Aclidinium inhibits cholinergic and tobacco smoke-induced MUC5AC in human airways

Mucus hypersecretion and mucin MUC5AC overexpression are pathological features of chronic obstructive pulmonary disease (COPD). This study examines the inhibitory effect of aclidinium, a new long-acting muscarinic antagonist, on MUC5AC expression in human airway epithelial cells. MUC5AC mRNA (RT-PCR) and protein expression (ELISA and immunohistochemistry) were studied in human bronchial tissue and differentiated human airway epithelial cells activated with carbachol (100 &mgr;M) or cigarette smoke extract in the absence or presence of aclidinium. Carbachol increased MUC5AC mRNA and protein expression in human bronchus and cultured epithelial cells. Aclidinium inhibited the carbachol-induced MUC5AC mRNA and protein expression with potency (half maximal inhibitory concentration) ∼1 nM in human bronchus and cultured airway epithelial cells. AG1478, a selective inhibitor of epidermal growth factor receptor (EGFR) tyrosine kinase, inhibited carbachol-induced MUC5AC responses, indicating EGFR transactivation. Aclidinium inhibited carbachol-induced phospho-EGFR and phospho-p44/42 MAPK expression. In cultured airway epithelial cells transfected with small interfering (si)RNA against muscarinic receptor subtypes, siRNA-M3 but not siRNA-M2 blocked carbachol-induced MUC5AC expression. Cigarette smoke-induced MUC5AC upregulation in cultured airway epithelial cells was suppressed by aclidinium. In conclusion, aclidinium decreases carbachol and tobacco smoke-induced MUC5AC overexpression in human airway epithelial cells. This effect may contribute to the clinical efficacy of aclidinium in mucus hypersecretory diseases including COPD.

characterise the effect of aclidinium on regulating carbacholinduced increase of MUC5AC expression in human isolated bronchus and well-differentiated human primary airway epithelial cells cultured in an air-liquid interface (ALI) system. In addition, since tobacco smoke exposure is associated with the pathogenesis of COPD and steroid resistance in COPD and severe asthma [10], we examined the effects of aclidinium on cigarette smoke extract (CSE)-induced increase of MUC5AC expression in human cultured airway epithelial cells. Aclidinium was found to effectively decrease carbachol and cigarette smoke-induced MUC5AC overexpression in human airway epithelial cells. Preliminary data from this study have been presented at the American Thoracic Society congress [11].

Human bronchial tissue experimental protocol
Human lung tissue was obtained from patients who had undergone surgery for lung carcinoma, as previously outlined [12]. Experiments were approved by the local ethics committee and informed consent was obtained. Tissue from a total of 12 patients was included in this study. All patients were smokers until the moment of lung carcinoma diagnosis (,1 month before lung surgery). None of the patients included in this study had COPD and none of the patients were chronically treated with theophylline, b-adrenoceptor agonists, corticosteroids or anticholinergic drugs. Clinical details of the different patients are provided in the online supplementary material.
Human bronchial tissues were pretreated with antagonists or their vehicles for 15 min prior to stimulation with carbachol and remained until termination of experiments. Carbachol was selected as a cholinergic agonist as it is widely used in the literature and is resistant to degradation by cholinesterases present in human bronchial epithelium [13]. In this study we used carbachol at 100 mM concentration, since we observed that it was near maximal responses, which is in keeping with other studies of goblet cell secretion and cell culture models using cholinomimetics [14]. Therefore, this carbachol concentration was selected for all subsequent experiments.

Differentiated human bronchial epithelial cells cultured in ALI
Human bronchial epithelial cells were cultured and differentiated in Transwell inserts (Corning Costar, High Wycombe, UK) under ALI conditions, as previously described [15]. In brief, a multilayered bronchial epithelium was obtained by seeding cells (8.25610 4

cells per insert) onto polyester inserts.
Cells were submerged in differentiation media (50% DMEM in basal epithelial growth media (BEGM; Clonetics, Wokingham, UK) for the first 7 days. Cells were then cultured for an additional 21 days with the apical surface exposed to air. Based on results from pilot experiments in airway epithelial ALI cultures that studied the time course of the expression of MUC5AC (1, 2, 6, 12 and 24 h), the carbachol response was obtained at 12 h, a time point similar to that studied in conventional culture models [12].
Mucin MUC5AC protein expression MUC5AC protein was measured by ELISA, as outlined previously [12]. In brief, for MUC5AC ELISA, 100 mg total protein extracted from human bronchial tissues was incubated at 40uC until dry. Plates were blocked with 2% bovine serum albumin for 1 h at room temperature. After three washes, plates were incubated with 50 mL of mouse monoclonal antibody (mAb) to MUC5AC (clone 45M1, 1:100; NeoMarkers, Fremont, CA, USA). After incubation with secondary horseradish peroxidase-goat anti-mouse (1:10,000) colour reaction was read at 450 nm. To confirm ELISA results, western blot analysis of MUC5AC was carried out in human bronchial homogenates, as previously reported [12].
Expression of phospho-EGFR and phospho-p44/42 MAPK Protein expression of phospho-EGFR (Tyr845) and phospho-p44/42 MAPK (Thr202/Tyr204) was determined in cultured airway epithelial cells by PathScan1 sandwich ELISA kit following the instructions of the manufacturer (Cell Signaling Technology, Beverly, MA, USA). A 15 min incubation time with carbachol was chosen because preliminary experiments indicated that there was little difference in phosphorylation between 5 and 15 min (data not shown). In some experiments, aclidinium was added 30 min before the addition of carbachol.

Real time RT-PCR
Total RNA was isolated from human bronchial tissue or human epithelial cells differentiated in ALI, as previously outlined [15]. cDNA was amplified with specific primers to MUC5AC, muscarinic receptors (M)1-5 and GAPDH (as endogenous control) using 7900 HT Fast Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). The primers used to amplify cDNA were obtained from TaqMan Gene Expression Assays (Applied Biosystem) to MUC5AC (catalogue number Hs01365616_m1), M1 (catalogue number Hs00912795_m1), M2 (catalogue number Hs00265208_m1), M3 (catalogue number Hs00325478_m1), M4 (catalogue number Hs00265219_s1), M5 (catalogue number Hs00255278_s1) and GAPDH (catalogue number 4352339E). Relative quantification of the different transcripts was determined with the 2 -DDCt method using GAPDH as endogenous control. Results were normalised to respective time control relative expression, as previously outlined [12].
Transfection of muscarinic receptor siRNA Cultured human bronchial epithelial ALI cells were transfected with a commercial 50 nM small interfering (si)RNA against M2 muscarinic receptor gene (PN 4392421; Ambion, Austin, TX, USA) or with 50 nM siRNA M3 muscarinic receptor gene (PN 4390815; Ambion) or with 50 nM siRNA control (Ambion, Huntingdon, UK), as outlined [16]. Briefly, ALI cultured bronchial epithelial cells were exposed to three successive solutions (4uC) containing: 1) EGTA 10 mM, KCl 120 mM, ATP 5 mM, MgCl 2 2 mM and TES 20 mM (pH 6.8; 20 min); 2) KCl 120 mM, ATP 5 mM, MgCl 2 2 mM, TES 20 mM and siRNA 50 nM (pH 6.8; 3 h); and 3) KCl 120 mM, ATP 5 mM, MgCl 2 10 mM, TES 20 mM and siRNA 50 nM (pH 6.8; 30 min). Subsequently, ALI epithelial cells were bathed in a fourth solution containing NaCl 140 mM, KCl 5 mM, MgCl 2 10 mM, glucose 5 mM and MOPS 2 mM (pH 7.1, 22uC), in which Ca 2+ concentration was gradually increased from 0.01 to 0.1 to 1.8 mM every 15 min. Then, cells were cultured in differentiation media. The mRNA expression for M2 and M3 transcripts was determined by real time RT-PCR after 72 h post-silencing and compared with siRNA control at the respective time to determine silencing efficiency. Furthermore, the M2 and M3 muscarinic receptor protein expression was measured by western blot after 72 h of silencing using commercial antibodies against muscarinic receptors M2 (catalogue number M9558; Sigma) and M3 (catalogue number M0194; Sigma), as previously described [15]. Protein expression was referred to bactin (1:1,000; Sigma) expression as internal control. All experiments were performed in triplicate.
Cigarette smoke exposure of differentiated human airway epithelial cells Differentiated bronchial epithelial cells were used to examine the MUC5AC responses after exposure to CSE prepared as previously outlined [17]. Aqueous CSE was applied to the apical side of epithelial cells to mimic the in vivo pattern of cigarette smoke exposure. Briefly, smoke of research grade cigarettes 2R4F (Tobacco Health Research, University of Kentucky, Lexington, KY, USA) was bubbled into a flask containing 25 mL of pre-warmed (37uC) DMEM medium obtaining a 100% CSE solution that was used within 30 min of CSE preparation. The CSE solutions were assessed by measuring absorbance at 320 nm. Stock solutions with absorbance values of 3.0¡0.1 were used. To test for cytotoxicity from CSE exposure, cultured cells were treated with CSE concentrations of up to 30% for 24 h. No significant increase of supernatant lactate dehydrogenase content above control levels were observed (colorimetric assay; Sigma; data not shown).

Statistics
Data are expressed as mean¡SEM of n experiments. In concentration-response experiments, the -log inhibitory concentration 50% (IC 50 ) was calculated by nonlinear regression to express compound potency (GraphPad Software Inc., San Diego, CA, USA). Statistical analysis was carried out by analysis of variance followed by appropriate post hoc tests, including Bonferroni correction. Significance was accepted when p,0.05.

Materials
Drugs used in this study were from Sigma (St Louis, MO, USA), except aclidinium bromide, which was synthesised at Almirall (Barcelona, Spain), and tyrphostin-AG1478 (Calbiochem, Nottingham, UK). All products were dissolved in distilled water mili-Q or in dimethyl sulphoxide (DMSO) at the concentration of 10 -3 M, except aclidinium, which was dissolved in HCl 1N 1% (v/v) at the concentration of 10 -3 M. Appropriate dilutions were obtained in freshly prepared medium. The DMSO final concentration in the assay solutions was f0.1% (v/v); this concentration was without significant effects on MUC5AC expression.

Carbachol-induced MUC5AC overexpression in human bronchus is blocked by aclidinium
Carbachol (100 mM) augmented the MUC5AC mRNA and protein expression with peak values reached at 1 h and 3 h after carbachol exposure, respectively ( fig. 1a). Furthermore, carbachol increased MUC5AC mRNA and protein expression in a dose-dependent manner, reaching a peak value at 100 mM ( fig. 1b). Thus, we selected this carbachol concentration for the rest of experiments. Immunohistochemistry experiments showed that MUC5AC immunoreactivity was localised in Alcian blue and PAS-stained goblet cells. The MUC5AC positive staining in airway epithelium was increased in carbachol-exposed preparations, and this augmentation was reduced in aclidinium-treated tissues ( fig. 2). Additional immunohistochemical experiments addressed to assert the activation of downstream elements of the EGFR pathway (see below) showed that carbachol augmented the expression of phospho-p44/42 in human bronchial tissue and this effect was also blocked by aclidinium ( fig. 2). Cigarette smoke-induced MUC5AC expression in cultured human ALI airway epithelial cells is blocked by aclidinium CSE increased MUC5AC mRNA and protein expression in a time and concentration-dependent manner ( fig. 4a and b). Based on these results, CSE (10%, 24 h) was selected for further experiments as producing consistent increases of MUC5AC mRNA and protein similar to those obtained with carbachol. CSE (10%, 24 h)-induced increase of MUC5AC mRNA and protein expression was suppressed by aclidinium and atropine in a dose-dependent fashion ( fig. 4c and d; table 1). Furthermore, CSE-induced MUC5AC expression was inhibited by aclidinium (100 nM) as well as atropine (1 mM) and tyrphostin-AG1478 (10 mM), but remained unaltered in the presence of hexamethonium (100 mM) ( fig. 4e).
M3 muscarinic receptor subtype appears involved in MUC5AC response to carbachol and CSE in human cultured airway epithelial cells Setting the expression level of M1 mRNA at 1, the amount of M2 mRNA expression in ALI airway epithelial cells was ,5.2 times higher and that of M3 was ,2.4-fold higher than that of M1 ( fig. 6a). Additional experiments demonstrated also the presence of M4 and M5 expression (data not shown). These data were corroborated at protein level, where M2 receptor was the most expressed followed by M3 ( fig. 6b).
Furthermore, following CSE exposure, a significant increase in the mRNA and protein expression of M3 was observed without significant changes of the relative levels of M1 and M2 ( fig. 6a and b).
Due to the low expression of M1, we studied the role of muscarinic M2 and M3 receptor subtypes in the MUC5AC secretion. To this respect, transfection of human ALI airway epithelial cells with siRNA-M2 or siRNA-M3 decreased the mRNA expression of M2 and M3 by ,62% and ,73%, respectively, with the corresponding protein downregulation ( fig. 7a and b). The siRNA-M2 did not affect the carbachol-or CSE-induced MUC5AC mRNA expression, while siRNA-M3 blocked the MUC5AC mRNA response to carbachol and CSE, thus indicating a preferential role of M3 receptors in this response ( fig. 7c and d). To further study the implication of M1, M2 and M3 receptors in MUC5AC upregulation, we preincubated differentiated human ALI epithelial cells with pirenzepine 1 mM (M1 antagonist), methoctramine 1 mM (M2 antagonist) and p-fluoro-hexahydro siladifenidol (pFHHSid) 1 mM (M3 antagonist). We detected that both pirenzepine and methoctramine did not modify significantly the effect of carbachol and CSE on MUC5AC expression. In contrast, pFHHSid effectively attenuated carbachol-and CSE-induced MUC5AC overexpression, confirming data from siRNA experiments ( fig. 7e and f).

DISCUSSION
In this study, we show that aclidinium potently inhibited the carbachol-induced MUC5AC overexpression in human bronchial tissue in vitro as well as the carbachol-and cigarette smoke-induced augmentation of MUC5AC expression in differentiated human airway epithelial cells cultured in ALI. This is the first report showing a direct effect of a cholinergic agonist in upregulating MUC5AC expression in human airway epithelial cells by activation of muscarinic receptors and transactivation of EGFR. Also, we showed that cigarette Cigarette smoke in human cultured airway epithelial cells  smoke-induced increase of MUC5AC expression in vitro can be effectively suppressed by the use of muscarinic antagonists. Since mucus hypersecretion is considered pathologically relevant in COPD and asthma, this inhibitory effect of aclidinium is of potentially added therapeutic value.
It is well established that goblet cell hypertrophy and hyperplasia occur in the large airways of habitual cigarette smokers with or without airway obstruction and result in epithelial mucin stores that are significantly higher than normal [18]. Goblet cells are located on the bronchial surface epithelium and MUC5AC is the best characterised mucin in this cell type, and its expression is directly correlated with airway obstruction [18][19][20]. Therefore, MUC5AC was selected for this study. While MUC5AC is increased directly by cigarette smoke in goblet cells [20], the role of other mucins such as MUC5B seems contradictory since MUC5B is not altered or, indeed, diminished in goblet cells from smokers and COPD patients [18,19]. However, the major place of mucin production is located in submucosal glands in the central airways, where both MUC5AC and MUC5B are abundant in smokers and COPD patients [19,21]. In this line, MUC5B is mainly produced in submucosal gland cells and is the most abundant in COPD sputum, while MUC5AC is mainly produced in goblets cells and is the most abundant in smoker sputum without airway obstruction [21]. However, it seems that only MUC5AC may be induced by different stimuli, while MUC5B appears more constitutively expressed and less sensitive to the inducible effect of irritants such as cigarette smoke [19]. This work is focused on MUC5AC production in goblet cells from the surface epithelium due to the inducible characteristic of MUC5AC and its correlation with human airway obstruction found in COPD. However, since the major mucins are released predominantly from submucosal gland cells, this may be considered a limitation of the present study.
The human bronchial tissue in vitro is a preparation that has previously been shown to have a basal secretion of mucin MUC5AC produced principally by goblet cells [12].
The rapid time course of MUC5AC secretion found in this work ( fig. 1) was in accord with previous reports for epidermal growth factor [12] and for carbachol in goblet cells from rat conjunctiva [14]. In this in vitro model, aclidinium inhibited the augmented expression of MUC5AC induced by carbachol in a concentration-dependent fashion, with potency values of ,1 nM, which is in the range of the potency of this muscarinic antagonist against cholinergic contraction in isolated trachea [22].
Immunohistochemical analysis of human bronchial tissues confirmed that carbachol exposure resulted in an augmented expression of MUC5AC-positive stained cells in the airway epithelium, which was paralleled by an increased presence of Alcian blue-positive and PAS-stained goblet cells. Treatment with aclidinium effectively prevented this carbachol-induced overexpression of MUC5AC. In this point, it is interesting to note that the histochemical analysis of goblet cells was not always correlated with the immunohistochemical staining of MUC5AC, as previously reported [19,20], which may be taken in account in the evaluation of this study.
EGFR and its tyrosine receptor kinase-based pathway mediate mucin production by airway epithelial cells in response to a variety of pathogenic and environmental insults [23]. The transactivation of EGFR following the stimulation of muscarinic receptors has been described in different cell types, including conjunctival goblet cells in relation to mucin secretion [24]. We confirmed and extended this finding by showing that carbacholinduced augmentation of MUC5AC mRNA and protein expression was mediated by the transactivation of EGFR and the subsequent phosphorylation of p44/42 MAPK. Thus, blockade of muscarinic receptors by aclidinium avoided transactivation of EGFR.
Because the human bronchial tissue contains a variety of different cell types, we also aimed to study the direct impact of carbachol on well-differentiated human primary airway epithelial cells with relation to MUC5AC expression. We found that carbachol-induced MUC5AC overexpression was abolished  by aclidinium and atropine, in a manner similar to that found in a tissue bronchium in vitro model, while hexamethonium did not reduce carbachol responses. Furthermore, nicotine (10 mM) did not evoke a MUC5AC response, which rules out nicotinic receptor activation in this mechanism. Consistent with the results observed in bronchial tissue, MUC5AC overexpression was a consequence of the transactivation of the EGFR and the subsequent triggering of the phosphorylated p44/42 MAPK signalling cascade, as previously suggested for goblet cells from rat conjunctiva [24].
Next we investigated the role of muscarinic receptor subtypes on carbachol-induced MUC5AC overexpression. The presence of the five subtypes of muscarinic receptors was detected but we focused our study on M1, M2 and M3 receptors, since only these subtypes exert well known physiological effects in the lungs [25]. We found a predominance of M2 and M3 muscarinic subtypes, with low amounts of M1 receptors. A similar pattern of expression was described for cultured respiratory epithelial cells [5,26]. In this sense, we found that siRNA specific for siRNA-M3 and the M3 antagonist pFHHSid attenuated the MUC5AC response to carbachol and CSE, while siRNA-M2 and the M2 antagonist methoctramine were without significant effects. These results suggest that the activation of muscarinic M3 receptor appears involved in the mucin MUC5AC response after cholinergic activation in differentiated human airway epithelium. The low expression of M1 receptors impeded the use of specific siRNA in the present study. However the use of the M1 antagonist pirenzepine was without effect, which discounts the M1 receptor in this process.
Cigarette smoke is widely used in in vitro studies due to its relevance in the pathogenesis of COPD [10]. The extent of the increases of MUC5AC expression obtained in the present study after cigarette smoke exposure, and the time point selected for this expression, are in keeping with findings from other studies [27]. In this work we observed that aclidinium with potency close to 1 nM, as well as atropine, inhibited the CSE-induced MUC5AC. In contrast, nicotinic receptors did not affect the MUC5AC upregulation. In this mechanism, the transactivation of EGFR after cigarette smoke exposure was also demonstrated by direct phosphorylation after CSE exposure and by the inhibition of MUC5AC expression obtained in the presence of tyrphostin-AG1478, a selective inhibitor of the EGFR tyrosine kinase. This result is consistent with the overproduction of mucin via EGFR found in response to cigarette smoke in the airway epithelial cell line NCI-H292 [28].
The activation of muscarinic receptors in response to cigarette smoke requires explanation. The epithelium of the airways possesses a non-neuronal cholinergic system and local release of acetylcholine may serve a variety of autocrine and paracrine functions [5]. Bronchial epithelial cells in culture express cholinesterases that inactivate acetylcholine [28]. Cigarette smoke is known to impose an important oxidative burden on exposed cells [10] that may degrade acetylcholinesterase and butyrylcholinesterase molecules, while cholineacetyltransferase is not affected, thus resulting in tissue accumulation of acetylcholine [29]. Conversely, we found an upregulation of muscarinic M3 receptors in airway epithelial cells exposed to CSE. This finding would be consistent with a recent study showing that human lung fibroblasts exposed to CSE dramatically increased cholineacetyltransferase and M3 muscarinic receptor expression [30], which may explain the effect of anti-muscarinics on tobacco smoke responses.
The results of this study indicate that aclidinium effectively decreased the MUC5AC overexpression elicited by cholinergic activation and cigarette smoke exposure in human bronchial epithelial cells. This direct inhibitory effect may have an additive effect in the clinical efficacy of aclidinium in mucus hypersecretory diseases, such as COPD.