Effects of External Potassium and Strophanthidin on Sodium Fluxes in Frog Striated Muscle

Unidirectional Na fluxes in isolated fibers from the frog''s semitendinosus muscle were measured in the presence of strophanthidin and increased external potassium ion concentrations. Strophanthidin at a concentration of 10^-5 M inhibited about 80 per cent of the resting Na efflux without having any detectable effect on the resting Na influx. From this it is concluded that the major portion of the resting Na efflux is caused by active transport processes. External potassium concentrations from 2.5 to 7.5 mM had little effect on resting Na efflux. Above 7.5 mM and up to 15 mM external K, the Na efflux was markedly stimulated; with 15 mM K the Na influx was 250 to 300 per cent greater than normal. On the other hand, Na influx was unchanged with 15 mM K. The stimulated Na efflux with the higher concentrations was not appreciably reduced when choline or Li was substituted for external Na, but was completely inhibited by 10^-5 M strophanthidin. From these findings it is concluded that the active transport of Na is stimulated by the higher concentrations of K. It is postulated that this effect on the Na "pump" is produced as a result of the depolarization of the muscle membranes and is related to the increased metabolism and heat production found under conditions of high external K.     

Strophanthidin-sensitive components of potassium and sodium movements in skeletal muscle as influenced by the internal sodium concentration

"Low sodium" muscles were prepared which contained around 5 mmoles/kg fiber of intracellular sodium. "High sodium" muscles containing between 15 and 30 mmoles/kg fiber of intracellular sodium were also prepared. In low sodium muscles application of 10^-5 M strophanthidin reduced potassium influx by about 5%. Potassium efflux was unaffected by strophanthidin under these conditions. In high sodium muscles, 10^-5 M strophanthidin reduced potassium influx by 45% and increased potassium efflux by 70%, on the average. In low sodium muscles sodium efflux was reduced by 25% during application of 10^-5 M strophanthidin while in high sodium muscles similarly treated, sodium efflux was reduced by about 60%. Low sodium muscles showed a large reduction in sodium efflux when sodium ions in the Ringer solution were replaced by lithium ions. The average reduction in sodium efflux was 4.5-fold. Of the amount of sodium efflux remaining in lithium. Ringer''s solution, 40% could be inhibited by application of 10^-5 M strophanthidin. The total sodium efflux from low sodium muscles exposed to Ringer''s solution in which lithium had been substituted for sodium ions for a period of 1 hr can be fractionated as 78% Na-for-Na interchange, 10% strophanthidin-sensitive sodium pump, and 12% residual sodium efflux. It is concluded that large strophanthidin-sensitive components of sodium and potassium flux can be expected only at elevated sodium concentrations within the muscle cells.     

The Dual Effect of Lithium Ions on Sodium Efflux in Skeletal Muscle

Sartorius muscle cells from the frog were stored in a K-free Ringer solution at 3°C until their average sodium contents rose to around 23 mM/kg fiber (about 40 mM/liter fiber water). Such muscles, when placed in Ringer''s solution containing 60 mM LiCl and 50 mM NaCl at 20°C, extruded 9.8 mM/kg of sodium and gained an equivalent quantity of lithium in a 2 hr period. The presence of 10^-5 M strophanthidin in the 60 mM LiCl/50 mM NaCl Ringer solution prevented the net extrusion of sodium from the muscles. Lithium ions were found to enter muscles with a lowered internal sodium concentration at a rate about half that for entry into sodium-enriched muscles. When sodium-enriched muscles labeled with radioactive sodium ions were transferred from Ringer''s solution to a sodium-free lithium-substituted Ringer solution, an increase in the rate of tracer sodium output was observed. When the lithium-substituted Ringer solution contained 10^-5 M strophanthidin, a large decrease in the rate of tracer sodium output was observed upon transferring labeled sodium-enriched muscles from Ringer''s solution to the sodium-free medium. It is concluded that lithium ions have a direct stimulating action on the sodium pump in skeletal muscle cells and that a significantly large external sodium-dependent component of sodium efflux is present in muscles with an elevated sodium content. In the sodium-rich muscles, about 23% of the total sodium efflux was due to strophanthidin-insensitive Na-for-Na interchange, about 67% being due to strophanthidin-sensitive sodium pumping.     

Effects of Sodium Azide on Sodium Fluxes in Frog Striated Muscle

Unidirectional Na fluxes from frog''s striated muscle were measured in the presence of 0 to 5 mM sodium azide. With azide concentrations of 2 and 5 mM the Na efflux was markedly stimulated; the Na efflux with 5 mM azide was about 300 per cent greater than normal. A similar increase was present when all but the 5.0 mM sodium added with azide was replaced by choline. 10^-5 M strophanthidin abolished the azide effect on Na²⁴ efflux. Concentrations of azide of 1.0 mM or less had no effect on Na efflux. The Na influx, on the other hand, was only increased by 41 per cent in the presence of 5 mM NaN₃. From these findings it is concluded that the active transport of Na is stimulated by the higher concentrations of azide. The hypothesis is advanced that the active transport of Na is controlled by the transmembrane potential and that the stimulation of Na efflux is produced as a consequence of the membrane depolarization caused by the azide.     

Effects of the Intracellular Ca Ion Concentration upon the Excitability of the Muscle Fiber Membrane of a Barnacle

The membrane excitability and contraction were examined in single barnacle muscle fibers with different internal Ca⁺⁺ concentrations by using buffer solutions made up with EGTA and Ca-gluconate in various proportions. During the passage of dc currents the membrane shows all-or-none spike potentials for internal Ca⁺⁺ concentrations below about 8 x 10^-8 M, oscillatory potential changes in the range between 8 x 10^-8 to 5 x 10^-7 M, but neither oscillatory nor spike potentials were seen for concentrations above 5 x 10^-7 M. All-or-none spike potentials were suppressed when the internal Mg⁺⁺ concentration exceeded 5 mM. The suppression threshold of the internal Ca⁺⁺ concentration for the Sr spike is much higher than that for the Ca spike. The threshold concentration of internal Ca⁺⁺ for contraction was about 8 x 10^-7 M.     

The Effect of Quinidine on Calcium Accumulation by Isolated Sarcoplasmic Reticulum of Skeletal and Cardiac Muscle

Quinidine potentiates twitch tension and (at higher concentrations) causes contracture of skeletal muscle whereas the same drug reduces tension development of cardiac muscle. To gain insight into the possible differences in the excitation-contraction coupling mechanism of the two types of muscle the effect of quinidine on calcium accumulation by isolated sarcoplasmic reticulum from skeletal and cardiac muscle was investigated. In a medium containing ATP, Mg⁺⁺, oxalate, and ⁴⁵Ca, pharmacologically active concentrations of the drug inhibited calcium accumulation by both skeletal and cardiac sarcoplasmic reticulum. The inhibition of the rates of calcium, uptake by the skeletal muscle preparation ranged from 11% with 10^-4 M quinidine to 90% with 10^-3 M quinidine. With the cardiac muscle preparation the inhibition ranged from 16% with 3 x 10^-6 M quinidine to 100% with 10^-3 M quinidine. With both preparations the inhibition of calcium transport was accompanied by an inhibition of the Ca⁺⁺-activated ATPase activity of the sarcoplasmic reticulum. The effect of quinidine on the skeletal sarcoplasmic reticulum supports the hypothesis that this compound produces twitch potentiation and contracture by interfering with intracellular calcium, sequestration. Its effect on cardiac sarcoplasmic reticulum. has been interpreted in terms of the hypothesis that cardiac contractility is a function of the amount of calcium released from the sarcoplasmic reticulum which is in turn dependent upon the absolute calcium content of the reticulum. Hence, following inhibition of calcium transport there would be less calcium available for coupling.     

Cholinesterase activity per unit surface area of conducting membranes

According to theory, the action of acetylcholine (ACh) and ACh-esterase is essential for the permeability changes of excitable membranes during activity. It is, therefore, pertinent to know the activity of ACh-esterase per unit axonal surface area instead of per gram nerve, as it has been measured in the past. Such information has now been obtained with the newly developed microgasometric technique using a magnetic diver. (1) The cholinesterase (Ch-esterase) activity per mm² surface of sensory axons of the walking leg of lobster is 1.2 x 10^-3 µM/hr. (σ = ± 0.3 x 10^-3; SE = 0.17 x 10^-3); the corresponding value for the motor axons isslightly higher: 1.93 x 10^-3 µM/hr. (σ = ± 0.41 x 10^-3; SE = ± 0.14 x 10^-3). Referred to gram nerve, the Ch-esterase activity of the sensory axons is much higher than that of the motor axons: 741 µM/hr. (σ = ± 73.5; SE = ± 32.6) versus 111.6 µM/hr. (σ = ± 28.3; SE = ± 10). (2) The enzyme activity in the small fibers of the stellar nerve of squid is 3.2 x 10^-4 µM/mm²/hr. (σ = ± 0.96 x 10^-4; SE = ± 0.4 x 10^-4). (3) The Ch-esterase activity per mm² surface of squid giant axon is 9.5 x 10^-5 µM/hr. (σ = ± 1.55 x 10^-5; SE = ± 0.38 x 10^-5). The value was obtained with small pieces of carefully cleaned axons after removal of the axoplasm and exposure to sonic disintegration. Without the latter treatment the figurewas 3.85 x 10^-5 µM/mm²/hr. (σ = ± 3.24 x 10^-5; SE = ± 0.93 x 10^-5). The experiments indicate the existence of permeability barriers in the cell wall surrounding part of the enzyme, since the substrate cannot reach all the enzyme even when small fragments of the cell wall are used without disintegration. (4) On the basis of the data obtained, some tentative approximations are made of the ratio of ACh released to Na ions entering the squid giant axon per cm² per impulse.     

Studies of cardiac muscle with a high permeability to calcium produced by treatment with ethylenediaminetetraacetic acid

Thin strips of frog ventricle were isolated and bathed for 15 min in a solution containing 140 mM KCl, 5 mM Na₂ATP, 3 mM EDTA, and 10 mM Tris buffer at pH 7.0. The muscle was then exposed to contracture solutions containing 140 mM KCl, 5 mM Na₂ATP, 1 mM MgCl₂, 10 mM Tris, 3 mM EGTA, and CaCl₂ in amounts to produce concentrations of free calcium from 10^-4.8 M to 10^-9 M. The muscles developed some tension at approximately 10^-8 M, and maximum tension was achieved in 10^-5 M Ca⁺⁺. They relaxed in Ca⁺⁺ concentrations less than 10^-8 M. The development of tension by the EDTA-treated muscles was normalized by comparison with twitch tension at a stimulation rate of 9 per min before exposure to EDTA. In 10^-5 M Ca⁺⁺ tension was always several times the twitch tension and was greater than the contracture tension of a frog ventricular strip in KCl low Na-Ringer. Tension equal to half-maximum was produced at approximately 10^-6.2 M Ca⁺⁺. Intracellular recording of membrane potential indicated that after EDTA treatment the resting potential of cells in Ringer solution with 10^-5 M Ca or less was between 5 and 20 mv. Contracture solutions did not produce tension without prior treatment with EDTA. The high permeability of the membrane produced by EDTA was reversed and the normal resting and action potentials restored in 1 mM Ca-Ringer. Similar studies of EDTA-treated rabbit right ventricular papillary muscle produced a similar tension vs. Ca⁺⁺ concentration relation, and the high permeability state reversed with exposure to normal Krebs solution.     

Effect of changes in transepithelial transport on the uptake of sodium across the outer surface of the frog skin

The unidirectional sodium, uptake at the outer surface of the frog skin was measured by the method described by Biber and Curran (8). With bathing solutions containing 6 mM NaCl there is a good correlation between sodium uptake and short-circuit current (SCC) measured simultaneously except that the average uptake is about 40% higher than the average SCC. The discrepancy between uptake and SCC increases approximately in proportion to an increase in sodium concentration of the bathing solutions. Amiloride inhibits the unidirectional sodium uptake by 21 and 69% at a sodium concentration of 115 and 6 mM, respectively. This indicates that amiloride acts on the entry step of sodium but additional effects cannot be excluded. The sodium, uptake is not affected by 10^-4 M ouabain at a sodium concentration of 115 mM but is inhibited by 40% at a sodium concentration of 6 mM. Replacement of air by nitrogen leads to a 40% decrease of sodium uptake at a sodium concentration of 6 mM. The results support the view proposed previously (8) that the sodium uptake is made up of two components, a linear component which is, essentially, not involved in transepithelial movement of sodium and a saturating component which reflects changes in transepithelial transport. Amiloride, seems largely to affect the saturating component.     

The Site of Calcium Binding in Relation to the Activation of Myofibrillar Contraction

Skeletal muscle myofibrils, in the presence of 2 mM MgCl₂ at pH 7.0, were found to have two classes of calcium-binding sites with apparent affinity constants of 2.1 x 10⁶ M ^-1 (class 1) and ∼3 x 10⁴ M ^-1 (class 2), respectively. At free calcium concentrations essential for the activation of myofibrillar contraction (∼10^-6 M) there would be significant calcium binding only to the class 1 sites. These sites could bind about 1.3 µmoles of calcium per g protein. Extraction of myosin from the myofibrils did not alter their calcium-binding parameters. Myosin A, under identical experimental conditions, had little affinity for calcium. The class 1 sites are, therefore, presumed to be located in the I filaments. The class 1 sites could only be detected in F actin and myosin B preparations which were contaminated with the tropomyosin-troponin complex. Tropomyosin bound very little calcium. Troponin, which in conjunction with tropomyosin confers calcium sensitivity on actomyosin systems, could bind 22 µmoles of calcium per g protein with an apparent affinity constant of 2.4 x 10⁶ M ^-1. In view of the identical affinity constants of the myofibrils and troponin and the much greater number of calcium-binding sites on troponin it is suggested that calcium activates myofibrillar contraction by binding to the troponin molecule.     

Amino Acid and Sugar Transport in Rabbit Ileum

L-Alanine and 3-O-methyl-D-glucose accumulation by mucosal strips from rabbit ileum has been investigated with particular emphasis on the interaction between Na and these transport processes. L-Alanine is rapidly accumulated by mucosal tissue and intracellular concentrations of approximately 50 mM are reached within 30 min when extracellular L-alanine concentration is 5 mM. Evidence is presented that intracellular alanine exists in an unbound, osmotically active form and that accumulation is an active transport process. In the absence of extracellular Na, the final ratio of intracellular to extracellular L-alanine does not differ significantly from unity and the rate of net uptake is markedly inhibited. Amino acid accumulation is also inhibited by 5 x 10^-5 M ouabain. 3-O-methyl-D-glucose accumulation by this preparation is similarly affected by ouabain and by incubation in a Na-free medium. The effects of amino acid accumulation, of ouabain, and of incubation in a Na-free medium on cell water content and intracellular Na and K concentrations have also been investigated. These results are discussed with reference to the two hypotheses which have been suggested to explain the interaction between Na and intestinal nonelectrolyte transport.     

Further studies of the effect of calcium on the time course of action of carbamylcholine at the neuromuscular junction

The rate at which the postjunctional membrane of muscle fibers becomes desensitized to the action of carbamylcholine is increased after the muscle has been soaked in solutions containing increased concentrations of calcium. Some further aspects of this effect of calcium were investigated by measuring changes in the input resistance of single fibers of the frog sartorius during local perfusion of the neuromuscular junction with 2.73 x 10^-3 M carbamylcholine in isolated muscles immersed in 165 mM potassium acetate. It was found that (a) sudden changes in the local concentration of calcium brought about by perfusing fibers with carbamylcholine solutions containing 20 mM calcium, 40 mM oxalate, or 40 mM EDTA were followed within 20 sec by marked changes in the rate of desensitization; (b) prior to 13 sec after the introduction of carbamylcholine, however, no effect on the input resistance could be detected even though the muscle had been presoaked in 10 mM calcium; (c) the ability of high concentrations of calcium to bring about rapid desensitization disappears when a lower concentration of carbamylcholine (0.137 x 10^-3 M) is applied to the muscle fiber. These findings suggest that calcium present in the extracellular fluid can act directly on the postjunctional membrane to promote the desensitization process and that an increased permeability of the membrane to calcium brought about by the presence of carbamylcholine is a factor which contributes to this action.     

Ionic channels and nerve membrane lipids. Cholesterol-tetrodotoxin interaction

Experiments were carried out to investigate possible interactions of tetrodotoxin (TTX) with lipid molecules isolated from nerve fiber plasma membranes of the squid Dosidicus gigas. TTX has a highly selective ability to block the channel normally used by Na⁺ to cross the axolemma during nervous impulse conduction. In order to investigate the interaction each lipid sample was spread on 5 x 10^-7 M TTX and TTX-free 0.15 M NaCl solutions adjusted to pH 7.4 with 7 x 10^-3 M phosphate buffer. The surface pressure-area diagrams of the lipid monolayers revealed that TTX interacts only with cholesterol. The expansion of the cholesterol monolayers at 5 x 10^-7 M TTX was 2 A²/molecule at zero pressure for the experiments at 20°C and 2.5 A²/molecule for those at 25°C. Similar results were obtained in KCl subphases. The apparent dissociation constant of the cholesterol-TTX complex calculated from dose-response experiments is 2.6 x 10^-7 M. Experiments at pH 10.1 revealed that the zwitter ionic form of TTX is less active. Experiments with cholesterol derivatives (cholesteryl acetate, cholesterol methyl ether, cholestanol, and cholestanyl acetate) indicate that for the interaction with TTX a partial negatively charged group at C-3 and a double bond between C-5 and C-6 on the steroid nucleus are required. Tetrodonic acid, a biologically inactive derivative of TTX, does not interact with cholesterol. The results lead us to propose that cholesterol is part of the Na⁺ channel.     

The Effects of Zinc on Contractility,Membrane Potentials,and Cation Content of Rat Atria

Zinc depresses the contractile force of electrically driven rat atria logarithmically with time. The threshold concentration is about 5 x 10^-6 M zinc and the half-time for contractile depression at 10^-4 M is about 25 minutes. Zinc also depresses spontaneous activity of atria and alters the transmembrane potential parameters in a manner similar to quinidine. Unlike quinidine, zinc causes an elevation of the resting potential and an elevation of cellular potassium which varies with time in the same way as the resting potential. Exposure to 10^-4 M zinc for 60 minutes causes a statistically significant fall in atrial calcium content and an amount of radioactively labeled zinc is taken up which is quantitatively equal to the calcium lost. Zinc has no effect on rigor caused by iodoacetate but inhibits rigor caused by 1-fluoro-2,4 dinitrobenzene. It is postulated that zinc depression of contractile force is not due to metabolic inhibition, probably not due to quinidine-like action on the cell membrane, but may be due to an interference in the handling of calcium by the cell.     

Effect of osmolality on the hydraulic permeability coefficient of red cells

The osmotic water permeability coefficient, L_p, for human and dog red cells has been measured as a function of medium osmolality, and found to depend on the osmolality of the bathing medium. In the case of human red cells L_p falls from 1.87 x 10^-11 cm³/dyne sec at 199 mOSM to 0.76 x 10^-11 cm³/dyne sec at 516 mOSM. A similar decrease was observed for dog red cells. Moreover, L_p was independent of the direction of water movement and the nature of the solute used to provide the osmotic pressure gradient; it depended only on the final osmolality of the medium. Furthermore, L_p was not affected by pH in the range of 6 to 8 nor by the presence of drugs such as valinomycin (1 x 10^-6 M) and tetrodotoxin (3.2 x 10^-6 M). The instantaneous nature of the response to changes in external osmolality suggests that the hydraulic conductivity of the membrane is controlled by a thin layer at the outer face of the membrane.     

Sodium fluxes in internally dialyzed squid axons

The effects which alterations in the concentrations of internal sodium and high energy phosphate compounds had on the sodium influx and efflux of internally dialyzed squid axons were examined. Nine naturally occurring high energy phosphate compounds were ineffective in supporting significant sodium extrusion. These compounds were: AcP, PEP, G-3-P, ADP, AMP, GTP, CTP, PA, and UTP.¹ the compound d-ATP supported 25–50% of the normal sodium extrusion, while ATP supported 80–100%. The relation between internal ATP and sodium efflux was nonlinear, rising most steeply in the range 1 to 10 µM and more gradually in the range 10 to 10,000 µM. There was no evidence of saturation of efflux even at internal ATP concentrations of 10,000 µM. The relation between internal sodium and sodium efflux was linear in the range 2 to 240 mM. The presence of external strophanthidin (10 µM) changed the sodium efflux to about 8–12 pmoles/cm² sec regardless of the initial level of efflux; this changed level was not altered by subsequent dialysis with large concentrations of ATP. Sodium influx was reduced about 50 % by removal of either ATP or Na and about 70 % by removing both ATP and Na from inside the axon.     

Control of growth hormone production by a clonal strain of rat pituitary cells. Stimulation by hydrocortisone

Addition of hydrocortisone to the medium of a clonal strain of rat pituitary cells (GH₃) stimulated the rate of production of growth hormone. The stimulation had a lag period of about 24 hr, reached a maximum at 70–100 hr, and was observed at a hydrocortisone concentration as low as 5 x 10^-8 M. Cells maximally stimulated with 3 x 10^-6 M hydrocortisone produced 50–160 µg growth hormone/mg cell protein/24 hr. These rates were four to eight times those observed in control cells. At maximum stimulation, intracellular levels of growth hormone in both stimulated and control cells were equal to the amount secreted into the medium in about 15 min. Removal of hydrocortisone from the medium of GH³ cells caused a return of the rate of growth hormone production to that in control cells. Addition of hydrocortisone to the medium of cells growing exponentially with a population-doubling time of 60 hr caused both an increase in the doubling time to 90 hr and a stimulation of growth hormone production. Cycloheximide (3.6 x 10^-5 M) and puromycin (3.7 x 10^-4 M) suppressed incorporation of labeled amino acids into protein by 93 and 98%, respectively, and suppressed growth hormone production by stimulated and control cells by at least 94%.     

The effects of varying concentrations of colchicine on the progression of grasshopper neuroblasts into metaphase

The effects of four concentrations of colchicine (2.5 x 10^-7, x 10^-5, x 10^-3, and x 10^-2 M) on the cell cycle of grasshopper neuroblasts have been determined by direct observations on living cells. The lowest concentration, 2.5 x 10^-7 M, does not completely disorganize the spindle but does retard its action. The three higher concentrations disorganize the spindle, so that all cells reaching metaphase are blocked in a c-mitotic condition throughout the period of observations (308 min at 38°C, the minimum duration of the cell cycle in untreated neuroblasts). Continuous treatment with all concentrations reduces the rate at which neuroblasts enter metaphase, the extent of the reduction being a function of increasing concentration and time of exposure. After a short exposure to 2.5 x 10^-5 M colchicine, the neuroblasts recover from the inhibiting effects on progression through the cycle to metaphase, but they show no recovery from the inhibiting effects on spindle formation for more than 3 hr. Apparent stimulation of progression rate occurs early in exposure to all concentrations and during recovery from a short exposure to 2.5 x 10^-5 M. Morphological alterations in the chromatin of telophase, interphase, and prophase cells are induced by the higher concentrations of colchicine. The data indicate that caution should be exercised in the use of colchicine for determining cell cycle duration and/or the effects of physical and chemical agents on the cycle.     

Potassium and sodium transport across single distal tubules of Amphiuma

The transport properties of potassium (K) and sodium (Na) were studied in single distal tubules of Amphiuma using free-flow micropuncture techniques and stationary microperfusion methods. The transepithelial movement of labeled potassium was measured utilizing a three-compartment system in series in which the time course of tracer disappearance from the lumen was followed. Under control conditions, in blood- and doubly-perfused kidneys, extensive active net reabsorption of sodium and potassium obtains along single distal tubules. Tubular potassium reabsorption is abolished by ouabain at a concentration of 5 x 10^-6 M. Significant net secretion of K can be induced by exposing Amphiuma to a high K environment (100 mM KCl) or by adding acetazoleamide (1 x 10^-4 M) to the perfusion fluid. Transepithelial movement of potassium involves mixing with only a small fraction of total distal tubular cell potassium. This transport pool of potassium increases significantly with the transition from tubular net reabsorption to net secretion. Indirect evidence is presented which indicates that increased active K uptake across the peritubular cell boundary may be of prime importance during states of net K secretion.     

An electrogenic sodium pump in Limulus ventral photoreceptor cells

A hyperpolarization can be recorded intracellularly following either a single bright light stimulus or the intracellular injection of Na⁺. This after-hyperpolarization is abolished by bathing in 5 x 10^-6 M strophanthidin or removal of extracellular K⁺. Both treatments also lead to a small, rapid depolarization of the dark-adapted cell. When either treatment is prolonged, light responses can still be elicited, although with repetitive stimuli the responses are slowly and progressively diminished in size. The rate of diminution is greater for higher values of [Ca⁺⁺]_out; with [Ca⁺⁺]_out = 0.1 mM, there is almost no progressive diminution of repetitive responses produced by either K⁺-free seawater or strophanthidin. We propose that an electrogenic Na⁺ pump contributes directly to dark-adapted membrane voltage and also generates the after-hyperpolarizations, but does not directly generate the receptor potential. Inhibition of this pump leads to intracellular accumulation of sodium ions, which in turn leads to an increase in intracellular Ca⁺⁺ (provided there is sufficient extracellular Ca⁺⁺). This increase in intracellular calcium probably accounts for the progressive decrease in the size of the receptor potential seen when the pump is inhibited.