Transient inhibition of the muscarinic actions of carbachol during reactivation of the electrogenic sodium pump in guinea pig taenia caeci smooth muscle

In guinea pig taenia caeci smooth muscle the muscarinic receptor stimulant carbachol evoked depolarization and contraction, which was followed by hyperpolarization and relaxation on its removal. Both the hyperpolarization and relaxation were inhibited by removal of K+ from the external medium. During Na+-pump blockade (K+-free solution) the depolarizing and contracting actions of carbachol decreased. When the Na+ pump was switched on again by readmission of 5.9 mmol/L K+ to K+-depleted and Na+-enriched preparations, electrogenic hyperpolarization and relaxation developed. During this period carbachol failed to produce depolarization and contraction.     

Na+, K+-ATPase of the canine mesenteric artery

Na+,K+-ATPase, the enzymatic moiety that operates as the electrogenic sodium-potassium pump of the cell plasma membrane, is inhibited by cardiac glycosides, and this specific interaction of a drug with an enzyme has been considered to be responsible for digitalis-induced vascular smooth muscle contraction. Although studies aimed at localization, isolation, and measurement of the Na+,K+-ATPase activity (or Na+, K- pump activity) indicate its presence in vascular smooth muscle sarcolemma, its characterization as the putative vasopressor receptor site for cardiac glycosides has depended on pharmacological studies of vascular response in vivo and on isolated artery contractile responses in vitro. More recently, radioligand-binding studies using [3H]ouabain have aided in the characterization of drug-enzyme interaction. Such studies indicate that in canine superior mesenteric artery (SMA), Na+,K+-ATPase is the only specific site of interaction of ouabain with resultant inhibition of the enzyme. The characteristics of [3H]ouabain binding to this site are similar to those of purified or partially purified Na+,K+-ATPase of other tissues, which suggests that if Na+,K+-ATPase inhibition is causally related to digitalis-mediated effects on vascular smooth muscle contraction, then therapeutic concentrations of cardiac glycosides could act to cause SMA vasoconstriction. The additional finding from radioligand-binding studies that Na+,K+-ATPase exists in much smaller quantities (density of sites per cell) in SMA than in either heart or kidney may have implications concerning its physiological, biochemical or pharmacological role in modulating vascular muscle tone.     

Inhibition of the acetylcholine-induced relaxation of canine isolated basilar artery by potassium-conductance blockers

Canine basilar artery rings precontracted with 5-hydroxytryptamine (0.1-0.5 microM) relaxed in the presence of acetylcholine (25-100 microM), sodium nitroprusside (0.1 microM), or stimulation of the electrogenic sodium pump by restoration of extracellular K+ (4.5 mM) after K(+)-deprivation. Acetylcholine-induced relaxation is believed to be caused by the release of endothelium-derived relaxing factor (EDRF) and is prevented by mechanical removal of the endothelium, while relaxations induced by sodium nitroprusside or restarting of the sodium pump are endothelium-independent. Acetylcholine-induced relaxation was selectively blocked by pretreatment of the tissue with the nonselective K+ conductance inhibitors, 4-aminopyridine (4-AP, 3 mM), Ba2+ (1 mM), and tetraethylammonium (20 mM), 4-AP also blocked ACh-mediated relaxation in muscles contracted with elevated external K+. Relaxation of 5-hydroxytryptamine-induced contraction by sodium nitroprusside, or by addition of K+ to K(+)-deprived muscle, was not affected by 4-AP. Relaxation of basilar artery with acidified sodium nitrite solution (containing nitric oxide) was reduced by 4-AP. These results suggest that 4-AP and possibly Ba2+ inhibit acetylcholine-induced endothelium-dependent relaxation by inhibition of the action of EDRF on the smooth muscle rather than through inhibition of release of EDRF. The increase in K+ conductance involved in acetylcholine-induced relaxation is not due to ATP-inhibited K+ channels, as it is not blocked by glyburide (10(-6) M). Endothelium-derived relaxant factor(s) may relax smooth muscle by mode(s) of action different from that of sodium nitroprusside or by hyperpolarization due to the electrogenic sodium pumping.(ABSTRACT TRUNCATED AT 250 WORDS)     

Tunicamycin reduces Na(+)-K(+)-pump expression in cultured skeletal muscle.

The purpose of this study was to examine effects of tunicamycin (TM), which inhibits core glycosylation of the beta-subunit, on functional expression of the Na(+)-K+ pump in primary cultures of embryonic chick skeletal muscle. Measurements were made of specific-[3H]-ouabain binding, ouabain-sensitive 86Rb uptake, resting membrane potential (Em), and electrogenic pump contribution to Em (Ep) of single myotubes with intracellular microelectrodes. Growth of 4-6-day-old skeletal myotubes in the presence of TM (1 microgram/ml) for 21-24 hr reduced the number of Na(+)-K+ pumps to 60-90% of control. Na(+)-K+ pump activity, the level of resting Em and Ep were also reduced significantly by TM. In addition, TM completely blocked the hyperpolarization of Em induced in single myotubes by cooling to 10 degrees C and then re-warming to 37 degrees C. Effects of tunicamycin were compared with those of tetrodotoxin (TTX; 2 x 10(-7) M for 24 hr), which blocks voltage-dependent Na+ channels. TM produced significantly greater decreases in ouabain-binding and Em than did TTX, findings that indicate that reduced Na(+)-K+ pump expression was not exclusively secondary to decreased intracellular Na+, the primary regulator of pump synthesis in cultured muscle. Similarly, effects of TM were significantly greater than those of cycloheximide, which inhibits protein synthesis by 95%. These findings demonstrate that effects were not due to inhibition of protein synthesis. We conclude that glycosylation of the Na(+)-K+ pump beta-subunit is required for full physiological expression of pump activity in skeletal muscle.     

Potassium effects on contraction in arterial smooth muscle mediated by Na+, K+-ATPase

A local increase in the extracellular potassium concentration [K+]o, up to about 8 meq/liter either by topical application or intra-arterial infusion of K+ salts, causes arteriolar dilation and decreased resistance to blood flow in systemic vascular beds. Isolated vascular smooth muscle responds to a similar increase in [K+] in the bathing fluid with relaxation if the preparation has some initial active tension. Reduction in [K+] over physiological ranges produces arteriolar constriction and increased resistance to blood flow. K+ vasodilation is accompanied by hyperpolarization of the smooth muscle cell whereas the vasoconstriction is accompanied by depolarization. All these responses can be blocked by ouabain, a potent Na+, K+-ATPase inhibitor. It is therefore thought that K+ vasodilation results from stimulation of the electrogenic Na+-K+ pump whereas the vasoconstriction results from inhibition of this pump. A number of conditions that alter resistance also alter interstitial fluid [K+]. These include exercise, myocardial ischemia, epileptic convulsions, and evoked electrical activity of the somatomotor cortex. Certain findings, including those during administration of ouabain, suggest that changes in [K+] contribute significantly to some of the changes in resistance.     

Modulation of plasma membrane Ca2+ pump by membrane potential in cultured vascular smooth muscle cells

We examined the effect of membrane potential (Em) on the activity of the plasma membrane Ca2+ pump in cultured rat aortic smooth muscle cells (VSMCs). Inside-negative K+ diffusion potential higher or lower than the resting Em (-46 mV) was artificially imposed on VSMCs with various concentrations of extracellular K+ (K+o) and 1 microM valinomycin. We found that the recovery phase of the intracellular Ca2+ transient elicited with 1 microM ionomycin was accelerated by depolarizing Em, whereas it was retarded by hyperpolarizing Em. The rate of extracellular Na+ (Na+o)-independent 45Ca2+ efflux from VSMCs stimulated with 1 microM ionomycin increased almost linearly with a change in Em from -98 to -3 mV. This effect of Em was abolished by extracellularly added LaCl3 or a combination of high pH (pH 8.8) and high Mg2+ (20 mM), conditions that presumably inhibit the plasma membrane Ca2+ pump (Furukawa, K.-I., Tawada, Y., & Shigekawa, M. (1988) J. Biol. Chem. 263, 8058-8065). Intracellular contents of Na+ and K+ and intracellular pH, on the other hand, were not influenced by the change in Em under the conditions used. These results indicate that alteration in Em can modulate the intracellular Ca2+ concentration in intact VSMCs by changing the rate of Ca2+ extrusion by the plasma membrane Ca2+ pump. The data strongly suggest that the plasma membrane Ca2+ pump in VSMCs is electrogenic.     

Solute transport process in intestinal epithelial cells

In rat small intestine, the active transport of organic solutes results in significant depolarization of the membrane potential measured in an epithelial cell with respect to a grounded mucosal solution and in an increase in the transepithelial potential difference. According to the analysis with an equivalent circuit model for the epithelium, the changes in emf's of mucosal and serosal membranes induced by active solute transport were calculated using the measured conductive parameters. The result indicates that the mucosal cell membrane depolarizes while the serosal cell membrane remarkably hyperpolarizes on the active solute transport. Corresponding results are derived from the calculations of emf's in a variety of intestines, using the data that have hitherto been reported. The hyperpolarization of serosal membrane induced by the active solute transport might be ascribed to activation of the serosal electrogenic sodium pump. In an attempt to determine the causative factors in mucosal membrane depolarization during active solute transport, cell water contents and ion concentrations were measured. The cell water content remarkably increased and, at the same time, intracellular monovalent ion concentrations significantly decreased with glucose transport. Net gain of glucose within the cell was estimated from the restraint of osmotic balance between intracellular and extracellular fluids. In contrast to the apparent decreases in intracellular Na+ and K+ concentrations, significant gains of Na+ and K+ occurred with glucose transport. The quantitative relationships among net gains of Na+, K+ and glucose during active glucose transport suggest that the coupling ratio between glucose and Na+ entry by the carrier mechanism on the mucosal membrane is approximately 1:1 and the coupling ratio between Na+-efflux and K+-influx of the serosal electrogenic sodium pump is approximately 4:3 in rat small intestine. In addition to the electrogenic ternary complex inflow across the mucosal cell membrane, the decreases in intracellular monovalent ion concentrations, the temporary formation of an osmotic pressure gradient across the cell membrane and the streaming potential induced by water inflow through negatively charged pores of the cell membrane in the course of an active solute transport in intestinal epithelial cells are apparently all possible causes of mucosal membrane depolarization.     

Veratridine-induced oscillations in membrane potential of cultured rat skeletal muscle: Role of the Na-K pump

1. The acute effects of veratridine on membrane potential (Em) and Na-K pump activity in cultured skeletal muscle were examined. 2. At a concentration of 10(-4) M, veratridine caused depolarization of Em and a decrease in Na-K pump activity. At concentrations of 10(-5) and 10(-6) M, veratridine caused oscillations of Em and an increase in Na-K pump activity compared to untreated, control cells. The oscillations consisted of depolarization to about -40 mV followed by hyperpolarization to about -90 mV; the level of hyperpolarization was higher at 37 than at 23 degrees C. 3. Veratridine-induced oscillations could be prevented by pretreatment with tetrodotoxin (10(-6) M) and blocked or prevented by ouabain, which depolarizes Em of cultured myotubes. In contrast, depolarization of Em to -60 mV by excess K+ did not alter the amplitude or frequency of the oscillations. 4. The results demonstrate that veratridine-induced increase in Na influx both depolarizes cultured myotubes and increases the activity of the Na-K pump, which repolarizes Em to levels higher than control. This sequence accounts for veratridine-induced oscillations in Em. High concentrations of veratridine cause only depolarization of Em and inhibition of Na-K pump activity.     

Evidence for a circulating endogenous Na+-K+ pump inhibitor in low-renin hypertension

It is now more than 10 years since we suggested that an endogenous Na+,K+-ATPase inhibitor might participate in the genesis of certain forms of ren hypertension. Although the question is not yet fully resolved, there has been much activity in the area. We here review that activity. In 1980 we reported that supernatant of boiled plasma from dogs with one-kidney, one wrapped hypertension reduces Na+-K+ pump activity when applied to an artery from another animal. Since then, we and a number of other investigators have described Na+-K+ pump inhibitory activity in the plasma of animals and humans with hypertension, particularly the low-renin varieties. The activity results from a heat-stable small molecule, but the chemical structure of the molecule is unknown. It appears to be released from the hypothalamus in response to pulmonary vascular distension and to act on blood vessels via electrogenic depolarization. Although it may be sufficient by itself to raise pressure, it may be most effective when superimposed on vascular smooth muscle cells that are abnormally permeable to Na+. Efforts to determine the chemical structure of the agent or agents should be intensified.     

Role of Na-K ATPase in regulation of resting membrane potential of cultured rat skeletal myotubes

The role of Na-K ATPase in the determination of resting membrane potential (Em) as a function of extracellular K ion concentration was investigated in cultured rat myotubes. The Em of control myotubes at 37 degrees C varied as a function of (K+)0 with a slope of about 58-60 mV per ten-fold change in (K+)0. Inhibition of the Na-K pump with ouabain or by reduced temperature revealed that this relation consists of two components. One, between (K+)0 of 10 and 100 mM, remains unchanged by alterations in enzyme activity; The second, between (K+)0 of 1 and 10 mM, is related to the amount of Na-K pump activity, the slope decreasing as pump activity decreases. Indeed, with complete inhibition of the Na-K pump, Em does not change over the range of (K+)0 1 to 10 mM. Measurements of 86Rb efflux and input resistance of individual myotubes showed that membrane permeability does not change as (K+)0 increases from 1 to 10 mM but increases as (K+)0 increases further. Monensin, which increases Na ion permeability, increases Em at values of external K+ below 10 mM, and is without effect at higher values of K+ concentration. The effect of monensin is blocked by ouabain. Tetrodotoxin, which blocks voltage-dependent Na+ channels, decreases Em at low (2-10 mM) K+. We conclude that changes in Em as a function of extracellular K+ concentration in the physiological range are not adequately explained by the diffusion potential hypothesis of Em, and that other theories (electrogenic pump, surface-absorption) must be considered.     

Relationship between vanadate induced relaxation and vanadium content in guinea pig taenia coli

Abstract: Vanadate has been known to induce a transient increase in high K+ induced contraction, and also gradually relax the high K+ contraction itself in guinea pig taenia coli. The relationship between the rate of relaxation and ion content of Na+, K+, and V ion at the cellular level was investigated when vanadate was applied to contracted muscle. Tissue Na+ and V ion content increased linearly, depending on the time after vanadate treatment, reaching maximum levels of approximately 50 mM x kg(-1) and 0.25 mM x kg(-1) wet weight, respectively. There was a positive correlation between the V ion and Na+ contents, while there was a negative correlation between both ions and the relaxed rate of the high K+ induced contraction. The uptake of V ion was affected by the external K+ concentration, and the maximum rate of V ion uptake decreased to 40% in the presence of 90 mM external K+. These results suggest that a small amount of V ion was enough to inhibit the Na+ pump activity and muscle contraction in the high K+ solution.     

Reactivity in human cerebral artery: species variation

It is becoming obvious that the reactivity of vascular smooth muscle to vasoactive agents is not homogeneous in different arteries (cerebral vs. other arteries) from the same animal species or in cerebral arteries from different species. In this communication, the reactivity of human cerebral arteries to norepinephrine (NE), dopamine (DA), small amounts of K+ and ouabain and their mechanisms of action are compared with those in monkey and dog cerebral arteries. NE produces moderate contractions in the primate arteries, mediated by alpha 1 adrenoceptors, and slight contractions in the dog arteries, possibly mediated by alpha 2 receptors. DA relaxes human and monkey cerebral arteries but contracts the dog arteries. The primate artery relaxation mediated by dopaminergic receptors is large enough to predominate over the alpha 1 receptor-mediated contraction. Minute amounts of K+ preferentially relax cerebral arteries from humans, monkeys, and dogs, possibly activating the electrogenic Na+ pump. Ouabain, a Na+ pump inhibitor, contracts these cerebral arteries with low concentrations. Human and monkey cerebral arteries respond similarly to vasoactive agents presented so far, and appear to share the same mechanisms of action; however, the responsiveness of dog cerebral arteries differs.     

Chloride in smooth muscle

Interest in the functions of intracellular chloride expanded about twenty years ago but mostly this referred to tissues other than smooth muscle. On the other hand, accumulation of chloride above equilibrium seems to have been recognised more readily in smooth muscle.

Experimental data is used to show by calculation that the Donnan equilibrium cannot account for the chloride distribution in smooth muscle but it can in skeletal muscle. The evidence that chloride is normally above equilibrium in smooth muscle is discussed and comparisons are made with skeletal and cardiac muscle. The accent is on vascular smooth muscle and the mechanisms of accumulation and dissipation.

The three mechanisms by which chloride can be accumulated are described with some emphasis on calculating the driving forces, where this is possible. The mechanisms are chloride/bicarbonate exchange, (Na+K+Cl) cotransport and a novel entity, “pump III”, known only from own work. Their contributions to chloride accumulation vary and appear to be characteristic of individual smooth muscles. Thus, (Na+K+Cl) always drives chloride inwards, chloride/bicarbonate exchange is always present but does not always do it and “pump III” is not universal.

Three quite different biophysical approaches to assessing chloride permeability are considered and the calculations underlying them are worked out fully. Comparisons with other tissues are made to illustrate that low chloride permeability is a feature of smooth muscle.

Some of the functions of the high intracellular chloride concentrations are considered. This includes calculations to illustrate its depolarising influence on the membrane potential, a concept which, experience tells us, some people find confusing. The major topic is the role of chloride in the regulation of smooth muscle contractility. Whilst there is strong evidence that the opening of the calcium-dependent chloride channel leads to depolarisation, calcium entry and contraction in some smooth muscles, it appears that chloride serves a different function in others. Thus, although activation and inhibition of (Na+K+Cl) cotransport is associated with contraction and relaxation respectively, the converse association of inhibition and contraction has been seen. Nevertheless, inhibition of chloride/bicarbonate exchange and “pump III” and stimulation of (K+Cl) cotransport can all cause relaxation and this suggests that chloride is always involved in the contraction of smooth muscle.

The evidence that (Na+K+Cl) cotransport more active in experimental hypertension is discussed. This is a common but not universal observation. The information comes almost exclusively from work on cultured cells, usually from rat aorta. Nevertheless, work on smooth muscle freshly isolated from hypertensive rats confirms that (Na+K+Cl) cotransport is activated in hypertension but there are several other differences, of which the depolarisation of the membrane potential may be the most important.

Finally, a simple calculation is made which indicates as much as 40% of the energy put into the smooth muscle cell membrane by the sodium pump is necessary to drive (Na+K+Cl) cotransport. Notwithstanding the approximations in this calculation, this suggests that chloride accumulation is energetically expensive. Presumably, this is related to the apparently universal role of chloride in contraction.     

Properties of smooth muscle hyperpolarization and relaxation to K+ in the rat isolated mesenteric artery

Smooth muscle membrane potential and tension in rat isolated small mesenteric arteries (inner diameter 100-200 microm) were measured simultaneously to investigate whether the intensity of smooth muscle stimulation and the endothelium influence responses to exogenous K+. Variable smooth muscle depolarization and contraction were stimulated by titration with 0.1-10 microM phenylephrine. Raising external K+ to 10.8 mM evoked correlated, sustained hyperpolarization and relaxation, both of which were inhibited as the smooth muscle depolarized and contracted to around -38 mV and 10 mN, respectively. At these higher levels of stimulation, raising the K+ concentration to 13.8 mM still hyperpolarized and relaxed the smooth muscle. Relaxation to endothelium-derived hyperpolarizing factor, released by ACh, was not altered by the level of stimulation. In endothelium-denuded arteries, the concentration-relaxation curve to K+ was shifted to the right but was not depressed. In denuded arteries, relaxation to K+ was unaffected by the extent of prior stimulation and was blocked with 0.1 mM ouabain but not with 30 microM Ba2+. The ability of K+ to stimulate simultaneous hyperpolarization and relaxation in the mesenteric artery is consistent with a role as an endothelium-derived hyperpolarizing factor activating inwardly rectifying K+ channels on the endothelium and Na+-K+-ATPase on the smooth muscle cells.     

Trans-membrane cationic flow and hemodialysis]

Abnormalities of intracellular ion concentrations and transmembrane fluxes were reported in uremia. In RBC from 12 chronically hemodialyzed patients (age 41 + 12, 7 men, 5 women; mean dialysis duration 31 + 24 months), we evaluated the acute effects of hemodialysis on intracellular Na and K concentrations, ouabain sensitive Na/K pump, furosemide sensitive Na/K cotransport, Na/Li countertransport, and passive permeability to Na. Six patients were normotensive and 6 were taking antihypertensive drugs which were withdrawn before the study. When compared to our normal reference group, uremic patients showed a significant increase in intracellular K concentration and a significant decrease in ouabain-sensitive Na/K pump. Intracellular sodium was not increased. No correlation was found between the activity of sodium-potassium pump and the duration of hemodialysis. The other transport systems were comparable to normal. No significant change was observed between the values measured before and after dialysis. Ouabain sensitive Na/K pump was lower in hypertensive as compared to normotensive patients, but this difference was not significant. Our data support the existence of ion transport derangements in uremia, which are not acutely affected by hemodialysis.     

Arachidonate has a positive effect on the electrogenic sodium-potassium pump in the mouse diaphragm

Sodium arachidonate 5 X 10(-5) mol X l-1 shortened the time course of hyperpolarization caused by the electrogenic Na+-K+ pump in intact muscle fibres in the mouse diaphragm preincubated in a K+-free physiological solution. Contrary to experiments on membrane fragments, no inhibition of the ouabain-sensitive Na+-K+ ATPase was observed. It is unlikely that the arachidonate may be identical with the endogenous "ouabain-like" substance (Bidard et al. 1984).     

Contraction, membrane potential, and calcium fluxes in rabbit pulmonary arterial muscle

The rabbit main pulmonary artery (RMPA) has frequently been used for studies of contraction, membrane properties, and ion fluxes. The resting membrane potential (Em) of the smooth muscle cells of the RMPA is close to -60 mV. The diffusion potential calculated from ion concentrations and permeabilities is -31 to -40 mV, which suggests that electrogenic ion pumping contributes to the actual Em. Circumferential strips of RMPA possess cablelike properties with a space constant lambda of 1.9 mm. Contraction of RMPA to high K+ depends on extracellular Ca2+, is associated with 45Ca influx, is inhibited by Ca2+ entry blockers, and occurs after depolarization of the membrane to -45 to -33 mV. Maximal contractile responses to K+ and norepinephrine (NE) were similar. At low concentrations (3 X 10(-8)-10(-6) M) NE and the alpha 1-agonist methoxamine induced concentration-dependent depolarization and contraction. Above 10(-6) M contraction occurred in the absence of further changes in Em. Membrane resistance, estimated from measurements of space constant, decreased over the entire concentration-contraction curve of alpha agonists. Blockade of potassium channels by tetraethylammonium unmasked depolarization at high NE concentrations. It is concluded that in the RMPA alpha 1-adrenoceptor stimulation is associated with changes in electrical membrane properties and may in this way trigger contraction.     

Estrogen receptors and effects of estrogen on membrane electrical properties of coronary vascular smooth muscle

The effect of estrogen stimulation in vitro on the electrical properties of vascular smooth muscle (VSM), and the concentration of estrogen receptors in VSM were measured in isolated coronary arteries. Microelectrode measurements of the dog coronary artery membrane potential (Em) showed quiescent values of -51 millivolts (mV) and an input resistance (rin) of 10 megohms. Addition by diethylstilbestrol (DES) at 10(-6) M hyperpolarized the membrane to -64 mV and reduced input resistance (rin) to 5 megohms within 15 minutes. Extrapolation of the Em vs. log [K]o curve to zero potential gave similar values of [K]i of around 170 mM in both normal and DES treated muscles suggesting that the DES induced hyperpolarization is not due to increased Na-K pump activity. The 0.5% ethanol vehicle alone had no effect on the membrane potentials. Tetraethylammonium ion (TEA) induced action potentials in the previously quiescent tissue. When DES was applied in the presence of TEA, the membrane potential increased and the action potentials were abolished. Scatchard analysis of the estrogen receptor binding demonstrated both a high and a low affinity receptor for estrogen in the VSM. These data indicate that DES hyperpolarizes the VSM cells by a mechanism other than an increased Na-K pump activity. The mechanism of this increased Em may be due to factors which increase K+ conductance either mediated directly through estrogen interaction with its cytosolic receptors or through some unidentified second mechanism.     

Sodium-calcium exchange in regulation of cardiac contractility. Evidence for an electrogenic, voltage-dependent mechanism

The origin and regulatory mechanisms of tonic tension (Ca current-independent component of contractility) were investigated in frog atrial muscle under voltage-clamp conditions. Tonic tension was elicited by depolarizing pulses of 160 mV (Em = +90 mV, i.e., close to E ca) and 400--600 ms long. An application of Na-free (LiCl) or Ca-free Ringer's solutions resulted in a fast (less than 120 s), almost complete abolition of tonic tension. When [Na]o was reduced (with LiCl or sucrose as the substitutes), the peak tonic tension increased transiently and then decreased below the control level. The transient changes in tonic tension were prevented by using low-Na, low-Ca solutions where the ratios [Ca]0/[Na]40 to [Ca]o/[Na]4o were kept constant (1.1 X 10(-8) mM-3 to 8.7 X 10(-13) mM-5). Na-free (LiCl) solution elicited contractures accompanied by a membrane hyperpolarization or by an outward current even when the Na-K pump was inhibited. 15 mM MnCl2 (or 3 mM LaCl3) inhibited the development of the Na-free contracture and the related part of hyperpolarization or the outward current. In conclusion, our results indicate that tonic tension is regulated by a Na-Ca exchange mechanism. Furthermore, they suggest that this exchange could be electrogenic (exchanging three or more Na ions for one Ca ion) and thus voltage dependent. The possible contribution of an electrogenic Na-Ca exchange in the maintenance of cardiac membrane potential is discussed.