INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

AMILORIDE-SENSITIVE Na⁺ channels are the primary pathway for the entry of Na⁺ into a large number of epithelial cells (9, 10, 28, 38). Electrophysiological studies have shown that Na⁺ channels display functional heterogeneity regarding their biophysical and pharmacological properties (reviewed in Ref. 10). Their classification is based according to their kinetics, pharmacology, and single-channel conductance (10, 36, 42). Presently, these channels are thought to be composed of three different subunits, referred to as the -, -, and -epithelial Na⁺ channels (ENaCs), cloned from the colon of salt-deprived rats and human lung tissue by a number of investigators (2, 3, 24, 48). Experiments utilizing point mutations suggest that all three subunits are involved in pore formation (40), although the exact stoichiometry is debated, with different groups reporting either four (7) or nine (43) subunits in the complex.

Existing evidence indicates that active Na⁺ transport across the adult alveolar epithelium plays an important role in maintaining the alveolar space free of fluid, especially after lung injury when alveolar permeability to plasma proteins has been increased (29, 34, 50, 51). Amiloride-sensitive Na⁺ channels, shown to be present in both fetal (32, 35, 48) and adult (17, 27, 51, 52) alveolar type II (ATII) cells, represent the major pathways for Na⁺ entry across these cells, with other pathways such as glucose and amino acid cotransporters accounting for only a minor fraction of the total Na⁺ flux (33, 39, 50). Although both fetal and adult ATII cells have been shown to contain mRNAs for the various ENaC subunits (for a review, see Ref. 28), there is considerable controversy as to whether the lung channels are composed of the same subunits as channels found in the kidney and colon (28).

Because of the overall importance of Na⁺ transport in lung fluid balance in both normal and pathological conditions, there is considerable interest in identifying the basic mechanisms responsible for the regulation of Na⁺ channels in adult ATII cells. These studies have been hampered by the fact that ATII cells are difficult to isolate and maintain in primary culture. In addition, during culture, ATII cells undergo dedifferentiation and lose their ability to secrete surfactant and express Na⁺-channel proteins (49). For these reasons, it is important to document the presence of amiloride-sensitive Na⁺ channels in lung epithelial cell lines and characterize their biophysical properties.

A number of these lines, including A549 cells that originated from a human alveolar cell carcinoma and possess many characteristics of type II cells including multilamellar cytoplasmic inclusion bodies and the ability to synthesize surfactant phospholipids (22), are routinely used for ion transport studies (19, 25). Herein we show that A549 cells express a 8.6-pS amiloride-sensitive Na⁺ channel with biophysical properties similar to those found in ATII cells in primary culture. Furthermore, our results demonstrate that increases in A549 cell cAMP levels upregulate whole cell amiloride-sensitive currents by increasing the product [channel activity (NP_o)] of the number of channels (N) times their open probability (P_o). Finally, nitric oxide (· NO), in concentrations likely to be found in the alveolar spaces of injured lungs, decreases whole cell current and NP_o by increasing A549 cell cGMP levels. These data offer new insight into the cellular mechanisms responsible for the regulation of Na⁺ transport across alveolar epithelial cells.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cell culture. A549 cells were purchased from American Type Culture Collection (Manassas, VA) in the 76th passage. They were suspended in DMEM-F-12 medium (Cellgro) supplemented with 1% penicillin-streptomycin and 10% fetal calf serum, plated on plastic tissue culture flasks (Corning Glass Works, Corning, NY), and placed in an incubator in 21% O₂, 5% CO₂, and balance N₂ at 37°C and 100% humidity. All measurements were conducted in cells between the 78th and 97th passages.

Patch-clamp recordings: whole cell measurements. In the first series of experiments, macroscopic currents were recorded from A549 cells in the whole cell recording mode of the patch-clamp technique (13). Twenty-four to thirty-six hours before any electrophysiological measurements, A549 cells were lifted from the tissue plates by treatment with 2.5% trypsin-EDTA (Sigma, St. Louis, MO) for 3-6 min at 37°C and then seeded on 12-mm-diameter glass coverslips in DMEM-F-12 medium. Just before the start of the experiment, each coverslip was rinsed with standard external solution (SES) with the following ionic composition (in mM): 145 NaCl, 2.7 KCl, 1.8 CaCl₂, 2 MgCl₂, 5.5 glucose, and 10 HEPES, pH 7.4, 323 mosmol. The coverslip was then transferred to the recording chamber and mounted on the stage of an inverted microscope (IMT-2 Olympus) for patch-clamp recordings.

Pipettes were made from LG16-type capillary glass (Dagan, Minneapolis, MN) with a vertical puller (model PB-7, Narishige). They were back-filled with standard internal solution (SIS; pH 7.2 at 22°C, 300 mosmol) with the following ionic composition (in mM): 135 potassium methylsulfonic acid, 10 KCl, 6 NaCl, 1 Mg₂ATP, 2 Na₃ATP, 5.5 glucose, 10 HEPES, and 0.5 EGTA. The pipette resistance varied from 3 to 5 M when filled with SIS. The pipette offset potential was corrected just before gigaseal formation. Series resistance and capacitance transients were compensated for with the patch-clamp amplifier (Axopatch 200, Axon). Junction potentials were corrected as previously described (1).

The cell membrane potential was held at 40 mV during all whole cell recordings. Inward and outward currents across the cell membrane were elicited by altering the membrane potential from the holding value (40 mV) by 100 to + 100 mV in 10-mV increments of either 450 or 900 ms duration every 10 s with the Clampex program (pCLAMP, Axon Instruments). Currents were digitized with a digital-to-analog and analog-to-digital converter (DigiData 1200A, Axon Instruments), filtered through an internal four-pole Bessel filter at either 0.5 or 1 kHz, and sampled at 2 kHz. Current-voltage (I-V) curves were constructed by measuring the steady-state current values 300 ms from the start of voltage pulses with the Clampfit Program (Axon Instruments) and Origin Software (Microcal Software, Northampton, MA). In some experiments designed to measure the relative permeability of these channels to various ions, NaCl in SES was replaced with equivalent amounts of LiCl, KCl, or N-methyl-D-glucamine chloride (NMDG); all other components were maintained at their respective concentrations.

To test the extent to which whole cell currents were inhibited by amiloride, we measured I-V relationships of cells in the whole cell mode and then repeated the measurements after perfusing cells with SES containing amiloride in concentrations ranging from 1 nM to 100 µM. We then calculated the amiloride-sensitive currents by digitally subtracting the currents in the presence of amiloride from its corresponding control value.

To study the short-term regulation of these Na⁺ channels by cAMP and · NO, amiloride-sensitive I-V relationships were measured before and after perfusion of the cells with SES containing forskolin, 8-bromo-cGMP (8-BrcGMP; both from Calbiochem, La Jolla, CA), or PAPA NONOate (Cayman Chemical, Ann Arbor, MI), a · NO donor. PAPA NONOate stocks were prepared by dissolving it in 43 mM phosphate buffer (pH 9) just before use. Evolution of · NO in the SES medium (pH 7.4, 22°C) was measured with an ISO-NO electrochemical probe (World Precision Instruments, Sarasota, FL) connected to an IBM-compatible computer equipped with an analog-to-digital converter. Mean · NO concentration values were calculated as previously described (12). To assess potential mechanisms by which · NO modulated whole cell currents, A549 cells were incubated with 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ; Tocris Cookson, St. Louis, MO), an inhibitor of guanylyl cyclase, for 30 min and then perfused with SES containing PAPA NONOate as described above.

Single channels. To further evaluate the biophysical properties of these channels, we patched A549 cells in the cell-attached mode and recorded single-channel currents. The ionic composition of the pipette solution was (in mM) 145 sodium glutamate or sodium aspartate, 1 CaCl₂, 5 MgCl₂, 5.5 glucose, and 10 HEPES, pH 7.4. Cells were depolarized to 0 mV by perfusing them with the following solution (in mM): 135 potassium glutamate, 10 KCl, 5 MgCl₂, 10 HEPES, and 5.5 glucose, pH 7.4. Single-channel currents were filtered at 1 kHz, sampled at 2 kHz, and analyzed with the Fetchan and pStat programs (pCLAMP, Axon Instruments). The amplitude and open probability (P_o) were calculated from all event histograms, constructed as previously described (21). The product of the number of channels (N) times the P_o in a patch (NP_o), which reflects the activity of channels, was calculated from single-channel recordings as follows
where T is the total recording time, i is the number of open channels, and t_i is the recording time during which i channels were open. Calculation of NP_o does not involve any assumptions about the N in a patch or the P_o of a single channel.

RNA isolation and semiquantitative RT-PCR. A549 cells were treated with a 2.5% trypsin-EDTA solution (Sigma) for 3-6 min at 37°C and seeded on 10-cm tissue culture plates for 24 h in DMEM-F-12 medium. At that time, 1 ml of TRI Reagent (Molecular Research Center, Cincinnati, OH) was added to each plate, and RNA was isolated from ~2 million cells according to the protocol supplied by the manufacturer (Molecular Research Center) with the method of Chomczynski and Sacchi (4). One microgram of total RNA in a total volume of 12 µl was denatured at 70°C for 10 min. Denatured RNA was chilled for 2 min in ice and used for reverse transcription as described by the standard protocol (GIBCO BRL, Bethesda, MD). In brief, 1 µl of Superscript II (GIBCO BRL), 20 U of RNasin, 1 µl of random hexamer, 2 µl of 100 mM dithiothreitol, and 1 µl of 10 mM deoxynucleotide triphosphate mixture in a total volume of 20 µl were added to the sample, mixed well, and incubated for 1 h at 42°C. The reaction mixture was heated to 70°C for 15 min to inactivate RT. Two microliters of the reaction mixture were used for PCR amplification in a Robocycler (Stratagene) with 1 µl (20 pmol) each of upstream and downstream primers specific for the -, -, and -subunits of the human ENaC (hENaC) gene and the housekeeping gene hypoxanthine phosphoribosyltransferase (HPRT; Table 1). Each primer was added into a 50-µl mixture containing 1 µl of 10 mM deoxynucleotide triphosphate mixture, 5 µl of 10× PCR buffer, 2 mM MgCl₂, and 2.5 U of Taq polymerase. The cycle parameters were initial denaturation at 95°C for 5 min, 62°C for 1 min, 72°C for 1 min, and 95°C for 1 min for a total of 35 cycles and final extension for 7 min at 72°C. Twenty microliters of the final amplified product were electrophoresed on a 1.2% agarose gel, and DNA was visualized after ethidium bromide staining.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Amiloride effect on whole cell currents in A549 cells. In the whole cell mode, A549 cells exhibited inward rectifying currents carried by Na⁺ (Fig. 1). Perfusion of the cells with SES containing 10 µM amiloride rapidly reduced the inward (Na⁺) but not the outward (K⁺) currents (Fig. 1, B-D). The amiloride-sensitive current reversed at +47 mV. Amiloride dose-response relationships of inward currents measured at 100 mV are shown in Fig. 2. The amiloride inhibition constant (K_i), calculated as described in the legend of Fig. 2, was 0.83 ± 0.07 (SE) µM (n = 6 cells). When the ionic gradients were reversed (i.e., bath, 145 mM K⁺; pipette, 145 mM Na⁺), perfusion with SES containing amiloride (10 µM) reduced the outward (Na⁺) but not the inward (K⁺) currents (data not shown).

View larger version (31K): [in this window] [in a new window]   Fig. 1.   Representative recordings (A-C) and mean values (D) of whole cell currents in A549 cells. A: cell was held at a potential of 40 mV. Currents were elicited by applying voltage steps between 100 and +100 mV in 10-mV steps lasting 450 ms every 10 s. Pipette was filled with standard internal solution (SIS), and cell was perfused with standard external solution (SES; see MATERIALS AND METHODS for details). B: currents were recorded after perfusion of cell with SES solution containing 10 µM amiloride. C: amiloride-sensitive currents obtained by subtracting residual current in B from total current in A. D: current-voltage (I-V) relationships of total and amiloride (Amil)-sensitive currents. Values are means ± SE; n = 30 cells.

View larger version (11K): [in this window] [in a new window]   Fig. 2.   Inhibition of whole cell Na+ currents by amiloride. For each cell, inward Na+ currents recorded at 100-mV voltage step were measured during perfusion with SES. Cells were then perfused with SES containing indicated concentrations of amiloride until a steady state was reached, usually within 3-5 min from onset of perfusion. Normalized amiloride-sensitive currents were calculated as follows: I = [1  (I0  Ix)/(I0  I10)] · 100, where I0 is control value, Ix is steady-state current during perfusion with concentration x, and I10 is current measured during perfusion with 10 µM amiloride. Values are means ± SE; n = 6 cells. Inhibition constant (Ki) was calculated by best fit of data points to the following equation: I = {1  [1/(1 + Ki/x)]} · Imax, where Imax is maximum current.

Replacement of Na⁺ in the bath with NMDG⁺ or K⁺ greatly decreased the inward amiloride-sensitive Na⁺ current (Fig. 3A). On the other hand, substitution of external Na⁺ with Li⁺ had no appreciable effect on the amiloride-sensitive currents. The relative permeabilities of Na⁺ to K⁺ (P_Na/P_K) and Na⁺ to Li⁺ (P_Na/P_Li), calculated from the constant-field equation with the reversal potentials of the amiloride-sensitive currents, were 6:1 and 1:1.2, respectively.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   A: cation selectivity of A549 ion channels. For each cell, amiloride-sensitive currents were computed by subtracting value of inward whole cell current obtained at 100-mV voltage step in presence of 10 µM amiloride from corresponding total current. Measurements were conducted with SES containing Na+, N-methyl-D-glucamine (NMDG+), K+, or Li+. Values are means ± SE; N, no. of cells. B: dependence of amiloride-sensitive currents on bath Na+ concentration ([Na+]out). Normalized amiloride-sensitive currents were computed by dividing amiloride-sensitive current at 100-mV voltage step at indicated [Na+]out by corresponding value at 145 mM Na+. Each point represents mean ± SE of 3 experiments; n = 6 cells. Solid line, Michaelis-Menten equation fitted through these data (half-saturation constant = 37 ± 3.5 mM).

Effect of extracellular Na⁺ concentration on the whole cell current. Values of normalized amiloride-sensitive Na⁺ currents at 100-mV voltage step (i.e., ratio of the amiloride-sensitive current measured at a known external Na⁺ concentration divided by the corresponding value at 145 mM Na⁺) were plotted against the Na⁺ concentration in the bath and fitted to the Michaelis-Menten equation (Fig. 3B). The half-saturation constant (K_m) was found to be 37 ± 3.5 (SE) mM (n = 6 cells).

Amiloride-sensitive single Na⁺ channels in A549 cells. Single-channel activity was seen in ~20% of 144 successful cell-attached patches. A characteristic recording is shown in Fig. 4A. Two current levels are seen. The single-channel conductance was determined with the histogram distribution fitted to a Gaussian equation (Fig. 4B). At a patch potential of 100 mV, the single-channel conductance was 8.6 ± 0.04 pS (n = 411 events). Single-channel open time for this record ranged from a few milliseconds to hundreds of milliseconds (395 ± 18 ms), with an P_o of 0.68 ± 0.02 for amplitude level 1 and 0.19 ± 0.02 for amplitude level 2. When amiloride (10 µM) was included in the pipette solution, no channel activity was seen in 12 cell-attached patches (Fig. 4C).

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Single-channel currents in A549 cells. Single-channel currents recorded in a cell-attached patch from an A549 cell at a holding potential of 100 mV. Pipette was filled with 145 mM Na+; cell was perfused with a solution containing 145 mM K+. Recording (A) and amplitude distribution (B), constructed from a 384-s recording, show 2 active channels with amplitude of 0.86 ± 0.004 pA. Single-channel conductance was 8.6 ± 0.04 (SE) pS (n = 411 events). No single-channel activity was seen when 10 µM amiloride was present in pipette solution (C). Results are from typical experiments that were repeated at least 6 times with different cells.

Modulation of whole cell current and single-channel activity by cAMP. Perfusion of A549 cells patched in the whole cell mode with SES containing 10 µM forskolin resulted in a large increase in the inward (Na⁺) current, which was rapidly reversed by amiloride. Mean values of the inward currents as a function of time after forskolin perfusion are shown in Fig. 5.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Effect of forskolin on A549 whole cell inward current as shown by continuous recording of inward Na+ currents across A549 cells elicited by 100-mV voltage pulse applied to cells every 10 s from a holding potential of 40 mV. Pipette was filled with SIS, and cell was perfused with SES. Perfusion of cells with SES containing 10 µM forskolin induced a significant increase in inward current; >90% of forskolin-induced current was inhibited by amiloride. Data points are means ± SE; n = 5 cells. Currents were normalized by dividing current in each patch by measured value of current 2 min from onset of perfusion with forskolin.

In the cell-attached mode, forskolin increased single-channel activity and NP_o of single channels without affecting their amplitude (Fig. 6, A and B). Forskolin increased NP_o in all patches. Mean NP_o values measured in five different cells with low spontaneous single-channel activity were 0.01 ± 0.04 for control cells and 1.5 ± 0.2 for forskolin-treated cells (n = 5). The fact that the effect was immediate in the cell-attached mode suggests that the delay observed in whole cell measurements was not due to a slow perfusion rate but rather to the dilution of necessary intracellular components by the intrapipette solution. When 10 µM amiloride was included in the pipette solution, no single-channel activity was seen in six separate patches after perfusion of the cells with forskolin. A typical record is shown in Fig. 6C.

View larger version (29K): [in this window] [in a new window]   Fig. 6.   Effects of forskolin on A549 single-channel currents recorded in a cell-attached patch from an A549 cell at a holding potential of 100 mV just before and during perfusion with forskolin. Pipette was filled with 145 mM Na+; cell was perfused with a solution containing 145 mM K+. Recording (A) and amplitude distribution (B) show at least 5 different channels. Effect of forskolin was totally abolished by presence of 10 µM amiloride in pipette (C). Results are from typical experiments that were repeated 6 times with different cells.

Effect of · NO and cGMP on whole cell and single-channel currents. Addition of 100 µM PAPA NONOate into the SES solution resulted in a rapid release of ~4 µM · NO, which persisted for >20 min. Addition of oxyhemoglobin (20 µM), a potent scavenger of · NO, into the bath solution decreased · NO release to zero. As shown in Fig. 7, perfusion of A549 cells with SES containing 100 µM PAPA NONOate inhibited the inward but not the outward whole cell currents in a rapid and reversible fashion; furthermore, the inward current returned to the baseline value when the cells (n = 6) were reperfused with SES alone. The reversal potential of the · NO-sensitive currents (obtained by subtracting the current remaining after perfusion with PAPA NONOate from the total current) was similar to that of the amiloride-sensitive current (+47 mV). Similar effects were achieved with considerably lower concentrations of PAPA NONOate (10 µM), releasing 300 nM · NO (results not shown). Preincubation of A549 cells with 3 µM ODQ (a potent inhibitor of soluble guanyl cyclase) for 30 min before perfusion with PAPA NONOate totally prevented the reduction of the inward Na⁺ currents (Fig. 8). Furthermore, perfusion of A549 cells with 100 µM 8-BrcGMP (n = 4) markedly inhibited the inward but not the outward currents (Fig. 9). Finally, 100 µM PAPA NONOate markedly decreased single-channel activity in the cell-attached patches of A549 cells (Fig 10).

View larger version (30K): [in this window] [in a new window]   Fig. 7.   Effects of nitric oxide (· NO) on whole cell A549 currents. A: time course recording of whole cell inward (Na+) current evoked by 100-mV voltage pulses every 10 s before, during (solid line), and after perfusion of an A549 cell with SES containing 100 µM PAPA NONOate. Pipette was filled with SIS. B and C: whole cell I-V relationships before and 5 min after, respectively, PAPA NONOate perfusion when steady-state currents were seen. Whole cell current inhibited by · NO (· NO sensitive) was calculated by digitally subtracting currents at steady-state effect of · NO (as shown in C) from current before perfusion with · NO containing SES (as shown in B). D: I-V relationships for total and · NO-sensitive (sensit) currents. Values are means ± SE; n = 6 cells.

View larger version (16K): [in this window] [in a new window]   Fig. 8.   1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ) prevents decrease of inward current by PAPA NONOate as shown by time-course recording of whole cell inward (Na+) current evoked by 100-mV voltage pulses every 10 s in an A549 cell incubated for 30 min with ODQ (an inhibitor of soluble guanylyl cyclase). Typical experiment was repeated 3 different times. Note that in this case PAPA NONOate did not decrease current.

View larger version (40K): [in this window] [in a new window]   Fig. 9.   Effect of 8-bromo-cGMP (100 µM) on whole cell currents of an A549 cell. Cell was held at a potential of 40 mV. Currents were elicited by applying voltage steps between 100 and + 100 mV in 10-mV steps lasting 900 ms every 10 s. Pipette was filled with SIS, and cell was perfused with SES (see MATERIALS AND METHODS for details). A: control whole cell currents. B: currents after perfusion with 8-bromo-cGMP (100 µM) for 4 min when a steady-state effect was seen. C: cGMP-inhibitable whole currents obtained by digitally subtracting records in B from those in A. D: total and cGMP-sensitive I-V relationships. Values are means ± SE; n = 4 cells.

View larger version (11K): [in this window] [in a new window]   Fig. 10.   Effects of PAPA NONOate on single-channel currents. Single-channel currents were recorded in cell-attached mode at 100 mV. Cell was perfused with depolarizing K+ solution (see MATERIALS AND METHODS) containing 100 µM PAPA NONOate. Note gradual inhibition and complete cessation of channel activity within 40 s from onset of perfusion with PAPA NONOate solution. Results are from a typical experiment that was repeated 6 times in different cells.

Functional relationship with ENaC gene expression. RT-PCR of total RNA obtained from rapidly growing A549 cells was performed with -, -, and -subunit-specific upstream and downstream primers of the hENaC gene (Table 1). These primers amplified the expected size for each subunit message of the hENaC genes (440 bp for , 542 bp for , 633 bp for ; Fig. 11). Two -bands were seen. The origin of the minor lower band has not been identified. A very faint -band was occasionally seen but is not present in Fig. 11. No signal was seen when RT was omitted during the first-strand cDNA synthesis, indicating no cellular DNA contamination in the RNA preparations. The results indicate that A549 cells contain at least - and -ENaC and possibly -ENaC mRNAs.

View larger version (52K): [in this window] [in a new window]   Fig. 11.   RT-PCR showing presence of - and -subunit mRNAs of epithelial Na+ channel (ENaC) genes and a housekeeping gene [hypoxanthine phosphoribosyltransferase (HPRT)] in A549 cells (see MATERIALS AND METHODS for detail). A faint -band was seen occasionally but is not clear here. This is a typical result that was reproduced 3 times.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The presence of amiloride-inhibitable ²²Na⁺ uptake across A549 cells was previously documented (25), and these cells have been used as a model of ion transport for ATII cells. To our knowledge, this is the first study characterizing the amiloride-sensitive pathway in the human A549 cell line. We have demonstrated the presence of the message for the three subunits of the ENaC (, , and ) with the RT-PCR technique and the presence of an amiloride-sensitive channel with a unitary conductance of 8.6 pS by patch-clamp techniques, which is moderately selective for Na⁺ over K⁺ (P_Na/P_K = 6:1). The single channel identified in the cell-attached mode mediates the inward whole cell current recorded in the presence of Na⁺ in the outside medium. Na⁺ channels with a similar unitary conductance were found in the A6 kidney cell line (8.4 pS) (14) and in rat macrophages (10.2 pS) (31)

A variety of channels have been described in freshly isolated and cultured adult and fetal ATII cells (for a review, see Ref. 28). Yue et al. (52) identified a 25-pS channel with P_Na/P_K = 6:1 in inside-out membrane patches of ATII cells in primary culture. This channel was inhibited by both amiloride and ethylisopropylamiloride. The same biophysical properties were seen in channels expressed by a putative Na⁺-channel protein isolated from ATII cells and reconstituted in planer lipid bilayers (41). Jain et al. (17) reported the existence of a 20-pS, cGMP-inhibited, nonselective cation channel in cell-attached patches of ATII cells cultured for 24-96 h. Marunaka et al. (26) found two channels in fetal ATII cells: a nonselective Ca²⁺-activated channel of 25 pS and a highly selective channel of 11.2 pS. A 25-pS, nonselective, Ca²⁺-activated Na⁺ channel has also been reported in cultured ATII cells (6). However, Voilley et al. (48) found a highly selective 4.4-pS channel in outside-out patches of fetal rat ATII cells. ATII cells have been shown to contain rat -ENaC (-rENaC) mRNA (51) and  and  mRNAs, although these last two subunits exist in very low abundance compared with -rENaC (5). Because the biophysical properties of ATII cell Na⁺ channels differ from those of ENaCs, there is a controversy as to whether ENaC is the major channel in fetal and adult alveolar epithelial cells (28). However, a recent study (18) demonstrated that treatment of ATII cells with antisense oligonucleotides targeting -rENaC resulted in decreased density of the nonselective 20-pS channel. Interestingly, inhibition of - or -rENaC had no effect on channel density.

Results shown herein indicate that the amiloride-sensitive whole cell Na⁺ currents in A549 cells are inwardly rectifying and exhibit no time or voltage dependence. When A549 cells were internally dialyzed with high Na⁺, the current became outwardly rectifying (results not shown). Similar modulation of amiloride-sensitive Na⁺ currents was observed in Madin-Darby canine kidney (MDCK) cells transfected with rENaC (15) and, in principle, cells of kidney collecting ducts (8). In contrast, Na⁺ channels in ATII cells have linear I-V relationships (52). This may be due to different stoichiometry of the Na⁺-channel subunit proteins. Oocytes transfected with various combinations of rENaC subunits (30) exhibited amiloride-sensitive Na⁺ channels with various unitary conductances and biophysical properties.

It is known that Na⁺ absorption via the amiloride-sensitive channels located in the apical membrane of epithelia is controlled by the fluctuations of the external Na⁺ concentration (10). However, the exact mechanisms involved have not been elucidated. Data presented herein show that the amiloride-sensitive whole cell current recorded in A549 cells saturate, with a K_m of 36.93 mM. It is unlikely that the saturation of the whole cell Na⁺ current reported here is due to a large modification of the intracellular ionic composition because the cytosol was dialyzed with a pipette solution with a well-defined ionic composition. Palmer and Frindt (37) described saturation of the ENaC at the single-channel level in the collecting duct, with a K_m of 25 mM, a value that is not very much different from our whole cell measurement. Two saturation values were reported for rENaC heterologously expressed in oocytes (30) and MDCK cells (15). The value reported in oocytes (4.9 mM) is different from the value reported here (37 mM) and from rENaC expressed in MDCK cells (24.4 mM) (15) and the collecting duct (25 mM) (37). This discrepancy could be due to the different experimental approaches used because the intracellular ionic composition cannot be controlled in oocytes with only two microelectrodes.

The exact mechanism by which single-channel activity is regulated by Na⁺ was not identified in this study. However, we can assume that single channels will saturate at the same rate as the whole cell current because the macroscopic current equals the NP_o. Direct evidence of ENaC regulation by Na⁺ was reported by Ishikawa et al. (15) in MDCK cells, in mouse submandibular duct cells (20), and by Ismailov et al. (16) using immunopurified bovine renal papillary Na⁺ channels reconstituted in planar lipid bilayers, although in this case, channel activation required the presence of Ca²⁺ in the bath.

The experimental data reported here demonstrate that the amiloride-sensitive Na⁺ channel in the human A549 cell line is inhibited by · NO in concentrations likely to be present in inflamed tissues. · NO has complex biological reactivity, and its physiological effects depend on its concentration and redox state, the nature of the target molecules, and the presence of other free radicals (44). There is evidence to indicate that · NO modulates cation-channel activity by increasing cGMP levels. Light et al. (23) demonstrated the presence of a 28-pS cation channel in rat renal inner medullary collecting duct cells, the activity of which was decreased both by cGMP per se and via cGMP kinase-induced phosphorylation. · NO released from bradykinin-stimulated endothelial cells or spermine NONOate decreased net ²²Na⁺ flux across isolated perfused cortical collecting ducts (47) and decreased Na⁺ short-circuit current across a cortical collecting duct (CCD) cell line while increasing their cGMP content (45, 46). Selective permeabilization of the apical membranes of the CCD cells with nystatin reversed the inhibition of short-circuit current. Based on these findings, it was concluded that · NO inhibited CCD apical Na⁺ channels by increasing their cGMP content (46). Jain et al. (17) reported that S-nitrosoglutathione and S-nitroso-N-acetylpenicillamine increased ATII cell cGMP content and significantly reduced the P_o of a 20-pS nonselective channel in cell-attached patches; pretreatment with a protein kinase G inhibitor prevented the inhibitory effects of S-nitrosoglutathione on this channel; incubation of ATII cells with a cell-permeable analog of cGMP (8-BrcCMP) also decreased the P_o. They concluded that · NO decreased the activity of this channel by activating a cGMP-dependent protein kinase. Our results show that ODQ suppressed the inhibitory effect of exogenous · NO on the amiloride-sensitive channels and that the incubation of A549 cells with 8-BrcGMP mimicked the effects of · NO and provided evidence that the inhibitory effect of · NO is mediated by an increase in cGMP. On the other hand, in a previous study, Guo et al. (11) showed that · NO decreased short-circuit currents across cultured ATII monolayers by inhibiting both the amiloride-sensitive Na⁺ channels and Na⁺-K⁺-ATPase through cGMP-independent mechanisms. Thus it is possible that · NO could modulate ion channels by a variety of mechanisms.

In summary, our results indicate that A549 cells contain mRNAs for -, -, and -hENaC and express Na⁺ channels in their plasma membranes. In agreement with what has been reported in ATII cells, the biophysical properties of these moderately selective Na⁺ channels differ from those expressed when the three ENaC subunits are expressed in oocytes and are regulated by both cAMP and cGMP. Although there are many differences between A549 and ATII cells, these studies help establish the A549 cells as a model to investigate regulation of alveolar epithelial Na⁺ channel by second messengers and reactive species.