Published ahead of print on November 15, 2007, doi:10.1164/rccm.200703-344OC
© 2008 American Thoracic Society doi: 10.1164/rccm.200703-344OC
Preclinical Evidence that Sildenafil and Vardenafil Activate Chloride Transport in Cystic FibrosisDepartments of 1 Clinical Chemistry, 2 Cell Physiology, and 3 Pediatric Pulmonology, Université Catholique de Louvain, Brussels, Belgium; and 4 Department of Biochemistry, Erasmus University Medical Center, Rotterdam, The Netherlands Correspondence and requests for reprints should be addressed to Teresinha Leal, M.D., Ph.D., Department of Clinical Chemistry, St. Luc University Hospital, 10 Av. Hippocrate, B-1200 Brussels, Belgium. E-mail: teresinha.leal{at}clin.ucl.ac.be
Rationale: Sildenafil has been implicated in the activation of cystic fibrosis transmembrane conductance regulator (CFTR) protein. The effect was observed in vitro and in the presence of doses roughly 300 times larger than those commonly used for treating erectile dysfunction. Objectives: To evaluate in vivo the therapeutic efficacy of clinical doses of sildenafil and vardenafil, two clinically approved phosphodiesterase 5 inhibitors, for activating ion transport in cystic fibrosis. Methods: We used transepithelial potential difference in vivo across the nasal mucosa as a measure of sodium and chloride transport. The effect of a single intraperitoneal injection of sildenafil (0.7 mg/kg) or vardenafil (0.14 mg/kg) was investigated in F508del, cftr knockout and normal homozygous mice. Measurements and Main Results: In F508del mice, but not in cftr knockout mice, the chloride conductance, evaluated by perfusing the nasal mucosa with a chloride-free solution in the presence of amiloride and with forskolin, was corrected 1 hour after sildenafil administration. A more prolonged effect, persisting for at least 24 hours, was observed with vardenafil. The forskolin response was increased after sildenafil and vardenafil in both normal and F508del mutant animals. In F508del mice, the chloride conductance in the presence of 200 µM 4-4'-diisothiocyanostilbene-2,2'-disulphonic acid, an inhibitor of alternative chloride channels, was much higher after sildenafil injection than after placebo treatment. No effect on the sodium conductance was detected in any group of animals. Conclusions: Our results provide preclinical evidence that both drugs stimulate chloride transport activity of F508del-CFTR protein.
Key Words: cystic fibrosis • phosphodiesterase inhibitors • sildenafil • chloride transport • nasal potential difference
Cystic fibrosis (CF) is the most common, life-threatening, recessively inherited disease in white populations, resulting from mutations affecting the CF transmembrane conductance regulator (CFTR) gene (1, 2). These mutations decrease the expression or the activity of the CFTR protein through a variety of mechanisms (3, 4). The most common CF mutation results in deletion of a single phenylalanine residue at position 508 (F508del), and causes defective synthesis and folding of the mutant protein that fails to escape from the endoplasmic reticulum and reach the apical membrane of many epithelial cell types (5). CFTR protein has been shown to be an ohmic, small conductance chloride channel, also functioning as a regulator of other transport mechanisms, most notably down-regulating the epithelial sodium channel (ENaC) (6, 7). CF epithelia are characterized by defective electrolyte transport—namely, reduced chloride conductance and increased sodium conductance, which can be assessed in vivo by measuring transepithelial nasal potential difference (PD) (8, 9). Cyclic nucleotide phosphodiesterases (PDEs) are enzymes that regulate the cellular levels of the second messengers, cAMP and/or cGMP, by controlling their rates of degradation. Eleven families of PDEs with varying selectivities for cAMP and/or cGMP have been identified in mammalian tissues. Within these families, over 50 isoforms are expressed either as products of different genes or as products of the same gene through alternative splicing (10–12). PDE5 is highly specific for cGMP and is involved in regulation of the intracellular concentration of cGMP in various tissues (13). It was recently reported that sildenafil, a PDE5 inhibitor, increases the expression of F508del-CFTR protein in nasal epithelial cells harvested from patients with CF (14). The effect, observed in the presence of 150 µM sildenafil, a dose roughly 300 times larger than that usually used in male erectile dysfunction (15), was associated with an increase in the activity of chloride transport. We hypothesized that in vivo treatment with a clinical dose of PDE5 inhibitors activates ion transport in respiratory epithelial cells in CF. To verify this hypothesis, we examined the effect of treatment with two PDE5 inhibitors, sildenafil and vardenafil, on the transepithelial sodium and chloride movement across the nasal mucosa of CF and wild-type mice. Some of these results have been previously reported in the form of abstracts (16, 17).
Animal Model CF mice homozygous for the F508del mutation in the 129/FVB outbred background (18) and their normal homozygous wild-type littermates were studied. Two strains (congenic FVB and mixed-background C57Bl/6/129) of cftr knockout mice were also investigated. Mouse age ranged from 3 to 4 months and weight ranged from 20 to 30 g. The animals were housed at the Animal Care Facility of the University of Louvain following recommendations of the Federation of European Laboratory Animal Science Associations (19). These studies and procedures were approved by the local ethics committee for animal welfare and conformed to the European Community regulations for animal use in research (CEE 86/609).
PDE Inhibitor Treatment
Nasal PD Measurements
Statistical Analysis
Baseline and Stimulated Nasal PD Values in Nontreated F508del, cftr Knockout, and Control Mice As illustrated in representative tracings (Figures 1A–1C), both CF animal models showed ion transport abnormalities similar to those observed in patients with CF (8, 9, 22–24): a large negative PD in baseline conditions; an increased depolarizing response to amiloride, which blocks the ENaC; and a reduced response to perfusion with chloride-free solution containing amiloride plus forskolin, a cAMP agonist.
Maximal baseline values of nasal PD were significantly different between F508del mice and their corresponding control littermates, but were identical between F508del and cftr knockout mice (P = 0.29). Mean values of PD changes in response to chloride-free solution in the presence of amiloride and with forskolin (   Cl), were reduced by almost half in F508del mice, and by 75% in cftr knockout mice, as compared with the corresponding values measured in control mice (Figure 2).
Effect of PDE Inhibitors on Nasal PD Parameters in F508del and Wild-Type Mice Treatment with PDE5 inhibitors was well tolerated and no adverse effect was observed.
Sildenafil treatment.
Sildenafil did not induce any noticeable effect on the amiloride response in either wild-type or F508del mice, either after 1 hour or after 24 hours of treatment with the PDE inhibitor as compared with placebo-treated animals (Figure 4A).
A significant effect on chloride transport was detected 1 hour after sildenafil treatment in the F508del mouse group as compared with the corresponding placebo group (Figure 4B). At this time point, the mean (±SEM) values obtained 1 hour after sildenafil were comparable to those recorded in placebo-treated wild-type mice (–14.0 ± 1.5 mV and –14.7 ± 1.0 mV, respectively; P = 0.70). At 24 hours after sildenafil treatment, no effect on chloride transport was detected in the F508del mouse group (Figure 4B). No significant effect on the chloride transport was detected in the control group, either 1 or 24 hours after treatment (Figure 4B).
Vardenafil treatment.
The duration of chloride transport activation in F508del mice was more prolonged after vardenafil administration. Indeed, at 1 hour after treatment, the magnitude of PD changes was similar for both PDE inhibitors (–14.3 ± 2.0 mV and –14.0 ± 1.5 mV for vardenafil and sildenafil, respectively; P = 0.90). Nevertheless, at 24 hours, chloride transport activation persisted after vardenafil administration (Figure 6B), whereas it did not after sildenafil administration (Figure 4B). Accordingly, mean PD changes in response to chloride-free solution plus forskolin perfusion in F508del mice 24 hours after vardenafil and sildenafil reached –17.5 (±1.8 mV) and –8.4 (±1.4 mV), respectively (P = 0.009). In the wild-type mouse group, no significant effect on chloride transport was observed 24 hours after vardenafil administration. However, a significant increase in chloride conductance was detected at 1 hour after vardenafil administration (Figure 6B).
Influence of PDE5 Inhibitors on the Separate Components of Chloride Transport In wild-type mice (Figure 7A), a significant effect of PDE5 inhibitors on the chloride-free response, representing the major component of the global chloride conductance, was only observed 1 hour after vardenafil administration as compared with placebo-treated animals. However, the forskolin response, although representing the minor component of the global chloride conductance, appeared to be significantly increased at both 1 and 24 hours after treatment with both PDE5 inhibitors as compared with placebo-treated animals (Figure 7A).
In the F508del mouse group (Figure 7B), the chloride-free response was significantly increased after treatment with PDE5 inhibitors as compared with the placebo group, except for that measured 24 hours after sildenafil treatment. At this time point, the contribution of the chloride-free response was almost equivalent to that of the forskolin response. No significant difference in forskolin response was observed between placebo wild-type and F508del mice (P = 0.08). After treatment with PDE inhibitors, significant effects were reached 1 hour after vardenafil and 24 hours after sildenafil when compared with placebo-treated animals (Figure 7A).
Influence of Alternative Chloride Transport Pathways on the Effect of PDE5 Inhibitors
In placebo F508del mice (Figure 8A), chloride transport in the presence of DIDS was changed from –7.5 (±1.0) to –2.0 (±0.0) mV, indicating that alternative chloride transport pathways may contribute to the largest part of the residual chloride conductance in the nasal mucosa of F508del mice in the129/FVB background. Alternatively, apically localized F508del–CFTR, indicated by DIDS-insensitive PD changes, corresponded to one-quarter of the residual chloride conductance. In sildenafil-treated F508del animals (Figure 8B), DIDS-sensitive PD changes were of the same magnitude in placebo- and sildenafil-treated animals. However, PD changes after DIDS were at least fourfold higher in sildenafil-treated than in placebo-treated animals (–9.5 ± 0.29 vs. –2.0 ± 0.0 mV; P < 0.0001). This finding indicates that the effect of sildenafil in F508del mice does not principally depend on alternative chloride transport pathways. Interestingly, the magnitude of the inhibiting effect of DIDS was similar when the inhibitor was perfused after amiloride or after chloride-free solution 1 hour after either placebo or sildenafil administration (data not shown).
Effect of PDE5 Inhibitors on Nasal PD in cftr Knockout Mice
The present study was designed to test the hypothesis that the PDE5 inhibitors, sildenafil and vardenafil, administered in vivo at clinical doses, stimulate ion transport in CF mice. Because both drugs are in clinical use, preclinical studies using animal models may be of great relevance for assessing the potential therapeutic efficacy of the drugs before they can be evaluated in randomized, controlled trials in patients with CF. The F508del mouse model used in our work (18) shows clear signs of residual CFTR activity; in addition, it displays nearly normal mRNA expression levels of the mutant CFTR protein, as well as high survival rates as compared with the two other F508del mouse models (25, 26). These three properties make it a useful in vivo system to assist in studying novel therapeutic strategies aiming at activating the protein. PDE activity is found in every cell in the body, although there is distinct cellular and subcellular distribution of the 11 mammalian isoforms. PDE5, originally named cGMP-PDE because of its high affinity for cGMP (13), is known to be abundant in most smooth muscle cells (10–12). However, its presence has also been characterized in lungs (27, 28) and in a variety of epithelial cells, such as those of pancreatic ducts, proximal tubules, and collecting renal ducts (29). Several compounds that potently inhibit PDE5 have been synthesized recently, and three of these are currently in clinical use for treatment of male erectile dysfunction. Sildenafil, the first compound of this class to be marketed, provides well-tolerated pharmacotherapy for erectile dysfunction (15, 30). The drug is also currently indicated in the clinical treatment of ailments related to smooth muscle tissues, such as pulmonary hypertension (31). Two newer selective PDE5 inhibitors, vardenafil and taladafil, have the same mechanism of action, as they are specific for binding to the catalytic site of the enzyme catalyzing the breakdown of cGMP to 5'-GMP, but have differing potencies (32). As taladafil is practically insoluble in water, we choose to investigate in vivo effects of sildenafil and vardenafil in CF and wild-type mice. It is generally agreed that, in intact cells, the classic cAMP–protein kinase A pathway is the major mechanism regulating CFTR activity (33). Pharmacological agents modulating cellular cAMP levels have been reported to activate wild-type and mutant CFTR molecules in epithelial cell lines and in CF mice, although demonstration of chloride secretion in human subjects has been variable (25, 34–37). A nonselective PDE inhibitor, isobutylmethylxanthine, has been shown to activate chloride transport in Xenopus oocytes expressing F508del-CFTR (36); however, no effect could be evidenced in CF airway epithelial cells expressing F508del-CFTR protein (35). Studies conducted with isotype-specific PDE inhibitors, such as milrinone, a PDE3 inhibitor displaying a high affinity for cAMP (dissociation constant [Kd], 0.2 µM), and rolipram, a PDE4 inhibitor also displaying a great affinity for cAMP (Kd, 4 µM), reported a CFTR-dependent chloride transport in polarized, CFTR-expressing airway epithelial cells (37). Milrinone increased nasal PD in CF mice, but not in patients with CF (34). Because the rises in cAMP concentration produced by these PDE inhibitors do not parallel the resulting increases in chloride conductance (37), it has been postulated that PDE inhibitors may affect CFTR through cAMP-independent mechanisms (38). The interest in PDE5 inhibitors in CF followed the description that PDE5 inhibition corrected defective mucin secretion response to the β-agonist, isoproterenol, in submandibular acinar cells in which CFTR activity was inhibited by antibodies directed against the protein to simulate a CF-like phenotype (39). Correction of the CFTR-mediated mucin secretion defect was shown to involve elevation of cellular levels of cGMP (39). More recently, Dormer and colleagues (14) reported that exposure of nasal epithelial cells harvested from patients with CF to 150 µM sildenafil resulted in recruitment of F508del-CFTR to the apical cell membrane and stimulation of chloride transport activity. We have shown here that treatment with a clinical dose of 0.7 mg/kg sildenafil, resulting in a maximal circulating concentration lower than 1 µM (data from Pfizer: www.viagra.com), restored the chloride transport function in the respiratory epithelium of F508del mice. The response was characterized by an early and complete correction of chloride conductance when assessed by the nasal PD. No effect of sildenafil was observed in either wild-type or cftr knockout mice.
A similar early activating effect on chloride conductance was observed in F508del mutant mice after treatment with a clinical dose of vardenafil. Distinct biochemical properties, chemical structures, and pharmacokinetic considerations may contribute to the longer-lasting effect of vardenafil. Indeed, vardenafil is known to be more potent and more selective than sildenafil at inhibiting PDE5 (30, 32). Relative potency studies, determined by 50% inhibitory concentration measurements and isotherm Kd binding to the PDE5 catalytic site, indicated a potency ratio of 1:13 to 1:41 for sildenafil:vardenafil, respectively (32). The higher biochemical potency of vardenafil, as compared with sildenafil, has been related to small structural differences within the double rings of the two compounds. Indeed, a nitrogen atom is present in the 7-position of the double ring of sildenafil, whereas vardenafil contains a nitrogen atom in the 5-position (32). Slower hepatic metabolization of drugs, as indicated by a longer elimination half-life of vardenafil (4–5 h) as compared with that of sildenafil ( The mechanism of action through which sildenafil and vardenafil activate F508del-CFTR chloride channel function is not clear. In human nasal epithelial cells, suprapharmacological dosages of sildenafil (up to 1 mM) resulted in rapid recruitment (2 h at 37°C) of F508del-CFTR to the apical membrane and the appearance of functional activity (14). A similar increase in CFTR function, but occurring at much lower concentrations (10 nM–10 µM) of sildenafil or structural analogs, was recently reported for an F508del-CFTR–expressing epithelial cell line, BHK (40). Chloride conductance after perfusion with DIDS, a well-described inhibitor of non-CFTR chloride channels (41) and electrogenic anion exchangers (42) in murine epithelia was much higher in F508del mice in the presence of sildenafil as compared with placebo. This observation, in addition to the finding that the effect was lost in cftr knockout mice, indicates that action of PDE5 inhibitors on nasal PD involves F508del-CFTR and not a CFTR bypass channel. Our data support the view that the short-term action of sildenafil on nasal PD is due, at least in part, to an acute activation of F508del-CFTR channel activity rather than to a correction of the trafficking defect of the mutant protein. Therefore cGMP selective PDE inhibitors should be included in the category of F508del-CFTR activity potentiators, such as genistein and NS004 (43, 44). Similar to genistein (44), PDE5 inhibitors could activate wild-type CFTR protein. However, whether this action depends on PDE5 inhibition and cGMP generation (45), or is due to a more direct interaction of the compounds with the F508del-CFTR protein itself, has not been clarified in these functional studies. In keeping with the assumption that cGMP blocks amiloride-sensitive sodium channels, including ENaC (46, 47), it has been reported that treatment of CF cells with cGMP PDE inhibitors, or with guanylate cyclase agonists, can reverse hyperacidification of endosomes, which has been ascribed to up-regulated sodium transport (48, 49). Indeed, endosomal pH was corrected from 6.1 to around 7.0 after treatment of CF cells with either 200 µM isobutylmethylxanthine or 300 nM sildenafil, but not with 10 µM rolipram, indicating a role for cGMP and not cAMP (48, 49). In our study, however, we did not observe any effect of PDE5 inhibitors on electrogenic sodium transport in the nasal mucosa, suggesting that ENaC inhibition by cGMP, if it occurs in mouse nasal epithelial cells, might be restricted to the endosomal compartment. The lack of effect of PDE5 inhibitors on amiloride response, reflecting ENaC-mediated electrogenic sodium transport, in F508del nasal epithelium under conditions in which the compound succeeded in activating CFTR-mediated chloride transport, might argue against a direct reciprocal relationship between CFTR and ENaC activity in native airways. Future studies of sildenafil stimulation in patients with CF are needed in order to determine whether this finding remains restricted to murine airways, or if it is also relevant for human nasal epithelium. An additional potential interest of PDE5 inhibitors in CF might be raised by the assumption that the drugs may have a possible antiinflammatory action (46, 49, 50). Actually, it has been reported that treatment with 1 mg/kg sildenafil, a dose quite similar to that used by us, significantly reduced neutrophil influx in a Guinea pig model of airways disease developed after exposure to LPS to induce acute lung inflammation (50). Moreover, application of 300 nM sildenafil for 24 hours to CF cells in culture reduced secretion of IL-8 in response to Pseudomonas aeruginosa (46), a crucial CF pathogen responsible for the progressive loss of lung function, which is the main cause of morbidity and mortality in CF. In conclusion, our results provide evidence that sildenafil and vardenafil, administered in vivo at clinical doses, activate chloride transport in CF respiratory epithelium, presumably acting as F508del-CFTR potentiators rather than as correctors of the trafficking defect of the mutant protein. In F508del mice, the major effect was a rapid (1 h) increase in global chloride transport. The effect was apparently not dependent on alternative chloride transport pathways, and was lost in CFTR-null mice. In addition, the drugs induced an increased forskolin response in both F508del and wild-type mice; sodium transport remained unaffected.
Sildenafil citrate was a gift from Pfizer (Sandwich, UK), and Vardenafil was a gift from Bayer (West Haven, Germany).
Supported by a grant from the French Cystic Fibrosis Association (VLM), and by an educational grant from Pfizer Belgium. Originally Published in Press as DOI: 10.1164/rccm.200703-344OC on November 15, 2007 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form March 1, 2007; accepted in final form November 15, 2007
Related articles in AJRCCM:
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
AT A GLANCE COMMENTARY
) voltmeter (Knick Portamess 913; Elektronische Meßgeräte, Berlin, Germany). In brief, a double-lumen catheter was placed in a nasal passage, one lumen being used for perfusion of isotonic saline–buffered solutions, and the other one serving to measure an Ag/AgCl electrode (SLE Instruments, South Croydon, UK) connected to the positive terminal of the voltmeter through an electrode cream (Signa cream; Parker Labs, Fairfield, NJ) diluted 1:1 (vol/vol) in 3 M KCl. Measured values were negative. A needle inserted in the subcutaneous space in a hind leg served as a reference bridge. Solutions were perfused at a constant rate of 12 µl/minute in the following succession: basal solution, basal solution containing 10–4 M amiloride (amil), chloride-free solution containing amiloride (0 Cl), and chloride-free solution with amiloride plus 10–5 M forskolin (forskolin). The basal isotonic saline solution consisted of 140 mM Na+, 120 mM Cl–, 5.2 mM K+, 25 mM HCO3–, 2.4 mM HPO4–, 0.4 mM H2PO4–, Ca2+ 1.2 mM, 1.2 mM Mg2+, pH 7.4. In the chloride-free solution, chloride was replaced isoosmotically by gluconate, generating a diffusion potential that is generally interpreted as an index of chloride transport and chloride conductance of the luminal membrane of the nasal epithelium (
1–2 h), could additionally contribute to the longer lasting effect of vardenafil. An additional practical advantage of vardenafil over sildenafil is that the former is less active against PDE6, a central effector enzyme found in the retina and involved in cone and rod phototransduction. Alteration of color perception due to inhibition of retinal PDE6 is a side effect sometimes described with sildenafil, whereas it has not been reported with vardenafil (
B, Schermuly RT, Weissmann N, Seeger W, Grimminger F. Differences in hemodynamic and oxygenation responses to three different phosphodiesterase-5 inhibitors in patients with pulmonary arterial hypertension: a randomized prospective study. J Am Coll Cardiol 2004;44:1488–1496.
