electrochemical gradient and the membrane permeability. In the present experiments, the electrochemical gradient was measured directly ( as Em - EX) and ionic conductance and permeability can be derived by means of the constant field equation and/ or Ohm's law (see p. 32). in these derivations: The following assumptions have been made a) that backflux was negligible,at least early in the course of K efflux or Cl influx, when the calculations were made; b) that the giant cell is roughly a sphere for purposes of determining the intracellular volume; c) that cellular surface area can be estimated by means of a specific capacitance of 1 ~F/cm2 (Katz, 1965); d) that the cellular surface area so measured is the same surface across which the ionic fluxes are taking place. The first assumption is based on findings in squid axon that show ionic influx and efflux are not independent processes (Hodgkin & Keynes, 1955b). The interaction between the two processes is such that, when there is a net movement in one direction, the reverse unidirectional flux is inhibited as though there were only single file columns in which ions could cross the membrane. of K, any random influx would be inhibited. Thus, during efflux Likewise, during Cl influx, random efflux would be inhibited. The second assumption is based on visual observations of the giant cell which usually has a spherical shape. The average diameter measured with an eyepiece reticule, accurate to within 25 ~, was 30 519 ± 19~. This gives an average cell volume of 73.5 nano1iters. The third assumption is necessary because the giant cell is not a smooth sphere (Coggeshall, 1967) but is highly invaginated. Total membrane capacitance (Cm), membrane resistance and the time constant (T) for charging the membrane capacitance are related as follows: T = RmCm' Since the time constant can be measured as that time re- quired for the cell to reach 63% of its change in Em when a squarewave pulse of long duration is passed intrace11ular1y and (~ calculated ~ may be = Em/1m)' we can calculate a value for Cm• Five cells were studied with an Em-sensing electrode and a current-passing electrode (see Methods). was 175.4 The average time constant 12.7 msec, the average Rm was 1.59 ± .2 Mn and the average membrane capaci tance was 1. 1 x 10- 1 ~F. For a speci fic ± capacitance of 1 ~F/cm2, the cellular surface area would be 1.1 x 10- 1 cm2• This is 13 times larger than the surface area calculated on the assumption that the cell is a smooth sphere. This agrees well with Carpenter's (1970) estimation that -invag-ination results in a 10-fold increase in surface area. As for the fourth assumption, one can estimate the range of errors involved in attributing the flux data to exchanges across the somal membrane alone, assuming that exchange across the axonal membrane makes a negligible contribution to the measured flux. estimates, an axon of 2 mm in length and 100 1967) connected to a soma 500 ~ ~ For these in diameter (Coggeshall, in diameter will be assumed. In case A, the surface area of the soma is 13 times that of a smooth sphere. 31 The ratio of volume to surface area, VIA, when the axon is included would be smaller by 30% than when the axon is excluded. Hence, the reported flux would be larger by this same per cent (see Equation 7 below). T In case B, the total surface area, estimated from the relation = ~Cm' included the axon, but the volume was exclusive of axonal volume. In this case, the VIA ratio would be larger by 24%, and the reported fluxes should be reduced by the same per cent. the errors involved in ignoring the axon range over ± In summary, 30%. With the foregoing assumptions, K efflux or Cl influx can be calculated from the following equation: i i Mx = a Xl a X2 V A (7) t where Mx is ionic flux of the Xth ion in M/sec·cm 2 ; (aiXl - ai X2) is the change in intracellular activity in M/L; t is time in seconds; V is cell volume in liters; and A is surface area in cm2• The ionic current is then calculated as: IX = Mx . F (8) where Ix is ionic current carried by the Xth ion in amp/cm 2 and F is Faraday's constant. Ionic chord conductance, that is, the ease with which the ion moves across the membrane in response to an electrochemical driving force, can be calculated in two ways. The first is based on a simple ohmic relationship: (9) where gx is the chord conductance for the Xth ion in mho/cm 2 , IX and Em are as previously defined, and EX is the equilibrium potential of 32 the Xth ion. A second method of calculating ionic conductance is derived from the constant field equation and depends upon EX having a value similar to E , a condition satisfied by C1 but not by K in the giant cell m (Hodgkin & Horowicz, 1959). Hence, gCl = PCl . F3 Em . aOC1 . aiCl (RT)2 ; - aO a Cl Cl (10) where PCl is the chloride permeability calculated according to equation 12, and the other symbols are as previously defined. Finally, membrane permeability (P x in cm/sec) can be calculated i from the directly measured Em' a x' aOx and the calculated Ix (Hodgkin &Horowicz, 1959). For PK the equation is: (11 ) and for PCl it is: RT 1- exp(-~!DF/RT) PCl -_ I Cl '"fCZF' -- . exp(-EmF/RT) m aOCl - a'Cl (12) The results of these calculations for cooled and ouabain-treated cells are summarized in Table 9. The PK reported here ;s about two orders of magnitude less than that reported by Hodgkin and Katz (1949) for squid axon and frog skeletal muscle bathed in normal potassium concentrations (Hodgkin & Horowicz, 1959). This should not be sur5 prising, since the membrane resistance of the giant cell (1 x 10 ohm' cm 2) is about 100 times greater than for squid axon (7 x 10 2 ohm'cm2) and skeletal muscle (4 x 10 3 ohm'cm 2) (Katz, 1965). Pa:t of the 33 difference between the values for the giant cell and the other two tissues may depend upon the observation that PK can vary with changes in the electrical driving force (Em - EK). It has been shown that when squid axon or frog skeletal muscle is depolarized such that Em - EK = 30 mV (a value similar to that in the giant cell), the PK for these tissues declines by about la-fold (Hodgkin & Keynes, 1955b; + Hodgkin &Horowicz, 1959). Thus, at least part of the differences in K permeabilities may be accounted for. In my experiments, gx and Px were determined from the measured net fluxes, MX' Brown et ~ (1970) determined gx and Px from electrical measurements. As noted earlier, aE m/a10gXo= 5BT x ' where Tx is the transport number of the Xth ion and equals the fraction of the total membrane current carried by this ion. For potassium, TK = 0.57. Using this value, one can predict a value for gK in the present experiments and compare the values measured chemically in the present experiments with the value predicted from the electrical ~easurements of Brown et ~ (1970). In cells cooled to 0.5-2.5 0C, Rm was 1.39 Mn (n=5) which means that gM (total membrane chord conductance) equals 7 7.2 x 10- mho. The average cell surface area was 1.1 x 10-lcm2 so specific membrane conductance was 6.5 x 10- 6 mho/cm2• Since gK = TK • gM' a gK of 3.7 x 10- 6 mho/cm2 would be predicted from these electrical measurements. Considering the differences in experimental methods between the experiments of Brown et ala (1970), and the present ones, the results for gK are in good agreement (Table 9). Brown et ala (1970) reported a PK of 1.42 x 10-Bcm/sec. This is in good agreement with the present data (Table 9). It is important 34 to note that the only assumption held in common by the two methods was the one concerning membrane surface area. There is a 100-fold difference between the PCl and gCl results of Brown et ~ (1970), corrected for membrane infoldings, and those of the present report. This seems too large a difference to attribute to experimental error. The gCl measured chemically in the present ex- periments is 21 times larger than the total membrane conductance. It therefore appears that more Cl is crossing the membrane than can be accounted for by electrical measurements. Put another way, some Cl may be crossing the membrane in a form that is not detected by electrical measurements. possibly as an electrically neutral complex with a carrier (Eisenman, Sandblom &Walker, 1967). An estimate of this electrically invisible Cl flux can be obtained by using the transport number for chloride obtained earlier (Table 8). TCl was estimated to be 0.31 of the total membrane current; thus gCl should be (0.31) (6.5 x 10- 6 mho/cm2), which equals 6 2.02 x 10- mhos/cm2. This figure represents only a little over 1% of the 9Cl calculated from the net influx of Cl (Table 9). DISCUSSION Intracellular activity coefficients Sato, Austin, Va; and Maruhashi (1968) report an internal concentration of K of 232 mM for neurons in the abdominal ganglion of Aplysia. i In the present experiments, a K was 170.7 ± 3.4 mM (n=52). Assuming ionic strengths of cytoplasm and ASW to be similar, the calculated activity coefficient for intracellular K (f i K) was 0.74. This agrees well with f OK for ASW of 0.69 measured directly or calculated from Davies' modification of the extended Debye-Huckel equation. This means that intracellular potaSSium exists freely in solution. The hypothesis of Ling and Ochsenfe1d (1966) that internal cations are bound to a lattice of fixed negative charges does not appear valid in this case. Calculations of the activity coefficient of Cl with the use of the present aiCl values and the Cl concentrations which Sa to et ~ (1968) found in cells giving a hyperpolarizing or H response to acetycholine (R 2 also has an H response to acetylcholine) gave fi Cl of 0.88 which was somewhat higher than expected for a free solution. The discrepancy is attributed to the different methods but is in a direction that mitigates against any sizable amount of bound Cl inside the cell. Evidence for active transport of K and Cl Active transport of an ion across cell membranes is likely when 36 1) the ion is not in electrochemical equilibrium across the cell membrane (i.e. Ex ; Em)' and 2) the ion can achieve electrochemical equilibrium when the metabolic processes of the cell are inhibited. In the giant cell, Em ; EK ; EC1 ' When the cell is cooled to 0.50 2.5 C, Em = EK = ECl (Fig. 3 & 5). From the above definition, it is concluded that both K and Cl are actively transported across the membrane of R2. Tbe.transpOrtsystems involved can be characterized to some extent from the results of the present studies. The rate constant for net efflux of K was essentially the same for cooled cells and cells treated with ouabain, DNP or CN. finding allows only two possibilities: This 1) either all active trans- port had been abolished by these various methods, or 2) a certain amount of transport persisted for a variable period after treatment in each instance. The persistent component would have to have the same magnitude in each case. This seems unlikely because of the different mechanisms of action of these various metabolic-inhibiting procedures. And its persistence would be temporary since K eventually was distributed passively across the cell membrane. Moreover, if a temporary component of active transport persisted, then the rate constant of K efflux should increase when this component is exhausted. i In fact, the decrease in a K after metabolic inhibition was monoexponential with time in each instance, which is strong evidence against there being such a residual component of active transport. the three cooling experiments in which the rate constant of net K efflux showed an increase after 2-3 hours of cooling, membrane In 37 conductance had doubled about 30 minutes previously (Fig. 5). There- fore, the change in membrane conductance probably explains the appearance of a second, faster component of K efflux. Taken altogether, the data are interpreted as indicating that the inhibition of active transport was virtually complete in each of the four methods used. Since the monoexponential decline in aiK began within 10-30 minutes of cell treatment, this implies that ATP stores must be small in the giant cell. The effect of ouabain permits some speculation regarding the K transport system. Ouabain blocks the Na-K ATPase in the membrane of red blood cells, squid axon, brain and kidney (Glynn, 1964). The finding that ouabain had the same effect on net efflux of K as the other procedures which completely blocked active transport suggests that a Na-K ATPase system, similar to those referred to above, accounts completely for active K influx. There does not appear to be an ouabain-insensitive component in R2 as has been reported for skeletal muscle (Sjodin &Beauge~ 1968) and squid axon (Baker, Blaustein, Hodgkin &Steinhardt, 1969). The difference between Em and ECl is smaller (2-18 mV) than that between Em and EK• It might be argued that the Em has been decreased by multiple electrode impalement and that the membrane permeability to chloride does not permit redistribution of Cl in the control time used in these experiments. This possibility can be checked since the rate constant for Cl is known. By the use of data from the experiment depicted in Figure 2, o i namely, Em = -52.5 mV, a Cl = 330 mM, a C1 = 35.2 mM, ECl = -56.4 mV; 38 and the average rate constant of -1.36 x 10- 4sec- 1 (see p. 19), the time required for C1 to attain electrochemical equilibrium can be calculated. When Em i EC1 ' a C1 must equal 41 mM. If t is the time i required to increase a C1 from 35.2 to 41 mM, then t = i~3:1~3~0-4 = = 20 minutes (Riggs, 1963). Thus, in 20 minutes, EC1 should equal Em if there is no outward transport of C1 from the cell. As shown in Figure 2, after 6 hours in control solutions EC1 remained more negative than Em' 0 Cooling to 1 C resulted in a rapid influx of C1 so that EC1 Em within 20-80 minutes (Fig. 3; Table 5). The results of the control and cooling experiments are evidence that Cl is actively transported across the cell membrane. The inter- nal C1 is too low to be accounted for by passive distribution and therefore C1 is transported out of the cell. The nature of the trans- port system is unknown but the results obtained using ouabain and cyanide indicate that it may not be ATP-dependent. Active transport of Cl has also been suggested in squid axon (Keynes, 1963), crayfish giant axon (Strickholm & Wallin, 1965), algae (Cornwall, Peterson, Kunze, Walker &Brown, 1970), D-cells in Helix aspersa (Kerkut & Meech, 1966), turtle thyroid follicle (Chow, Kunze, Brown &Woodbury, 1970) and Aplysia ca1ifornica (Brown et a1., 1970). Active Cl transport into the cerebrospinal fluid has also been postulated and there is evidence that the carbonic anhydrase inhibitor, acetazolamide, can inhibit this active transport (Woodbury, 1965b); Maren, 1967; Maren & Broder, 1970). Therefore, it is possible that a carbonic anhydrase-dependent system is responsible for the active transport of Cl in the giant cell of the Ap1ysia. = 39 As a minimum, the active transport systems for K and Cl must perform enough work to prevent the passive fluxes that have been described. Energy requirements for active transport of potassium and chloride can be calculated from the present data. For potassium the equation is: . where WK is the transport power for potassium and the remaining symbols have meanings already given. On the assumption that the cell has a specific gravity of 1.0,0.019 calories/hour·gram would be required to just offset the net passive flux measured in these experiments. For chloride the equation is: WCl = -(ZC1'FoEm - RT ln aiCl/aoCl)MCl The energy requirement for chloride transport in the giant cell is 0.0044 calories/hour' gram. If one mole of ATP releases 7000 calories, potassium transport would require 2.7 x 10- 6 moles of ATP per hour. Chloride transport, as stated earlier, does not seem to be ATP dependent. Since oxygen consumption by the giant cell is not known, it is impossible to know if these calculated energy requirements exceed the cell's energy production. Mechanisms of passive permeation of K and Cl across the membrane of Rz Potassium is actively transported into R2 and Cl is actively transported out of R2 , but the transport systems for the two ions are 40 almost certainly different. When the transport systems were turned off, K leaked out of the cell down its electrochemical gradient. I have argued that the measurements of net K efflux and net C1 influx probably represent a close approximation of the unidirectional flux of either ion. Knowing the driving forces and fluxes for K and C1, we can measure the rates at which these ions permeate the membrane chemically or carry current across it electrically in the absence of the active transport systems. The close agreement between K con- ductance and permeability calculated either from electrical or chemical measurements is interpreted as indicating that K permeates as an ion. However, Cl conductance and permeability calculated from electrical measurements were two orders of magnitude smaller than analogous values calculated from chemical measurements. This indicates that a considerable amount of C1 may cross the membrane of R2 in a non-ionic form, possibly as a neutral complex with some carrier molecule. Another possib"ility is that the electrical measurements included a component of outward moving K current which offset the Cl-current measurements. REFERENCES Adrian, R. H. (1956). The effect of internal and external potassium concentration on the membrane potential of frog muscle. J. Physiol. 133,631-658. Baker, P. F., Blaustein, M. P., Hodgkin, A. L. & Steinhardt, R. A. (1969). The influence of calcium on sodium influx in squid axons. J. Physiol. 200, 431-458. Brown, A. M. &Berman, P. R. (1970). Mechanism of excitation of AtlYSia neurons by carbon dioxide. J. gen. Physio1. 56,543- 5 8. Brown, A. M., Walker, J. L. & Sutton, R. B. (1970). Increased chloride conductance as the proximate cause of hydrogen ion concentration effects in Aplysia neurons. J. gen. Physio1. 56, 559-582. Caldwell, P. C., Hodgkin, A. L., Keynes, R. D. & Shaw, T. I. (1960). The effects of injecting lenergy-rich l phosphate compounds on the active transport of ions in the giant axons of Loligo. J. Physiol. 152,561-590. D. o. (1970). Membrane potential produced directly by . the Na+ pump in Aplysia neurons. Compo Biochem. Physiol. 35, 371-385. Cq~penter, Carpenter, D. O. & Alving, B. O. (1968). A contribution of an electrogenic pump to membrane potential in Aplysia neurons. J. gen. Physiol. 52, 1-20. Chow, S. Y., Kunze, D., Brown, A. M. &Woodbury, D. M. (1970). Chloride and potassium activities in luminal fluid of turtle thyroid follicles as determined by selective ion-exchanger microelectrodes. Proc. natn. Acad. Sci. U.S.A. 67,998-1004. Coggeshall, R. E. (1967). A light and electron microscope study of the abdominal ganglion of Ap1ysia californica. J. Neurophysiol. 30, 1263-1287. Cornwall, M. C., Peterson, D. F., Kunze, D. L., Walker, J. L. & Brown, A. M. (1970). Intracellular potassium and chloride activities measured with liquid ion exchanger microelectrodes. Brain Res. 23, 433-436. Cross, S. B., Keynes, R. D. & Rybova: R. (1965). The coupling of sodium efflux and potassium influx in frog muscle. J. Physiol. 181,865-880. 42 Eisenman, G., Sandbloom, J. P. &Walker, J. L. (1967). Membrane structure and permeation. Science, N.Y., 155,965-974. Frazier, W. T., Kandel, E. P., Kupferman, I., Waziri, R. & Coggeshall, R. E. (1967). Morphological and functional properties of identified neurons in the abdominal ganglion of Aplysia' californica. J. Neurophysiol. 30, 1288-1351. Glynn, I. M. (1964). The action of cardiac glycosides on ion movements. Pharmac. Rev. 16, 381-409. Gomori, G. (1948). Histochemical demonstration of sites of choline esterase activity. Proc. Soc. expo Biol. Med. 68,354-358. Handbook of Chemistry and Physics. (1966). O. C. Cleveland: Chemical Rubber. p. 0173, ed. Weast, Hayes, F. R. & Pelluet, O. (1947). The inorganic constitution of molluscan blood and muscle. J. mar. biol. Ass. U.K. 26, 580-589. Hidalgo, C. &Latorre, R. (1970). Temperature dependence of nonelectrolyte and sodium permeability in giant axon of squid. J. Physiol. 211, 173-191. Hodgkin, A. L. (1958). Ionic movements and electrical activity in giant nerve fibers. Proc. R. Soc. B, 148, 1-36. Hodgkin, A. L. & Katz, B. (1949). The effect of sodium ions on the electrical ~ctivity of the giant axon of the squid. J. Physiol. 108, 37-77. Hodgkin, A. L. & Keynes, R. O. (1955a). Active transport of cations in giant axons from Sepia and Loligo. J. Physiol. 128, 28-60. Hodgkin, A. L. & Keynes, R. O. (1955b). The potassium permeability of a giant nerve fiber. J. Physio1. 128,61-88. Hodgkin, A. L. & Horowicz, P. (1959). The influence of potassium and chloride ions on the membrane potential of single muscle fibres. J. Physiol. 148, 127-160. Katz, B. (1965). Nerve, Muscle and Synapse, p. 47. McGraw-Hill. New York: Kerkut, G. A. &Meech, R. W. (1966). Microelectrode determination of intracellular chloride concentration in nerve cells. Life Sci. 5, 453-456. Kerkut, G. A. &Thomas, R. C. (1965). An electrogenic sodium pump in snail nerve cells. Compo Biochem. Physiol. 14, 167-183. 43 Keynes, R. D. (1963). 690-705. Chloride in squid giant axon. J. Physio1. 169, Lev, A. A. (1964). Determination of activity and activity coefficients of potassium and sodium ions in frog muscle fibres. Nature, Lond. 201, 1132-1134. Ling, G. N. &Ochsenfe1d, M. M. (1966). Studies on ion accumulation in muscle cells. J. gen. Physiol. 49,819-843. Marchiafava, P. L. (1970). The effects of temperature change on membrane potential and conductance in Aplysia giant nerve cell. Compo Biochem. Physio1. 34, 847-852. Maren, T. H. (1967). Carbonic anhydrase: chemistry, physiology and inhibition. Physio1. Rev. 47, 599-781. Maren, T. H. & Broder, L. E. (1970). The role of carbonic anhydrase in anion secretion into cerebrospinal fluid. J. Pharmac. expo Ther. 172,197-202. Martin, A. R. & Pilar, G. (1963). Dual mode of synaptic transmission in the aviary ciliary ganglion. J. Physio1. 168,443-463. Marmor, M. F. &Gorman, A. L. F. (1970). Membrane potential as a sum of ionic and metabolic components. Science, N.Y. 167,65-67. Nicolsky, B. P. (1937). J. Phys. Chem. U.S.S.R. 10, 495-506. by Lev, A. A. (1964). Nature, Lond. 201, 1132-1134. Cited Pinsker, H. & Kandel, E. R. (1969). Synaptic activation of an electrogenic sodium pump. Science, N.Y. 167, 65-67. Riggs, D. s. (1963). The Mathematical Approach to Physiological Problems, p. 126. Cambridge: M.l.T. Press. Robinson, R. A. & Stokes, R. H. (1968). Electrolyte Solutions, 2nd edn., p. 231. London: Butterworth. Sato, M., Austin, G., Yai, H. &Maruhashi, J. (1968). The ionic permeability changes during acetylcholine-induced responses of Aplysia ganglion cells. J. gen. Physio1. 51,321-345. Shanes, A. M. & Berman, M. D. (1955). Kinetics of ion movement in the squid giant axon. J. gen. Physio1. 39, 279-300. Sjodin, R. A. & Beaug/, L. A. (1968). Strophanthidin-sensitive components of potassium and sodium movements in skeletal muscle as influenced by the internal sodium concentration. J. gen. Physiol. 52, 389-407. 44 Strickholm, A. &Wallin, B. G. (1965). activity in crayfish giant axons. Intracellular chloride Nature, Lond. 208, 790-791. Walker, J. L. (1971). Ion specific liquid ion exchanger microelectrodes. Analyt. Chern. 43, 89A-93A. White, A., Handler, P. &Smith, E. L. (1964). Principles of Biochemistry, 3rd edn., p. 330. New York: McGraw-Hill. Woodbury, D. M. (1965a). Physiology of body fluids. In Physiology and BiOPh~SiCS, T9th edn., ed. Ruch, T. C. &Patton, H. D., pp. 889-8 O. Philadelphia: Saunders. Woodbury, D. M. (1965b). Blood-cerebrospinal fluid-brain fluid relationships. Tn Physiology and BiO§hYSics,.19th ed~., ed. Ruch, T. C. & Patton, H. D., pp. 942- 43. Phl1adelphla: Saunders. 45 FIG. 1.. Diagram of the experimental set-up showing the constant temperature bath, placement of the ganglion, and arrangement of the recording electrodes. Em Electrode ~ Drain ~ Light Source 47 FI G. 2. Em =e Intracellular ionic activity, ai K = c6.; aiCl =0 ; EK =A and ECl = II duri ng a 6-hour peri od of conti nuous impalement of R2 by 3 electrodes. The Em was measured in this cell with a 3M KC1-filled microelectrode having a resistance of 7M n. o (Temp. = 20 ± 1 C) 40 70 60 120 180 MINUTES 240 300 360 49 Effects of cooling R2 to 1.5 0 C on intracellular ionic i activity, a K =6, a\l =0; Em=.; EK=.& and Eel = • . The rate constant for aiC1 increase during the first 40 minutes of cooling was -1.04 x 10-4sec- 1• aiK fell for the first two hours FIG. 3. with a rate constant of 4.94 x 10- 5sec- 1. It then began to decline with a much faster rate constant, 16.5 x 10 -5 sec -1 • Notice that Em began decreasing about 20 minutes before the second decline in aiK began. Although not shown on the graph, Rm fell at the same time as Em did, from 1.1 MQ to 0.6 MQ. -----20°C ... 1......- - - - 1.5°C - - - - - - + I..... I I I I I I I 60 80 o 20 40 60 18 38 58 78 MINUTES 98 118 138 158 178 51 FIG. 4. Effects of cooling to 1.0 0 C on intracellular ionic acti vi ty, ai K =6 ; Em =. and EK = 6. . The ce 11 was monitored for 4 hours after cooling and showed no indication of the second, faster component of K loss seen in Figure 3. The -5 -1 rate constant for K loss in this cell was 5.01 x 10 sec • I 20°C ..I.. I I I I I I I 50 ~__~____~____~__~____~__~90 50 0 50 100 MINUTES 150 200 250 53 FIG. 5. o Effects of cooling to 2.5 C and then later re-warm·ing to o i 23 C on, a K = t::,. ; Em = • and EK =~ . Again, as in Fig. 3, there was approximately a 50% decrease in Rm which occurred 10-15 minutes before the second phase of K loss. The rate constant for the decline in aiK during the first 3 hours of cooling was 4.53 x 10-5 sec -1. During the faster phase of intracellular Kloss the rate constant was 18.9 x 10-5sec- 1• Rewarming the cell caused a reaccumul ation of "intracellular K at such a rate that EK > Em (more negative) from the earliest times at which Em and aiK were , measured (cf. Cross, Keynes, &Rybova, 1965). I 10 I I ° ...."!'. I -20C . 25°C I ---23OC \.. • 30 10 50 :E 60 E >- .> .U 40 > 70 <t 90 E 55 FIG. 6. Effects of cooling to 100C on intracellular ionic acti vi ti es, ai K = ~ , a; Cl = 0; Em = e ; EK =.& and ECl =II · Cooling caused the cell to depolarize by about 6 mV in the first 20 minutes, a depolarization comparable to that seen in cells cooled o to 0.5-2.5 C. i ; a K and a Cl changed very little during the 2 hours of cooling to 100C. The majority of the changes in EK and ECl are due simply to the effects of the lower temperature alone on the electrochemical equilibrium. 0> e 80 27 57 MINUTES 87 117 57 FIG. 7. Plot showing the relation between control Em and the amount of initial depolarization caused by ouabain (()) before significant . changes 1n ai K occur. In addition, the same relationship is shown for DNP-treated cell s (D) and cyani de-treated cell s (6.). The regression line is drawn with the use of the data of the ouabaintreated cells. The slope of the line is 0.99 and has a correlation coefficient of 0.86. -60 00 -58 -56 0 -54 -52 ~o 0 > e -50 e UJ -48 -l 0 ~ z ~ 0 -46 0 u -44 -42 -40 0 0 59 FIG. 8. Effects of 2 x 10- 4 Mouabain on intracellular ionic ac t i vi ties. ai K = 6 ; a i C1 = 0 ; e ; EK =A Em = and EC1 =II Decline in ai K had a rate constant of 4.22 x 10 -5 sec -1 and continued unchanged for 5 hours. aiC1 was little affected by ouabain in this particular cell (but see Table 6) (Temp. = 20 ± 10 C) . '-----OUABAIN 2 x10- 4 M--------.- 30 40 50 > E 60 70 30 27 87 147 MINUTES 201 261 61 FIG . 9. Effects of 0.2 mM DNP on aiK =6 ; Em =e and EK =. Notice that membrane depolarization and the decline in aiK began immediately after the application ofDNP. aiK fell with a rate constant of 5.95 x 10- 5 sec- l • (Temp. = 20 ± 10 C) ~-DNP 0.2 m M - - -...... 20 60 8 28 48 MINUTES 68 88 108 40 63 FIG. 10. Effects of 5 mM cyanide on intracellular ionic activities, ai K = 6. , aiel = D ; Em =e ; EK = A and Eel = II " Notice the large decrease in Em within 10 minutes of cyanide application. The rate constant of net K efflux was 3.46 x 10- 5sec- 1• (Temp. = 20 ± 1°C) .....---CN 5 m M - - - - -.. ·'40 50 150 70~ 20 90 20 40 60 80 7 27 MINUTES 47 67 87 107 127 147 65 FIG. 11. Effect of cooling on membrane resistance, Rm. Cell was impaled with one electrode to measure Em and with a second one to 8 pass current through a 5 x 10 ohm resistor. During the measurement of Rm, Em was held constant at -52 mV by passing the appropriate constant current from one stimulator, then testing Rm with a squarewave hyperpolarizing pulse of 6 sec duration from a second stimulator. Each point required about 2 minutes of steady intracellular current injection to maintain Em at -52 mV long enough to test with 3 test pulses. 0 Notice that at 6 C no effect on Rm was observed; but, immediately upon further cooling to 1.5 0 C, ~ fell by 19% followed by a rapid increase back to control values. The last 40 minutes depicts the steady increase in membrane resistance typically seen in these experiments. The effects of cooling on the Em of this cell were small, consisting of a slow, steady depolarization which amounted to 5 mV by the end of the experiment. MEMBRANE RESISTANCE Mn • en d I I 0 ~ I 0 P <.N o 0 0 1 I ~ ,en , P o I ~ 3:1') - z (J'I c:: --t rr1 (J) U1 (J'I - • U1 en (J'1 o o Table 1 Solutions NaCl KC1 mM mM CaC1 2 MgC1 2 MgS0 4 Tris- * maleate mM mM mM mM Na 2S0 4 Ca proI;!ionate mM mM Control ASW 494 10 10 20 30 10 ASW to be 0 cooled to 1 494 10 10 20 30 10 1/2 Cl- 247 10 10 50 10 215 1/4 Cl- 124 10 50 10 330 1/2 cr cooled to 10 247 10 50 10 215 1/4 Clcooled to 10 124 10 50 10 330 5mM KCN 494 5 10 mM KCN 494 10 10 20 30 10 10 20 30 10 10 *Tris-(hydroxymethy1}-aminomethane-ma1eate (Tris-~leate) buffer made according to Gomor; (1948). 0'1 ........ 68 Table 2 Chloride electrode selectivity constants, K·. lJ Interfering ion II = 0.1 II = 1.0 Propionate 0.5 0.7 Isethionate 0.2 0.2 Sulfate 0.03 0.02 Bicarbonate 0.05 0.05 69 Table 3 Potassium electrode selectivity constants, Kij Interfering ion II = 0.1 1l=1.0 Sodium 0.02 0.02 Hydrogen 0.02 0.03 Calcium 0.002 0.03 Magnesium 0.001 0.001 Table 4 Effects of electrolyte used to fill microelectrodes on Em. aiKt aiCl' EK and ECl Expts. Electrode Resistance Mo Experiment Duration hrs I Em Initial Values i ai K EK a C1 Final Values Em EC1 I ai K EK i a Cl ECl 1-29 5 4 -35.7 mM mV mV mV 3 M KC1-filled Microelectrodes 154.7 -77 .9 33.7 -56.7 -44.6 1-30 7 6 -35.5 176.0 -81.2 34.5 -56.8 -52.2 157.5 -78.4 38.1 -54.4 4-21 12 3 -47.5 168.5 -80.2 32.8 -60.2 -47.2 166.5 -79.7 32.5 -60.5 8-19 5 n = 4 X" ± S.E. 5la 30.4 -79.0 35.1 ± .4 ± 2.2 -60.5 -57.1 ± 2.0 4-6 8 5 -48.0 29.6 -62.2 -49.6 -41. 7 166.4 -79.8 32.7 -59.0 -47.2 161.9 ± 3.5 ± 6.2 ± 1.0 ± 1.1 ± 1.4 ± 1.4 ± 2.6 2 MNa Citrate-filled Microelectrodes -59.5 199.5 -86.0 -58.3 206.5 4-26 9 4 6 3-31 n = 3 X" ± S.E. 4 mV , -35.0 mM mM mV mM 161.7 -79.0 39.4 -52.8 161.2 -79.4 48.3 -49.8 -47.5 197.5 0.6 M Na S04-fi11ed Microe1ectrodes -53.6 174.0 -80.8 -53.2 187.0 -49.4 178.2 -82.1 -53.0 197.0 ± 7.4 ± 11.3 ± 2.0 ± 3.1 ± 5.6 mV -86.6 -84.0 46.7 -51. 1 -82.5 -84.4 ± 1.2 ...... 0 71 Table 5 Effects of cooling on Em' ai C1 and EC1 Control (200 C) 40-80 minutes after cooling ; Em i a Cl EC1 Temp. Em a Cl ECl mV mM mV °c mV mM mV 7-27 -50.6 30.2 -60.8 3.2 -48.3 36.0 -48.5 8-3 -53.1 34.4 -59.8 4.0 -46.3 43.0 -47.0 8-4 -45.6 29.5 -62.3 3.8 -53.0 34.2 -53.0 8-11 -42.0 31.0 -60.6 0.4 -51.3 33.7 -52.0 10-27 -44.2 42.9 -53.2 1.5 -39.0 61.3 -38.8 11-9 -52.2 29.0 -62.4 1.0 -46.0 50.8 -45.9 11-5 -45.4 38.5 -55.6 0.8 -42.7 44.8 -46.5 11-12 -47.1 26.5 -64.6 0.5 -49.0 39.2 -50.3 11-16 -47.2 36.6 -55.4 0.8 -50.5 46.0 -50.7 11-18 -49.0 44.6 -51.9 1.0 -44.8 51.8 -44.5 11-23 -53.9 32.7 -57.5 0.9 -49.7 37.6 -49.5 -48.2 ± 1.2 ± 1.8 Expts. n = 11 X ± S.E. 34.2 -58.6 ± 1.2 -47.8 ± 1.4 43.5 ± 2.6 -47.9 ± 1.2 Each point after cooling was chosen as the first point when ECl = Em' Thereafter, ECl remained equal to Em for the duration of the experiment, up to 8 hours. 72 Table 6 i Effects of ouabain on Em' a C1 and ECl Maximal Effect * Control ECl Em a C1 EC1 Tlme exposed to ouabain mM mV mV mM mV minutes -50.6 37.6 -55.5 -38.3 43.3 -51.9 297 10-21 -51.8 39.1 -54.1 -38.3 48.2 -48.8 302 10-22 -58.5 39.0 -53.0 -44.2 46.4 -48.6 173 11-11 -42.8 28.8 -64.3 -35.0 46.6 -52.2 153 1-25 -44.0 37.0 -55.9 -37.0 42.7 -52.3 153 36.3 1.9 ± 2.0 -38.6 1.53 ± 1.05 Expts. i Em a C1 mV 10-20 n= 5 ± r S.E. -50. 1 ± 2.6 ± -56.6 ± i 45.4 -50.8 ± .8 *The time of maximal effect was, in each case, at the end of the experiment. Table 7 Effects of cyanide on Emt aiKt aiC1 and ECl Control Expts. 60 minutes after cyanide ; Cyanide Cone. Em ai K a C1 ECl Em ai K i a C1 EC1 kefflux of K mM mV mM mM mV mV mM mM mV x 10- 5 sec- 1 1-18 5 -51.8 185.3 49.0 -49.5 -45.7 169.2 42.9 -52.9 2.61 1-19 5 -50.8 132.0 44.3 -50.6 -43.5 112.1 52.4 -46.4 3.62 1-20 5 -51.8 186.0 43.9 -53.0 -45.0 158.0 42.1 -54. 1 3.46 =3 X ± S.E. -51.5 ± 0.3 167.8 ± 17.9 45.7 ± 1.6 -51.0 ± 1.0 -44.7 ± 0.6 146.4 ± 17.5 45.8 ± 3.3 -51. 1 ± 2.4 3.23 ±0.3 1-22 10 -48.0 168.8 34.5 -56.1 -48.3 144.9 33.4 -56.9 4.22 n ........ w 74 Table 8 Effects of cooling on gCl Control TC1 Rm gm Mn Hyperpolarizing cells n =3 Oepo1arizing cells n =3 0.31 0.31 1.25 1.5 Cooled gC1 TC1 Mn 1Jmho 80 67 Rm 25 21 0.38 0.17 1.85 1.2 gm gC1 % Change 21 -18 llmho 54 , 83 f4!., 1 , -33 Table 9 Summary of effects of inhibiting active transport on net transmembrane movements of K and C1 Treatment n MK Cooled to 0.5-2.5 0 C 7 Ouabain 2 x 10-4M 13 IK Potassium gK Chloride PK M/cm2si~ x lO- amp/cm~ x lO- mho/cm~ x lO- cm/se x lO- 5.06 ±.89 4.81 ±.85 1.70 ±.35 1.32 ±.21 4.51 ±.62 4.53 ±.6 1.22 ±.14 1.77 ±.2 n a MC1 IC1 M/cm2si~ amp/c, x lO- x lO- 4 4.39 ±1.23 4.24 ±1.28 g~l ( q. 3) 2 x 10- 1.38 ±.32 gC1 (eq. 4) ~ho/cm2 x 10- 4 1, 16 ±,33 PC1 cm/se, x 10- 3.69 ±1.] All figures are averages ± standard error of values taken 20-30 minutes after cooling was begun or ouabain applied to the giant cell. It should be pointed out that, as long as the rate constant remained unchanged, the above calculated figures changed very little during the course of an experiment. "-J U'I