Published ahead of print on October 3, 2003, doi:10.1165/rcmb.2003-0056OC
© 2004 American Thoracic Society DOI: 10.1165/rcmb.2003-0056OC Airway Epithelial Integrity Is Protected by a Long-Acting ß2-Adrenergic Receptor AgonistINSERM UMRS 514 and EA 2070, IFR 53, Laboratoire de Pharmacologie, Centre Hospitalier Universitaire Maison Blanche, Reims, France Address correspondence to: Edith Puchelle, UMRS INSERM 514, Hôpital Maison Blanche, 45, rue Cognacq Jay, 51092 Reims Cedex, France. E-mail: edith.puchelle{at}univ-reims.fr
Airway epithelial integrity may be impaired by bacterial exoproducts, which are able to degrade tight junction–associated proteins such as zonula occludens 1 (ZO-1). We have investigated the protective effect of salmeterol, a long-acting ß2-adrenergic agonist, on Pseudomonas aeruginosa–induced alteration of the epithelial junctional barrier. We demonstrate in human airway epithelial cells (HAEC) that salmeterol induces a time-dependent increase in ZO-1 protein, although no significant change in ZO-1 transcripts was observed. When HAEC cultures were exposed to P. aeruginosa (PAO1) supernatants, apical expression of ZO-1 protein was maintained in salmeterol-pretreated HAEC cultures, whereas it disappeared after PAO1 exposure in cultures not pretreated with salmeterol. Western blot experiments showed that the 220-kD ZO-1 protein was decreased after PAO1 incubation but was still present in salmeterol-pretreated HAEC extracts. The functional activity of ZO-1 protein was monitored by measuring transepithelial resistance and analyzing the diffusion of a low molecular weight tracer through the intercellular spaces. After PAO1 incubation, the epithelial integrity of HAEC was impaired, as shown by a decrease in transepithelial resistance and increased paracellular permeability, but was not significantly altered after salmeterol preincubation. These results demonstrate that salmeterol may contribute to the protection of the airway epithelium barrier against bacterial virulence factors.
Abbreviations: ß2-receptor agonist, ß2-RA • cyclic AMP, cAMP • cystic fibrosis, CF • colony-forming unit, cfu • chronic obstructive pulmonary disease, COPD • glyceraldehyde-3-phosphate dehydrogenase, GAPDH • human airway epithelial cells, HAEC • N-2-hydroxyethylpiperazine-N'-2-ethane sulfonic acid, HEPES • matrix metalloproteinase, MMP • phosphate-buffered saline, PBS • sodium dodecyl sulfate, SDS • transepithelial resistance, TER • tight junction, TJ • Trypticase Soy Broth, TSB • zonula occludens 1, ZO-1
The integrity of the barrier formed by the epithelium lining the airways is directly dependent upon the continuity of the superficial epithelial cell layer of the airway surface and the effectiveness of the adhesion molecules present between neighboring cells. Tight junctions (TJs) are located at the apicolateral borders of adjacent cells of diverse epithelial systems such as the airways, kidney, and intestine and serve as selective barriers that regulate the transepithelial movement of solutes and water (1).
In addition to sealing the intercellular space, TJs also provide an intramembrane barrier restricting the diffusion of lipids and proteins between the apical and basolateral domains (2). TJs undergo dynamic modulation by agents such as calcium (3), protein kinase C (4), cyclic AMP (cAMP) (5), and when subjected to an inflammatory environment (6, 7). They have been implicated in the pathogenesis of inflammatory diseases such as ulcerative colitis (8). Although most studies on the effects of chronic cytokine exposure on epithelial cells have focused on nonairway epithelial cells, a recent study by Coyne and coworkers (9) on primary human airway cells clearly demonstrated that TJs are regulated by proinflammatory cytokines and that combined exposure to tumor necrosis factor and interferon- Apart from inflammatory cytokines, bacterial toxins are also able to induce TJs and epithelial integrity alteration. This has been reported for intestinal bacterial toxins (10) as well as for bacterial virulence factors from organisms such as Pseudomonas aeruginosa. Azghani (11) has reported that P. aeruginosa elastase and exotoxin A are able to degrade two major components of TJs: the Zonula occludens (ZO-1 and ZO-2). Several TJ-associated proteins have been identified, such as ZO-1, -2, and -3, occludin, claudins, and junctional adhesion molecules. ZO-1 and ZO-2 are found just beneath the cell membrane and form a network for the anchorage of other TJ proteins, whereas occludin, claudins, and junctional adhesion molecules are positioned as integral transmembrane bridges that are associated with the network of ZO-1 and other TJ-associated proteins (1, 12, 13). In this work, we have chosen to analyze ZO-1 expression and distribution following bacterial injury because it has been extensively used as a marker of epithelial integrity. The regulation of TJ function is not completely understood, and may vary among cell types and possibly in response to bacterial virulence factors. Azghani has reported, on type II pneumocytes exposed to P. aeruginosa elastase and exotoxin A, an increased epithelial permeability and damaged TJ-associated proteins (ZO-1 and ZO-2) (11). This author suggested that these two P. aeruginosa exoproducts act in concert to promote bacterial dissemination in airway tissues. Moreover, P. aeruginosa is an opportunistic pathogen that causes infections in immunocompromised patients as well as in patients with cystic fibrosis (CF). The infection is characterized by inflammation of the airways associated with an influx of neutrophils and chronic bacterial infection with P. aeruginosa. The association of bacterial virulence factors and inflammatory cytokines represents optimal conditions for inducing severe alterations of TJ-associated proteins and functional changes in the airway epithelial barrier function. It has been reported that a long-acting ß2-receptor agonist (ß2-RA), such as salmeterol, exhibits cytoprotective effects on the airway epithelium by reducing the damage to respiratory epithelium during P. aeruginosa infection of upper airway epithelial cells in culture (14). The objectives of the present study were to analyze the specific effect of salmeterol on the expression of ZO-1 in human airway epithelial cells, and to examine its protective role in limiting the degradation of ZO-1 protein induced by P. aeruginosa exoproducts, thus preventing alterations to tightness and permeability of the airway epithelium.
Primary Human Airway Epithelial Cell Culture The human airway tissue samples used here were nasal polyps obtained from patients undergoing nasal polypectomy. Specimens were immediately immersed in RPMI 1640 culture medium containing 20 mM N-2-hydroxyethylpiperazine-N'-2-ethane sulfonic acid (HEPES) and antibiotics (200 U/ml of penicillin and 200 µg/ml of streptomycin). The tissue was then digested with 0.1% pronase E (Sigma Aldrich, St. Quentin Fallavier, France) for 12 h at 4°C, and the dissociated epithelial cells were seeded in 12-well plates (2 x 106 cells/ml) coated with type I collagen (2.5 mg/ml in 0.016 mM acetic acid) prepared from rat tail tendons according to the technique of Chambard and coworkers (15). The cells were grown to confluence in RPMI 1640 culture medium supplemented with epidermal growth factor (10 ng/ml), insulin (1 µg/ml), hydrocortisone (0.5 µg/ml), apotransferrin (1 µg/ml), retinoic acid (10 ng/ml), and antibiotics (200 U/ml penicillin, 200 µg/ml streptomycin) at 37°C in a humidified 5% CO2 atmosphere.
Preparation of Bacterial Culture Supernatant
Viability of PAO-1–Treated Cells For the real-time monitoring of airway epithelial cell viability, a fluorescence staining method using the LIVE/DEAD Cell Viability Kit (Molecular Probes Inc., Eugene, OR) was used. LIVE/DEAD kit is composed of two nucleic acid binding stains: SYTO 10 and DEAD Red. SYTO 10 penetrates into cells with either intact or damaged membranes (green-stained cells), whereas DEAD Red only penetrates into cells with damaged membranes (red-stained cells). Cells were cultured in Transwell as previously described. At cell confluence, the medium was removed from the wells and cells were incubated with 2 µl of SYTO 10 and 2 µl of DEAD Red in 1 ml of culture medium for 0.5 h in the dark at 37°C. The culture wells were then placed on the stage of an inverted microscope (Axiovert 200M; Zeiss, Le Pecq, France) equipped with an environmental chamber (37°C, 5% CO2, 100% relative humidity) and with a CCD video camera (Coolsnap Fx; Roper Scientific, Evry, France). Using Metamorph (Universal Imaging, Downington, PA) software, time-lapse images were recorded every 5 min for 0.5 h. At each time point, a phase-contrast image and 2 fluorescent images were recorded at x32 magnification. Green fluorescence staining corresponding to SYTO 10 and red fluorescence staining corresponding to DEAD Red were obtained through a 480-nm excitation filter/500-nm emission filter and a 490-nm excitation filter/635-nm emission filter. After the first 0.5-h period, PAO1 supernatant at 10% final concentration was added to the apical compartment of the culture well and time-lapse images were recorded as previously described. For the MTT assay, cells treated or not with PAO1 supernatant for 0.5 h were incubated with MTT at 1 mg/ml in culture medium for 1 h. Supernatants were then removed and cells were treated with isopropanol to dissolve the formazan crystals formed in alive metabolically active cells. The percentage of viability of PAO1-treated epithelial cells was calculated by determining the absorbance of treated and untreated cells with an automatic microplate scanning spectrophotometer (Bio-Rad Laboratories, Richmond, CA). The viability of PAO1-trated cells was performed in triplicate on three different airway epithelial cell cultures.
Preparation of the ß2-RA Salmeterol A highly specific ß2 receptor antagonist, ICI 118,551 hydrochloride, was purchased (Tocris, Ballwin, MO). This drug was prepared as a 10–2 M stock solution in dionized water and kept at –20°C. The stock solution was used at a final concentration of 100 nM in complete medium (18).
Western Blot Analyses of ZO-1
Reverse Transcriptase–Polymerase Chain Reaction Analyses
Immunolocalization of ZO-1 Protein by Immunofluorescence
Measurement of Intracellular cAMP
Transepithelial Resistance Measurement
Permeability to Lanthanum Nitrate and Transmission Electron Microscopy
Experimental Protocol
Statistical Analyses
Salmeterol Increases ZO-1 Protein but Not mRNA Expression To analyze whether a ß2-RA agonist could prevent the deleterious effects of PAO1 exoproducts by increasing the endogenous expression of ZO-1 in HAEC, we first analyzed by Western blot, the time-dependent effect of salmeterol at a concentration of 2 x 10–7 M after incubation with HAEC for 0.5, 16, and 24 h, respectively. In control HAEC, both isoforms of ZO-1,  – and + (19), were quite visible (Figure 1). A significant increase of ZO-1 level was observed at 16 h (P = 0.0002).
To demonstrate the specific effect of salmeterol on the increased ZO-1 protein expression, we coincubated cells with salmeterol and a specific ß2-selective competitive antagonist (ICI 118,151 hydrochloride) at a final concentration of 100 nM. A significantly increased expression of ZO-1 protein was observed after 0.5 h and 16 h of salmeterol (P = 0.03 and 0.01, respectively) that was not observed when the HAEC were incubated with the specific ß2-selective antagonist. RT-PCR semiquantitative analysis of ZO-1 mRNA expression before and after exposure to salmeterol (0.25, 0.5, 1, 3, 16, and 24 h) did not show any significant change in the ZO-1 transcripts (Figure 2), suggesting that the increase in ZO-1 protein was a post-transcriptional effect.
Effect of PAO1 Supernatant on ZO-1 Protein Expression To further investigate the time-dependent effect of PAO1 supernatant on the ZO-1 protein expression of HAEC, we analyzed ZO-1 expression levels of HAEC treated with PAO1 (10%) for 0.5, 1, 2, and 3 h, respectively. As shown in Figure 3, we observed a significant (P = 0.003) time-dependent effect of PAO1 on the decreased expression of ZO-1 protein. The effect was significant (P = 0.005) as early as 0.5 h after the onset of incubation, whereas after 3 h, a complete loss of the ZO-1 protein expression was observed.
Viability of PAO1-Treated Cells The imaging of airway epithelial cell viability over the course of 0.5 h of interaction with PAO1 supernatant is shown in Figure 4. No alteration of the fluorescent staining of the airway epithelial cells was observed. The airway cells exhibited a green fluorescent staining at the onset of the incubation with the PAO1 supernatant (Figure 4A). The green fluorescent staining was maintained for the 0.5-h incubation period with the PAO1 supernatant, showing no alteration in the integrity of the airway epithelial cell plasma membranes (Figure 4B). It was noteworthy that throughout the 0.5-h incubation period with PAO1 supernatant, no cells were extruded from the culture.
Using the MTT assay, the viability of the cells treated with PAO1 supernatant for 0.5 h was not significantly different from that measured in nontreated cells (absorbance at 570 nm: 0.83 ± 0.10 and 0.87 ± 0.19, respectively).
Effect of PAO1 Supernatant on Immunolocalization of ZO-1 Junctional Protein
Salmeterol Prevents PAO1-Induced Alteration of ZO-1 Protein To analyze whether salmeterol could prevent the alteration of ZO-1 protein expression, HAEC cultures were incubated for 0.5 h with PAO1 exoproducts either immediately or after a 16-h pre-incubation (optimal incubation time for the increased expression of ZO-1 protein) with 2 x 10–7 M salmeterol. As shown in Figure 6, we observed a decrease of ZO-1 protein expression (P = 0.003) after PAO1 supernatant incubation as compared with control values. In presence of salmeterol, PAO1 exoproducts do not induce a significant decrease of ZO-1 protein level.
Salmeterol Increases the Intracellular cAMP Level of HAEC Intracellular cAMP changes were monitored to assess whether salmeterol might be associated with an increase in intracellular cAMP. Although the intracellular cAMP level was not affected by exposure to PAO1 exoproducts at 0.5, 16, and 24 h (Figure 7A), salmeterol alone or in association with PAO1 exoproducts induced a transient, rapid (0.5 h) significant increase (P < 0.001) in the intracellular cAMP level of HAEC that was not observed after 16 and 24 h (Figure 7B).
Salmeterol Prevents the PAO1-Induced Loss of Epithelial Integrity To evaluate both the functionality of the tight junctions after PAO1 incubation and the protective effect of salmeterol on epithelial integrity, we analyzed changes in TER of the HAEC cultures. As shown in Figure 8, after 0.5 h incubation with PAO1 exoproducts, a significant (P < 0.001) decrease in TER (1,167.2 ± 312.2  · cm–2) as compared with control values (which reached a mean value of 7,207.8 ± 720.6 · cm–2 2 d after confluence) was observed. The preincubation of cells with salmeterol prevented the PAO1-induced decrease of TER, maintaining levels that were not significantly different from control values (5,924.2 ± 595.2 · cm–2). To further confirm the protection of HAEC integrity by salmeterol, the diffusion of a low-molecular-weight tracer, lanthanum nitrate was analyzed by transmission electron microscopy (Figure 9). In control HAEC cultures, lanthanum nitrate was located apically and was limited to the first layer of epithelial cells (Figure 9A). In PAO1 exoproducts–treated HAEC, lanthanum nitrate penetrated through the cell layers and was identified in the intercellular spaces of most basal cells (Figure 9B). In contrast, after a 16-h preincubation with salmeterol, lanthanum nitrate remained excluded from most of the intercellular spaces. It was only identified, either at the apical surface of the cells or limited, as observed under control conditions, to some areas of the most apical layers of the HAEC cultures (Figure 9C).
These results demonstrate that the protective activity of a long-acting ß2-adrenergic agonist, salmeterol, against P. aeruginosa virulence factors is associated with an increased expression of the TJ-associated protein ZO-1 that allows the human airway epithelial barrier to be preserved. Interestingly, we show that, under our experimental culture conditions, the human airway epithelial cells exhibit a high degree of polarity and tightness as shown by the high TER observed under basal conditions. This clearly demonstrates that culture of human surface airway epithelial cells represents a valid model for analyzing the capacity of salmeterol, to preserve the integrity of airway epithelial barrier tight junctions and function following bacterial injury. We ensured that in our experimental conditions, PAO1 exoproducts did not induce airway cell death as assessed by the cleavage of MTT and by a cell viability dynamic fluorescence microscopy assay. This result confirms the data reported by Rajan and coworkers (20) and by Plotkowski and coworkers (21) who have shown that a polarized and tight airway epithelium is highly resistant to P. aeruginosa–induced apoptosis and cell death whereas airway cells that do not establish TJ complexes are more susceptible to the cytotoxic effect of bacterial exoproducts. We have previously observed that P. aeruginosa is often associated with the injury and remodeling of the airway epithelium (22, 23) and that P. aeruginosa virulence factors present in PAO1 stationary phase culture supernatants, are able to delay the restoration of the airway epithelial integrity after wound closure (24). P. aeruginosa virulence factors, particularly elastase, have been reported to increase alveolar epithelial permeability by damaging the TJ-associated proteins ZO-1 and ZO-2 (11, 25, 26). Nevertheless, no data have been described on primary cultures of well-differentiated and junctional HAEC cultures. Dowling and coworkers (14, 27) have previously shown that salmeterol reduced P. aeruginosa– and Haemophilus influenzae–induced epithelial damage. They have observed a reduction of the epithelial damage assessed by a decrease in TJ leakiness, epithelial stripping, and the preservation of the number of both ciliated and unciliated cells. This effect was also associated with a decreased number of bacteria adherent to airway mucosa, consistent with the observation that P. aeruginosa, H. influenzae, and Staphylococcus aureus adhere preferentially to damaged epithelial cell surfaces (28). Our present results demonstrate that the ZO-1 protein degradation occurs as early as 0.5 h post PAO1 incubation and induces a very significant decrease of ZO-1 protein level with a nearly complete degradation after 3 h. These results confirm the data of Azgahni obtained in epithelial cell lines (11). Moreover, we show that salmeterol, used at a concentration previously described as allowing epithelial protective activity in vitro (14, 27), is able to induce a time-dependent increase in ZO-1 protein expression which is, after 16 h of salmeterol treatment, 1.5- to 2-fold higher as compared with untreated HAEC. This "cytoprotective" effect appeared to be specific as it was partially inhibited by the selective ß2-receptor antagonist ICI 118,551 (26). This epithelial protective effect of salmeterol was also demonstrated by the maintenance of a homogeneous ZO-1 protein expression after bacterial exposure as well as by the functional maintenance of the integrity of the epithelium barrier as shown by the absence of a decrease in TER and a maintained impermeability to low-molecular-weight lanthanum diffusion through the intercellular spaces. Among the virulence factors present in PAO1 supernatants, it was earlier shown that the elastase concentration of these supernatants attained 2 µg/ml and could impede airway epithelial repair by altering cell motility and causing an imbalance between pro- and activated forms of matrix metalloproteinases 2 (MMP-2) (24). This elastase concentration is in fact far below the elastase concentrations found in CF sputum samples, which can range from 3–110 µg/mg sputum, and have earlier been shown to severely alter epithelial barrier function (29). This protective effect of salmeterol on epithelial barrier function may render patients less prone to acute bacterial exacerbations. It is well known that most of the bacterial adhesins display high affinity for surface receptors present at the basolateral domains of airway cells and that opening of the TJs may facilitate adhesion of bacteria to the newly exposed basolateral receptors (23). It was recently reported that salmeterol significantly delayed the onset of exacerbation in patients with chronic obstructive pulmonary disease (COPD) and reduced the incidence of bronchitis (30, 31). Our results suggest that increased ZO-1 junctional protein may induce a decrease in bacterial adherence and colonization and offer protection against respiratory infections. There is no in vivo clear demonstration that salmeterol protects against airway bacterial colonization. Mahler and coworkers (31) have reported that a significantly higher percentage of salmeterol-treated patients completed the study without experiencing a COPD exacerbation than did placebo-treated patients. They have also found that salmeterol-treated patients had a delayed onset of exacerbations compared with placebo-treated patients, but the bacterial or viral origin of these exacerbations was not described. Another important question that remains to be elucidated concerns the specific mechanism of action of salmeterol leading to the increased ZO-1 expression. The increased expression of ZO-1 protein following salmeterol treatment was time-dependent and appeared maximal after 16 h. This effect was post-transcriptional because salmeterol did not induce a significant change in ZO-1 gene transcription. It has been suggested that the capacity of salmeterol to increase cAMP is responsible for the protection of respiratory epithelium from damage caused by bacterial infection (14). However, although this increased cAMP is well known to regulate epithelial TJ permeability (5), there is no direct evidence of a close relationship between cAMP and TJ structure (32). The role of the adenylate cyclase protein kinase A pathway in this cytoprotective function is not clear because we and others observed a rapid peak of cAMP in airway cells following salmeterol incubation even after PAO1 treatment, but this effect was no longer observed after 16 h, a period identified with a maximal increase of ZO-1 protein. In fact, we recently reported that ß2-RA stimulation induced a time-dependent increase in CF transmembrane conductance regulator protein expression in human airway epithelial cells through a cAMP protein kinase–independent pathway (33). Whether ß2-RA activation could modulate the expression of other junctional proteins and could induce other effects on ion and water transport, polarization, and differentiation of the airway cells remains to be determined. Moreover, we do not know whether the maintenance of tight junction and transepithelial resistance after a challenge with PAO1 products in presence of salmeterol only reflects a direct action of this long-acting ß2 agonist on ZO-1 or whether other effects on airway epithelial cells could be involved. In summary, this study provides evidence that salmeterol induces a cytoprotective effect by increasing ZO-1 TJ-associated protein and therefore by limiting the loss of airway epithelium integrity. Our findings suggest that the stimulation of TJ-associated proteins by salmeterol may be relevant to the mechanism of action of this drug in the protection against airway bacterial colonization. The increased ZO-1 expression associated with ß2-RA stimulation may be of major importance in pathological situations associated with bacterial infections such as COPD and CF.
Comparison of Data from Solid and Permeant Supports In our conditions of culture, the epithelial cells are confluent and well-polarized, as shown by the homogeneous network of ZO-1 protein observed by immunocytochemistry in cultures not treated with salmeterol. To compare the response of the epithelial cells to PAO1, salmeterol, and PAO1 with a preincubation of salmeterol, in experimental conditions similar to that used for TER and transmission electron microscopy (Transwells coated with collagen), we performed ZO-1 immunocytochemistry on Transwells coated with collagen 1. Although the images are less fine (probably due to the noise of the Transwell filter), we could observe results similar to those reported in the legend for Figure 5. A homogeneous labeling of ZO-1 is observed in nontreated airway epithelial cells (A) and in salmeterol-treated cells (B), whereas it decreases after exposure to PAO1 (C). When cells are pretreated by salmeterol before exposure to PAO1, the ZO-1 immunolabeling was similar to that observed in A (D) (Figure 10).
In human airway epithelial cells cultured on collagen gel, the TER exhibits high values ( These data demonstrate that the two model systems (permeant and solid support) give rather similar results.
This work was supported by INSERM, GlaxoSmithKline, and by the Association Vaincre la Mucoviscidose. The authors thank Pr. Malcolm Johnson (Uxbridge, UK) for his helpful comments, suggestions, and support of their work. They particularly thank all of the ENT physicians: Dr. J. M. Klossek, Jean Bernard Hospital (Poitiers, France), Dr. C. Ruaux, Clinique Mutualiste de la Sagesse (Rennes, France), and Dr. Corlieu, Tenon Hospital (Paris, France), who sent us human nasal tissues.
* Participated equally in this work.  Received in original form February 19, 2003 Received in final form September 22, 2003
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
induced drastic effects on TJ expression and barrier function, with significant alterations in the airway epithelial permeability. 
3,000 
