Inhibition of Na+,K+-ATPase by ouabain triggers epithelial cell death independently of inversion of the [Na+]i/[K+]i ratio

Treatment with ouabain led to massive death of principal cells from collecting ducts (C7-MDCK), indicated by cell swelling, loss of mitochondrial function, an irregular pattern of DNA degradation, and insensitivity to pan-caspase inhibitor. Equimolar substitution of extracellular Na(+) by K(+) or choline(+) sharply attenuated the effect of ouabain on intracellular Na(+) and K(+) content but did not protect the cells from death in the presence of ouabain. In contrast to ouabain, inhibition of the Na(+)/K(+) pump in K(+)-free medium increased Na(+)(i) content but did not affect cell survival. In control and K(+)-free medium, ouabain triggered half-maximal cell death at concentrations of approximately 0.5 and 0.05 microM, respectively, which was consistent with elevation of Na(+)/K(+) pump sensitivity to ouabain in K(+)-depleted medium. Our results show for the first time that the death of ouabain-treated renal epithelial cells is independent of the inhibition of Na(+)/K(+) pump-mediated ion fluxes and the [Na(+)](i)]/[K(+)](i) ratio.     

Transmembrane K+, Na+, and Cl- permeability and conductance changes in the frog sartorius fibres following ouabain and zero [K+]o treatment

Changes in the K+, Na+, and Cl- permeabilities (P) and conductances (g) of the intact frog sartorius fibre membrane following ouabain or zero [K+]o treatment were calculated from intrafibre activity and whole muscle electrolyte changes. Conventional equations relating ionic fluxes to resting potential (E), ionic gradient potential, and internal and external ionic activities were used. Both treatments produced a three- to five-fold increase in PNa and gNa. In addition, ouabain produced a fivefold increase in PK (and gK) and a small decrease in PCl (and gCl), whereas zero [K+]o produced a 60% reduction in PK, a 90% reduction in gK, and a threefold increase in PCl (and gCl). When the two treatments were combined, the P and g changes were paradoxical, suggesting that the ouabain-induced increase in gK and the zero [K+]o-induced decrease in gK were occurring but in different channels (or carriers). During ouabain treatment, E reflects mainly the transmembrane K+ gradient potential; during zero [K+]o treatment, E reflects mainly the Cl- gradient potential. Despite channel (or carrier) specificity, it appears that all three ionic permeabilities are altered during the perturbations.     

Coordinate expression of piretanide receptors and Na+,K+,Cl- cotransport activity in Madin-Darby canine kidney cell mutants

Two receptor sites for [3H]piretanide, a sulfamoylbenzoic acid loop diuretic, have been identified in intact Madin-Darby canine kidney cells, an epithelial cell line derived from dog kidney. The two receptor sites differed in their affinity for piretanide (KD1 = 2.1 +/- 1.4 nM and KD2 = 264 +/- 88 nM) and the maximal number of sites (Bmax1 = 11 +/- 4 and Bmax2 = 120 +/- 80 fmol/mg of protein). Madin-Darby canine kidney cells are known to possess a tightly coupled and highly cooperative Na+,K+,Cl- cotransporter which is sensitive to loop diuretics. Under ionic conditions identical to those used to study piretanide binding (30 mM Na+, 30 mM K+, 30 mM Cl-), the Ki for inhibition of the initial rate of 86Rb+ uptake by piretanide was 333 +/- 92 nM, a value not significantly different from the KD of the low affinity receptor site. [3H]Piretanide binding to three low K+-resistant mutants derived from this cell line was also studied. These mutants had been previously characterized as being partially or completely defective in Na+,K+,Cl- cotransport activity (McRoberts, J. A., Tran, C. T., and Saier, M. H., Jr. (1983) J. Biol. Chem. 258, 12320-12326). One of these mutants had undetectable levels of Na+,K+,Cl- cotransport activity and low to undetectable levels of specific piretanide binding. The second mutant had low but measurable levels of cotransport activity (11% of the wild-type levels) and displayed very low affinity (KD approximately 8000 nM) specific piretanide binding. In the third mutant, expression of Na+,K+,Cl- cotransport activity and both piretanide receptors was cell density-dependent. Subconfluent to just-confluent cultures of this mutant lacked detectable cotransport activity as well as specific piretanide binding, whereas very dense cultures displayed both piretanide receptors and had intermediate to nearly normal levels of cotransport activity. These results demonstrate that the Na+,K+,Cl- cotransporter is a receptor for loop diuretics, but they also raise questions about the functional significance of the two piretanide receptor sites.     

Role of cyclic GMP in atrial natriuretic factor stimulation of Na+,K+,Cl- cotransport in vascular smooth muscle cells

Atrial natriuretic factor (ANF) has been shown to bind to specific receptors on vascular smooth muscle cells (VSMC) and to cause an increase in intracellular cyclic GMP (cGMP) content. We have recently demonstrated that a prominent Na+,K+,Cl- cotransport system is present in VSMC and that a permeable cGMP analog (8-bromo-cGMP) stimulates activity of the cotransporter. We have also shown that the ANF peptide, rat atriopeptin III, stimulates Na+,K+,Cl- cotransport and elevates intracellular cGMP levels in VSMC. In the present study, we tested the hypothesis that ANF stimulation of Na+,K+,Cl- cotransport occurs via an increase in cGMP levels. When the quinolinedione, 6-anilo-5,8-quinolinedione (LY83583) (10 microM), was used to block formation of cGMP in VSMC from primary cultures of rat thoracic aorta, it was found that both basal and rat atriopeptin III (100 nM)-stimulated Na+,K+,Cl- cotransport were significantly inhibited. The effect of LY83583 was dose-dependent and the half-maximal inhibitory concentration was 0.5 microM. LY83583 also inhibited cotransport in the presence of a maximal concentration of 8-bromo-cGMP. However, this inhibition was not seen in cells also treated with 2-O-propoxyphenyl-8-azapurin-6-one (M&B 22,948), an inhibitor of cGMP phosphodiesterase. M&B 22,948 alone also increased levels of cotransport. Since inhibition of cGMP formation blocks ANF-stimulated Na+,K+,Cl- cotransport and inhibition of cGMP breakdown enhances Na+, K+, Cl- cotransport, we conclude that ANF stimulation of Na+,K+,Cl- cotransport in VSMC is mediated via increase in intracellular cGMP levels.     

Na+, K+, and Cl- transport in resting pancreatic acinar cells

To understand the role of Na+, K+, and Cl- transporters in fluid and electrolyte secretion by pancreatic acinar cells, we studied the relationship between them in resting and stimulated cells. Measurements of [Cl-]i in resting cells showed that in HCO3(-)-buffered medium [Cl- ]i and Cl- fluxes are dominated by the Cl-/HCO3- exchanger. In the absence of HCO3-, [Cl-]i is regulated by NaCl and NaK2Cl cotransport systems. Measurements of [Na+]i showed that the Na(+)-coupled Cl- transporters contributed to the regulation of [Na+]i, but the major Na+ influx pathway in resting pancreatic acinar cells is the Na+/H+ exchanger. 86Rb influx measurements revealed that > 95% of K+ influx is mediated by the Na+ pump and the NaK2Cl cotransporter. In resting cells, the two transporters appear to be coupled through [K+]i in that inhibition of either transporter had small effect on 86Rb uptake, but inhibition of both transporters largely prevented 86Rb uptake. Another form of coupling occurs between the Na+ influx transporters and the Na+ pump. Thus, inhibition of NaK2Cl cotransport increased Na+ influx by the Na+/H+ exchanger to fuel the Na+ pump. Similarly, inhibition of Na+/H+ exchange increased the activity of the NaK2Cl cotransporter. The combined measurements of [Na+]i and 86Rb influx indicate that the Na+/H+ exchanger contributes twice more than the NaK2Cl cotransporter and three times more than the NaCl cotransporter and a tetraethylammonium-sensitive channel to Na+ influx in resting cells. These findings were used to develop a model for the relationship between the transporters in resting pancreatic acinar cells.     

On the inhibition of [Na+,K+]-ATPases by the components of Naja mossambica mossambica venom: evidence for two distinct rat brain [Na+,K+]-ATPase activities

We have compared the effects of highly purified preparations of cardiotoxins and phospholipases A2 from Naja mossambica mossambica venom on rat brain [Na+,K+]-ATPase activity. The results were the following: (i) micromolar concentrations of cardiotoxin preparations were required to inhibit [Na+,K+]-ATPase activity to the extent achieved by picomolar concentrations of phospholipases A2; i.e., the inhibitory effect of cardiotoxins appeared to be related to the contamination of the preparations by trace amounts of phospholipase A2; (ii) comparing phospholipases A2 from varied origins, a correlation was observed between [Na+,K+]-ATPase inhibition, isoelectric point, and toxicity for mice; (iii) when rat brain membranes were used, incubation for extended times with the most basic N. mossambica mossambica phospholipase A2 resulted in a biphasic [Na+,K+]-ATPase inhibition, suggesting that two distinct [Na+,K+]-ATPases were affected differentially. In contrast, incubation of rat brain membranes with either porcine pancreatic phospholipase A2, notexin, or beta-bungarotoxin and also incubation of erythrocyte membranes with the most basic N. mossambica mossambica phospholipase A2 produced monophasic [Na+,K+]-ATPase inhibitions. We discuss a possible specific action of toxic, basic phospholipase A2 on one of the [Na+,K+]-ATPase isoforms of excitable membranes.     

Furosemide-sensitive salt transport in the Madin-Darby canine kidney cell line. Evidence for the cotransport of Na+, K+, and Cl-

Confluent monolayer cultures of the Madin-Darby canine kidney (MDCK) cell line have been shown to possess a furosemide and bumetanide-sensitive (Na+,K+)-cotransport system. We have studied the effect of anion substitutions on (Na+,K+)-cotransport. In Na+-depleted cells, bumetanide-sensitive uptake of 22Na+ or 86Rb+ exhibited an absolute requirement for extracellular Cl-. Chloride could be replaced in the buffers by Br-, but not by F-, I-, acetate, nitrate, thiocyanate, sulfate, or gluconate. The effect of Cl- was saturating, and Na+-stimulated 86RB+ uptake as well as K+-stimulated 22Na+ uptake was shown to be dependent on the square of the Cl- concentration. The concentration of Cl- which gave half-maximal stimulation of cation cotransport varied between 58 and 70 mM. There was a small degree of cooperativity between the binding affinities for Cl- and K+ at constant Na+ concentrations. Bumetanide-sensitive 36Cl- uptake could be demonstrated when extracellular Na+ and K+ were present simultaneously. Uptake through this system was unaffected by changes in the membrane potential or by the imposition of pH gradients. Together these data strongly suggest that the bumetanide-sensitive transport system in Madin-Darby canine kidney cells co-transports Na+, K+, and Cl- in a ratio of 1:1:2.     

Transient activation and delayed inhibition of Na+,K+,Cl- cotransport in ATP-treated C11-MDCK cells involve distinct P2Y receptor subtypes and signaling mechanisms

In C11-MDCK cells, which resemble intercalated cells from collecting ducts of the canine kidney, P2Y agonists promote transient activation of the Na+,K+,Cl- cotransporter (NKCC), followed by its sustained inhibition. We designed this study to identify P2Y receptor subtypes involved in dual regulation of this carrier. Real time polymerase chain reaction analysis demonstrated that C11-MDCK cells express abundant P2Y1 and P2Y2 mRNA compared with that of other P2Y receptor subtypes. The rank order of potency of agents (ATP approximately UTP > 2-(methylthio)-ATP (2MeSATP); adenosine 5'-[beta-thio]diphosphate (ADPbetaS) inactive) indicated that P2Y2 rather than P2Y1 receptors mediate a 3-4-fold activation of NKCC within the first 5-10 min of nucleotide addition. NKCC activation in ATP-treated cells was abolished by the intracellular calcium chelator 1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid, calmodulin (CaM) antagonists trifluoroperazine and W-7, and KN-62, an inhibitor of Ca2+/CaM-dependent protein kinase II. By contrast with the transient activation, 30-min incubation with nucleotides produced up to 4-5-fold inhibition of NKCC, and this inhibition exhibited a rank order of potency (2MeSATP > ADPbetaS > ATP > UTP) typical of P2Y1 receptors. Unlike the early response, delayed inhibition of NKCC occurred in 1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid-loaded cells and was completely abolished by the P2Y1 antagonists MRS2179 and MRS2500. Transient activation and delayed inhibition of NKCC in C11 cell monolayers were observed after the addition of ATP to mucosal and serosal solutions, respectively. NKCC inhibition triggered by basolateral application of ADPbetaS was abolished by MRS2500. Our results thus show that transient activation and delayed inhibition of NKCC in ATP-treated C11-MDCK cells is mediated by Ca2+/CaM-dependent protein kinase II- and Ca2+-independent signaling triggered by apical P2Y2 and basolateral P2Y1 receptors, respectively.     

Regulation of ion concentration by Drosophila hydei Na+, K+, Ca2+, and Cl-

The ionic composition of the haemolymph of osmotically unencumbered larvae of Drosophila hydei shows a pattern that is typical (Bone, 1944) for highly developed phytophagous insect larvae: 36 mval/l. Cl^?; 56 mval/l. Na⁺; 31 mval/l. K⁺; approximately 18 mval/l. Ca²⁺ at an osmolality of 299 mOsmol/l.The larvae are able to maintain their most favourable ionic concentrations in the haemolymph after experimental osmotic stress in hypertonic as well as in hypotonic media. The reactions are most distinct with an increase or decrease of Cl^? concentration of the external medium. Characteristic regulating processes begin, and the Cl^? concentration of the haemolymph adjusts to the ‘standard’ again. The principal lapse shows the functional representation of an intensely suppressed oscillation. There seems to be a two-point regulation which requires the existence of a Cl^? ion depot and the existence of Cl^?-sensitive receptors.     

Removal of Na+ channels in squid giant axons by perfusion with trypsin

The irreversible effects of the proteolytic enzyme trypsin on ionic and gating currents of voltage-clamped squid axon membranes have been studied. At physiological pH, internal perfusion of the fibre with trypsin was found to be very effective in removing Na+ channels leaving the potassium system almost unaltered. At T = 13 degrees C the rates of channel-cleavage averaged 1/10 min-1 for the Na+ and 1/128 min-1 for the K+ channel, respectively. As estimated by the decrement of peak sodium conductance, the rate of loss of Na+ channels correlates well with the rate of decrease of the total charge associated with the ON component of gating currents, indicating that trypsin probably interacts with an essential proteic portion of the channel whose removal might prevent both the displacement of gating charges and the subsequent opening of the channel. Intracellular pH remarkably influences the action of the enzyme. A plot of the pH-dependence of the rate of cleavage of Na+ channels suggests the involvement of a positively charged group (either lysine or arginine) in the substrate region of the trypsin catalytic reaction.     

Cardiotonic steroids differentially affect intracellular Na+ and [Na+]i/[K+]i-independent signaling in C7-MDCK cells

Recently, we reported that ouabain kills renal epithelial and vascular endothelial cells independently of elevation of the [Na(+)](i)/[K(+)](i) ratio. These observations raised the possibility of finding cardiotonic steroids (CTS) that inhibit the Na(+),K(+) pump without attenuating cell survival and vice versa. To test this hypothesis, we compared CTS action on Na(+),K(+) pump, [Na(+)](i) content, and survival of Madin-Darby canine kidney cells. At a concentration of 1 microM, ouabain and other tested cardenolides, as well as bufadienolides such as bufalin, cinobufagin, cinobufotalin, and telobufotoxin, led to approximately 10-fold inhibition of the Na(+),K(+) pump, a 2-3-fold decrease in staining with dimethylthiazol-diphenyltetrazolium (MTT), and massive death indicated by detachment of approximately 80% of cells and caspase-3 activation. In contrast, Na(+),K(+) pump inhibition and elevation of [Na(+)](i) seen in the presence of 3 microM marinobufagenin (MBG) and marinobufotoxin did not affect MTT staining and cell survival. Inhibition of the Na(+),Rb(+) pump in K(+)-free medium was not accompanied by a decline of MTT staining and cell detachment but increased sensitivity to CTS. In K(+)-free medium, half-maximal inhibition of (86)Rb influx was observed in the presence of 0.04 microM ouabain and 0.1 microM MBG, whereas half-maximal detachment and decline of MTT staining were detected at 0.03 and 0.004 microM of ouabain versus 10 and 3 microM of MBG, respectively. Both ouabain binding and ouabain-induced [Na(+)](i),[K(+)](i)-independent signaling were suppressed in the presence of MBG. Thus, our results show that CTS exhibit distinctly different potency in Na(+),K(+) pump inhibition and triggering of [Na(+)](i)/[K(+)](i)-independent signaling, including cell death.     

Regulation of Na+/K+/Cl- cotransport and [3H]bumetanide binding site density by phorbol esters in HT29 cells

The involvement of protein kinase C in the regulation of Na+/K+/Cl- cotransport was investigated in cultured HT29 human colonic adenocarcinoma cells. We have demonstrated previously the presence of a Na+/K+/Cl- cotransport pathway in HT29 cells (Kim, H.D., Tsai, Y-S., Franklin, C.C., and Turner, J.T. (1989) Biochim. Biophys. Acta 946, 397-404). Treatment of cells with the phorbol esters phorbol 12-myristate 13-acetate (PMA) and phorbol 12,13-dibutyrate (PDBu) caused an increase in membrane-associated protein kinase C activity that was accompanied by a concomitant decrease in cytosolic protein kinase C activity. PMA also produced a rapid transient increase in cotransport to 137% of control values by 5 min followed by a progressive decrease to 19% of control values by 2 h. To determine the underlying mechanism for the reduction in Na+/K+/Cl- cotransport, changes in cotransporter number and/or affinity were determined in radioligand binding studies using [3H]bumetanide. PMA and PDBu produced essentially identical time- and dose-dependent decreases in specific [3H]bumetanide binding that were similar to the observed decreases in cotransport. Analysis of saturation and competition binding data indicated that the decrease in binding was due to a lowered Bmax with no change in affinity. Both the decrease in binding and the changes in cotransport elicited by PMA were prevented by the protein kinase inhibitor H7. These findings suggest that phorbol esters cause a decrease in the number of cotransporters in HT29 cells, resulting in a reduction in Na+/K+/Cl- cotransport activity.     

Comparison of changes evoked by GABA (gamma-aminobutyric acid) and anoxia in [K+]o, [Cl-]o, and [Na+]o in stratum pyramidale and stratum radiatum of the guinea pig hippocampus

Ion-selective microelectrode recordings were made to assess a possible contribution of extracellular gamma-aminobutyric acid (GABA) accumulation to early responses evoked in the brain by anoxia and ischemia. Changes evoked by GABA or N2 in [K+]o, [Cl-]o, [Na+]o, and [TMA+]o were recorded in the cell body and dendritic regions of the stratum pyramidale (SP) and stratum radiatum (SR), respectively, of pyramidal neurons in CA1 of guinea pig hippocampal slices. Bath application of GABA (1-10 mM) for approximately 5 min evoked changes in [K+]o and [Cl-]o with respective EC50 levels of 3.8 and 4.1 mM in SP, and 4.7 and 5.6 mM in SR. In SP 5 mM GABA reversibly increased [K+]o and [Cl-]o and decreased [Na+]o; replacement of 95% O2 -5% CO2 by 95% N2 -5% CO2 for a similar period of time evoked changes which were for each ion in the same direction as those with GABA. In SR both GABA and N2 caused increases in [K+]o and decreases in [Cl-]o and [Na+]. The reduction of extracellular space, estimated from levels of [TMA+]o during exposures to GABA and N2, was 5-6% and insufficient to cause the observed changes in ion concentration. Ion changes induced by GABA and N2 were reversibly attenuated by the GABA(A) receptor antagonist bicuculline methiodide (BMI, 100 microM). GABA-evoked changes in [K+]o in SP and SR and [Cl-]o in SP were depressed by > or =90%, and of [Cl-]o in SR by 50%; N2-evoked changes in [K+]o in SP and SR were decreased by 70% and those of [Cl-]o by 50%. BMI blocked delta [Na+]o with both GABA and N2 by 20-30%. It is concluded that during early anoxia: (i) accumulation of GABA and activation of GABA(A) receptors may contribute to the ion changes and play a significant role, and (ii) responses in the dendritic (SR) regions are greater than and (or) differ from those in the somal (SP) layers. A large component of the [K+]o increase may involve a GABA-evoked Ca2+-activated gk, secondary to [Ca2+]i increase. A major part of [Cl-]o changes may arise from GABA-induced g(Cl) and glial efflux, with strong stimulation of active outward transport and anion exchange at SP, and inward Na+/K+/2Cl- co-transport at SR. Na+ influx is attributable mainly to Na+-dependent transmitter uptake, with only a small amount related to GABA(A) receptor activation. Although the release and (or) accumulation of GABA during anoxia might be viewed as potentially protectant, the ultimate role may more likely be an important contribution to toxicity and delayed neuronal death.     

Inhibition of K/HCO(3) cotransport in squid axons by quaternary ammonium ions

Previous squid-axon studies identified a novel K/HCO₃ cotransporter that is insensitive to disulfonic stilbene derivatives. This cotransporter presumably responds to intracellular alkali loads by moving K⁺ and HCO⁻ ₃ out of the cell, tending to lower intracellular pH (pH_i). With an inwardly directed K/HCO₃ gradient, the cotransporter mediates a net uptake of alkali (i.e., K⁺ and HCO⁻ ₃ influx). Here we test the hypothesis that intracellular quaternary ammonium ions (QA⁺) inhibit the inwardly directed cotransporter by interacting at the intracellular K⁺ site. We computed the equivalent HCO⁻ ₃ influx (J _HCO3) mediated by the cotransporter from the rate of pH_i increase, as measured with pH-sensitive microelectrodes. We dialyzed axons to pH_i 8.0, using a dialysis fluid (DF) free of K⁺, Na⁺ and Cl⁻. Our standard artificial seawater (ASW) also lacked Na⁺, K⁺ and Cl⁻. After halting dialysis, we introduced an ASW containing 437 mm K⁺ and 0.5% CO₂/12 mm HCO⁻ ₃, which (i) caused membrane potential to become transiently very positive, and (ii) caused a rapid pH_i decrease, due to CO₂ influx, followed by a slower plateau-phase pH_i increase, due to inward cotransport of K⁺ and HCO⁻ ₃. With no QA⁺ in the DF, J _HCO3 was ∼58 pmole cm⁻² sec⁻¹. With 400 mm tetraethylammonium (TEA⁺) in the DF, J _HCO3 was virtually zero. The apparent K _i for intracellular TEA⁺ was ∼78 mm, more than two orders of magnitude greater than that obtained by others for inhibition of K⁺ channels. Introducing 100 mm inhibitor into the DF reduced J _HCO3 to ∼20 pmole cm⁻² sec⁻¹ for tetramethylammonium (TMA⁺), ∼24 for TEA⁺, ∼10 for tetrapropylammonium (TPA⁺), and virtually zero for tetrabutylammonium (TBA⁺). The apparent K _i value for TBA⁺ is ∼0.86 mm. The most potent inhibitor was phenyl-propyltetraethylammonium (PPTEA⁺), with an apparent K _i of ∼91 μm. Thus, trans-side quaternary ammonium ions inhibit K/HCO₃ influx in the potency sequence PPTEA⁺ > TBA⁺ > TPA⁺ > TEA⁺≅ TMA⁺. The identification of inhibitors of the K/HCO₃ cotransporter, for which no inhibitors previously existed, will facilitate the study of this transporter. Received: 21 November 2000/Revised: 14 May 2001     

Coupled Na/K/Cl efflux. "Reverse" unidirectional fluxes in squid giant axons

Studies of unidirectional Cl-, Na+, and K+ effluxes were performed on isolated, internally dialyzed squid giant axons. The studies were designed to determine whether the coupled Na/K/Cl co-transporter previously identified as mediating influxes (Russell. 1983. Journal of General Physiology. 81:909-925) could also mediate the reverse fluxes (effluxes). We found that 10 microM bumetanide blocked 7-8 pmol/cm2 X s of Cl- efflux from axons containing ATP, Na+, and K+. However, if any one of these solutes was removed from the internal dialysis fluid, Cl- efflux was reduced by 7-8 pmol/cm2 X s and the remainder was insensitive to bumetanide. About 5 pmol/cm2 X s of Na+ efflux was inhibited by 10 microM bumetanide in the continuous presence of 10(-5) M ouabain and 10(-7) M tetrodotoxin if Cl-, K+, and ATP were all present in the internal dialysis fluid. However, the omission of Cl- or K+ or ATP reduced the Na+ efflux, leaving it bumetanide insensitive. K+ efflux had to be studied under voltage-clamp conditions with the membrane potential held at -90 mV because the dominant pathway for K+ efflux (the delayed rectifier) has a high degree of voltage sensitivity. Under this voltage-clamped condition, 1.8 pmol/cm2 X s of K+ efflux could be inhibited by 10 microM bumetanide. All of these results are consistent with a tightly coupled Na/K/Cl co-transporting efflux mechanism. Furthermore, the requirements for cis-side co-ions and intracellular ATP are exactly like those previously described for the coupled Na/K/Cl influx process. We propose that the same transporter mediates both influx and efflux, hence demonstrating "reversibility," a necessary property for an ion-gradient-driven transport process.