Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Under normal conditions, the airway epithelium functions as a highly selective barrier that protects submucosal and deeper inner tissues from inhaled irritants and mediators. Upon encountering appropriate stimuli, the epithelium produces inflammatory mediators, such as interferon- (IFN-) or tumor necrosis factor (TNF), which activate inflammatory cells residing in the airways. In turn, the epithelium itself becomes a target for factors released by the infiltrating inflammatory cells, resulting in further pathophysiological alterations, such as hypersecretion of mucus or a propagation of the inflammatory process.

The biological effects of cytokines are pleiotropic and may affect expression of genes, such as the inducible nitric-oxide synthase (iNOS), that are crucial for the immune response to infection. Although iNOS expression can be activated by individual cytokines, generally human cells require a cytokine mixture for maximal iNOS activity. For example, in human lung epithelial cell line A549, maximal expression of both iNOS mRNA and protein activity occurs after stimulation with a mixture of cytokines composed of INF-, TNF, and interleukin-1 (IL-1) (Asano et al., 1994).

The expression of iNOS exerts both cytotoxic and protective effects (Kroncke et al., 1997). The cytotoxic action of NO includes production of a highly oxidizing agent, peroxynitrite (ONOO), which reacts with various molecular targets, including hydroxyl- and sulfhydryl-containing proteins and may directly cause tissue damage (Beckman and Koppenol, 1996). The protective role of iNOS in inflammation is suggested by experiments with iNOS knockout mice, which indicate that the absence of iNOS leads to more severe inflammatory response than in the wild-type mice in intestinal inflammation (McCafferty et al., 1997).

When cultured with cytokines, epithelial cell lines can be used as an in vitro system to investigate inflammatory processes in the airways. In particular, A549 cells possess many characteristics of alveolar type II alveolar cells (Lieber et al., 1976), and have been frequently used to study induction of iNOS expression by proinflammatory cytokines (Asano et al., 1994; Geller et al., 1995; Berkman et al., 1996). In the present study, we investigated the effects of cytokines on the regulation of whole-cell current in A549 cells. Our data indicate that cytokine-induced iNOS expression leads to generation of ONOO and modification of the NO/cGMP-dependent channel activation pathway.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cells. The A549 cells were obtained from the American Type Culture Collection (Manassas, VA). The cells were grown in Ham's F-12 K medium (Sigma, St. Louis, MO) supplemented with 5% (v/w) fetal bovine serum (HyClone Laboratories, Logan, UT), gentamycin sulfate (60 µg/ml), penicillin-G (60 µg/ml), and streptomycin (100 µg/ml), and maintained in a humidified incubator (95% air and 5% CO₂) at 37°C. Cell monolayers were either harvested for biochemical assays or were subcultured using trypsin (50 µg/ml; Sigma), and used for patch-clamp recordings. To study the effect of cytokine treatment, cells were incubated in a serum-free cell medium for 24 h before adding IL-1, TNF, and IFN- (10 ng/ml of each cytokine), and were incubated with cytokines for up to 24 h before being used.

Patch-Clamp Recordings. Whole-cell current measurements were obtained using the amphotericin-perforated patch-clamp configuration. Pipette electrodes were made from thin-walled borosilicate glass using a P-87 (Sutter Instruments, Novato, CA) electrode puller. Electrode tips were fire polished, resulting in a final resistance of 3 to 7 M. The pipette solution contained 5 mM NaCl, 140 mM KCl, 1 mM MgCl₂, 0.2 mM CaCl₂, 10 mM HEPES (pH 7.4), and 0.5 mM EGTA. The electrode tip was filled by quick dipping (less than 1 s) into pipette solution, which resulted in the movement of filling solution ~400 µm up into the tip. The electrodes were then backfilled with the pipette solution containing 240 µg/ml amphotericin B. The bath solution contained 140 mM NaCl, 5 mM KCl, 0.5 mM MgCl₂, 1 mM CaCl₂, and 10 mM HEPES (pH 7.4). The pipette-offset potentials were compensated before forming a seal, and once a seal was formed, the pipette potential was set to 40 mV and voltage pulses were delivered to monitor incorporation of amphotericin B and changes in the access resistance. At least 10 min was allowed after gigaohm seal formation to achieve equilibration of Cl across the perforated patch. Recording of currents was initiated when the access resistance had stabilized at 8 to 20 M at which time much of the cell capacitance and series resistance (~75%) was compensated electronically. All currents were recorded within 1 to 5 min after the addition of tested reagents. The series resistance of the patch and cell capacitance was measured directly by the compensation circuitry of the patch-clamp amplifier (EPC-7; List Medical, Darmstadt, Germany) and was equal to 9.8 ± 5.2 M and 16.4 ± 8.9 pF (n = 26), respectively. The current-voltage relationship was obtained from the mean current during the final 10 ms of each voltage step. The recordings were acquired and analyzed using patch-clamp software developed by Dr. A. S. French (Department of Physiology and Biophysics, Dalhousie University, Halifax, Nova Scotia, Canada). Currents were filtered at 1 kHz and sampled at 5 kHz. To accommodate for the variations in cell size, currents were normalized as a current density and expressed as pA/pF. All experiments were performed at room temperature.

NOS Assay. The activity of NOS was assayed by measuring the rate of conversion of L-[¹⁴C]arginine to L-[¹⁴C]citrulline as described previously (Radomski et al., 1998). Briefly, the samples were incubated at 37°C with L-[¹⁴C]arginine (Amersham, Indianapolis, IN) in assay buffer containing 50 mM KH₂PO₄, 1 mM MgCl₂, 0.2 mM CaCl₂, 50 mM L-valine, 1 mM L-citrulline, 20 µM L-arginine, 0.1 mM NADPH, 10 µM tetrahydrobiopterin, and 1.5 mM dithiothreitol, in the presence or absence of 1.5 mM N^G-monomethyl-L-arginine. EGTA (2 mM) was used to differentiate between Ca²⁺-dependent and Ca²⁺-independent NOS. After a 20-min incubation the reaction was terminated by dilution and removal of nonreacted L-arginine using AG 50W-X8 resin (Bio-Rad, Mississauga, Ontario, Canada) and the remaining radioactivity counted using a liquid scintillation counter.

Western Blot. Samples (25 µg of protein/lane) were subjected to 7.5% SDS-polyacrylamide gel electrophoresis under reducing conditions. Proteins were transferred onto polyvinylidene difluoride membranes (Schlecher & Schuell, Keene, NH) using a Trans-Blot cell system (Bio-Rad, Montreal, Canada). The endothelial NOS (eNOS), neuronal NOS, and iNOS isoforms were identified using respective polyclonal antibodies (0.1 µg/ml; Santa Cruz Biotechnology, La Jolla, CA). Blots were developed using an ECL kit (Amersham) and the density of bands was quantified using a ScanJet 3c scanner (Hewlett Packard, Boise, ID) and SigmaGel measurement software (Jandel Corporation, San Rafael, CA).

Measurement of cGMP. Cells were seeded in 24-well plates (Becton Dickinson Labware, Franklin Lakes, NJ) at a density of approximately 65,000 cells/well and grown to confluence as described above. Monolayers of control and cytokine-treated cells were incubated for 10 min at 37°C with 100 µM 3-isobutyl-1-methylxanthine (IBMX) in the presence or absence of S-nitroso-glutathione (GSNO) and/or 1H-[1,2,4]oxadiazole [4,3-]quinoxalin-1-one (ODQ). After incubation, 5 mM EGTA was added and the cells were homogenized and used for the measurement of cGMP content by the dual range acetylation enzyme immunoassay system (Amersham). Protein was measured by using the Bio-Rad protein assay kit (Bio-Rad, Hercules, CA) using bovine serum albumin as a standard. The amount of cGMP in picomoles was normalized for milligrams of protein.

Measurement of ONOO. Intracellular generation of ONOO was measured using the 2,7-dihydrodichlorofluorescein-diacetate (DCF-DA) assay (Possel et al., 1997). DCF-DA enters cells passively, where it is enzymatically deacetylated by esterases to the nonfluorescent 2,7-dihydrodichlorofluorescein (DCF-H), and oxidizing molecules such as ONOO, covert DCF-H to the highly fluorescent DCF. For DCF fluorescence calibration DCF-H was prepared by adding 2 ml of 0.1 M NaOH to 0.5 ml DCF-DA (in methanol) and left at room temperature for 30 min. The reaction was stopped by adding 7.5 ml 0.1 M phosphate-buffered saline, producing 50 µM DCF-H solution. DCF-H oxidation by ONOO was performed by adding different amounts of ONOO (0-5 µM) to 1 µM DCF-H solution. DCF fluorescence intensity was measured using PTI spectrofluorimeter (Photon Technology International, Princeton, NJ), set at the excitation and emission wavelengths of 488 and 525 nm, respectively. A linear relationship between the DCF fluorescence and ONOO concentration was observed up to ~5 µM ONOO. Theoretical detection limit (fluorescent signal greater than 3 S.D. over the baseline) was ~400 nM.

Lactate Dehydrogenase (LDH) Release Assay. To assess the cytotoxic potential of cytokine treatment on A549 cells, we measured LDH release into the medium, using an in vitro toxicology assay kit (product no. KR430; Sigma). Confluent monolayers were washed three times with Dulbecco's modified Eagle's medium (without fetal bovine serum and phenol red) and treated with cytokines for 0 to 24 h. Supernatant was collected after 2, 4, 6, 8, 12, and 24 h, and LDH activity in the medium was measured according to the manufacturer's instructions. Results were expressed as percentage of control (nontreated cells).

Chemicals. GSNO, 8-bromo-cGMP (8-Br-cGMP), IBMX, SOD, and catalase were purchased from Sigma; ODQ from Tocris Cookson (St. Louis, MO); DCF-DA from Molecular Probes (Eugene, OR); and ONOO and its negative control (decomposed ONOO) from Alexis Biochemicals (San Diego, CA). GSNO and 8-Br-cGMP were prepared ex tempore in the bath buffer as 10 and 50 mM stock solutions, respectively. Stock solutions of IBMX (100 mM in ethanol) and ODQ [1 mM in 50% (v/v) dimethyl sulfoxide] were stored at 4°C. SOD and catalase were freshly prepared in the bath solution. DCF-DA (5 mM) was prepared in ethanol and stored at 80°C. ONOO and its negative control were added to the cell bath directly from the stock solution (100 mM) to obtain the required final concentrations. Recombinant cytokines IL-1, TNF, and IFN- were purchased from R&D Systems (Minneapolis, MN). Stock solutions of IL-1 (1 µg/ml) and TNF (10 µg/ml) were prepared in sterile phosphate-buffered saline containing 0.1% human serum albumin (Sigma). INF- (10 µg/ml) was prepared in 10 mM acetic acid containing 0.1% human serum albumin. Amphotericin B was prepared as a 50-mg/ml stock solution in dimethyl sulfoxide immediately before use.

Statistical Analysis. Data are expressed as mean ± S.D. Student's t test and ANOVA were performed for paired variates and multiple variates, respectively, and P < .05 was considered statistically significant. The analysis was performed using Origin Technical graphics and Data Analysis software (Microcal Inc., Northampton, MA).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of Cytokines on Activity and Expression of eNOS and iNOS. Under control conditions, A549 cells showed predominantly Ca²⁺-dependent NOS activities, as determined by measurement of citrulline production (Fig. 1A). Cytokine treatment (IL-1, TNF, and IFN-, 10 ng/ml each) caused an increase in Ca²⁺-independent but not Ca²⁺-dependent NOS activities. Similarly, Western blot analysis showed the presence of eNOS and small amounts of iNOS under control conditions (Fig. 1B). Incubation of cells with cytokines for 12 h did not affect eNOS levels, but strongly increased the amount of iNOS protein (Fig. 1B). Western blot analysis did not detect neuronal NOS expression in either control or cytokine-treated cells (n = 3, data not shown), indicating that eNOS accounts for all Ca²⁺-dependent NOS activity in A549 cells. Cytokine treatment had no significant effect on the cell viability for at least 24 h, as determined by measurement of LDH release (P > .05, n = 6).

View larger version (32K): [in this window] [in a new window]   Fig. 1.   A, effect of cytokines on Ca2+-dependent and Ca2+-independent NOS activities in A549 cells. Data are mean ± S.D. (n = 4). *P < .05, Ca2+-independent NOS in control versus cytokine-stimulated cells, at different time intervals. B, Western blot showing the effect of cytokines on eNOS and iNOS expression (n = 4).

Effect of GSNO and Cytokine Treatment on Whole-Cell Current. In 16 different recordings, the basal whole-cell current ranged from 80 to 380 pA at 70 mV, and the average membrane capacitance was 16.4 ± 4.9 pF. To facilitate comparison between experiments, current traces were normalized using the capacitance determined for the individual cell, and the mean current density was 11.2 ± 1.8 pA/pF. After cytokine treatment the mean current density was 10.4 ± 1.3 pA/pF (n = 9), a value that was not significantly different from the control basal current (P > .05).

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Effect of GSNO on whole-cell current in control and cytokine-treated cells. A, representative recording showing activation of the whole-cell current by GSNO under control conditions. B, GSNO has no significant effect on the whole-cell current after cytokine treatment. C, current-voltage relationships corresponding to recordings shown in A and B. Because there was no significant effect of GSNO on the whole-cell current of cells that had been previously treated with cytokines, only one I-V relationship is shown. The data are representative of 16 recordings obtained with control cells and 9 recordings after cytokine treatment. D, voltage protocol used to obtain current-voltage relationships. The same voltage protocol was used in all perforated patch-clamp recordings shown in this study.

Role of NO/cGMP Pathway in Activation of Whole-Cell Current. GSNO increases the whole-cell Cl current via activation of the NO/cGMP-dependent pathway (Kamosinska et al., 1997). Because GSNO had no effect on the whole-cell current after cytokine treatment, we have investigated its effects on [cGMP]_i. GSNO (100 µM) produced ~3-fold increase in [cGMP]_i level (Fig. 3A). This increase was abolished by prior treatment with ODQ (10 µM), a specific inhibitor of the soluble guanylyl cyclase. In cytokine-treated cells, the level of [cGMP]_i was not significantly different that in control cells, but GSNO had no effect on [cGMP]_i (n = 4, P > .05) (Fig. 3A).

View larger version (33K): [in this window] [in a new window]   Fig. 3.   A, effect of GSNO (100 µM) on cGMP generation in control and cytokine-treated cells. The data are mean ± S.D. of four measurements. **P < .01 (ANOVA). The involvement of the soluble guanylyl cyclase in this process has been shown using a specific inhibitor, ODQ (10 µM). B, representative recordings showing the effect of 8-Br-cGMP on the whole-cell current in the control and cytokine-treated cells. C, summary of experiments showing the effect of 8-Br-cGMP on the whole-cell current in the control and cytokine-treated cells (V = 70 mV). The data are mean ± S.D. of five control recordings and seven after cytokine treatment. *P < .05.

Inhibition of iNOS Restores Activation of Whole-Cell Current by GSNO. The cytokines significantly increased the expression and activity of iNOS in A549 cells (Fig. 1). Therefore, we have asked whether iNOS activation could contribute to the alterations of current activation caused by cytokines. To address this question we treated cells with cytokines in the presence of 1400W (10 µM), a specific inhibitor of iNOS (Garvey et al., 1997). Figure 4 shows activation of the whole-cell current by GSNO in cells that had been treated with cytokines and 1400W (Fig. 4). These results suggest that cytokine-induced expression of iNOS affects activation of the whole-cell current by GSNO.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   A, representative recording showing activation of the whole-cell current by GSNO (100 µM) in cells that had been exposed to both cytokines and iNOS inhibitor 1400W (10 µM). B, current-voltage relationships showing activation of the current by GSNO. The data are shown as mean current density ± S.D. (n = 3).

ONOO Modifies Current Activation by GSNO. Some of the cytotoxic effects of NO are mediated by the potent oxidant ONOO that is produced by the near-diffusion limited reaction of NO with O₂ (Beckman and Koppenol, 1996). We assessed the involvement of ONOO in this study by measuring ONOO formation in control and cytokine-treated cells using DCF-H fluorescent method (Possel et al., 1997). Although DCF-H could be oxidized by different oxidants, it has been shown recently that this reagent displayed a much higher sensitivity to ONOO than to NO^·, H₂O₂, or O₂ (Possel et al., 1997). In addition, ONOO has been shown to be the only oxidant that oxidized DCF-H within minutes. We have measured DCF fluorescence in cells treated with cytokines in the presence and absence of 1400W (10 µM) or SOD (100 U/ml)/catalase (1000 U/ml). The relative fluorescence of cytokine-treated and control cells is shown in Fig. 5A. Concentration of ONOO in cytokine-treated cells was 841 ± 59 nM (n = 6). The concentrations of ONOO in control cells, or cells treated in the presence of 1400W or SOD/catalase were at the detection limit and could not be reliably determined.

View larger version (30K): [in this window] [in a new window]   Fig. 5.   A, relative DCF fluorescence measured in control cells and cells treated with cytokines in the presence and absence of 1400W (10 µM) or SOD (100 U/ml)/catalase (1000 U/ml). The data are mean ± S.D. of six measurements in each group. *P < .05, compared with control. B, representative recordings showing that ONOO has no significant effect on the whole-cell current, but prevents subsequent current activation by GSNO. Similar recordings were obtained in three different experiments.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Changes in epithelial permeability have been documented in several clinical conditions associated with inflammation and the release of proinflammatory cytokines (Madara, 1989). Several studies have shown that cytokines could alter the barrier function of epithelial monolayers by changing the permeability and ionic conductance of the paracellular pathway (Marano et al., 1998; Fish et al., 1999). Although the effect of cytokines on the paracellular pathway is clearly established, there is also significant evidence suggesting modification of the transcellular ion permeation pathway by cytokines. Transepithelial measurements have shown that cytokines selectively inhibited the short-circuit current (Zund et al., 1996; Marano et al., 1998; Fish et al., 1999). Similarly, culturing of epithelial cells in the presence of cytokines has been shown to alter the expression of at least two proteins associated with the transcellular ion transport, the cystic fibrosis transmembrane conductance regulator and the Na⁺-K⁺-2Cl cotransporter (Besancon et al., 1994; Colgan et al., 1994). The results of our study expand these observations by showing that activation of the Cl current in A549 cells via NO/cGMP-dependent pathway is altered by cytokines.

How could cytokines alter the mechanisms of agonist activated whole-cell current? Several hypotheses could be formulated to address this question. The simplest explanation is that proinflammatory cytokines exert toxic effects on cell metabolism, making them unable to respond to subsequent stimulation of the NO/cGMP pathway. However, measurements of cell viability using the lactate dehydrogenase assay did not show that cytokine treatment had any significant effect on cells viability for up to 24 h. Therefore, the results reported in this study are not due to the cytotoxic effects of cytokines. It has been previously shown that cytokines, such as TNF, could form pH- and voltage-dependent ion channels when inserted into lipid bilayers (Kagan et al., 1992). However, this intrinsic ion channel-forming activity of TNF would likely increase the whole-cell current, in contrast to the results of this study. TNF has been shown to affect cell membrane potential and to stimulate the Na⁺/H⁺ exchanger (Lee et al., 1992). Although the expression of the Na⁺/H⁺ exchanger in A549 cells has not been studied, the Cl currents in A549 cells are not voltage gated (Kamosinska et al., 1997), suggesting that this mechanism is not likely the explanation of our findings. Alternatively, proinflammatory cytokines are important regulators of gene expression, and this action could significantly affect the regulation of the whole-cell current. In particular, cytokines were shown to regulate the expression of apical Cl channels, resulting in diminished electrolyte secretion (Zund et al., 1996).

Although many different effects could contribute to the alteration of the whole-cell current activation by cytokines, this study has focused on increased iNOS expression and the role of NO in this process. We have shown that cytokine-treated cells generated significantly less cGMP in response to GSNO stimulation than control cells. This effect is surprising because increased expression of iNOS would be expected to increase cGMP production. The reduced generation of cGMP in cytokine-treated cells could be caused by either a decrease in soluble guanylyl cyclase activity, by an increase in cGMP phosphodiesterase activity, or by a reaction of endogenously produced NO with O₂, leading to formation of ONOO. ONOO could further affect the activity of soluble guanylyl cyclase, by oxidizing functional groups crucial for this enzyme function. Interestingly, abnormal function of the NO/cGMP signal transduction pathway was also shown in the rat lungs exposed to lipopolysaccharide (Kurrek et al., 1995).

The fact that 8-Br-cGMP activated the whole-cell current in control but not in cytokine-treated cells would indicate that cytokines caused not only abnormal generation of cGMP, but also altered the upstream events involved in the current activation via NO/cGMP pathway. There are at least three observations that suggest important role for ONOO in this process. First, ONOO levels in the cytokine-treated cells were significantly higher than in control cells. Second, cytokine treatment in the presence of a specific iNOS inhibitor, 1400W, restored activation of the whole-cell current by GSNO. Third, GSNO was unable to activate the whole-cell current in cells pretreated with exogenous ONOO.

How could ONOO alter NO/cGMP-dependent activation of the whole-cell current in A549 cells? Oxidation by ONOO of amino acid residues crucial for ion channel function has been shown to make channels unable to respond to activating agents in several studies. For example, oxidation of sulfhydryl groups by ONOO inhibits epithelial sodium channels (DuVall et al., 1998). Other studies have shown that ONOO and other cellular oxidants are capable of depleting intracellular ATP levels (Lipton, 1999). These agents can interfere with ATP production by inhibition of glycolysis (Le Goffe et al., 1999), activation of poly(ADP-ribose) synthase (Jung et al., 2000), inhibition of mitochondrial electron transport components such as succinate dehydrogenase and NADH dehydrogenase, as well as the mitochondrial ATPase (Radi et al., 1994), or by increasing the degradation of aconitase and other cellular proteins by proteasome (Grune et al., 1998). Although ATP depletion is known to affect the paracellular ion permeation pathway through the regulation of tight junctions (Madara, 1998), it may also affect the transcellular pathway. ONOO has also been shown to increase [Ca²⁺]_i (Packer and Murphy, 1994). However, this effect is not likely to play a significant role in this study because an increase in [Ca²⁺]_i usually activates electrolyte secretion in epithelial cells (Boucher, 1994). Further work will be needed to better delineate the mechanisms of ONOO effects on regulation of ion channels.

Although this study has shown that increased expression of iNOS and the generation of ONOO affects Cl channel function, it is likely that other ion channels would also be affected by excessive generation of NO. Indeed, amiloride-sensitive Na⁺ channels were inhibited by increased production of NO by iNOS (Ding et al., 1998), and by ONOO (DuVall et al., 1998). These studies suggested that under inflammatory conditions NO and/or ONOO would promote lung edema formation by reducing the rate of alveolar fluid absorption. Interestingly, other studies have shown that inhaled NO prevented IL-1-induced edema formation in rat lung (Guidot et al., 1996), and NO has been found beneficial in high-altitude pulmonary edema by improving arterial oxygenation (Scherrer et al., 1996). All these studies show that the effects of NO on epithelial ion transport are complex, and are likely to affect other ion channels. Full understanding of the role of NO in the regulation of ion channels may provide further insight into the mechanism of ion channel function as well as information regarding the pathogenesis of pulmonary edema.