Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Shortly after birth, active re-absorption of sodium (Na⁺) across the alveolar epithelium creates an osmotic gradient favoring the re-absorption of fetal lung fluid. Na⁺ ions enter the apical membranes of alveolar epithelial cells through amiloride-sensitive Na⁺ channels (ENaC) and are extruded across the basolateral membranes by the ouabain-sensitive Na,K-ATPase (1). The importance of active Na⁺ transport in fetal fluid re-absorption was clearly elucidated by the work of Hummler and coworkers (2) who showed that newborn mice lacking the alpha (pore-forming) subunit of the amiloride-sensitive channel (ENaC) failed to clear their lung fluid and died within 48 h from respiratory failure. Additional studies have shown that active Na⁺ transport plays an important part in decreasing alveolar fluid, thus optimizing gas exchange, in both the neonatal and adult lung (3).

However, in a number of pathologic situations, active alveolar epithelial Na⁺ transport may be compromised. During lung inflammation, reactive oxygen and nitrogen species, generated by epithelial and inflammatory cells (4) may damage apical and basolateral Na⁺ transporters. In addition, volutrauma during mechanical ventilation (7) was also found to decrease lung liquid clearance (8). Both oxygen toxicity and volutrauma contribute to the pathogenesis of respiratory distress syndrome and bronchopulmonary dysplasia.

For these reasons, there have been several attempts to enhance Na⁺ re-absorption across the newborn and adult alveolar epithelium. For example, mineralocorticoids and glucocorticoids have been shown to increase electrogenic Na⁺ absorption and fluid clearance by transcriptional regulation of both ENaC and Na,K-ATPase genes (9, 10). The improvement of lung function in premature infants by prenatal administration of glucocorticoids may in part be due to increased pulmonary Na⁺ absorption (11, 12). Heterologous gene transfer may be another approach to increase levels of Na⁺ transporters and thus increase alveolar fluid clearance. Indeed, Factor and colleagues (13) recently demonstrated that adenovirus-mediated transfer and overexpression of a gene encoding for the ₁ subunit of the Na,K-ATPase increases Na,K-ATPase activity in alveolar epithelial type II (ATII) cells in suspension, as measured by ⁸⁶Rb uptake, and clearance of intratracheally instilled isotonic fluid in isolated adult rat lungs. Furthermore, transfecting confluent monolayers of airway cells with genes encoding for a ₁ subunit of the Na,K-ATPase resulted in a 1 µA/cm² increase of amiloride-sensitive and ouabain-sensitive short circuit currents (I_SC), but changes in total current of intact monolayers were not reported (13). Similar studies have not yet been performed in monolayers of either adult or fetal ATII cells, the cells thought to be responsible for lung liquid clearance. The demonstration of increased vectorial transport across confluent monolayers of ATII cells would be important to ensure that the additional Na,K-ATPase subunits are properly assembled into functional proteins and integrated into the basolateral membrane. Presently, it remains an open question whether such gene transfer can actually increase ion transport across monolayers of ATII cells.

Herein we isolated distal epithelial cells from the lungs of fetal rats and transfected them with an adenoviral gene transfer vector containing either the ₁ or ₁ subunit of rat Na,K-ATPase. Once they formed confluent monolayers, we mounted them in Ussing chambers and measured their ability to vectorially transport Na⁺ ions under short circuit conditions. We then selectively permeabilized either the apical or basolateral membranes of the monolayers and measured the effects of the transfection on the ability of the apical and basolateral pathways to transport Na⁺ ions. Finally, we investigated whether transfection alleviated the impairment of Na⁺ transport seen after exposure of fetal distal lung epithelial (FDLE) monolayers to hydrogen peroxide.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

FDLE Cell Isolation

The isolation procedure has been described previously (14). Briefly, the lungs of 19- to 20-d gestation fetal rats (term = 22 d) were digested in a solution containing 0.125% trypsin and 0.4 mg/ml DNAse in Eagle's minimum essential medium (MEM) for 10 min. Digestion was stopped by the addition of MEM containing 10% fetal bovine serum (FBS). Cells were collected by centrifugation and resuspended in 15 ml MEM containing 0.1% collagenase and DNAse. This solution was incubated for 15 min at 37°C. The collagenase activity was neutralized by the addition of 15 ml MEM containing 10% FBS. The cells were plated twice for 1.5 h to remove contaminating fibroblasts. The supernatant contained epithelial cells with > 95% purity (14). Cells were counted and seeded on permeable Transwell culture inserts (Corning Inc., Corning, NY) with 0.33 cm² surface area and 0.4 µm pore size. Cells were seeded at a density of 5 × 10⁴ cells/filter and cultured in Dulbecco's minimal essential medium with 10% FBS and 1% penicillin/streptomycin.

Adenovirus Construction and Transfection

Generation of the adenoviral vectors used in this study has been described previously (13). Briefly, shuttle vectors containing a cytomegalovirus (CMV) immediate early promoter driven expression cassette with rat complementary DNAs (cDNAs) for the ₁ and ₁ Na,K-ATPase subunits or -galactosidase (-gal) were cotransfected with a plasmid that contains a human type 5 adenovirus genome (pJM17) into HEK-293 cells (American Type Culture Collection, Rockville, MD), resulting in ad₁, ad₁, and ad-gal, respectively. Cells from plates showing cytopathologic effect (CPE) were collected and disrupted by six cycles of freezing and thawing. This crude viral lysate was expanded in HEK-293 cells. After repeat development of a CPE, polymerase chain reaction (PCR) was used to confirm the presence of the CMV promoter and appropriate cDNAs. PCR-positive cultures were plaque-purified three times in HEK-293 cells before large-scale amplification and purification through serial CsCl density gradient centrifugations. The resultant virus was dialyzed against 10 mM Tris-HCl pH 7.4/1 mM MgCl/10% glycerol to remove CsCl before storage in 10% glycerol at 70°C. Viral titers were ascertained by the enumeration of plaques produced in HEK-293 cells grown under agarose. Viral purity was assessed by testing for E1a and E1b genes by PCR and plaque production in A549 cells grown under agarose. All virus preparations used in the studies reported herein were free of signs of wild-type adenovirus and endotoxin (Endosafe; Charles River Industries, Charleston, SC).

Transfection of FDLE cells was carried out 24 h after seeding. At that time, measurements of electrical transepithelial resistance indicated that confluent monolayers had not yet been formed. The culture medium was aspirated and replaced with fresh solution. Thereafter, viruses at nominal concentrations (based on number of seeded cells) of 5, 10, 50, and 100 plaque-forming units (pfu) suspended in the same culture medium were added to both the apical and the basolateral sides. Control filters were transfected with either null vectors (ad0) or viruses carrying a -gal gene (ad-gal). The actual ratio of viruses per cell was estimated from the cell number determinations by DNA measurement in 24-h cultures (see subsequent text).

Transfection efficiency was estimated by infecting FDLE cells with various plaque-forming units of ad-gal 24 h after seeding on flat-bottom tissue culture plates (Falcon 3846; Becton Dickinson, Franklin Lakes, NJ). Two days later, the plates were placed on ice and the cells were washed twice with phosphate-buffered saline (PBS) at 0°C. Thereafter, the cells were fixed with 0.25% glutaraldehyde in PBS for 10 min at 4°C, rinsed again twice with PBS, and stained overnight at 37°C with a solution containing 2.5 mM 5-bromo-4-chloro-3-indolyl--D-galactopyranoside (Boehringer Mannheim Corp., Indianapolis, IN), 3.0 mM K₃Fe(CN)₆, 3.0 mM K₄Fe(CN)₆, 80 mM Na₂HPO₄, 20 mM NaH₂PO₄, and 1.3 mM MgCl₂. A total number of at least 400 cells per plate in at least four randomly selected visual fields at ×320 magnification was counted, and the proportion of stained cells was determined.

DNA Measurements and Cell Number Determination

Because not all seeded cells may attach and others may divide in culture, the number of cells at the time of transfection was determined as follows: cells in the seeding suspension were counted with a standard hemocytometer. They were then centrifuged and resuspended in PBS, lysed by sonication, and incubated with H33258 dye (Hoechst Marion Roussell, Kansas City, MO). Fluorescence was measured at _ex 356 nm and _em 458nm with a luminescence spectrometer (model no. LS 50B; Perkin Elmer, Norwalk, CT). The DNA content of the samples was calculated by comparison to a set of DNA standards, and the amount of DNA per cell was determined by dividing total DNA by the cell number. Subsequently, in a number of filters, the culture medium was replaced by PBS 24 h after seeding, the cells were disrupted on the filter by sonication, the amount of DNA was measured, and the total cell number was calculated as described previously.

Western Blot Analysis

Membrane proteins were obtained by homogenizing FDLE cells in 300 mM mannitol, 10 mM N-2-hydroxyethylpiperazine-N'-ethane sulfonic acid (Hepes)-Tris (pH 7.4) with 3 mM ethyleneglycol-bis-(-aminoethyl ether)-N,N'-tetraacetic acid/1 mM ethylenediaminetetraacetic acid and protease inhibitors (0.01 mg/ml N-tosyl-L-phenylalanine chlorylmethyl ketone, 0.1 mM phenylmethylsulfonyl fluoride, 0.01 mg/ml leupeptin; all from Sigma, St. Louis, MO). Cell debris was removed by centrifugation at 10,000 × g for 20 min at 4°C. The resultant supernatant was centrifuged at 100,000 × g at 4°C. The pellet was resuspended in 100 µl of homogenization buffer and protein content was quantified (Bio-Rad protein assay; Bio-Rad, Hercules, CA). A total of 10 µg of protein was separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis and electropheretically transferred to nitrocellulose (Optitran; Schleicher & Schuell, Keene, NH). Nitrocellulose blots were blocked for 2 h at room temperature in blotto (2.5% wt/vol nonfat dry milk, 20 mM Tris [pH 7.4], 150 mM NaCl, 0.1% Tween-20; all from Sigma). A polyclonal rabbit antihuman ₁ Na,K-ATPase antibody (Dr. Martin Vasalo, University of Tenerife, Tenerife, Spain) and a monoclonal antihuman ₁ (Upstate Biotech, Upsala, NY) were used as primary antibodies. An immunoperoxidase-based chemiluminescent detection system (ECL+plus; Amersham Corp., Arlington Heights, IL) was used to allow visualization of proteins of interest after exposure to Hyperfilm (Amersham).

Preparation of Basolateral Membranes

Cells were disrupted in homogenization buffer (HB) (300 mM mannitol, 12 mM Hepes-Tris, pH 7.4) and basolateral membranes were isolated using the method described by Hammond and coworkers (15). Briefly, 0.8 ml of suspended cells (approximately 3 to 5 × 10⁶ cells) were homogenized for 90 s with a motorized pestle before centrifugation (200 × g for 8 min at 4°C). The supernatant was transferred to a fresh tube. The pellet was resuspended in 0.8 ml of HB and homogenized for 60 s. This supernatant was cleared as described previously and combined with the first supernatant. Samples were then centrifuged at 9,000 × g for 20 min at 4°C to remove mitochondrial and nuclear components. The resultant supernatant was centrifuged at 48,000 × g at 4°C for 30 min in a Beckman TLX-120.2 rotor (Beckman, Fullerton, CA) to collect the total membrane fraction. This pellet was resuspended in 0.8 ml of HB with 0.2 ml of Percoll (Sigma) followed by centrifugation in a Beckman TLS-55 rotor (48,000 × g for 30 min at 4°C). Samples were maintained on ice through all steps of isolation. Protein thus obtained was separated and analyzed as described previously.

Measurement of Bioelectrical Properties of FDLE Cells

All experiments were performed 48 to 72 h after viral transfection. Filters with transepithelial resistance exceeding 0.8 k·cm² were mounted in Ussing chambers, as previously described. The chambers were filled with a solution containing (in mM): 145 Na⁺, 5 K⁺, 1.2 Ca²⁺, 1.2 Mg²⁺, 125 Cl, 25 HCO₃, 3.3 H₂PO₄, 0.8 HPO₄² (pH 7.4). Furthermore, the basolateral side contained 10 mM glucose and the apical side continued 10 mM mannitol to minimize the contribution of a putative apical Na⁺-glucose co-transporter to Na⁺ influx. The ionic composition was identical on both sides unless stated otherwise. The solutions were continuously bubbled with a mixture of 95% O₂ and 5% CO₂, and heated to 37°C. The I_SC was continuously measured with a transepithelial voltage clamp (EC-825; Warner Instruments, Hamden, CT). Square wave pulses (2 mV, 5 s) were applied across the monolayers every 60 s, allowing calculation of the transepithelial resistance (Rte) from the current change using Ohm's law. Amiloride- and ouabain-sensitive currents were calculated as the difference in the steady-state values of I_SC before and after addition of either amiloride (10 µM) or ouabain (1 mM) in the apical and basolateral compartments, respectively.

In a number of monolayers, after recording the basal I_SC, the apical membrane was permeabilized by adding 10 µM amphotericin B (a pore-forming antibiotic; Sigma) to the apical side of the Ussing chamber, thereby loading the cytosol with Na⁺ and eliciting the maximum Na⁺ transport by the Na,K-ATPase (16). When the I_SC had risen to its maximum value, ouabain (1 mM; Sigma) was added to the basolateral compartment of the Ussing chamber and the ouabain-sensitive component of the amphotericin-induced maximal I_SC (ouab_max) was calculated.

To determine whether transfection with ad₁ elicited an increase in Na⁺ transport across apical pathways, we established a 145:5 mM Na⁺ gradient across some monolayers (apical to basolateral compartments) by replacing 140 mM of basolateral Na⁺ with 140 mM N-methyl-D-glucamine, an impermeant cation, thus reducing the Na⁺ concentration in the basolateral compartment to 5 mM. After recording the basal I_SC, 100 µM amphotericin B was added into the basolateral side of the Ussing chamber. In this case, I_SC was due to the passive Na⁺ flux from the apical to the basolateral side through apical Na⁺ conductive pathways down a favorable Na⁺ concentration gradient (5). When the I_SC had attained its maximum value, the amiloride-sensitive component was determined by adding 10 µM amiloride (Sigma) into the apical compartment.

To assess the sensitivity of the Na⁺ transport to reactive species (5, 17), we added 500 µM hydrogen peroxide (Sigma) to both sides of the Ussing chambers 15 min before permeabilizing the apical membrane with 10 µM amphotericin B. Experiments were conducted in control monolayers and those transfected with 10 and 50 pfu (nominal values). To avoid a systematic error arising from a time-dependent decay of the I_SC in the presence of H₂O₂, the maximum I_SC after addition of amphotericin B was taken as the best possible estimate of ouabain-sensitive current (ouab_max) in these experiments because ouabain invariably decreased the current close to zero.

Statistical Analysis

Significant differences among group means were determined by one- or two-way analysis of variance (ANOVA) and Dunnett's post hoc test or, if data were not normally distributed, the Kruskal-Wallis one-way ANOVA with Dunn's post hoc test. A P < 0.05 was considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

The DNA content of freshly isolated FDLE cells was 10.5 ± 0.6 pg/cell (mean ± standard deviation [SD]; n = 12). Based on this value, we calculated that each filter contained approximately 2.2 ± 0.4 × 10⁵ cells (mean ± SD, n = 12) 24 h after seeding, which was approximately 4.4 times higher than at time zero. Therefore, the actual numbers of plaque-forming units per cell were 4.4-fold lower than their nominal value. FDLE cells grown on transparent plates and infected with 0.2, 1, and 2 pfu of ad-gal demonstrated transfection efficiencies of approximately 15, 45, and 60%, respectively (mean values of two plates for each plaque-forming unit number, 400 cells/plate counted). Typical photomicrographs are shown in Figure 1.

View larger version (121K): [in this window] [in a new window]   View larger version (121K): [in this window] [in a new window]   View larger version (128K): [in this window] [in a new window]   Figure 1.   Photomicrographs of FDLE cells infected with ad-gal (A, 0.2 pfu; B, 1 pfu; C, 2 pfu) 48 h before staining. Blue color indicates expression of the -galactosidase gene incorporated in the ad-gal viruses. Typical results were repeated on two different cell isolations.

Western blot studies using whole cell proteins showed an increased expression of ₁ or ₁ Na,K-ATPase subunit expression after transfection with either ad₁ or ad₁, respectively (Figure 2). The endogenous levels of ₁ and ₁ Na,K-ATPase subunits were below the level of detection using whole cell membrane fractions. On the other hand, Western blot studies using membrane fractions enriched for basolateral membrane components showed endogenous as well as transgenic proteins. Cells infected with ad₁ showed increased amounts of the ₁ subunit, whereas isolated basolateral membranes of cells infected with ad₁ showed increased amounts of both ₁ and ₁ subunits (Figure 2).

View larger version (89K): [in this window] [in a new window]   Figure 2.   Western blot analysis of 1 and 1 subunit expression in whole cells (A) and basolateral membranes (B) of FDLE cells infected with either adNull, ad1, or ad1. The lack of signal in null transfected FDLE cells is due to the low ambundance of these subunits. As in prior studies (13), the transgenic 1 (1tr) found in basolateral membranes is smaller than endogenous 1 protein, presumably due to differences in glycosylation. Each blot was repeated at least twice with similar results.

All FDLE cells used in these studies were obtained from 33 different cell isolations. Material from at least four different cell isolations was used in all experiments except for the one shown in Figure 3 (two isolations). Mean value for Rte for all monolayers was 1.16 ± 0.5 k · cm² (mean ± SD, n = 444). Seventy-six percent of all monolayers had a resistance of more than 0.8 k · cm². In the first set of experiments, we assessed the effects of transfection with either ad0 or ad₁ (2 pfu) on baseline, ouabain-, and amiloride-sensitive currents across intact monolayers. Mean values for I_SC for ad0-transfected monolayers in this group of experiments was 5.1 ± 0.3 (µA/cm² (mean ± 1 standard error; n = 17). Addition of ouabain (1 mM) in the basolateral side or amiloride (10 µM) in the apical side decreased I_SC significantly (Figure 3). These data indicate that more than 75% of the I_SC was due to Na⁺ absorption. Transfection with ad₁ significantly increased baseline I_SC by about 20% to 6.1 ± 0.3 µA/cm² (Figure 3 and Table 1). The ouabain-sensitive components of these I_SC values were 3.5 ± 0.28 (n = 17) and 4.3 ± 0.3 µA/cm² (n = 18) (X ± 1; n = number of filters; P < 0.05), whereas the amiloride-sensitve components were 3.5 ± 0.3 (n = 17) and 3.7 ± 0.3 µA/ cm² (n = 18) (X ± 1 standard error of the mean [SEM]; n = number of filters; P > 0.05).

View larger version (26K): [in this window] [in a new window]   Figure 3.   ISC currents of nonpermeabilized FDLE monolayers infected with 2 pfu of null (ad0) vector or ad1 on baseline ISC as well as ISC after addition of ouabain (1 mM) or amiloride (10 µM) into the basolateral and apical compartments, respectively. Values are given as means ± SEM; numbers of pfu are shown on the abscissa and total number of measured monolayers within the bars. After transfection with 2 pfu of ad1, the total ISC and its amiloride-insensitive component were significantly different from their corresponding control values transfected with null vectors (P < 0.05) (indicated with an asterisk above the the appropriate bar).

In the second set of experiments, we assessed the effects of transfection on I_SC across permeabilized monolayers. Typical I_SC tracings before and after permeabilization of the apical membrane in control and ad₁-transfected monolayers are shown in Figure 4. Mean I_SC values (± 1 SEM) at baseline and after addition of amphotericin B, ouabain, and the ouabain-sensitive component of the amphotericin B-elicited I_SC (ouab_max) are shown in Table 1. Basal values of I_SC were increased significantly after transfection with 2 pfu of ad₁ (Figure 2 and Table 1). Transfection with 11 pfu of ad₁ also increased baseline I_SC; however, the mean value was not statistically different from control or 2 pfu. Addition of amphotericin B into the apical compartment led to a large increase of I_SC due to the removal of the apical barrier and increased cytosolic [Na⁺]. The ouab_max currents were significantly higher after transfection with 2, 11, or 22 pfu ad₁ in a dose-dependent manner (Table 1). Transfection of FDLE cells with 2, 11, or 22 pfu ad₁ alone had no effect in any of the measured variables (Table 1), whereas transfection with both ad₁ and ad₁ increased both baseline and ouab_max, but not more than transfection with the ad₁ alone (Table 1). Finally, there was no difference in any of the measured currents between noninfected monolayers and those infected with 2 pfu of ad0 or ad-gal (data not shown).

View larger version (21K): [in this window] [in a new window]   Figure 4.   ISC currents across two FDLE monolayers. The tracings show the values of the basal, amphotericin-elicited (after addition of 10 µM amphotericin B in the apical compartment), and ouabain-sensitive (after addition of 1 mM ouabain on the basolateral side) currents. The solid line indicates a control monolayer, whereas the dotted line indicates one infected with 11 pfu/ cell of ad1. Results of typical experiments.

Despite the large increase in the ouab_max currents, the conductance of apically located Na⁺ permeable pathways remained unchanged after transfection with ad₁ as shown by controlling the driving force across the apical membrane by an apical to basolateral Na⁺ gradient of 145:5. With the basolateral membrane permeabilized by amphotericin B, both baseline and amiloride-sensitive currents were not different from their corresponding control values (Figure 5).

View larger version (36K): [in this window] [in a new window]   Figure 5.   Effects of the transfection of FDLE cells with ad1 on amiloride-sensitive currents. Numbers are mean ± SEM; numbers of pfu and total number of monolayers (numbers in parentheses) are also shown. All monolayers were derived from three different cell isolations. All currents were measured in the presence of an apical to basolateral Na+ gradient of 145:5 mM. After the baseline measurement (open bars), monolayers were permeabilized by the addition of 10 µM amphotericin B in the basolateral side (dotted bars) and the amphotericin B-elicited current was then calculated. Amiloride (10 µM; hatched bars) was then added to the apical side of the monolayer causing the ISC to decrease. The amiloride-sensitive component of ISC (solid bars) was calculated as the difference of the current before and after addition of amiloride. Values are mean ± SEM; numbers of pfu and monolayers (in parentheses) are shown in the abscissa.

Data shown in Figure 6 indicate that addition of 500 µM H₂O₂ to both compartments of the Ussing chambers significantly decreased the basal I_SC of cells infected with 0, 2, or 11 pfu of ad₁ to 69, 89, and 64% of controls without H₂O₂, respectively, and ouab_max to 47, 65, and 55%. Monolayers infected with either 2 or 11 pfu of ad₁ had higher residual baseline and ouab_max currents than did noninfected cells. In fact, the levels were similar to those of noninfected cells in the absence of H₂O₂. The relative decrease of I_SC by H₂O₂ in transfected monolayers (basal, 35%; ouabain-sensitive, 45%) was similar to nontransfected monolayers.

View larger version (13K): [in this window] [in a new window]   View larger version (17K): [in this window] [in a new window]   Figure 6.   (A) Effects of H2O2 on ISC currents of ad1 monolayers transfected with 0, 2, and 11 pfu of ad1. Boluses of hydrogen peroxide (500 µM; solid bars) were added at time zero into both the apical and basolateral compartments of the Ussing chambers and ISC values were recorded continuously. In control studies (open bars), an equal amount of buffer was instilled in both compartments of the Ussing chamber. Bar graphs are mean ± SEM measured at 15 min after addition of H2O2 or buffer; numbers of pfu are shown on the abscissa and total number of measurements within the bars. (B) After 15 min, 10 µM amphotericin B was added into the apical membrane and ouabain-sensitive currents were measured as described previously. All monolayers were derived from five different cell isolations. ISC before permeabilization and ouabain-sensitive currents were significantly increased by ad1 transfection and significantly decreased by H2O2 (two-way ANOVA, P < 0.05). Despite H2O2, ouabain-sensitive current remained higher in infected cells and equaled the current of noninfected cells in the absence of H2O2. *Significantly different from control with 0 pfu.  Significantly different from control with same pfu. § Significantly different from 0 pfu with H2O2. Baseline ISC values in control monolayers infected with 0, 2, and 11 pfu of ad1 were 3.6 ± 0.3, 4.8 ± 0.3, and 5.2 ± 0.5 µA/cm2, respectively (mean ± SEM ); ISC baseline values of FDLE monolayers subsequently treated with H2O2 were 3.8 ± 0.2, 5.2 ± 0.4, and 5.3 ± 0.6 µA/cm2, respectively.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Na,K-ATPases are complex transmembrane proteins containing multiple subunits (at least , , and ), each one of them consisting of multiple isoforms (18). In the mammalian lung, both the  and  subunits are necessary for Na,K-ATPase to function properly (19). The larger  subunit is responsible for ion translocation across the basolateral membrane and adenosine triphosphate hydrolysis, whereas the smaller  subunit is a glycoprotein that controls heterodimer assembly and insertion of the complex into the basolateral membrane. Active Na⁺ absorption across fetal and adult ATII cells creates an osmotic pressure gradient that contributes to the removal of fetal fluid and the re-absorption of edema fluid in adult injured lungs (3). Inhibition of the Na,K-ATPase with ouabain has been shown to impair ion transport in cultured ATII cells and edema clearance in isolated lungs (20).

We used fetal distal lung epithelia in our experiments. Although these cells are generally considered to be secretory rather than absorptive in vivo, we found the majority of their I_SC to be amiloride sensitive (Figure 3), indicative of Na⁺ absorption. It is possible that the cells change their properties during the time in cell culture, which must be considered when interpreting our results.

Our results show that both the baseline I_SC and the maximum Na⁺ transport capacity of FDLE cells were enhanced by adenovirus-mediated transfer and overexpression of a gene encoding for the ₁ but not the ₁ subunit of the Na,K-ATPase (Figures 2, 3, and 6, and Table 1). This effect is specific for the ₁ subunit of the Na,K-ATPase because viral transfection per se did not change the bioelectric properties of FDLE monolayers, as shown by the lack of an effect after infection with ad0, ad-gal, and ad₁ viruses, in agreement with previous findings in airway cells (13), and suggests enhanced Na⁺ transport by an increased number of Na,K-ATPases. The lack of baseline I_SC increase after transfection with 22 pfu may result from toxicity due to high capsid and DNA loads (21), which may impair Na⁺ entry mechanisms and thereby prevent a higher baseline I_SC even with very high levels and function of the Na,K-ATPases (Table 1). The optimal virus dose appears to be between 2 and 11 pfu. It is interesting to note that about 25 to 40% of the maximal current was not inhibited by 1 mM ouabain (Table 1). It is possible that higher concentrations of ouabain are needed to totally inhibit Na,K-ATPase activity. Alternatively, the remaining current may be due to an existence of an H,K-ATPase (22).

The ability of ad₁ and ad₁ to overexpress their respective messenger RNAs (mRNAs) and proteins has been tested in several experimental models. Specifically, they have been shown to be capable of transgene activation and processing in isolated adult rat alveolar type 2 cells, human bronchial epithelial cells, and a lung epithelial cell line (A549) (13, 23). In other studies, these viruses have produced high-level transgene expression for at least 7 d after transfection in normal rat lungs (13), and ad₁ has been shown to improve survival in hyperoxic rats (24). Furthermore, we have demonstrated a high transfection efficiency with ad-gal and a markedly increased ₁ and ₁ subunit expression after infection with ad₁ and ad₁, respectively. Thus, dysfunctional transgene activation, processing, or function are unlikely explanations for the absence of functional changes after ₁ overexpression in the studies presented herein.

Our data provide additional support for the hypothesis that the ₁ subunit is a rate limiting subunit in the assembly of Na,K-ATPase enzymes (25). Whereas ad₁ and ad₁ increased protein expression of the respective protein subunits in whole cells, only infection with ad₁ and overexpression of the ₁ subunit resulted in an increase of Na⁺ transport. This is in accordance with our observation of a simultaneous increase of ₁ and ₁ subunit expression in the basolateral membrane after infection with ad₁. In addition, rat FDLE cells obtained between 17 and 22 d gestational age have been shown to have much more mRNA for the ₁ subunit than for the ₁ subunit, with an ₁:₁ ratio of 10 on gestation Day 17, 4 on Day 19, 3 on Day 20, and 4 on Day 22 (26), whereas the absolute amounts of mRNA for both subunits as well as the Na,K-ATPase activity increased toward term (22 d) (26). The constitutive excess of ₁ subunits over ₁ subunits would explain why transgenic overexpression of additional ₁ subunits alone is sufficient for increasing the number of functional Na,K-ATPases containing ₁ and ₁ subunits in the basolateral membrane, whereas overexpression of ₁ subunits does not lead to any additional ion transport. Whether additional overexpression of ₁ subunits may have additional effects under oxidative stress, however, was not studied here.

In ATII cells of rats exposed to hyperoxia, ouabain-sensitive ⁸⁶Rb⁺ uptake was not inhibited despite a significant decrease of ₁ subunit protein level and hydrolytic activity. The ₁ subunit protein level did not change and correlated best with Na,K-ATPase activity (27), indicating again that the ₁ subunit protein level determines the amount of Na,K-ATPase activity in alveolar cells. In A549 cells, however, increased ⁸⁶Rb⁺ uptake was observed after ad₁ infection (23), indicating that cancer cell lines may be different in this respect.

In intact monolayers, basal and ouabain-sensitive I_SC values were significantly increased after transfection with 2 pfu of ad₁, as compared with the corresponding values in ad0-infected monolayers. The smaller magnitude of I_SC changes in nonpermeabilized monolayers after transfection as compared with the current changes after permeabilization were most likely due to the fact that the majority of resistance to sodium influx across most epithelial cells is encountered across the apical membrane. It is very unlikely, but possible, that a small amount of apically instilled amphotericin B may have gained access to the basolateral membrane and thus decreased measurable Na,K-ATPase activity. In this case, our measurements of the ouab_max in both control and infected monolayers will underestimate the true Na,K-ATPase activity, but this does not invalidate our conclusions. It should be stressed that an increase of baseline I_SC of about 1 µA/cm², as observed in nonpermeabilized monolayers after transfection with 2 pfu of ad₁, will considerably increase water re-absorption. With the elemental charge being 1.602 × 10¹⁹ Coulomb (C), the charge of 1 mol Na⁺ amounts to 96,488 C. Thus, a current increase of 1 µA/cm², equivalent to 10^-6 C/s/cm² corresponds to an increase of Na⁺ transport by about 10¹¹ mol/ s/cm². With a Na⁺ concentration of 140 mmol/liter, osmotic forces will result in an additional movement of 7.4 × 10⁵ µl/s/cm² of water. In a human adult with a pulmonary surface area of 90 m², 3% of which are ATII cells, the water absorption would be increased by 120 µl/min.

We speculate that the observed increase in basal I_SC after transfection with 2 pfu resulted from hyperpolarization of the apical membrane due to lower intracellular Na⁺ because of higher Na,K-ATPase activity. Hyperpolarization and lowering the intracellular Na⁺ concentration would increase the driving force for Na⁺ entry and thus increase transcellular Na⁺ transport and I_SC even without an increase in ENaC expression or function. This interpretation is supported by the fact that infected and control monolayers had similar values of amiloride-sensitive currents when we controlled the driving force across the apical membrane by setting a transepithelial Na⁺ concentration gradient and permeabilizing the basolateral membrane with amphotericin B. These experiments also argue against an increase in amiloride-insensitive Na⁺ absorption because there was no difference between control and ad₁-infected monolayers in the maximum current elicited by amphotericin B (Figure 5). However, an increase of amiloride- insensitive current was seen in nonpermeabilized monolayers after ad₁ infection (Figure 3), which may be related to different amiloride-insensitive pathways, including chloride secretion.

Oxidative stress was instituted by adding 500 µM H₂O₂, a biologically relevant concentration that can be achieved in vivo near activated cellular oxidases (28), which acutely diminished Na⁺ transport, possibly by damaging the Na,K-ATPase (17). Similarly, instillation of H₂O₂ to the basolateral side only has been shown to decrease short circuit current in ATII cells with an IC₅₀ of only 40 µM (29). Further, 500 µM H₂O₂ was found to decrease the ouabain-sensitive current in colonic epithelial cells (30). In our ad₁-infected cells, all currents were diminished by H₂O₂, but higher abundance of Na,K-ATPase resulted in higher residual basal as well as ouabain-sensitive currents as compared with noninfected controls. The increased number of possible targets for H₂O₂ probably increased the number of proteins remaining undamaged by a constant concentration of H₂O₂. Long-term responses to hyperoxia observed in ATII cells in other studies were more complicated. A transient decrease of Na,K-ATPase activity was followed by overexpression and activation, presumably as part of the defense against oxygen radicals and pulmonary edema due to increased permeability (6, 20, 31).

Enhancing lung liquid clearance may be important, especially in premature infants whose pulmonary epithelium may not have completed the transition from secretion to absorption (32) and may be vulnerable to oxidative stress because of underdeveloped antioxidant systems (35). Because Na⁺ absorption is diminished in premature animals (32) and infants (33), and in addition, volutrauma (8) and oxidative stress (4) appear to further inhibit this process, especially in preterm subjects with increased vulnerability (35), overexpression of Na,K-ATPase subunits by gene transfer with increased Na,K-ATPase function and Na⁺ absorption in alveolar epithelium could be of great benefit. The increase of amphotericin B-elicited ouabain-sensitive currents, reflecting the maximum ability of Na,K-ATPase to pump Na⁺ ions after transfection, may be especially beneficial in pathologic conditions when apical membrane integrity may be impaired, resulting in increased levels of intracellular Na⁺.

We conclude that adenovirus-mediated transfer of a gene encoding for a ₁ subunit of Na,K-ATPase increased Na,K-ATPase expression and function, and enhanced the baseline I_SC and the maximum capacity of Na,K-ATPase in both control and H₂O₂-treated FDLE monolayers. Because the treatment of premature infants often requires high inspired oxygen concentrations, which increase formation of superoxide and H₂O₂, and thereby exposes their lungs to significant oxidative stress, we speculate that overexpression of genes encoding for Na,K-ATPase may enhance lung liquid clearance, thereby improving lung function and possibly reducing chronic lung disease.