Abstract
One of the major therapeutic strategy in cystic fibrosis aims at developing modulators of cystic fibrosis transmembrane conductance regulator (CFTR) channels. We recently discovered methylglyoxal α-aminoazaheterocycle adducts, as a new family of CFTR inhibitors. In a structure–activity relationship study, we have now identified GPact-11a, a compound able not to inhibit but to activate CFTR.
Here, we present the effect of GPact-11a on CFTR activity using in vitro (iodide efflux, fluorescence imaging and patch-clamp recordings), ex vivo (short-circuit current measurements) and in vivo (salivary secretion) experiments.
We report that GPact-11a: 1) is an activator of CFTR in several airway epithelial cell lines; 2) activates rescued F508del-CFTR in nasal, tracheal, bronchial, pancreatic cell lines and in human CF ciliated epithelial cells, freshly dissociated from lung samples; 3) stimulates ex vivo the colonic chloride secretion and increases in vivo the salivary secretion in cftr+/+ but not cftr−/− mice; and 4) is selective for CFTR because its effect is inhibited by CFTRinh-172, GlyH-101, glibenclamide and GPinh-5a.
To conclude, this work identifies a selective activator of wild-type and rescued F508del-CFTR. This nontoxic and water-soluble agent represents a good candidate, alone or in combination with a F508del-CFTR corrector, for the development of a CFTR modulator in cystic fibrosis.
- Airway epithelial cells
- cystic fibrosis
- epithelial ion transports
- F508del-CFTR activator
- knock-out mice
- methylglyoxal-9-propyladenine
The cystic fibrosis transmembrane conductance regulator (CFTR) protein is a cAMP-dependent and ATP-gated chloride channel, localised at the apical surface, mediating Cl- transepithelial transport in airways, intestine and other fluid transporting tissues 1–4. Defective function of CFTR is responsible for cystic fibrosis (CF), one of the most common, lethal autosomal recessive human disorders in Caucasians. The principal clinical problem in CF is recurrent lung infections, resulting in progressive lung deterioration. More than 1,500 mutations in CFTR have been identified that cause CF phenotypes, with ∼90% of CF patients having the F508del mutation in one or both CFTR alleles. The deletion of F508 induces a trafficking- and gating-defect of CFTR protein leading to an impaired channel activity even when present at the cell plasma membrane 5.
In developing chloride-channel enhancement therapy for CF, it is assumed that restoring Cl- permeability would correct the underlying cellular defect that causes lung disease. Therefore, searching for reagents modifying the activity of CFTR is an active area of research and one of the major therapeutic strategies in CF. A number of pharmacological agents able to potentiate the cAMP-dependent CFTR function have been identified 6, 7 and are termed “potentiators” (for example genistein, VX-770, alkylxanthines). Moreover, some classes of potentiators, although being very potent on wild-type (wt) CFTR, have limited efficacy on F508del-CFTR mutants, such as forskolin (Fsk), β2-adrenergic or A2B-adenosine receptor agonists 6, 8, 9, some of them lack selectivity 6, 10 and have low affinity 11, 12. Another class of CFTR modulators, called “activators” of CFTR, were selected on the basis of their potency (low micromolar to submicromolar range) and their ability to induce CFTR-dependent electrogenic Cl- transport in a way independent of cAMP.
In a previous study, we described a new family of CFTR inhibitors: the methylglyoxal α-aminoazaheterocycle adducts; among them, the nucleoside 5a (termed “GPinh-5a”; fig. 1a) is a very potent CFTR inhibitor 13, 14. Here, following a structure–activity study, we identified methylglyoxal-9-propyladenine adducts (fig. 1b), an analogue of 5a named GPact-11a (Grenoble Poitiers CFTR activator-11a), as a selective, water-soluble and nontoxic activator of wt and mutated F508del-CFTR.
MATERIAL AND METHODS
Chemistry
All starting materials were commercially available research-grade chemicals and used without further purification. Reactions were monitored by analytical thin layer chromatography with fluorescent indicator UV254 from Macherey-Nagel (Duren, Germany). 1H and 13C nuclear magnetic resonance (NMR) spectra were recorded on a Bruker avance (Madison, WI, USA) 400 (400 and 100 MHz). Chemical shifts are reported in ppm relative to the residual signal of the solvent, and the signals are described as singlet (s), broad singlet (bs), doublet (d), triplet (t), doublet of doublet (dd), quartet (q), multiplet (m); coupling constants are reported in Hz. To the commercially available concentrated aqueous solution of methyglyoxal (40%; 1.5 mL, 9.0 mmol), 9-propyladenine (480 mg, 1.8 mmol) and water (5 mL) were added. Argon was flushed through the solution and the mixture was heated at 50°C for 17 h. After evaporation under reduced pressure, the residue was chromatographed on C18 reversed phase (10 g) eluting with H2O and then H2O-methanol (MeOH) (95:5) to give GPact-11a (24%) and GPact-11b (17%).
Chemical properties of GPact-11 isomers
GPact-11a. Melting point: 152–154°C; 1H NMR (400 MHz, D2O): 8.84 (1H, s, CHAr), 8.20 (1H, s, CHAr), 4.53 (1H, s, CH), 4.20 (2H, t, J = 7.0 Hz, CH2), 1.87–1.81 (2H, m, CH2), 1.75 (3H, s, CH3), 1.67 (3H, s, CH3), 0.85 (3H, t, J = 7.0 Hz, CH3); 13C NMR (100 MHz, D2O): 176.6 (COOH), 146.8 (CIV), 145.1 (CHAr), 142.7 (CHAr), 117.6 (CIV), 89.2 (CIV), 70.6 (CH), 63.3 (CIV), 46.1 (CH2), 25.7 (CH3), 22.8 (CH2), 22.2 (CH3), 10.2 (CH3); LRMS (FAB [+], glycerol): m/z 322 ([M+H]+); elemental analysis calculated for C14H19N5O4, 1/3 H2O: C 51.37, H 6.06, N 21.39, found C 51.34, H 5.98, N 21.24.
GPact-11b. mp: 102°C; 1H NMR (400 MHz, D2O pH = 7, phosphate buffer): 8.72 (1H, s, CH), 8.23 (1H, s, CH) 4.21 (3H, m, CH2 and CH), 1.86 (5H, m, CH2 and CH3), 1.74 (3H, s, CH3), 0.84 (3H, t, J = 7.0 Hz, CH3), 13C NMR (50 MHz, D2O, pH = 7, phosphate buffer) 176.1 (COO), 146.5 (CIV), 144.2 (CHAr), 144.0 (CHAr), 117.5 (CIV), 74.1 (CH), 60.7 (CIV), 45.5 (CH2), 26.0 (CH3), 24.4 (CH2), 18,0 (CH3), 9.6 (CH3), HRMS (EI) calc. for C14H19N5O4: [M+H]+ 322.1515, found 322.1507, [M+Na]+ 344.1335, found 344.1343.
Other chemicals
3-[(3-trifluoromethyl)-phenyl]-5-[(4-carboxyphenyl)methylene]-2-thioxo-4-thiazolidinone (CFTRinh-172) and N-(2-naphthalenyl)-[(3,5-dibromo-2,4-dihydroxyphenyl)methylene]glycine hydrazide (GlyH-101) were from VWR International (Fontenay-sous-Bois, France). GPinh-5a was synthesised as previously described 13. Miglustat was from Toronto research chemicals (Toronto, ON, Canada). All other chemicals were from SIGMA (St Louis, MO, USA). All chemical agents were dissolved in DMSO, with the exception of GPact-11a, GPinh-5a, amiloride, miglustat and GlyH-101, which were dissolved in H2O, and isoprenaline, which was dissolved in NaCl 0.9%.
Cell culture
Cell lines were cultured as previously described: Chinese hamster ovary (CHO), Calu-3, JME/CF15 15, CF-KM4 and MM39 16, NuLi and CuFi-1 17 and CFPAC-1 18. HEK293 cells were cultured and transiently transfected with the pEGFP-CFTR wt or F508del as already described 19, 20.
Freshly isolated human ciliated epithelial cells
The study was approved by our local institutional ethics committee. Human lung tissues were obtained from seven patients: three non-CF males with a mean age of 61 yrs, a non-CF female aged 65 yrs, a F508del/F508del-CFTR male aged 37 yrs, a F508del/R1066C female aged 26 yrs and a F508del/H1085R-CFTR female aged 23 yrs. Following lobectomy, lung samples, distant from the malignant lesion, were quickly dissected. After removal of connective tissues, cartilage and smooth muscle tissues, epithelial tissues were cut out in small segments. Single epithelial cells were mechanically dissociated by passing the bronchial tissue repeatedly through fire-polished Pasteur pipettes. Dissociated ciliated human epithelial cells were plated onto culture dishes for 6 h in culture medium (Dulbecco's modified Eagle medium /HAM-F12, insulin 5 μg·mL−1, transferin 7.5 μg·mL−1, hydrocortisone 10−6 M, endothelial cell growth supplement 2 μg·mL−1, epithelial growth factor 25 ng·mL−1, T3 3.10−5 M, l-glutamine 200 mM, and penicillin/streptomycin 100 μg·mL−1).
Animals
Wt (cftr+/+) and cftr knock-out (cftr−/−) female mice (C57Bl/6 129-CFTRtm1Unc) obtained from CDTA (Centre de Distribution, Typage et Archivage Animal, Orléans, France), were studied between the age of 3 and 4 weeks. To prevent lethal intestinal obstruction, mice were fed with movicol® (30 g·L−1) in drinking water.
Functional analysis of CFTR activity in vitro
CFTR Cl- channel activity was assayed by iodide (125I) efflux 21 or by single-cell fluorescence imaging, using the potential-sensitive probe, bis-(1,3-diethylthiobarbituric acid)trimethine oxonol (DiSBAC2(3); Molecular Probes, Eugene, OR, USA) 22. Iodide efflux curves were constructed by plotting rate of 125I versus time. All comparisons were based on maximal values for the time-dependent rates (k = peak rates, min−1) excluding the points used to establish the baseline (k peak-k basal, min−1). The results of fluorescence imaging are presented as transformed data to obtain the percentage signal variation (Fx) relative to the time of addition of the stimulus, according to the equation: Fx = ((Ft-F0)/F0)×100 where Ft and F0 are the fluorescent values at the time t and at the time of addition of the stimulus, respectively. For histogram representation, the values correspond to the level of stable variation of fluorescence induced by each drug. CFTR Cl- currents were measured in the broken-patch, whole-cell configuration of the patch-clamp technique 20.
cAMP concentration measurements
The intracellular cAMP was evaluated with the cAMP HTS Immunoassay Kit (Millipore, Billerca, MA, USA) following the manufacturer instructions. Luminescence was measured with TopCount NXT (PerkinElmer, Waltham, MA, USA) microplate reader.
Short-circuit current measurements
The apical membrane Cl- current was measured on mice colonic epithelium as already described 15 or after a depolarisation protocol 23 using a serosal-to-mucosal gradient with the following bath solution: apical, 107 mM K-gluconate, 4.5 mM KCl, 25 mM NaHCO3,1 mM MgSO4, 1.8 mM Na2HPO4, 0.2 mM NaH2PO4, 5.75 mM Ca-gluconate, 12 mM d-glucose; basolateral, 111.5 mM KCl, 25 mM NaHCO3, 1 mM MgSO4, 1.8 mM Na2HPO4, 0.2 mM NaH2PO4, 1.25 mM CaCl2, 12 mM d-glucose 23.
Salivary secretion assay
The salivary secretion was collected in response to subcutaneous injection of different pharmacological agents as described 24, 25. Basal salivary secretion due to cholinergic response was inhibited by atropine (1 mM).
Data analysis
Data are presented as mean±sem, where n refers to the number of isolated cell (single cell measurements), cell populations (iodide efflux and cAMP measurements) and N to the number of animals. Datasets were compared with an unpaired t-test. All graphs were plotted with GraphPad Prism 5.0 for Windows (GraphPad Software, San Diego, CA, USA). A p-value <0.05 were considered as statistically significant.
RESULTS
Effect of GPact-11a on wt-CFTR activity
We first examined the effect of GPact-11a on CHO cells stably expressing wt-CFTR with our robotic cell-based primary screening assay using iodide efflux measurement. This assay allows a rapid detection of CFTR channel activation or inhibition. The adenylate cyclase activator Fsk (fig. 1c) stimulated an iodide efflux in wt-CFTR CHO cells (relative rate = kpeak-kbasal = 0.44±0.01 min−1, n = 4; fig. 1c). This efflux was inhibited by GPinh-5a (fig. 1c). In the experiment presented in figure 1c, GPact-11a potentiated the Fsk-induced response to a relative rate of 0.38±0.02 min−1 (n = 4). The stimulation of CFTR by GPact-11a, in the presence of Fsk followed a dose-dependent relationship described by a half-maximal effective concentration (EC50) of 2.1±1.3 μM (n = 4; fig. 1d). Similar experiments were performed with the minor isomer GPact-11b leading to an EC50 of 2.9±0.9 μM (data not shown, n = 4). The effect of GPact-11a was fully inhibited by CFTRinh-172, GlyH-101 and GPinh-5a, three selective CFTR inhibitors 13, 26, 27 but was not affected by DIDS and calixarene, two non-CFTR inhibitors 28 (fig. 1e). We extended our study to evaluate the effect of GPact-11a on endogenous wt-CFTR in the human pulmonary epithelial cell line Calu-3, and found an EC50 of 1.8±1.3 μM in presence of Fsk (fig. 1f).
GPact-11a is an activator of wt-CFTR chloride channels
We then explored the cellular cAMP level in wt-CFTR CHO in response to several concentrations of GPact-11a. Figure 2a shows that GPact-11a has no effect on the basal cellular cAMP level. However, because we measured a low level of cAMP with Fsk, we wished to confirm these results in a human tracheal airway epithelial cell line (MM39) endogenously expressing wt-CFTR. Figure 2b shows the strong increase of cAMP level stimulated by Fsk but confirmed that GPact-11a had no effect on cAMP signalling pathway.
We compared the effect of GPact-11a on CFTR activity, in the presence or absence of Fsk, using single-cell fluorescence imaging technique applied to MM39. As expected, we observed that Fsk increased the recorded fluorescence signal. The signal was then further increased by GPact-11a and fully inhibited by CFTRinh-172 (fig. 3a). Interestingly, as shown figure 3b, GPact-11a increased the recorded fluorescence signal in the absence of Fsk and this effect was inhibited by CFTRinh-172. Figure 3c summarises the result obtained by a stimulation of Fsk or GPact-11a compared with the amplitude of the response generated by the co-application of GPact-11a+Fsk. These results demonstrated an additive and not a synergic effect suggesting two different mechanisms of action for Fsk and GPact-11a. We determined the concentration-response effect of GPact-11a, in the presence or absence of Fsk, and calculated an EC50 of 2.8±0.1 μM and 2.1±0.1 μM, respectively (fig. 3d). We determined the effect of GPact-11a on human ciliated epithelial cells freshly dissociated from lung samples (freshly isolated HBE; fig. 3f). In these cells, GPact-11a induced a strong increase of the fluorescence inhibited by CFTRinh-172 (fig. 3e). Finally, GPact-11a activated wt-CFTR on HEK293 cells transiently expressing EGFP-wt-CFTR proteins, in tracheal (MM39) and bronchial (NuLi) cell lines and in freshly isolated HBE cells (fig. 3f). Importantly, the CFTR response in HBE was 6-fold higher than in cell lines.
We used whole-cell patch-clamp technique to record CFTR currents on HEK293 cells transiently expressing EGFP-wt-CFTR proteins. After stimulation with GPact-11a, the whole-cell current increased linearly and reversed at -40 mV (fig. 4a and b) confirming its Cl- nature. This Cl- current was inhibited by the extracellular perfusion of CFTRinh-172 (fig. 4).
GPact-11a is an activator of rescued F508del-CFTR chloride channels
We determined by single-cell fluorescence imaging, the effect of GPact-11a on human CF tracheal glandular epithelial cells endogenously expressing F508del-CFTR (CF-KM4 cells). The defective trafficking characteristic of F508del-CFTR can be corrected by miglustat 22, 29. A sharp increase of fluorescence, inhibited by CFTRinh-172 (fig. 5a) or GlyH-101 (data not shown), was recorded after addition of GPact-11a on miglustat-corrected CF-KM4 cells. On the contrary, GPact-11a had no effect on untreated CF-KM4 cells (fig. 5a), confirming the CFTR selectivity of GPact-11a. On miglustat-corrected CF-KM4 cells, GPact-11a induced a concentration-dependent elevation of fluorescence with an EC50 of 34.3±1.8 μM (fig. 5b). Complementary studies performed on several human epithelial cell lines endogenously or transiently expressing F508del-CFTR are presented in figure 5c. Again, following the rescue of F508del-CFTR to the plasma membrane by miglustat, GPact-11a activated F508del-CFTR on HEK293 cells transiently expressing EGFP-F508del-CFTR proteins and in nasal (CF15), tracheal (CF-KM4), bronchial (CuFi) and pancreatic (CFPAC) CF cell lines (fig. 5c). Finally, we evaluated the effect of GPact-11a on ciliated epithelial cells freshly dissociated from human lungs of different CF genotype. The mutation R1066C is a missense mutation in the second transmembrane domain of CFTR 30 and causes a severe CF phenotype. R1066C-CFTR protein is not correctly processed and, unlike F508del-CFTR, this defect can not be corrected in reduced temperature or butyrate-treated cells 31. The mutation H1085R is a severe and rare missense mutation identified by Mercier et al. 32. H1085R-CFTR protein presents a trafficking defect which can be corrected by F508del-CFTR corrector 33. Figure 5d shows that in miglustat-pretreated cells from 3 CF patients with different genotypes, GPact-11a induced a strong increase of the fluorescence.
Whole-cell patch-clamp currents were recorded in HEK293 cells transiently expressing EGFP-F508del-CFTR proteins (fig. 6). Whereas GPact-11a has no effect in these cells cultured at 37°C (left traces; fig. 6a and b), a linear non-voltage-dependent Cl- current was recorded for cells treated by miglustat (middle traces; fig. 6a and b) or cultured at low temperature (right traces; fig. 6a and b). F508del-CFTR Cl- currents activated by GPact-11a were inhibited by CFTRinh-172 (fig. 6a and b).
Ex vivo evaluation of GPact-11a
We determined the effect of GPact-11a on the transepithelial ion transport under short-circuit current (Isc) conditions on cftr+/+ and cftr−/− mice colonic epithelium. In these conditions, cftr+/+ tissues had a mean transepithelial resistance (Rte) of 40.1±1.5 Ω·cm2 and a mean spontaneous Isc of 42.1±2.7 μA·cm−2 (n = 88). For cftr−/− tissues, Rte and Isc were 34.4±2.5 Ω·cm2 and -2.7±2.9 μA·cm−2 (n = 24), respectively. All experiments were conducted in presence of 100 μM amiloride to inhibit the resting Na+ current set by the activity of the epithelial sodium (ENaC) channel. The contribution of the transepithelial Cl- transport (ICl) was investigated after depolarisation of the basolateral membrane with a solution containing high potassium concentration 23. This procedure allowed measurement of changes in the apical anion conductance, avoiding contamination in the current response from, for instance, charybdotoxin-sensitive apical K+ channels 23. Following depolarisation, the average Rte and Isc in cftr+/+ colon were 40.2±1.4 Ω·cm2 and 147±7 μA·cm−2 (n = 12), respectively. Adding increasing concentrations of GPact-11a to both sides of cftr+/+ colonic epithelium induced a dose-dependent elevation of Isc with an EC50 of 175.6±1.1 μM (n = 3; fig. 7a and c). It was inhibited by glibenclamide (fig. 7a). On the contrary, application in the same experimental conditions of GPact-11a had no effect on cftr−/− colonic epithelium (fig. 7b). Addition of glucose increased Isc showing the viability of the tissues (fig. 7b).
Further experiments were carried out without the depolarisation protocol to determine the specificity of GPact-11a. Firstly, we permeabilised either apical or basolateral membrane with 200 μg·mL−1 of nystatin for 30 min and then measured Isc. As shown in figure 7d, permeabilising the apical membrane with nystatin prevented the GPact-11a effect (n = 5) whereas basolateral permeabilisation had no significant effect on GPact-11a Isc response (n = 4). Secondly, in cftr+/+ colonic epithelium, the elevation of Isc by GPact-11a was inhibited by CFTRinh-172 and glibenclamide but not affected by DIDS (fig. 7e). Finally, we found, with this protocol, no effect of GPact-11a on cftr−/− colonic epithelium (n = 4, data not shown).
In vivo evaluation of GPact-11a
To evaluate the in vivo effect of GPact-11a on cftr+/+ and cftr−/− mice, we used a noninvasive measurement technique of salivary secretion 25. First, we determined the secretory capacity of salivary secretion in cftr+/+ mice by stimulating the salivary gland with GPact-11a alone. GPact-11a by itself did not elevate salivary secretion (n = 3; fig. 8a). On the contrary, isoprenaline stimulated salivary secretion with a mean of 36.9±5.2 μg·min−1·g−1 (n = 19; fig. 8a). GPact-11a potentiated isoprenaline-induced secretion (70.2±9.0 μg·min−1·g−1, n = 7; fig. 8a). Figure 8b summarises the results obtained with increasing concentrations of GPact-11a and showed that GPact-11a, in the presence of isoprenaline, induced a dose-dependent salivary secretion in cftr+/+ mice with an EC50 of 7.1±1.1 μM (5<n<8; fig. 8c). The isoprenaline/GPact-11a-induced salivary secretion was inhibited by GPinh-5a and GlyH-101 (fig. 8d and e). Finally, we found no effect of isoprenaline and GPact-11a on the salivary secretion of cftr−/− mice (fig. 8f).
DISCUSSION
In the present study, we report on the discovery of methylglyoxal-9-propyladenine adducts named GPact-11a, as a selective, water-soluble and nontoxic activator of wt and F508del-CFTR. The most important results of our present investigation are: 1) GPact-11a is a novel activator of CFTR chloride channels in human airway epithelial cell lines; 2) GPact-11a activates rescued F508del-CFTR in nasal, tracheal, bronchial and pancreatic human CF epithelial cells; 3) GPact-11a is an activator of wt-CFTR and rescued F508del-CFTR in human ciliated epithelial cells freshly dissociated from CF and non-CF lung samples; 4) ex vivo, GPact-11a stimulates colonic chloride secretion in cftr+/+ mice; 5) in vivo, it increases salivary secretion in cftr+/+ mice; 6) GPact-11a is selective for CFTR because, in all our experimental conditions, we inhibited the effect of GPact-11a by the established CFTR inhibitors CFTRinh-172, GlyH-101, glibenclamide or GPinh-5a. In addition, we did not observe activation of any non-CFTR channels in uncorrected F508del-CFTR cells and on cftr−/− mice.
Structural determinants for CFTR activation
In the chemical structure of GPact-11a (fig. 1b), three main elements can be distinguished: 1) the purine aromatic heterocycles composed of fused pyrimidine and imidazole rings; 2) a hydrophobic propyl side chain attached to a nitrogen atom of the purine; and 3) the hydrophilic rings formed by condensation of methylglyoxal with 9-propyladenine. This latter ring probably includes the main pharmacophore because 9-propyladenine is not active on the CFTR channel (data not shown). The structure of the activator GPact-11a is close to the structure of the inhibitor GPinh-5a. In both structures, the first and third elements are present. This observation suggests a potential common binding site on the target involving the hydrophilic pharmacophore (i.e. the third element). The 9-propyl hydrophobic substituent in GPact-11a is replaced by a 9-(2-deoxyribos-1-yl) substituent which is hydrophilic in GPinh-5a. This substitution produces a shift from an activator to an inhibitor. GPinh-5a and GPact-11a are both zwitterionic at physiological pH, carrying a carboxylate function and a protonated imino group, two properties explaining their high water solubility.
In comparison with the activators previously described, some similarities can be mentioned. The xanthines activators are purine heterocycles substituted with hydrophobic substituents attached to a nitrogen atom on the purine (methyl, cyclohexyl, cyclopentadienyl, isopropyl) 6. Hydrophobic substituents attached to a nitrogen atom also are present in pyrrolo[2,3-b]pyrazines (butyl in RP107 and RP108) 15 and in the benzimidazolones activators (ethyl in 1-EBIO and DCEBIO) 6. These activators are made of two fused 6-member and 5-member rings like the purine heterocycle in GPact-11a. Some benzoquinolizinium activators also present a hydrophobic side chain, butyl in MPB-91 and MPB-104, linked to an aromatic heterocycle composed of three fused 6-member rings 6.
Towards a new generation of CFTR activators
Compared with other ion channels, such as voltage-dependent channels or neurotransmitter-activated channels 34–36, CFTR pharmacology is still in its infancy. No selective ligands, activators or inhibitors have been approved for clinical use. The CFTR modulators of the first generation were originally identified after investigations of the intracellular signalling pathways. For example, CFTR can be activated by agents that elevate cellular cAMP or stabilise phosphorylated CFTR (phosphatase inhibitors) 6. However, two receptor signalling pathways (β2-adrenergic receptors and A2B-adenosine receptors) efficiently activate wt-CFTR by stimulating adenyl cyclase and raising cellular cAMP 9, 11, 37, 38, but fail to activate F508del-CFTR in human airway epithelial cells. In recent years, a series of compounds that show activity in vitro has been described. CFTR openers include flavonoids, xanthines 39, 40, benzoquinoliziniums 6 or fluorescein derivatives 41. However some of these compounds lack selectivity and have low affinity. CFTR activation requires elevation of cAMP and the consequent phosphorylation at the regulatory domain (R domain) by protein kinase A 6, 34. CFTR activators can act directly on CFTR protein, or indirectly by inhibiting phosphodiesterases (thus elevating cAMP) or inhibiting phosphatases (thus increasing CFTR phosphorylation) 6, 34. Indirect effects are not expected to be useful for CF pharmacotherapy, as the basic defect is intrinsic to the CFTR protein and not to upstream regulatory pathways. In this study, we identified GPact-11a as a selective and cAMP-independent activator of wt and rescued F508del-CFTR. Therefore, while the mechanism of action of GPact-11a remains unknown, GPact-11a represents an interesting and potential candidate for the development of a future pharmacological therapy in CF because of its water solubility and absence of toxicity.
A corrector plus an activator for a CF bi-therapy?
F508del-CFTR activators are not expected to be effective unless they are associated with a corrector of F508del mistrafficking. Although a single compound with both types of activity is preferable, a bi-therapy involving a corrector plus an activator could be considered. We believe that the activator GPact-11a is one such valuable candidate. The water-solubility and apparent absence of toxicity of GPact-11a make this agent very attractive. Although a complete pre-clinical safety study will be required, in our preliminary in vivo experiments we did not observe serious adverse effects or mortality of mice. The selectivity of a pharmacological agent is a prerequisite before a further industrial and/or clinical development. GPact-11a seems to be selective for CFTR in epithelial cells because we did not observe activation of any types of ionic channels in uncorrected F508del-CFTR cells and in the colon of cftr−/− mice. Furthermore, in our models with a functional CFTR, the effect of GPact-11a was systematically inhibited by well-established CFTR inhibitors.
In summary, activation of surface F508del-CFTR will require combinational therapies including agents tailored for rescuing F508del-CFTR processing, coupled to strategies restoring F508del-CFTR function/regulation at the plasma membrane. Such bi-therapy could include, for example, miglustat with GPact-11a and will be considered in our future studies as a potential therapeutic in CF.
Acknowledgments
The authors would like to thank G. Cabrini, J. Glokner-Pagel and M. Merten for providing NuLi and CuFi-1, CFPAC, and MM39 and CF-KM4 cell lines, respectively. We thank P. Corbi and P. Bonnette for access to non-cystic fibrosis and cystic fibrosis human lung samples, respectively.
Footnotes
Support Statement
J. Bertrand was supported by a studentship from Mucovie. B. Boucherle and A. Billet were supported by studentships from Vaincre la mucoviscidose (VLM) and the French ministry of Research, respectively. L. Dannhoffer was supported by a postdoctoral fellowship from VLM and P. Melin-Heschel by a grant from VLM.
Statement of Interest
None declared.
- Received July 31, 2009.
- Accepted January 12, 2010.
- ©ERS 2010