Enhanced neuronal K+ conductance: a possible common mechanism for sedative-hypnotic drug action

It is commonly thought that central nervous system depressant drugs exert their actions through enhancement of gamma-aminobutyrate (GABA)-mediated mechanisms. Recently, the cellular electrophysiological evidence from this laboratory and others suggests that both sedative hypnotics and general anaesthetics inhibit central neurons by increasing potassium conductance (GK). We have utilized the mammalian in vitro hippocampal and cerebellar slice preparations at 34-36 degrees C. Intracellular recordings from CA1, CA3, and cerebellar Purkinje cells were obtained. Low dose (sedative) concentrations of ethanol (less than or equal to 20 mM), two different benzodiazepines (midazolam and clonazepam in low nanomolar concentrations), and pentobarbital (10(-6) to 10(-4) M) were applied by pressure ejection or were bath perfused. All drugs caused a hyperpolarization with decreased spontaneous activity, and enhanced post spike afterhyperpolarizations (AHPs). These long-lasting AHPs are presumably due to enhanced calcium-mediated GK. Increased responsiveness to focally applied GABA was only seen at higher doses (ethanol, 100 mM; midazolam, 10(-7) M; pentobarbital, 10(-4) M). These data suggest that the above neurodepressant drugs, when applied at sedative doses to hippocampal pyramidal cells, enhance GK and not the actions of GABA.     

Phencyclidine actions measured intracellularly in hippocampal CA1 neurons

The electrophysiological effects of phencyclidine (PCP) were measured intracellularly in guinea pig hippocampal CA1 neurons in vitro. At all doses tested (0.2 microM - 10 mM), PCP increased the width of action potentials (APs). Doses of 10 microM and higher were associated with decreased action potential amplitude. PCP decreased inhibitory postsynaptic potentials and excitatory postsynaptic potentials but did not alter responses to focally applied GABA. At the lowest dose (0.2 microM), PCP decreased the input resistance (Rin), while at all other doses Rin was increased. PCP decreased post-spike train afterhyperpolarizations at low and medium doses. PCP effects persisted in low calcium medium and also in medium containing 10(-6) M tetrodotoxin. It is concluded that in these central neurons, PCP primarily blocks potassium conductances at all doses and, at anesthetic doses, depresses sodium-dependent spikes.     

Metabolism of extracellular phosphocreatine during changes in the ionic composition of the medium in the perfused rat heart]

Perfusion of isolated rat hearts with a phosphocreatine (10(-4) M) containing solution to which strophanthin or KCl had been added up to a concentration of 27 mM as well as Ca2+ depletion decreased phosphocreatine concentration in the perfusate with a simultaneous increase in creatine and phosphocreatine concentrations in the myocardium. Neither high extracellular concentrations of Na+ (200 mM), nor phosphocreatine increased creatine and phosphocreatine levels in the myocardium. The effect of high sodium perfusion media was completely reversed by strophanthin. Phosphocreatine decreased the lactate content in the perfusate. Strophanthin or potassium chloride enhanced the effect of phosphocreatine on the lactate release. Conversely, creatine augmented the lactate content in the perfusate. A high specificity of the phosphocreatine effect on the myocardium independently of the ionic composition of the perfusate was postulated. A mechanism of protective effects of phosphocreatine and high sodium perfusion media on "calcium paradox" is proposed.     

Cationophore properties of the new polyether antibiotic Salinomycin investigated in distal rabbit colon in vivo and in vitro

The distal rabbit colon was used as a model to investigate the influence of the cationophore Salinomycin in vivo with a single-pass perfusion, and in vitro with a modified Ussing chamber technique. For in vivo experiments with labelled 14C-PEG as a volume marker in the perfusate, a dose of 10E-4 mol/l Salinomycin was used. Net water (53 microliters/h/cm), net chloride (3 mumol/h/cm) and net sodium (3.6 mumol/h/cm) absorption was not significantly influenced, but net potassium secretion (-3 mumol/h/cm) was decreased to zero and transepithelial potential (PD) reduced from -45 mV to -33 mV. 10E-4 mol/l Salinomycin, applied in vitro on the muscosal side, decreased PD in 80 min and 10E-3 mol/l in 30 min from 18 mV to zero. Both concentrations decreased the short-circuit current (Isc = 77 microA/cm2) in 60 min, respectively 30 min to 40 microA/cm2. After 60 min mucosal 10E-4 mol/l Salinomycin the Isc increased, resulting from a transepithelial conductance (Gt) increase from 3 to 40 mS/cm2. A dose-related effect was present on PD, Isc and Gt at concentrations between 10E-7 and 10E-6 mol/l. The unidirectional 22Na fluxes were increased to 20 times the control values and net Na transport disappeared. We conclude that Salinomycin when applied in usual doses (10E-4 mol/l) as a coccidiostatic feed additive, profoundly affects colonic electrolyte transport.     

Inhibition of small conductance K+ -channels attenuated melatonin-induced relaxation of serotonin-contracted rat gastric fundus

The aim of this study was to investigate the effects of melatonin on rat gastric fundus smooth muscle. Melatonin (10(-4) to 10(-3) M) had no effect on the basal tone of gastric smooth muscle. After precontraction with carbachol (10(-6) M) or serotonin (10(-7) M), melatonin caused a concentration dependent inhibitory action. The half maximal effect on serotonin-induced contraction was found with 1.12 +/- 0.86 x 10(-5) M of melatonin. Increasing concentrations of melatonin (10(-5) to 10(-3) M) resulted in a right shift of the serotonin concentration response curve (10(-10) to 10(-5) M). This inhibitory effect of melatonin was partially blocked in the presence of apamin (10(-10) to 10(-7) M), a specific blocker of the small conductance calcium-dependent potassium channel, but not in the presence of other potassium channel blockers like charybdotoxin (10(-8) M), glibenclamide (l0(-5) M), or tetraethylammonium (ODQ, 10(-4) M). The inhibitory effect was not changed in the presence of the neuronal blocker tetrodotoxin (10(-6) M), the selective P2-receptor antagonist pyridoxalphosphate-6-azophenyl-2',4'-disulphonic acid (3 x 10(-5) M), the nitric-oxide synthase inhibitor N-nitro-L-arginine (3 x 10(-4) M), or the guanylate cyclase inhibitor 1H-[1,2,4]oxadiazolo[4,3-a]-quinoxalin-1-one (10(-4) M), suggesting that neither the purinergic, nitrergic, nor guanylate cyclase pathways were involved. We further investigated inhibitory responses to electrical field stimulation (EFS) at different frequencies under non-adrenergic, non-cholinergic (NANC) conditions on a serotonin-induced contraction in the presence of melatonin (10)-5 to 10(-4) M). Melatonin significantly reduced these inhibitory NANC responses in higher (8-32 Hz), but not lower (05-4 Hz), frequencies (16 Hz without melatonin, 103 +/- 6.3%; melatonin 10(-5) M, 80.4 +/- 7.5%; melatonin 10(-4) M, 39.1 +/- 17.1%). Melatonin had no effect on contractile responses induced by EFS under basal tone. These results demonstrate that the inhibitory effect of melatonin in rat gastric fundus smooth muscle is apamin sensitive, but is not affected by other potassium channel blockers. This suggests that melatonin may be another transmitter candidate for the apamin sensitive responses within the gastrointestinal tract.     

Thyrotropin-releasing hormone (TRH) and its analog (DN-1417): interaction with pentobarbital in oxygen consumption and cyclic AMP formation of rat cerebral cortex slices

Effects of TRH or its analog DN-1417 (gamma-butyrolactone-gamma-carbonyl-L-histidyl-L- prolinamide ) and pentobarbital, alone or in combination, on oxygen consumption and cyclic AMP formation in rat cerebral cortex slices were investigated. The oxygen consumption of rat cerebral cortex slices as measured with a Warburg apparatus, increased linearly over time (0 to 60-min incubation at 37C). Addition of pentobarbital (1 to 7 x 10-4M) inhibited oxygen consumption, in a concentration-dependent manner, up to 45% of control. A concomitant application of DN-1417 (10-5M) or TRH (10-4M) and pentobarbital (5 x 10-4M) led to a partial recovery of the pentobarbital effect. The similar anti-pentobarbital effects were observed with the addition of carbachol (10-4M) or dibutyryl cyclic AMP (10-3M), but not norepinephrine (10-4M) or dopamine (10-4M). DN-1417, TRH, carbachol, norepinephrine or dopamine at 10-4M stimulated cyclic AMP formation in the cerebral cortex slices. Addition of pentobarbital (1 to 7 x 10-4M) inhibited the cyclic AMP formation, in a concentration-dependent manner. DN-1417, TRH or carbachol at 10-4M but not norepinephrine or dopamine at 10-4M significantly reversed the reduction of cyclic AMP formation induced by pentobarbital (5 x 10-4M). Atropine (10-4M) almost completely abolished DN-1417-, TRH- and carbachol-induced cyclic AMP formation in the presence and absence of pentobarbital.     

Effects of epidermal growth factor and calcium omission on cholecystokinin-stimulated Cl- conductance in rat pancreatic zymogen granules.

Evidence suggests that cholecystokinin-octapeptide (CCK-8)-induced activation of a Cl- conductance in the membrane of zymogen granules (ZG) is closely related to pancreatic enzyme secretion. Following stimulation of isolated pancreatic acinar cells with increasing concentrations of CCK-8, the Cl- conductance in the ZG from these acini increased, reached a maximum of 40 +/- 7% above basal Cl- conductance at 10(-12) M CCK-8, and then decreased at CCK-8 concentrations higher than 10(-9) M to a level comparable to the basal Cl- conductance. We had interpreted the inhibitory action of high CCK-8 concentrations to be due to the generation of high concentrations of diacylglycerol and/or its metabolites by an "overstimulation" of phospholipase C at supramaximal CCK-8 concentrations. We now show that EGF abolishes the downstroke of the dose response curve for CCK-8-induced ZG Cl- conductance and shifts the stimulatory response to higher CCK-8 concentrations. Similarly in a nominally "Ca(2+)-free buffer" (free [Ca2+] approximately 0.2 nM), stimulated Cl- conductance at 10(-12) M CCK-8 is nearly abolished and the decreased Cl- conductance at 10(-8) M CCK-8 is increased to the level of maximal stimulation at 10(-12) M CCK-8. We conclude that both EGF and low [Ca2+] affect CCK-8-induced ZG Cl- conductance by decreasing phospholipase C activity.     

Interaction between pentobarbital and GABA-activated ionic channels in rat cerebellar neurons

Using techniques of voltage clamping at the membrane, intracellular perfusion, and concentration clamping, GABA- and barbiturate-activated currents were investigated in single neurons isolated from the rat cerebellum. The dissociation constant for interaction between GABA and GABA receptors was measured at 3±0.8 × 10^–5 M. The presence of pentobarbital in the bathing solution exerts a potentiating effect on GABA-induced conductance in isolated neurons, shifting the dose-response curve for GABA towards lower concentration values without increasing peak chloride conductance. The concentrations at which GABA effects are potentiated range between 10^–6–10^–4 M. High concentrations of pentobarbital inhibit GABA-activated conductance; at concentrations in excess of 5 × 10^–4 M, it also brings about activation of chloride conductance, depressed by bicuculline and picrotoxin, in the absence of GABA. A short-term increase in membrane conductance is produced by rapid pentobarbital washout.A. A. Bogomolets Institute of Physiology, Academy of Sciences of the Ukrainian SSR, Kiev. Translated from Neirofiziologiya, Vol. 22, No. 1, pp. 93–98, January–February, 1990.     

Nitric oxide selectively enhances cAMP levels and electrical coupling between identified RPaD2/VD1 neurons in the CNS of Lymnaea stagnalis (L.)

The isolated CNS of the freshwater mollusc Lymnaea stagnalis was used as a model to study the role of cAMP in NO-mediated mechanisms. The NO donor, DEA/NO (10(-5)-10(-3) M) increased cAMP concentrations in the cerebral, pedal, pleural, parietal and visceral ganglia. In contrast, in the buccal ganglia the same doses of DEA/NO decreased the level of cAMP production. The NOS inhibitor, L-NNA (10(-4) M) increased cAMP concentrations in all areas of the CNS. L-arginine (1 mM), a metabolic precursor of NO, mimicked the action of the NO-donor. The coefficient of electrical coupling between two viscero-parietal peptidergic neurons (VD1/RPaD2) was enhanced by both DEA/NO (10(-4) M) and 8-Br-cAMP (10(-4) M) whereas 8-Br-cGMP (2x10(-4) M) reduced the coupling. We suggest that cAMP-dependent mechanisms are involved in neuronal NO signaling in this simpler nervous system.     

Potassium conductance in straight proximal tubule cells of the mouse. Effect of barium, verapamil and quinidine

The present study has been performed to test for the influence of verapamil and quinidine on the potential difference across the basolateral cell membrane (PDbl) and on the basolateral potassium conductance of isolated perfused segments of the mouse proximal tubule. PDbl was recorded continuously with conventional microelectrodes during rapid alterations of bath or luminal perfusate composition. The contribution of the basolateral potassium conductance to the conductance of both cell membranes (tk) was estimated from the effects of altered bath potassium concentration on PDbl. Under control conditions tk approaches 0.8, i.e. the basolateral cell membrane is mainly conductive to potassium. Neither quinidine nor verapamil affect PDbl at concentrations below 10 mumol/l. At higher concentrations both substances depolarize the basolateral cell membrane mimicking the effect of 1 mmol/l barium. In the presence of 0.1 mmol/l verapamil tk is virtually abolished at 5 to 10 mmol/l bath potassium concentration but is almost unaffected at bath potassium concentrations between 20 and 40 mmol/l. 1 mumol/l ionophore A-23187 does not change the depolarizing effect of 0.1 mmol/l verapamil on cell membrane potential. In the presence of 0.1 mmol/l quinidine, tk is reduced to some 50%, irrespective of the bath potassium concentration. It is concluded that the potassium conductance in straight proximal tubules is inhibited not only by barium but as well by high concentrations of verapamil and quinidine. The effect is probably direct and not related to alterations in the intracellular calcium activity.     

Calcium- and voltage-activated potassium channels in human macrophages.

Single calcium-activated potassium channel currents were recorded in intact and excised membrane patches from cultured human macrophages. Channel conductance was 240 pS in symmetrical 145 mM K+ and 130 pS in 5 mM external K+. Lower conductance current fluctuations (40% of the larger channels) with the same reversal potential as the higher conductance channels were noted in some patches. Ion substitution experiments indicated that the channel is permeable to potassium and relatively impermeable to sodium. The frequency of channel opening increased with depolarization and intracellular calcium concentration. At 10(-7) M (Ca++)i, channel activity was evident only at potentials of +40 mV or more depolarized, while at 10(-5) M, channels were open at all voltages tested (-40 to +60 mV). In intact patches, channels were seen at depolarized patch potentials of +50 mV or greater, indicating that the ionized calcium concentration in the macrophage is probably less than 10(-7) M.     

Reduction of the voltage-dependent calcium current inAplysia neurons by pentobarbital

Effects of pentobarbital on the calcium current of Aplysia neurons were investigated under current- and voltage-clamp conditions using the conventional two-microelectrode technique. Pentobarbital attenuated the progressive broadening of repeated action potentials of somata, suggesting a reduction in the calcium current. When calcium ion was replaced with barium ion in the perfusing solution, in which neither sodium nor potassium ions carried transmembrane currents, the barium current (IBa) which flowed through the calcium channel of the cell membrane was generated by depolarizing pulses of several hundred milliseconds applied every 1 min from a holding potential of -50 mV. The IBa was not affected by tetrodotoxin (30 microM). The current was decreased by pentobarbital (0.1-5 mM) in a dose-dependent manner. The inhibition was much greater at a lower pH of the perfusate, indicating that the uncharged form of the agent was responsible. The voltage-dependent inactivation of the IBa proceeded with two time constants [190 +/- 21 and 2020 +/- 146 msec (N = 4) at -10 mV], both of which were shortened by adding 1 mM pentobarbital [to 120 +/- 18 and 540 +/- 51 msec (N = 4), respectively]. The IBa recovered from the inactivation with two time constants [60 +/- 7 and 871 +/- 76 msec (N = 3) at -50 mV]. The anesthetic (1 mM) prolonged both of them, to 124 +/- 20 and 1480 +/- 172 msec (N = 3), respectively, resulting in a use-dependent depression of the current at 2-Hz stimulation. Pentobarbital reduced the IBa to a greater extent when the holding potential was more positive (-30 instead of -50 mV), indicating a higher affinity of the drug to the inactivated state of the channel. These findings suggest that the attenuation of the progressive broadening of successive spikes by pentobarbital is due to a decrease in the voltage- and time-dependent calcium current, ending in depression of transmitter release from the nerve terminal.     

Electrophysiological properties of cellular and paracellular conductive pathways of the rabbit cortical collecting duct

Summary Microelectrode techniques were applied to the rabbit isolated perfused cortical collecting duct to provide an initial quantitation and characterization of the cell membrane and tight junction conductances. Initial studies demonstrated that the fractional resistance (ratio of the resistance of the apical cell membrane to the sum of the resistances of the apical and basolateral membranes) was usually independent of the point along the tubule of microelectrode impalement—implicating little cell-to-cell coupling—supporting the application of quantitative techniques to the cortical collecting duct. It was demonstrated that in the presence of amiloride, either reduction in the luminal pH or the addition of barium to the perfusate selectively reduced the apical membrane potassium conductance. From the changes inG ^te and fractional resistance upon reducing the luminal pH or addition of barium to the perfusate, the transepithelial, apical membrane, basolateral membrane and tight junction conductances were estimated to be 9.3, 6.7, 8.1 and 6.0 mS cm^–2, respectively. Ninety to ninety-five percent of the apical membrane conductance reflected the barium-sensitive potassium conductance in the presence of amiloride with an estimated potassium permeability of 1.1×10^–4 cm sec^–1. Reduction in the perfusate pH to 4.0 caused a 70% decrease in the apical membrane potassium conductance, implying a blocking site with an acidic group having a pK _a near 4.4. It is concluded that both the transcellular and paracellular pathways of the cortical collecting tubule have high ionic conductances, and that the apical membrane conductance primarily reffects a high potassium conductance. Furthermore, both reduction in the perfusate pH and addition of barium to the perfusate selectively block the apical potassium channels, although the site of inhibition likely differs since the two ions display markedly different voltage-dependent blocks of the channel.     

Conductance increases produced by bath application of cholinergic agonists to Electrophorus electroplaques

When solutions containing agonists are applied to the innervated face of an Electrophorus electroplaque, the membrane's conductance increases. The agonist-induced conductance is increased at more negative membrane potentials. The "instantaneous" current-voltage curve for agonist-induced currents is linear and shows a reversal potential near zero mV; chord conductances, calculated on the basis of this reversal potential, change epsilon-fold for every 62-mV change in potential when the conductance is small. Conductance depends non-linearly on small agonist concentrations; at all potentials, the dose-response curve has a Hill coefficient of 1.45 for decamethonium (Deca) and 1.90 for carbamylcholine (Carb). With agonist concentrations greater than 10(-4) M Carb or 10(-%) M Deca, the conductance rises to a peak 0.5-1.5 min after introduction of agonist, then declines with time; this effect resembles the "desensitization" reported for myoneural junctions. Elapid alpha-toxin, tubocurarine, and desensitization reduce the conductance without changing the effects of potential; the apparent dissociation constant for tubocurarine is 2 X 10(-7) M. By contrast, procaine effects a greater fractional inhibition of the conductance at high negative potentials.     

delta 9-Tetrahydrocannabinol and cannabidiol: dose-dependent effects on evoked potentials in the hippocampal slice

The effects of (-) trans-delta 9-tetrahydrocannabinol (THC) and its metabolite cannabidiol (CBD) were investigated on evoked responses in the CA1 and dentate gyrus regions of the guinea pig transverse hippocampal slice. In both areas orthodromically evoked responses were enhanced by 10(-7) M THC, while 10(-6) M THC caused depression. Antidromic responses were not significantly affected. Antidromically-evoked inhibition in the CA1 region was decreased at low doses and unaffected at higher doses, while the facilitation by orthodromic interaction was unaffected at both dose ranges. The early part of the orthodromic field potential corresponding to the excitatory postsynaptic potential (EPSP) was enhanced at 10(-7) M in both areas. CBD (10(-6) M) decreased facilitation in CA1, and caused delayed excitation in the dentate granule layer. This study supports the conclusion that the biphasic effects of THC are dose dependent.     

Effects of a New Calcium Channel Blocker, KB-2796, on Protein Synthesis of the CA1 Pyramidal Cell and Delayed Neuronal Death Following Transient Forebrain Ischemia

The effects of a new calcium channel blocker, 1-[bis(4-fluorophenyl)methyl]-4-(2,3,4-trimethoxybenzyl)-piperazine dihydrochloride (KB-2796), on delayed neuronal death (DND) in the hippocampus were examined in gerbils in comparison with those of pentobarbital and flunarizine. The neuronal density in the hippocampal CA1 subfield was counted on the seventh day of recirculation following 5 min of bilateral carotid occlusion, and protein biosynthesis in the brain was also determined at 1, 2, 4, 24, and 72 h following occlusion. The drugs were intraperitoneally administered after recirculation. KB-2796 (10 mg/kg) significantly prevented DND in the CA1 subfield. Pentobarbital (40 mg/kg), but not flunarizine (3 and 10 mg/kg), inhibited DND. Protein synthetic activity in the CA1 subfield was reduced by ischemia and the reduction was not restored even at 72 h after recirculation. KB-2796 did not ameliorate the reduction of protein synthesis in the CA1 subfield by 24 h after recirculation, but in one of three animals restoration of protein synthesis was observed at 72 h of recirculation. Pentobarbital also restored the reduced protein synthesis in two of three animals at 72 h. These results suggest that calcium influx into neurons participates in the pathogenesis of DND, and also that KB-2796 might prevent both morphological and functional cell damage in CA1 neurons induced by transient ischemia.     

Carbonic anhydrase activity of rabbit lungs

Pulmonary carbonic anhydrase (CA) activity was studied in rabbit lungs perfused with solutions containing no CA. Measurements were made of the amount of 14CO2 appearing in the expired gas following injections of H14CO3(-), 14CO2, or a 20:1 mixture of each into the pulmonary artery. The fraction of the injected label in the expired gas was only 17% greater for 14CO2 than for the mixture, suggesting that equilibration between H14CO3(-) and 14CO2 was nearly complete during the capillary transit time. Inhibition of pulmonary CA decreased excretion of H14CO3(-) and the mixture by 40 and 49% and increased the excretion of 14CO2 by 96%. Addition of CA to the perfusate had no effect. Thus, CO2 exchange is not significantly limited by pulmonary CA if inhibitors are absent. Tissue binding of [3H]acetazolamide injected into the pulmonary artery was diminished by 50% when acetazolamide concentrations reached 0.13 x 10(-6) M. Each liter of extravascular lung water contained 1.25 x 10(-6) mol of receptors for acetazolamide that were accessible to plasma during a single circulation. Binding of [3H]acetazolamide was also observed in lungs of anesthetized rabbits, suggesting that pulmonary CA is accessible to plasma in vivo as well as in situ.     

Delayed Neuronal Death After Brief Histotoxic Hypoxia In Vitro

Abstract: The effect of three metabolic inhibitors—iodoacetate, potassium cyanide, and potassium arsenate—on neuronal viability was studied in primary rat cortical and hippocampal CA1 neuronal cultures. Iodoacetate (0.1 m M ) applied for 5 min to 8-day-old cultures resulted in delayed neuronal death within 3–24 h in cortical and hippocampal CA1 neurons. Neuronal degeneration was preceded by transient inhibition of energy metabolism to ∼40% and a permanent inhibition of protein synthesis to ∼50%. The inhibition of protein synthesis and the neuronal death were prevented by the free radical scavenger vitamin E but not by the glutamate antagonist MK-801. Removal of calcium during iodoacetate exposure could not protect against toxicity, and there was no increase of intracellular calcium concentration during and shortly after iodoacetate treatment. Cyanide and arsenate produced only partial neuronal degeneration, even at a dose of 10 m M . These observations demonstrate that brief exposure of neurons to low concentrations of iodoacetate produces a delayed type of neuronal death that is not mediated by either calcium or glutamate. The therapeutic effect of vitamin E points to a free-radical mediated injury and suggests that this type of pathology may also be involved in delayed neuronal death after transient energy depletion in vivo.     

New evidence for active sodium transport from fluid-filled rat lungs

The hypothesis that fluid reabsorption from the air spaces is mediated at least in part by active transport of Na+ was investigated in six sets of experiments conducted in isolated fluid-filled rat lungs. Fluid reabsorption was monitored by following the changes in the air space concentration of labeled albumin. We found that incorporation of bicarbonate rather than a nonvolatile buffer (N-2-hydroxy-ethylpiperazine-N'-2-ethanesulfonic acid) in the air space solution more than doubled the rate of fluid reabsorption. Addition of 10(-4) M amiloride to the air space solution reduced the rate of fluid reabsorption over a 2-h experiment from 1.2 +/- 0.1 to 0.7 +/- 0.1 ml and decreased reabsorption of both labeled and unlabeled Na+ from the air spaces. To show that Na+ could be reabsorbed from the air spaces even if the concentrations of Na+ in the perfusate increased above those in the air space, mannitol (150 mM) was added to the perfusate and air space solutions and the concentrations of Na+ and Cl- were reduced to 90 and 60 mM, respectively. Mannitol diffuses across the pulmonary epithelium very slowly, and it osmotically restrained the movement of water out of the air spaces. Na+ concentrations in the perfusate increased by 10 +/- 2 mM, but concentrations in the air space remained unchanged. Despite an increasingly unfavorable concentration gradient for Na+, 0.2 mmol Na+ and 0.6 ml water were reabsorbed from the air spaces in 2 h. Ouabain (10(-4) M) did not appear to slow fluid reabsorption in the presence of mannitol, but it reduced K+ secretion into the air spaces and increased K+ appearance in the perfusate in a manner consistent with inhibition of Na+-K+-adenosinetriphosphatase at the basolateral surface of the epithelial cells. Fluid reabsorption was not altered when the lungs were exposed to a hypotonic solution (185 mM), but secretion of K+ into the air spaces was accelerated and K+ was lost from the perfusate. These experiments are consistent with active Na+ transport from the air spaces.