Potassium transport across rat alveolar epithelium: evidence for an apical Na+-K+ pump

Alveolar epithelial transport in the adult lung

The alveolar surface comprises >99% of the internal surface area of the lungs. At birth, the fetal lung rapidly converts from a state of net fluid secretion, which is necessary for normal fetal lung development, to a state in which there is a minimal amount of alveolar liquid. The alveolar surface epithelium facing the air compartment is composed of TI and TII cells. The morphometric characteristics of both cell types are fairly constant over a range of mammalian species varying in body weight by a factor of approximately 50,000. From the conservation of size and shape across species, one may infer that both TI and TII cells also have important conserved functions. The regulation of alveolar ion and liquid transport has been extensively investigated using a variety of experimental models, including whole animal, isolated lung, isolated cell, and cultured cell model systems, each with their inherent strengths and weaknesses. The results obtained with different model systems and a variety of different species point to both interesting parallels and some surprising differences. Sometimes it has been difficult to reconcile results obtained with different model systems. In this section, the primary focus will be on aspects of alveolar ion and liquid transport under normal physiologic conditions, emphasizing newer data and describing evolving paradigms of lung ion and fluid transport. We will highlight some of the unanswered questions, outline the similarities and differences in results obtained with different model systems, and describe some of the complex and interweaving regulatory networks.

Spectrum of ion channels in alveolar epithelial cells: implications for alveolar fluid balance

The efficient transition from placental to atmospheric delivery of oxygen at birth is critically dependent on rapid reabsorption of fetal lung fluid. In the perinatal period, this process is driven by active transepithelial sodium transport and is almost exclusively dependent on expression and modulation of the amiloride-sensitive epithelial sodium channel (ENaC). However, later in development, the amiloride sensitivity of the reabsorptive response, which must be sustained to keep the lungs effectively dry, wanes as a function of postnatal age. This Featured Topic (Experimental Biology Meeting, Washington, DC, April, 2004) presented exciting new evidence to demonstrate that, in addition to ENaC, the adult alveolar epithelium expresses a plethora of amiloride-insensitive ion channels, including cystic fibrosis transmembrane conductance regulator, proton channels, voltage-dependent potassium channels, and cyclic nucleotide-gated cation channels. Furthermore, important evidence for selective modulation of ENaC subunits in the lung in response to cardiovascular disease was demonstrated. Finally, it is clear that newly emerging models of human alveolar epithelium in combination with the novel lung slice electrophysiological preparation will ensure that the ascription of function to specific ion channels in the in situ human lung will soon be a real possibility.

Resistance of the pulmonary epithelium to movement of buffer ions

Exposure of the apical surfaces of alveolar monolayers to acidic and alkaline solutions has been reported to have little influence on intracellular pH compared with basolateral challenges (Joseph D, Tirmizi O, Zhang X, Crandall ED, and Lubman RL. Am J Physiol Lung Cell Mol Physiol 282: L675-L683, 2002). We have used fluorescent pH indicators and a trifurcated optical bundle to determine whether the apical surfaces are less permeable to ionized buffers than the membranes that separate the vasculature from the tissues in intact rat lungs. In the first set of experiments, the air spaces were filled with perfusate containing FITC-dextran (mol wt 60000) or 2',7'-bis(2-carboxyethyl)-5(6)-carboxyfluorescein (BCECF). Air space pH fell progressively from 7.4 to 6.61 +/- 0.03 (mean +/- SE, n = 11, air space buffers at 10 mM). Perfusion for 2 min with 2 mM NH4Cl increased air space pH by 0.142 +/- 0.019 unit, without a subsequent acidic overshoot. Infusions of NaHCO3 and sodium acetate reduced pH without a subsequent alkaline overshoot. In the second set of experiments, cellular pH was monitored in air-filled lungs after perfusion with BCECFAM. Injections of NH4Cl caused a biphasic response, with initial alkalinization of the cellular compartment followed by acidification after the NH4Cl was washed from the lungs. Subsequent return of pH to normal was slowed by infusions of 1.0 mM dimethyl amiloride. These studies suggest that lung cells are protected from air space acidification by the impermeability of the apical membranes to buffer ions and that the cells extrude excess H+ through basolateral Na+/H+ exchangers.

Adult alveolar epithelial cells express multiple subtypes of voltage-gated K+ channels that are located in apical membrane

Whole cell perforated patch-clamp experiments were performed with adult rat alveolar epithelial cells. The holding potential was -60 mV, and depolarizing voltage steps activated voltage-gated K(+) (Kv) channels. The voltage-activated currents exhibited a mean reversal potential of -32 mV. Complete activation was achieved at -10 mV. The currents exhibited slow inactivation, with significant variability in the time course between cells. Tail current analysis revealed cell-to-cell variability in K(+) selectivity, suggesting contributions of multiple Kv alpha-subunits to the whole cell current. The Kv channels also displayed steady-state inactivation when the membrane potential was held at depolarized voltages with a window current between -30 and 5 mV. Analysis of RNA isolated from these cells by RT-PCR revealed the presence of eight Kv alpha-subunits (Kv1.1, Kv1.3, Kv1.4, Kv2.2, Kv4.1, Kv4.2, Kv4.3, and Kv9.3), three beta-subunits (Kvbeta1.1, Kvbeta2.1, and Kvbeta3.1), and two K(+) channel interacting protein (KChIP) isoforms (KChIP2 and KChIP3). Western blot analysis with available Kv alpha-subunit antibodies (Kv1.1, Kv1.3, Kv1.4, Kv4.2, and Kv4.3) showed labeling of 50-kDa proteins from alveolar epithelial cells grown in monolayer culture. Immunocytochemical analysis of cells from monolayers showed that Kv1.1, Kv1.3, Kv1.4, Kv4.2, and Kv4.3 were localized to the apical membrane. We conclude that expression of multiple Kv alpha-, beta-, and KChIP subunits explains the variability in inactivation gating and K(+) selectivity observed between cells and that Kv channels in the apical membrane may contribute to basal K(+) secretion across the alveolar epithelium.

Alveolar epithelial type I cells contain transport proteins and transport sodium, supporting an active role for type I cells in regulation of lung liquid homeostasis

Transport of lung liquid is essential for both normal pulmonary physiologic processes and for resolution of pathologic processes. The large internal surface area of the lung is lined by alveolar epithelial type I (TI) and type II (TII) cells; TI cells line >95% of this surface, TII cells <5%. Fluid transport is regulated by ion transport, with water movement following passively. Current concepts are that TII cells are the main sites of ion transport in the lung. TI cells have been thought to provide only passive barrier, rather than active, functions. Because TI cells line most of the internal surface area of the lung, we hypothesized that TI cells could be important in the regulation of lung liquid homeostasis. We measured both Na(+) and K(+) (Rb(+)) transport in TI cells isolated from adult rat lungs and compared the results to those of concomitant experiments with isolated TII cells. TI cells take up Na(+) in an amiloride-inhibitable fashion, suggesting the presence of Na(+) channels; TI cell Na(+) uptake, per microgram of protein, is approximately 2.5 times that of TII cells. Rb(+) uptake in TI cells was approximately 3 times that in TII cells and was inhibited by 10(-4) M ouabain, the latter observation suggesting that TI cells exhibit Na(+)-, K(+)-ATPase activity. By immunocytochemical methods, TI cells contain all three subunits (alpha, beta, and gamma) of the epithelial sodium channel ENaC and two subunits of Na(+)-, K(+)-ATPase. By Western blot analysis, TI cells contain approximately 3 times the amount of alphaENaC/microg protein of TII cells. Taken together, these studies demonstrate that TI cells not only contain molecular machinery necessary for active ion transport, but also transport ions. These results modify some basic concepts about lung liquid transport, suggesting that TI cells may contribute significantly in maintaining alveolar fluid balance and in resolving airspace edema.

Effects of SNP, ouabain, and amiloride on electrical potential profile of isolated sheep pleura

The fluid and solute transport properties of pleural tissue were studied by using specimens of intact visceral and parietal pleura from adult sheep lungs. The samples were transferred to the laboratory in a Krebs-Ringer solution at 4 degrees C within 1 h from the death of the animal. The pleura was then mounted as a planar sheet in a Ussing-type chamber. The results that are presented in this study are the means of six different experiments. The spontaneous potential difference and the inhibitory effects of sodium nitroprusside (SNP), ouabain, and amiloride on transepithelial electrical resistance (R(TE)) were measured. The spontaneous potential difference across parietal pleura was 0.5 +/- 0.1 mV, whereas that across visceral pleura was 0.4 +/- 0.1 mV. R(TE) of both pleura was very low: 22.02 +/- 4.1 Omega. cm2 for visceral pleura and 22.02 +/- 3.5 Omega. cm2 for parietal pleura. There was an increase in the R(TE) when SNP was added to the serosal bathing solution of parietal pleura and to the serosal or mucosal bathing solution in visceral pleura. The same was observed when ouabain was added to the mucosal surface of visceral pleura and to either the mucosal or serosal surface of parietal pleura. Furthermore, there was an increase in R(TE) when amiloride was added to the serosal bathing solution of parietal pleura. Consequently, the sheep pleura appears to play a role in the fluid and solute transport between the pleural capillaries and the pleural space. There results suggest that there is a Na+ and K+ transport across both the visceral and parietal pleura.

A large conductance, Ca2+-activated K+ channel in a human lung epithelial cell line (A549)

A large conductance, Ca2+-activated K+ channel in a human lung epithelial cell line (A549) was identified using the single channel patch clamp technique. Channel conductance was 242 +/- 33 pS (n = 67) in symmetrical KCl (140 mM). The channel was activated by membrane depolarization and increased cytosolic Ca2+. High selectivity was observed for K+ over Rb+(0.49) > Cs+(0.14) > Na+(0.09). Open probability was significantly decreased by Ba2+ (5 mM) and quinidine (5 mM) to either surface, but TEA (5 mM) was only effective when added to the external surface. All effects were reversible. Increasing cytosolic Ca2+ concentration from 10(-7) to 10(-6) M caused an increase in open probability from near zero to fully activated. ATP decreased open probability at approximately 2 mM, but the effect was variable. The channel was almost always observed together with a smaller conductance channel, although they could both be seen individually. We conclude that A549 cells contain large conductance Ca2+-activated K+ channels which could explain a major fraction of the K+ conductance in human alveolar epithelial membranes.

Ozone toxicity in the rat. III. Effect of changes in ambient temperature on pulmonary parameters

Pulmonary toxicity of ozone (O3) was examined in adult male Fischer 344 rats exposed to 0.5 parts/million O3 for either 6 or 23 h/day over 5 days while maintained at an ambient temperature (Ta) of either 10, 22, or 34 degrees C. Toxicity was evaluated by using changes in lung volumes and the concentrations of constituents of bronchoalveolar lavage fluid that signal lung injury and/or inflammation. Results indicated that toxicity increased as Ta decreased. Exposures conducted at 10 degrees C were associated with the greatest decreases in body weight and total lung capacity and the greatest increases in lavageable protein, lysozyme, alkaline phosphatase activity, and percent neutrophils. O3 effects not modified by Ta included increases in residual volume and lavageable potassium, glucose, urea, and ascorbic acid with exposure at 34 degrees C. Most effects were attenuated during the 5 exposure days and/or returned to normal levels after 7 air recovery days, regardless of prior O3 exposure or Ta. It is possible that Ta-induced changes in metabolic rate may have altered ventilation and, therefore, the O3 doses among rats exposed at the three different Ta levels.

Identification and properties of pathways for K+ transport in guinea-pig and rat alveolar epithelial type II cells

86Rb+ was used to study potassium uptake and efflux in type II pneumocytes freshly isolated from adult guinea-pig and rat lung. Both species exhibited a substantial ouabain-sensitive component of potassium influx. In rats, most of the ouabain-resistant influx was abolished by bumetanide and removal of extracellular chloride elicited no further effect. In contrast, only a proportion of the ouabain-insensitive uptake was inhibitable by bumetanide in guinea-pigs and this species showed an additional component of influx, which was chloride dependent and which was reduced by either the K(+)-H(+)-ATPase inhibitor, omeprazole, or by the stilbene derivative, 4,4'-diisothiocyanostilbene-2,2'-disulphonate (DIDS). The chloride-dependent component was also apparent in efflux experiments in guinea-pigs, but was absent in rats. Ouabain-insensitive ATPase activity was assayed in highly purified apical membranes from guinea-pig type II pneumocytes. This activity was inhibitable by omeprazole (apparent inhibition constant, Ki, was approximately 40 microM), was potassium dependent (apparent activation constant, Ka, was approximately 200 microM) and was doubled by the addition of nigericin. While potassium transport in rat type II cells is adequately accounted for by Na(+)-K(+)-ATPase and Na(+)-K(+)-2Cl- cotransport, our data suggest the additional presence of K(+)-Cl- cotransport and K(+)-H(+)-ATPase in guinea-pig type II pneumocytes. A model of how alveolar subphase acidification may occur is proposed.

Tripeptide transport in rat lung

Transport of L-alanyl-D-phenylalanyl-L-alanine was investigated with an in situ vascular perfusion preparation of rat lung and brush border membrane vesicles prepared from type II pneumocytes. In the perfused lung 1 mM tripeptide was transported intact from the alveolar lumen to the vascular perfusate at a mean rate of 25.1 +/- 1.29 (3) nmol/min per g dry weight. D-Phenylalanine also appeared in the vascular perfusate at a rate of 21.9 +/- 1.74 (3) nmol/min per g dry weight indicating that 47% of the absorbed tripeptide was split during passage across the epithelial layer. No dipeptide could be detected in the vascular effluent during perfusions with tripeptide. Rapid L-alanyl-D-phenylalanyl-L-alanine uptake occurred with fresh apical membrane vesicles prepared from type II pneumocytes and this was abolished by treatment with 0.1% triton. The related tripeptide, D-alanyl-L-phenylalanyl-D-alanine, was taken up significantly more slowly by the vesicles. D-phenylalanyl-L-alanine and D-phenylalanyl-D-alanine, were also studied with the vascularly perfused preparation; the mixed dipeptide appeared in the vascular perfusate significantly faster than L-alanyl-D-phenylalanyl-L-alanine whereas D-phenylalanyl-D-alanine appeared more slowly and was not hydrolysed.

Mechanisms and regulation of ion transport in adult mammalian alveolar type II pneumocytes

The adult alveolar epithelium consists of type I and type II (ATII) pneumocytes that form a tight barrier, which severely restricts the entry of lipid-insoluble molecules from the interstitial to the alveolar space. Current in vivo and in vitro evidence indicates that the alveolar epithelium is also an absorptive epithelium, capable of transporting Na+ from the alveolar lumen, which is bathed by a small amount of epithelial lining fluid, to the interstitial space. The in situ localization of Na(+)-K(+)-ATPase activity in ATII cells and the fact that these cells are involved in a number of crucial functions, such as surfactant secretion and alveolar remodeling after injury, led investigators to examine their transport characteristics. Radioactive flux studies, in both freshly isolated and cultured cells, and bioelectric measurements in ATII cells grown on porous supports indicate that they transport Na+ according to the Koefoed-Johnsen and Ussing model of epithelial transport. Na+ enters the apical membrane, because of the favorable electrochemical gradient, through Na+ cotransporters, a Na(+)-H+ antiport, and cation channels and is pumped across the basolateral membrane by a ouabain-sensitive Na(+)-K+ pump. Na+ transport is enhanced by substances that increase intracellular adenosine 3',5'-cyclic monophosphate. In addition to Na+ transporters, ATII cells contain several transporters that regulate their intracellular pH, including a H(+)-ATPase, which may explain the low pH of the epithelial lining fluid. The absorptive properties of ATII cells may play an important role in regulating the degree of alveolar fluid in health and disease.

Electrolyte transport across the pleura of rabbits

The amounts of Na+ and Cl- in the right pleural space of anesthetized rabbits were determined 10 and 60 min after a 2 ml hydrothorax with the following solutions: Ringer, Ringer with an inhibitor of the Na(+)-Cl- coupled transport or of the Na+/K+ pump, Ringer with gluconate instead of Cl- or with methylglucamine instead of Na+. During the 10-60 min period: (a) with Ringer Na+ and Cl- decreased (P less than 0.01) along with an iso-osmotic liquid absorption, (b) with disulfonic-stilbene (0.1 mM), amiloride (0.7 mM), acetazolamide (0.1 mM), or ouabain (0.5 mM) Na+ did not change and Cl- decreased less (P less than 0.01) than with Ringer. With gluconate-Ringer or methylglucamine-Ringer the liquid flow reversed: in the former case Cl- and, to a smaller extent, Na+ increased (P less than 0.01); in the latter only Na+ increased (P less than 0.01). These findings suggest: (1) the occurrence of a Na+/H+ and Cl-/HCO3- double exchange on the serosal side and of a Na+/K+ pump on the interstitial side of the pleural mesothelium; (2) a slow efflux from the pleural space of gluconate or methylglucamine relative to the corresponding influx of Cl- or Na+, respectively; this drags liquid into the space by osmotic gradient.

Development of the lung liquid reabsorptive mechanism in fetal sheep: synergism of triiodothyronine and hydrocortisone

1. Thyroidectomy was performed on twelve fetal sheep between 111 and 115 days gestation. Measurement of fetal lung liquid secretion and absorption rates (Jv) were made at rest and during short (45 min) and long (5 h) infusions of adrenaline (0.5 micrograms/min) in a total of thirty-seven experiments, some in the absence of triiodothyronine (T3) and hydrocortisone and some at set times after the administration of the two hormones. 2. T3 was given either as an I.V. infusion (60 micrograms/24 h) or as a bolus of 30 micrograms; hydrocortisone was given as an infusion of 10 mg/24 h. Both hormones were administered together. 3. Before T3 and hydrocortisone were given short infusions of adrenaline had no effect on Jv but 4 h after exposure to the hormones secretion rate was reduced to near zero (Jv = -0.5 +/- 1.6 ml/h, n = 4) by adrenaline; after 24 h of hormone exposure, absorption of fetal lung liquid was produced by adrenaline (Jv = -3.6 +/- 2.2 ml/h, n = 4) which was even greater after 72 h, (Jv = -11.2 +/- 2.2 ml/h, n = 4). 4. During long infusions of adrenaline when T3 and hydrocortisone were given at the start of the experiment, an effect on lung liquid secretion was evident at 2 h and absorption was produced at 4 h (Jv = -4.2 +/- 2.5 ml/h, n = 3). The effect was significantly different from control long infusions of adrenaline performed the previous day in the absence of hormones. 5. After 24 or 48 h of stopping T3 and hydrocortisone administration, adrenaline no longer produced absorption of lung liquid, indicating that the effect of the two hormones was reversible within 24-48 h. 6. The protein synthesis inhibitor cycloheximide put into lung liquid (4 x 10(-5) to 3 x 10(-4) M) blunted the effect of the hormones at 4 h and prevented absorption of lung liquid at 24 h. Jv during adrenaline was -3.6 +/- 1.5 ml/h in control experiments but was +3.3 +/- 0.9 ml/h after cycloheximide, n = 4, P < 0.01. This indicated that the two hormones produced their effect through protein synthesis.

Whole-cell K+ currents in type II pneumocytes freshly isolated from rat lung: pharmacological evidence for two subpopulations of cells

The patch clamp technique was used to record whole cell K+ currents in type II pneumocytes freshly isolated from adult rats. Depolarizing voltage steps evoked outward K+ currents which were distinguished into low and high threshold types, only one type being apparent in any one cell. Low-threshold (LT) currents were activated at test potentials of -40 mV to -20 mV and were reduced in amplitude by 20 mM tetraethylammonium (TEA). High-threshold (HT) currents were activated only at test potentials positive to -20 mV and current noise was always greater than for LT currents. HT currents were also significantly more sensitive than were LT currents to block by TEA. Quinine (1 mM) blocked LT currents reversibly at all activating test potentials. HT currents were also reversibly blocked by 1 mM quinine, but in a voltage-dependent manner, the degree of block increasing with increasing test potential. 4-Aminopyridine (2 mM) further distinguished the two current types: it was virtually without effect on HT currents but caused large reductions in LT current amplitudes, apparently by acting on the open channels underlying this current. These data clearly distinguish type II pneumocytes into two subpopulations and suggest that they may play separate roles in the functioning of the intact alveolar epithelium.

Cellular effects of beta-adrenergic and of cAMP stimulation on potassium transport in rat alveolar epithelium

Alveolar fluid absorption is greatly enhanced by cAMP and by beta-adrenergic agonists via an increase in Na+ transport. Little is known about K+ homeostasis under these circumstances. We studied K+ transport across alveolar epithelium in isolated perfused rat lungs stimulated either by dibutyryl-cAMP or isoproterenol. K+ fluxes and the apparent permeability of 86Rb across the epithelium (alveoli to plasma) were interpreted according to a model involving two types of cells, B and L, distinguished by the location of Na+-K+-ATPases (basal and luminal). Water is considered to be absorbed by B cells in a solute-coupled process energized by a basolateral Na+-K+-ATPase that is stimulated by isoproterenol and cAMP. K+ transport out of the alveoli is due to the activity of a Na+-K+-ATPase located in the apical membrane of L cells. In the present study net transport rate of K+ was -0.5 +/- 0.15 nmol/s, n = 20 (out of alveoli) in control conditions. When the epithelium was stimulated by dibutyryl-cAMP (10(-4) mol/l) net absorption of K+ reversed to net 'secretion' into alveoli (3.2 +/- 0.31 nmol/s), fluid absorption was not stimulated. K+ 'secretion' was abolished by apical Ba2+, indicating it was due to opening of apical K+ channels. Basolateral ouabain reversed net K+ 'secretion' to net absorption indicating that K+ entry into alveoli was dependent on activity of B cell basolateral Na+-K+-ATPase (masking simultaneous K+ removal by apical L cell Na+-K+-pump). When larger concentrations of dibutyryl-cAMP (10(-3) mol/l) or when isoproterenol were used to stimulate the epithelium there was a tripling of fluid absorption.(ABSTRACT TRUNCATED AT 250 WORDS)