Mechanisms of Cystic Fibrosis Transmembrane Conductance Regulator Activation by S-Nitrosoglutathione*

We investigated the mechanisms by which S-nitrosoglutathione (GSNO) alters cystic fibrosis transmembrane conductance regulator (CFTR) mediated chloride (Cl–) secretion across Calu-3 cells, an extensively used model of human airway gland serous cells. Confluent monolayers of Calu-3 cells, grown under an air-liquid interface, were mounted in Ussing chambers for the measurements of chloride short circuit current (Isc) and trans-epithelial resistance (Rt). Addition of GSNO into the apical compartment of these chambers resulted in significant and sustained increase of Isc with an IC50 of 3.2 ± 1 μm (mean ± 1 S.E.; n = 6). Addition of either glibenclamide or pre-treatment of Calu-3 cells with the soluble guanylate cyclase inhibitor 1H-(1,2,4)-oxadiazolo[4,3-a]quinoxalin-1-one totally prevented the GSNO-induced increase of Isc. Conversely, BAY 41-2272, a sGC stimulator, increased Isc in a dose-response fashion. The GSNO increase of Isc was reversed by addition of two phosphatases (PP2A1, PP2A2) into the apical compartment of Ussing chambers containing Calu-3 monolayers. Oxy-myoglobin (oxy-Mb, 300 μm) added into the apical compartment of Ussing chambers either prior or after GSNO either completely prevented or immediately reversed the increase of Isc. However, smaller concentrations of oxy-Mb (1–10 μm), sufficient to scavenge NO in the medium (as assessed by direct measurement of NO in the Ussing chamber using an ISO-NO meter) decreased Isc partially. Oxy-Mb did not reverse the increase of Isc following addition of GSNO and cysteine (50 μm). These findings indicate that GSNO stimulates Cl secretion via both cGMP-dependent and cGMP-independent mechanisms.

The cystic fibrosis transmembrane conductance regulator (CFTR) 2 is a 1480-amino acid protein and a member of the traffic ATPase family (1). It functions as a cAMP-regulated Cl Ϫ channel (2) and controls other ion conductive pathways, including epithelial Cl Ϫ and Na ϩ channels (3,4). In most epithelial cells, phosphorylation of the CFTR regulatory domain by cAMP-dependent protein kinase A increases its activity as a Cl Ϫ channel (5). Although earlier studies localized CFTR mainly in airway cells, more recent studies provide convincing evidence for the presence of both immunoreactive and functional CFTR at the apical membranes of both fetal and adult ATII cells (6).
There is considerable evidence that reactive oxygen nitrogen species, generated by environmental pollutants and activated inflammatory or airway cells, may alter the properties of a number of ion channels either by direct modification (such as nitrosation, nitration, and oxidation) or via signal transduction pathways (7). Our previous studies indicate that prolonged exposure of confluent monolayers of Calu-3, 16 human bronchial epithelial, or mouse tracheal epithelial cells to physiological levels of NO, generated by the chemical donor DETA NONOate, decrease levels of CFTR in their apical membranes and impaired chloride (Cl Ϫ ) secretion in response to cAMP, a well established regulator of CFTR (8,9). These changes were found to be due to post-translational modification of CFTR by reactive oxygen nitrogen intermediates. Decreased levels and function of normal CFTR may account for some of the cystic fibrosis-like symptoms that occur in chronic inflammatory lung diseases associated with increased NO production.
In several systems, the biological effects of NO on transport proteins have been associated with the formation of nitrosothiols (7). Micromolar concentrations of S-nitrosoglutathione (GSNO) have been detected in the airway fluid of normal subjects, and substantially higher levels were observed in the lungs of patients with pneumonia or during inhalation of 80 ppm of NO (10). Conversely, lower concentrations of GSNO were found in the lower airways of patients with cystic fibrosis (11) and asthma (12). GSNO may directly trans-nitrosate proteins, or release NO via reduction or homolysis. Oxidized glutathione may oxidize or glutathionylate a number of proteins, events known to alter protein function. NO, generated from GSNO, can stimulate cGMP or react with superoxide to form peroxynitrite (ONOO Ϫ ), providing an alternate pathway for directly modulating ion channel function (13). More recently, amino acid transporters have been identified as potential mediators of S-nitrosothiol transport without the intermediate formation of NO, revealing a novel mechanism through which S-nitrosothiol metabolism is controlled (14).
Currently, the effects of GSNO on CFTR function are in dispute: Zaman et al. (15) reported that GSNO increases maturation and function of both wild-type and ⌬F508 CFTR via post-transcriptional regulation, most likely via S-nitrosation. However, Wang et al. (16) reported that oxidized forms of glutathione (including GSNO) decreased CFTR function by oxidative modifications (glutathionylation). However, none of these studies utilized confluent monolayers of airway cells and thus were unable to ascertain the effects of GSNO on vectorial Cl Ϫ transport. Herein we exposed Calu-3 cells, an extensively used model of human airway gland serous cells, grown under an air liquid interface, to physiological concentrations of GSNO for up to 6 h and measured Cl Ϫ short circuit currents (I sc ) and trans-epithelial resistance (R t ). We have chosen to use Calu-3 cells, because they do not contain epithelial sodium channels, which are also affected by NO adducts (17)(18)(19), complicating the interpretation of the results. Our results indicate that addition of GSNO into the apical compartment of Ussing chambers, containing either intact or basolaterally permeabilized monolayers of Calu-3, resulted in sustained increases of I sc without significant changes of R t . These increases were partly due to (i) stimulation of soluble guanylyl cyclase and possible phosphorylation of CFTR via cGMP-dependent mechanisms and (ii) nitrosation of CFTR or other chaperon proteins by reactive intermediates of GSNO and cysteine (S-nitrocysteine), entering the cytoplasm via amino acid transport system L, where they are capable of nitrosating proteins without the intermediate formation of NO (20).

MATERIALS AND METHODS
Cell Culture-Human airway mucosal gland cells Calu-3 cells (HTB-55, ATCC, Manassas, VA) were grown using minimal essential medium (Invitrogen) supplemented with 10% fetal bovine serum, penicillin (100 units/ml), streptomycin (100 g/ml), and maintained in T-75 tissue culture flasks (Costar, Cambridge, MA) at 37°C in a humidified atmosphere of 5% CO 2 in air. Confluent cell layers were washed with phosphate-buffered saline two times and then incubated with 0.05% trypsin and 0.53 mM EDTA in saline for 30 min to lift them from the flasks. Cells were pelleted by centrifugation (125 ϫ g for 10 min), resuspended in the culture medium, and seeded at a density of 10 6 cells/cm 2 onto Type 3470 Costar Transwell inserts (Corning Inc., Corning, NY, 0.45-m pore size, 0.33-cm 2 surface area, 0.4 m), for mounting in Ussing chambers. The culture media on the basolateral side of the filters was replaced every 48 h. After 3 days, the apical fluid was removed, and when monolayers were able to exclude fluid for the apical surface they were cultured for an additional 3-4 days using an air (apical) and fluid (basolateral) interface (21). They were then used for bioelectric studies as described below.
NO Donors and Inhibitors-GSNO (Alexis, San Diego, CA) was dissolved in phosphate-buffered saline containing 10 M diethylenetriaminepentaacetic acid (Acros, Fairlawn, NJ). GSNO concentrations were determined by measuring the absorbance of the stock solution at 333 nm (⑀ 336 nm ϭ 900 mM Ϫ1 ϫ cm Ϫ1 ) with a Beckman DU 7400 spectrophotometer (Beckman Coulter, Fullerton, CA). Oxy-myoglobin (oxy-Mb) was prepared as previously described (22). Briefly, myoglobin from horse heart (Sigma) was dissolved in phosphate-buffer saline containing 10 M diethylenetriaminepentaacetic acid. Met-myoglobin was reduced to oxy-Mb by addition of sodium hydrosulfite and subsequently purified by gel-filtration chromatography using an NAPT 10 column (Amersham Biosciences). Oxy-Mb concentrations were determined by a Beckman DU 7400 spectrophotometer (Beckman Coulter) using the extinction coefficient per heme group ⑀ 582 nm ϭ 14.4 mM Ϫ1 ϫ cm Ϫ1 . Diethylamine NONOate (DEA/NO) (t1 ⁄ 2 ϭ 2 min at 37°C) was purchased from Cayman Chemical (Ann Arbor, MI) and prepared fresh each day according to manufacturer's instructions.
Measurements of Nitrosothiols and NO Release-GSNO (10 M) was added into the apical compartment of Ussing chambers housing confluent monolayer of Calu-3 cells (see below). In a number of cases, NO concentration into the medium was monitored continuously using an NO-specific electrode (ISO-NO, World Precision Instruments, Sarasota, FL). S-Nitrosothiol concentrations in the medium following addition of GSNO were measured by I 3 -reductive chemiluminescence as previously described (23).
To ascertain the contribution of Cl Ϫ and HCO 3 Ϫ ions to the baseline and GSNO, induced I sc , NaCl, or NaHCO 3 were replaced with equimolar amounts of Na ϩ -gluconate or Na-HEPES, respectively. In a third set of ion substitution experiments, both Cl Ϫ and HCO 3 Ϫ were replaced with Na ϩ -gluconate and Na ϩ -HEPES. In experiments conducted in the absence of HCO 3 Ϫ solutions were bubbled with 100% O 2 instead of 95%  and R t were recorded. In some experiments, Calu-3 cell monolayers were pre-treated with ODQ (5 M, Tocris Cookson, St. Louis, MO), a specific inhibitor of soluble guanylate cyclase, every 12 h for 72 h prior to being mounted in the Ussing chambers.
To evaluate effects of GSNO on apical membrane Cl Ϫ conductance, Calu-3 monolayers were mounted in Ussing chambers under shortcircuit conditions in the presence of a basolateral to apical (125:5 mM) Cl Ϫ gradient, and the pore-forming antibiotic amphotericin (10 M) was added into the basolateral compartment. Under these conditions, the basolateral membrane is eliminated as a barrier to the flow of monovalent ions, and I sc provides a direct measure of the apical membrane Cl Ϫ conductance. GSNO (10 M) was then added into the apical compartment followed by glibenclamide or CFTR inh-172.
Measurements of cAMP and cGMP-Filters containing Calu-3 cells were mounted in Ussing chambers; GSNO (10 M), along with cysteine (50 M) or an equivalent amount of vehicle, were added into the apical compartment for either 3 or 30 min. At the end of these times, filters were incubated with 0.1 M HCl plus 1% Triton X-100, and cells were pelleted by centrifugation at 600 ϫ g and kept at Ϫ80°C. cAMP or cGMP levels were measured using the Direct cAMP or cGMP enzymeimmunoassay kits (Assay Designs Inc., Ann Arbor, MI), according to the manufacturer's specifications.
Statistical Analysis-All data were expressed as means Ϯ 1 S.E.; statistical significance among multiple means was assessed by analysis of variance, followed by the Bonferroni modification of the t test, adjusted for multiple comparisons; p values of Ͻ 0.05 were considered significant.  and R t (right axis) were measured as described in the text. Once I sc stabilized, three boluses of glibenclamide (0.1 mM each) were added into the apical compartment. Results are shown of typical experiments. See Table 1 for mean values. Middle panel, CFTR inh -172 inhibits the GSNO-induced I sc in Calu-3 cells. Following addition of GSNO (10 M) into the apical compartment, addition of the specific CFTR inhibitor CFTR inh -172 inhibited I sc in a dose-response fashion. Results are shown for a typical experiment. For mean values see Table 1. ing confluent monolayers of Calu-3 cells, the NO concentration increased only up to ϳ0.3 M and rapidly declined to baseline values (Fig. 1). Similar levels of nitrosothiols were observed in the presence and absence of Calu-3 cells (Fig. 1). These data indicate that, under our experimental conditions, GSNO decomposed to produce free NO, which was then consumed by Calu-3 cells.  Table 1). Transepithelial resistance decreased somewhat, but stabilized to Ͼ600 ⍀ ϫ cm 2 (Fig. 2). Addition of glibenclamide into the apical compartment prior to GSNO decreased I sc and prevented the sustained increase of I sc in response to GSNO (see Table 1). The IC 50 for the GSNO induced increase of I sc was 3.6 Ϯ 1.3 M (n ϭ 6) (Fig. 2). Corresponding amounts of vehicle, nitrite, GSSG, or GSH had no effect on I sc or R t (data not shown).

GSNO Increases Cl
Because Calu-3 cells transport both Cl Ϫ and HCO 3 Ϫ , we investigated the contribution of both ions to the transient and steady response of the I sc to GSNO. As seen in Fig. 3, the majority of the peak increase was due to movement of HCO 3 Ϫ , whereas both Cl Ϫ and HCO 3 Ϫ contributed equally to the steady-state increases. GSNO Increases I sc by Stimulating CFTR-Because Cl Ϫ secretion across Calu-3 cells involves the coordinated entry of Cl Ϫ into the cytoplasm via basolateral transporters and their subsequent secretion across the apical membrane through CFTR pathways, we investigated the effects of GSNO across basolaterally permeabilized monolayers in the presence of secretory Cl Ϫ gradients. As shown in Fig. 4 and Table 2, GSNO increased I sc in a manner similar to what was observed in intact monolayers. Addition of glibenclamide (Fig. 4), oxy-Mb (Fig. 4) or CFTR inh -172 (Table 2) into the apical compartments returned I sc to its pre-GSNO levels.
Role of NO in GSNO-dependent Stimulation of CFTR-As shown in Fig. 5, addition of oxy-Mb (1 M) into an apical chamber decreased the GSNO induced I sc by Ͻ20%. At higher concentrations (300 M), oxy-Mb either completely prevented or immediately decreased the GSNO increase of I sc (Fig. 5 and Table 3). This was not due to nonspecific inhibition of CFTR, because the subsequent addition of 10 M forskolin readily increased I sc to the same extent as in control cells (Fig.  6). These data indicate that NO must be present to stimulate Cl Ϫ secretion. Indeed, as shown in Fig. 6 addition of diethylamine NONOate, a NO donor with a t1 ⁄ 2 of Ͻ3 min, caused only a transient increase of I sc .
Mechanisms of GSNO Action: Activation of sGC and Increased cGMP Levels-Pre-treatment of Calu-3 cells with the sGC inhibitor ODQ (5 M twice a day for 3 days) totally prevented the GSNO-induced increase in I sc

TABLE 1 GSNO (10 M) increases I sc across confluent monolayers of Calu-3 cells mounted in Ussing chambers
Calu-3 cells were grown on semi-permeable supports using an air-liquid interface and mounted in Ussing chambers. Nitrosoglutathione (GSNO, 10 M), glibenclamide (Glib., 0.3 mM), or CFTR inh .-172 (20 M) was added into the apical compartments (normal Ringer's) of the Ussing chambers; ouabain (Ouab., 1 mM) was added into the basolateral compartment. ⌬ Peak and ⌬ Plat. ϭ difference in peak and plateau values of I sc after addition of GSNO (10 M) from the corresponding baseline value (prior to GSNO addition); ⌬ inh ϭ differences in plateau values of I sc before and after addition of the indicated inhibitors. Values are mean I sc Ϯ 1 S.E.; n ϭ number of monolayers.  ( Fig. 7 and Table 3), indicating that sGC was involved in the GSNO-induced increase of I sc. Subsequent addition of forskolin or cpt-cAMP into the apical compartment of an Ussing chamber containing ODQ-treated cells resulted in a significant increase of I sc (Fig. 7), indicating that the cells remained viable and able to respond to agents that increase cellular cAMP content. Additional evidence for the involvement of sGC is rendered by the fact that BAY 41-2272, a non-NO-based direct stimulator of sGC (24), also increased I sc in a dose-response fashion (Fig. 7). Steady-state concentrations of cGMP levels were increased considerably, both at 3 and 30 min post GSNO addition (Fig. 8). In contrast, cAMP values remained at baseline levels (Fig. 5E). Addition of oxy-Mb at 30 min post GSNO decreased GMP to control levels within 2 min, corresponding to the decrease of I sc seen in Fig. 6.

Baseline
In subsequent experiments, we investigated the possibility that the observed increases of I sc secondary to addition of GSNO were cGMPmediated. Addition of cpt-cGMP into the apical compartment

TABLE 2 GSNO(10 M) increases Cl ؊ secretion across basolaterally permeabilized monolayers
Calu-3 cells were grown on confluent monolayers and mounted in Ussing chambers containing 5 mM in the apical and 125 mM Cl Ϫ in the basolateral compartment. Amphotericin B (10 M) was added into the basolateral compartment followed 10 -20 min later by ouabain (1 mM). Once I sc and R t had stabilized, GSNO (10 M) was added into the apical compartment. Subsequently, glibenclamide (glib., 0.3 mM), CFTR inh -172 (20 M), or Oxy-myoglobin (oxy-Mb, 0.3 mM) were added into the apical compartment, and changes in I sc and R t (data not shown) were recorded. ⌬GSNO Plat. ϭ difference in I sc plateau values before and after addition of GSNO; ⌬inh. ϭ difference in I sc plateau values before and after addition of the glibenclamide, CFTR inh -172, or oxy-Mb as noted (in the presence of GSNO). Values are mean I sc Ϯ 1 S.E.; n ϭ number of monolayers.   Table 2.

Baseline
increased I sc (Fig. 9) No additional increase of I sc was noted after subsequent addition of GSNO and/or forskolin, indicating that maximum stimulation of CFTR was achieved.
To assess a putative mechanism responsible for the activation of CFTR, we added two CFTR-specific phosphatases (PP2A1 and PP2A2, 2 milli-units/ml each) into the apical compartment of Ussing chambers containing Calu-3 cells, following the addition of GSNO. As seen in Fig. 10 PP2A1 and PP2A2, two phosphatases known to act on CFTR but not equivalent vehicle

TABLE 3 Pretreatment of confluent Calu-3 monolayers with ODQ or Oxy-Mb prevented the GSNO (10 M) increase of Cl ؊ secretion
Calu-3 cells were grown on confluent monolayers and mounted in Ussing chambers. Oxy-myoglobin (0.3 mM) was added into the apical compartment for at least 60 min. Once I sc and R t had stabilized, GSNO (10 M) was added into the apical compartment. Calu-3 cells were incubated with 5 M ODQ for 72 h (the solution was changed every 12 h). ODQ (5 M) was also added into the apical compartment of the Ussing chambers prior to the addition of GSNO.   Table 3. solutions, decreased the GSNO-induced I sc, shortly after being added into the apical compartment. Subsequent addition of GSNO increased the I sc , indicating that the action of phosphatases was not due to modulation of NO release from GSNO (data not shown). These findings are consistent with NO-induced phosphorylation of CFTR and concomitant increase of its activity. Activation of CFTR via NO-independent Mechanisms-Previous studies have shown that S-nitrocysteine may cross the plasma membranes via amino acid transport system L and trans-nitrosate intracel-lular proteins without the intermediate formation of NO (20). We thus added cysteine (50 M) into the apical compartment of the Ussing chamber following GSNO. Addition of cysteine resulted in a transient increase of NO (peak value: 1 M), which lasted ϳ2 min (data not shown). However, addition of cysteine after GSNO did not increase I sc , perhaps because I sc was already maximally stimulated by GSNO. However, in this case, addition of 0.3 mM oxy-Mb, failed to decrease the GSNO-induced I sc (Fig. 11).

DISCUSSION
The main findings of this study are that apical GSNO, in concentrations likely to be present in the normal and injured airway and alveolar fluids (1-10 M) (10), increases Cl Ϫ secretion (as assessed by changes of I sc ) across confluent monolayers of Calu-3 cells, by both cGMP-dependent and -independent mechanisms. Our data indicate that GSNO  Table 3 for mean values. Middle panel, the sGC activator BAY 41-2272 stimulates I sc across Calu-3 cells. Increasing concentrations of BAY 41-2272 (0.01-10 M) or an equivalent amount of vehicle were added into the apical compartment till new steady-state values of I sc were achieved. Glibenclamide (0.3 mM) was also added apically at the indicated time. Results of a typical experiment are shown. Lower panel, mean values of ⌬I sc Ϯ 1 S.E. of (calculated as the difference in I sc before and after applications of BAY or vehicle). n ϭ 12 for BAY and 7 for Veh. (Me 2 SO); *, p Ͻ 0.05 compared with the corresponding vehicle control (t-tests). releases sufficient NO, to activate sGC, which increases cGMP and possibly phosphorylates either CFTR or other chaperon proteins important in its function as a Cl Ϫ channel. Indeed, the time course of I sc following apical addition of GSNO paralleled the NO formation profiles in the Ussing chamber as measured by an NO electrode (an immediate transient followed by a sustained increase). Also, addition of the short acting NO-donor diethylamine NONOate resulted in only a transient increase of I sc . A role for NO is further supported by the observation that oxy-Mb, the ferrous heme group of which reacts rapidly (k of ϳ10 7 M Ϫ1 s Ϫ1 ) with NO oxidizing it to nitrate, inhibited and prevented both the initial and sustained changes of I sc .
Experiments across basolaterally permeabilized monolayers show that GSNO added into the apical compartment activates apical pathways of Cl Ϫ secretion. The fact that the induced I sc was inhibited both by glibenclamide (a K ATP and CFTR blocker) and CFTR-inh-172, a cellpermeable thiazolidinone compound that acts as a potent, reversible, rapid, and voltage-independent inhibitor of CFTR (25), strongly suggests activation of CFTR by GSNO.
Calu-3 cells also contain Ca 2ϩ -activated Cl Ϫ channels; however, addition of the Ca 2ϩ ionophore ionomycin into the apical compartment of Ussing chambers containing Calu-3 cells caused only a transient stimulation of I sc lasting no more than a few minutes (data not shown); thus activation of Ca ϩ2 -activated Cl Ϫ channels, even if it occurred, cannot account for the sustained GSNO-induced increase of I sc observed in these studies. Also, additional studies have reported that other NO donors such as nitroprusside or S-nitroso-N-acetylpenicillamine cause a dose-dependent decrease of both apical and basolateral Na ϩ -K ϩ -2Cl Ϫ co-transporters in a renal epithelial cell line (26). Thus, it is unlikely that the observed increase of I sc by GSNO was due to activation of basolateral Cl Ϫ transporters.
It is widely accepted that phosphorylation of the regulatory domain of CFTR by protein kinase A, followed by binding and hydrolysis of ATP at its nucleotide binding domain, is the main event leading to its activation (27). Agents that increase cAMP increase vectorial Cl Ϫ secretion across both native secretory epithelial (6,9,28,29) and heterologous expressing CFTR cells (30). However, contrary to what has been reported in endothelial cells (31), GSNO did not increase cAMP levels in Calu-3  cells; furthermore, ODQ, a well known specific inhibitor of sGC, prevented the GSNO increase of I sc across Calu-3 cells. These findings indicate that activation of sGC/cGMP were responsible for the noted effects in our experiments.
In previous studies, Berger et al. (30) showed the type Ia cGMP dependent protein kinase phosphorylated CFTR in vitro, and PP2A dephosphorylated CFTR. However, in contrast to protein kinase A, phosphorylation of CFTR by cGKI failed to activate Cl Ϫ secretion across human airway cells. Another membrane isoform of cGK (cGKII) has been isolated from intestinal epithelium by de Jong et al. (32), which is not present in airway epithelial cells. In subsequent studies it was shown that, although both isoforms phosphorylate CFTR, only cGKII activated CFTR channels in membrane patches and Cl Ϫ currents across cells patched in the whole cell mode (33). Furthermore, transfection of cells expressing CFTR with cGKII, but not cGKI, allowed cGMP-activation of CFTR (34). Thus it is likely that Calu-3 but not mouse tracheal epithelial cells express cGKII allowing phosphorylation and activation of CFTR (or critical chaperon proteins) by agents that increase cGMP.
We also failed to see a significant increase of Cl Ϫ secretion across confluent monolayers of mouse tracheal epithelial cells by GSNO and cGMP. On the other hand, addition of 3-morpholinosydnominine (SIN-1, 1 mM), a peroxynitrite (ONOO Ϫ ) donor, into the apical compartment of amiloride-treated monolayers of mouse tracheal epithelial cells resulted in a transient increase of I sc . Forskolin (10 M) augmented the SIN-1 effect when added at the peak of the SIN-1 response but not when I sc had returned to its baseline value. Perfusion of mouse tracheal epithelial cells with SIN-1 also increased whole cell Cl Ϫ currents 4-fold and the open probability of CFTR-type single channel currents from 0.041 to 0.92 in a transient fashion (21). The mechanism by which ONOO Ϫ , but not NO, stimulates Cl Ϫ secretion across mouse tracheal epithelial cells is not clear at this time.
Further insights into the cGMP-independent mechanism of GSNO activation of Cl Ϫ secretion were provided by adding both GSNO and L-cysteine in the apical compartments of Ussing chambers containing Calu-3 cells. Recent studies (14,20) have shown that L-cysteine provides a substrate for trans-nitrosation from GSNO, with the ensuing product S-nitrosocysteine readily being transported into cells via the amino acid transport system L (L-AT). Zhang and Hogg (20) showed an almost 50-fold increase in S-nitrosothiol content of RAW 264.7 cells incubated with GSNO and L-cysteine and significant inhibition of this process by BCH and L-leucine, inhibitors of the L-AT transporter. CFTR contains 18 intracellular cysteines (16) that may become nitrosated by these species. Post-translational modification (oxidation or nitrosation) of ⌬F508CFTR by GSNO has been associated with increased maturation (15). In our experiments, L-cysteine also caused a transient release of NO from GSNO as measured by an NO Ϫ electrode in the presence of cells, consistent with thiol-mediated 1-electron reduction of GSNO to NO, as has been previously observed with a number of nitrosothiols, including S-nitroso-N-acetylpenicillamine (14,36). However, no additional increase of I sc was noted, most likely because CFTR was already maximally stimulated by NO.
Oxy-Mb, even in high concentrations (0.4 mM), failed to modulate the GSNO plus cysteine-induced increase of I sc , suggesting that either formation of NO within Calu-3 cells prevents scavenging by oxy-Mb or intracellular S-nitrosothiols are stimulating Cl Ϫ currents independent of NO formation. On the basis that NO is relatively diffusible, it is unlikely that NO formation within cells can preclude scavenging by an efficient trap present outside the cells, a fact consistent with the hypertensive effects of cell-free hemoglobin (37,38). More importantly, in the presence of L-cysteine, GSNO stimulated cGMP formation declined to baseline upon addition of oxy-Mb. The fact that this was not associated with inhibition of I sc strongly suggests that, in the presence of L-cysteine, GSNO stimulates I sc independent of NO formation, and sGC stimulation likely occurs via intracellular S-nitrosothiol-dependent effects that could include direct S-nitrosation or S-glutathioylation of critical protein thiols. Ongoing experiments are evaluating the potential targets of S-nitrosothiols that could modulate I sc in Calu-3 cells.
As mentioned above, post-translational modifications of both wildtype and mutant CFTR has been reported in a number of studies. Most recently Wang et al. (16) reported that oxidized forms of glutathione (such as GSNO) decreased CFTR currents in membrane-excised patches from Baby hamster kidney cells, stably transfected with CFTR (BHK-CFTR) within a few minutes by glutathionylating Cys-1344. Based on the fact that S-nitroso-N-acetylpenicillamine did not decrease CFTR function, they concluded that the effects of GSNO were not due to the release of NO. However, NO release by S-nitroso-N-acetylpenicillamine involves trans-nitrosation reactions requiring the presence of significant levels of cysteine (36). Contrary to their observations, in our experiments oxidized glutathione (10 M) did not alter Cl Ϫ secretion across Calu-3 cells. Zaman and Palmer (15,39) reported that exposure of a variety of cells to GSNO causes an increase in mature CFTR after 4 h. This change was prevented by dithiothreitol and thus was attributed to oxidative modification of CFTR. In our studies we found no change in surface expression of CFTR of Calu-3 exposed to 50 M GSNO for up to 6 h (35). However, prolonged exposure of Calu-3 cells to DETA NONOate (100 M for 48 -72 h) resulted in nitration of CFTR, decreased apical CFTR levels, and cAMP-regulated Cl channel current (9).
In summary, our results provide direct evidence that GSNO stimulates Cl Ϫ secretion across Calu-3 cells by both cGMP-dependent and -independent mechanisms. Nitrosothiols may be contributing to basal activation of CFTR in vivo.