Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Staphylococcus aureus is a major respiratory pathogen in patients with the hereditary disorder cystic fibrosis (CF) (1), causing lung infections in about 36% of CF patients of all age groups (2, 3). These infections are difficult to treat, and the consequences of chronic lung inflammation as a response to the persistent pathogen may lead to considerably reduced lung function with deleterious effects for the patients. Genotyping of S. aureus strains from infected CF patients (4) reveals a high degree of strain identity between nose and sputum isolates (5), suggesting that initial colonization of the nasal mucosa with S. aureus sets the stage for lower respiratory tract infection. Such a route of transmission has also been proposed for other types of S. aureus infection (6).

The mechanisms leading to the high incidence of S. aureus lower respiratory tract infections in CF patients are largely unclear, but CF-specific factors are most probably involved. Evidence supporting this contention includes the observation that although approximately 30% of healthy individuals are nose-carriers of S. aureus, lower respiratory tract infection with S. aureus does not occur in the general population (9, 10). One of the factors facilitating S. aureus infection in CF may be the increased sulfation of the glycocalix of CF epithelial cells (11, 12) because S. aureus binds strongly to sulfated glycolipid (13) and heparan sulfate (14) in vitro. Furthermore, undersialylation of apical proteins and increased concentrations of asialoganglioside 1 (aGM1) were detected in apical membranes of CF bronchial epithelial cells compared with control cells (15). The glycoconjugate aGM1 contains the binding sequence Gal NAc1-4Gal for S. aureus and other pulmonary pathogens (16). Binding of S. aureus to cultured bronchial CF cells was significantly greater than for control cells (15). However, Schwab and coworkers (17) did not find significant differences in binding of S. aureus to freshly isolated epithelial cells from CF patients and healthy individuals. Thus, at present, the notion of an increased adherence of S. aureus to structurally modified epithelial membrane components in the CF respiratory tract remains hypothetical.

In general, a major difficulty in extrapolating in vitro results of bacterial adherence to the in vivo situation is the lack of suitable cell culture models. For example, in vitro, primary epithelial cells may degenerate after short periods of time, ciliated respiratory epithelial cells may rapidly lose their cilia, and most cell lines do not produce mucus. Particularly, mucus production is essential for an investigation of S. aureus adherence to the CF respiratory epithelium, since mucin binding of S. aureus has been demonstrated in animals (18) and in vitro (18), secreted mucins are oversulfated in CF (10, 11, 21), and inflammatory mechanisms in CF airways (22) provoke a pronounced mucus hypersecretion (23).

Thus, the objective of the present study was to directly locate S. aureus in lung tissue specimens of infected CF patients. Furthermore, we used primary respiratory epithelial cells displaying functional cilia and actively secreting mucus for 6 mo, in order to compare S. aureus adhesion to cells from CF patients with those obtained from healthy individuals.

	Material and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

CF Patients and Healthy Individuals

Lung material from the lobes of three CF patients (two males, one female; mean age: 12 yr) attending the Service de Pédiatrie, Centre Hospitalier Lyon-Sud, Pierre-Bénite, France, were studied. All patients had been infected with S. aureus for at least 6 yr. Additionally, a nasal polyp from one CF patient who underwent polypectomy for recurrent nasal polyposis at the Ear Nose and Throat Clinic, Klinikum Ludwigshafen, Ludwigshafen/Rhein, Germany was obtained. Immediately after lobectomy, tissues were cut in approximately 0.5 cm³ cubes. The cubes were fixed in 2.5% glutaraldehyde and shock-frozen in liquid nitrogen or fixed in fresh 10% formaldehyde for the different microscopic procedures. Additionally, nasal polyps from five CF patients (mean age: 13 yr) attending the Ear Nose and Throat Clinic, Klinikum Ludwigshafen, and from five healthy individuals (mean age: 40 yr) attending the Ear Nose and Throat Clinic, University of Tübingen (Tübingen, Germany), were used for suspension cell cultures. All subjects gave informed consent. The use of the clinical material was approved by the Comité Consultatif de Protection des Personnes en Recherche Biomédicale, Lyon A, France; and the Ethic Committees of the University of Tübingen and University of Giessen (Giessen, Germany).

Cell Culture System

The primary cell culture system described by Jorrisson and colleagues (24) was used with modifications. After repeated washings, polyps obtained from CF patients and healthy individuals were digested with 0.1% pronase (Sigma, Deisenhofen, Germany) in Dulbecco's modified Eagle's medium (DMEM) and HAM's F12 1:1 medium (GIBCO BRL, Eggenstein, Germany) supplemented with penicillin (50 µg/ml) (GIBCO), streptomycin (50 µg/ml) (GIBCO), and Nystatin (1,000 U/ml) (GIBCO) overnight at 4°C under continuous rotation. Washed cells were incubated in T-25 flasks (Becton Dickinson, Heidelberg, Germany) up to 10 d in culture medium supplemented with Ultoser G (2%; GIBCO), cholera toxin (10 ng/ml; Sigma), and retinoic acid (10⁷M; Sigma) at 37°C and 5% CO₂ under continous rotation for cell ball formation. Later they were kept in a T-25 flask at 37°C and 5% CO₂ without shaking for up to 6 mo. Cell balls reexpressed cilia with normal ultrastructure and coordinated beating after 4 wk. The cell balls maintained definite epithelial characteristics of respiratory cells as demonstrated by morphology (polarization, microvilli, tight junctions), positive reactions with epithelial specific antibodies (antikeratins), and a redifferentiation into a normal respiratory-type epithelium (24). Independent of the cell ball size (mean diameter: ~ 150 µm), cell balls moved at a speed of 8-12 rpm. The surface epithelium of the polyps from CF patients and normal individuals was mainly normal respiratory epithelium. Cell balls did not grow further in vitro. A typical scanning electron micrograph of cell balls is shown in Figure 3A.

View larger version (192K): [in this window] [in a new window]   Figure 3.   Scanning electron micrograph of primary nasal epithelial cells of cystic fibrosis patients (A, B, D) and a healthy individual (C) as three-dimensional cell balls. Note highly ciliated cell balls (A, inset). (B, C) Unwashed cell balls were inoculated with S. aureus for 2 h. Arrows point to S. aureus adhering to mucus on cell balls. Note S. aureus-free cell membranes. (D) Washed, mucus-depleted cell balls were inoculated with S. aureus for 2 h. Note S. aureus-free cell membranes. Magnifications: (A) Upper part: ×300, inset: ×3,000. (B) ×5,000. (C) ×8,000. (D) ×1,300. Bars: (A) Upper part: 25 µm, inset: 2.5 µm. (B) 1.5 µm. (C) 1.3 µm. (D) 5.8 µm.

Adhesion of S. aureus to Washed and Unwashed Primary Nasal Epithelial Cells

Cell balls, 3 to 5 mo old, from five CF patients and five healthy individuals, suspended in DMEM/HAM's F12 medium, depleted of antibiotics and antimycotics for 5 d, were used for the S. aureus adhesion assay without prior washing. Cell balls were washed after incubation with S. aureus using a cell strainer (Becton Dickinson). In other experiments, cell balls were washed with DMEM/HAM's F12 medium prior to the adhesion assay in order to remove secreted mucus on the cell ball surface. Cell balls were washed five times until mucicarmine staining (Sigma) was negative. In intercellular spaces, however, mucicarmine staining remained positive. For adhesion experiments an S. aureus strain, isolated from a CF patient and grown in tryptone soy broth at 37°C to an optical density of 2, was used. Cell balls were incubated with S. aureus at cell:bacteria ratios of 1:1,000 or 1:100 for 2 h at 37°C and 5% CO₂. For evaluation of bacterial adherence, 20 to 30 cell balls from each individual (~ 50 cells per cell ball) were analyzed by scanning electron microscopy (SEM).

Adhesion of S. aureus to Hypersecreting Primary Nasal Epithelial Cells

For the induction of hypersecretion of primary nasal epithelial cells, purified human neutrophil elastase (NE) was used. NE was isolated and purified from pooled expectorated CF sputum as previously described (25). The purity of the enzyme was checked with sodium dodecyl sulfate-polyacrylamide gel electrophoresis (26), revealing only isoenzyme bands between 25,000 and 30,000 Da. Enzyme activity was measured photometrically with the chromogenic peptide substrate methoxysuccinyl-L-alanyl-L-alanyl-L-prolyl-L-valin-p-nitroanilide (Bachem AG, Bubendorf, Switzerland) (25). Protein concentrations were measured with the method of Lowry and coworkers (27) using bovine serum albumin as a standard. Three- to 5-mo-old cell balls from CF patients and healthy individuals were treated with NE (1 µg/ml medium) for 30 min. NE activity was then inhibited by the addition of 100 µg/ml ₁-antitrypsin (Sigma). Thereafter, cells were incubated with S. aureus as described above. In other experiments, NE (1 µg/ml medium) was added for 30 min to cell balls which had been preincubated with S. aureus at cell:bacteria ratios of 1:100 for 2 h at 37°C and 5% CO₂.

Electron Microscopy

Glutaraldehyde fixed lung tissue samples were cut in 6 × 5 × 5-mm pieces with a bronchus segment of approximately 5- 10 mm² and prepared for SEM. After washing with phosphate-buffered saline (PBS), lung tissue specimens were postfixed with 1% OsO₄ for 1 h on ice. Glutaraldehyde fixed cell balls and lung tissue specimens were fixed with a commercial glue (UHU GmbH and Henkel KGaA, Bühl, Germany) on a 4 × 5 mm piece of polyvinyl chloride using the method of König with modifications (28) and postfixed with 1% OsO₄. Specimens were washed with PBS, dehydrated in a graded ethanol series, and critical point dried (Polaron E3000; Plano, Marburg, Germany). The dehydrated specimens were mounted on a stub and sputter coated with gold-palladium (200Å) (Gleichstromkathodenzerstäuber, MED010; Balzers Union, Lichtenstein). For transmission electron microscopy (TEM), glutaraldehyde fixed and OsO₄ postfixed specimens were embedded first in agar and then in Epon. Thin sections were mounted on copper or nickel grids and stained with uranyl acetate and lead citrate.

For the quantitation of damaged bronchial tissue and adherence of S. aureus by SEM, large areas, 5-10 mm², of the epithelium of three bronchi were divided into squares of approximately 600 µm² and analyzed at a magnification of ×3,000 using an image analyzer (Analysis; Soft Imaging Software GmbH, Münster, Germany). A total of 44 different squares were analyzed for the percentage of unciliated or microvilli-containing cells to ciliated cells, the length and distribution of cilia, and S. aureus adhesion.

Immunofluorescence

S. aureus was also localized in obstructed bronchioli and bronchi of three CF patients by indirect immunofluorescence using a rabbit antiserum against teichoic acid and using rabbit polyclonal antibodies (Sigma) against protein A (29). Cryostat thin sections (5-10 µm) were prepared (Kryostat 2800 Frigocut E; Reichert-Jung, Heidelberg, Germany) from shock-frozen lung tissue material. The thin sections were fixed on cover slips with acetone for 10 min and cover slips were preincubated with swine antiserum, 1:5 diluted. A fluorescein isothiocyanate (FITC)-conjugated swine antibody against rabbit IgG, diluted 1:40, was used (Dako, Hamburg, Germany). Thirty-nine thin sections from parts of the infected lobes were analyzed for the distance of S. aureus from the bronchial epithelium using the Kontrol Imaging System, KS300 (Kontron Electronic GmbH, Eching, Germany). The distance of 10 organisms most proximal to the epithelium was evaluated. Thin sections may contain artifacts at the border of the epithelium and the sputum due to the preparation method, which results in black interspaces. To allow for this, the distance of the artifacts was subtracted from the distance of the bacteria from the epithelium using the image analysis system.

In other experiments, thin sections of obstructed airways were also stained with hematoxylin eosin. Additionally, S. aureus was identified in mucus from obstructed airways, isolated from the CF lung tissue by gram stain.

Statistics

For the statistical analysis of the data the unpaired Student's t test (Excel 5.0; Microsoft Corporation, Redmont, WA) was used.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Localization of S. aureus in CF Airways

The majority of the larger and smaller bronchi of the three CF patients were obstructed as indicated by the sputum filled airways (Figure 1A). Although totally obstructed, the epithelium of the bronchioli appeared almost intact and ciliated (Figure 1B). S. aureus was detectable in thin sections of obstructed bronchioli by indirect immunofluorescence using antibodies to protein A or teichoic acid, major components of the S. aureus cell wall in CF airways (29) (Figure 1C) or by gram stain (Figure 1D). S. aureus was found to be uniformly distributed within the mucus and clustering was not observed.

View larger version (155K): [in this window] [in a new window]   Figure 1.   Thin sections of a bronchus filled with sputum of an S. aureus-infected lung of a CF patient, stained with hematoxylin eosin; staining was by indirect immunofluorescence using a rabbit antiserum against teichoic acid and FITC-conjugated swine antibody against rabbit IgG. Magnifications: (A) ×100; bar: 100 µm. (B) ×1,000; bar: 10 µm. (C) ×400; bar: 25 µm. Note intense cell infiltration (predominantly polymorphonuclear leukocytes) in A and B. In B, note the intact, ciliated epithelium (arrow). In C, note the epithelium (E) devoid of adhering S. aureus; the arrow points to bright yellow-stained S. aureus in the airways lumen filled with sputum (S). (D) Typical gram stain of sputum material, isolated from the bronchus of this CF patient, showing gram-positive clusters of S. aureus; magnification: ×1,000; bar: 10 µm.

In order to locate S. aureus in the bronchioli, 39 different thin sections from eight different sites of the three infected lobes, randomly selected, were analyzed by indirect immunofluorescence for the distance of the 10 most proximal S. aureus organisms to the epithelium. The distances of 390 bacterial cells from the epithelium were determined using an image analyzing system. From these, only 4.3% were found at a distance of zero from the epithelium. A similar result was obtained by analyzing 12 thin sections of one polyp (not shown). Furthermore, using SEM, an area of 33 mm² bronchial tissue from three bronchi was screened for the adherence of S. aureus to the epithelium. S. aureus was found adhering to neither ciliated (Figure 2A) nor unciliated (Figure 2B) cells. Additionally, in screening nine different ultrathin sections by TEM, S. aureus was found to be neither adhering to the bronchial epithelium nor present intracellularly (Figure 1C). The percentage of unciliated cells, cells with microvilli, and cells with an uneven distribution of cilia or short cilia ranged from 1.1 to 46% (mean: 26.3%) as demonstrated by SEM when a 33 mm² area of bronchial tissue was scanned.

View larger version (181K): [in this window] [in a new window]   Figure 2.   Scanning electron micrograph of a bronchus of an S. aureus-infected lung of a cystic fibrosis patient (A, B, D) and a transmission electron micrograph (C) of the same patient. Note the intact, ciliated epithelium (A, inset; C), devoid of adhering S. aureus (A-D). Note naked epithelium (B, inset), devoid of adhering S. aureus. Note epithelial cells, devoid of intracellular S. aureus (C). Note preserved sputum within the bronchus (D). Magnifications: (A) Upper part: ×300, inset: ×3,000. (B) Upper part: ×300, inset: ×600. (C) ×1,950. (D) ×150. Bars: (A) Upper part: 25 µm, inset: 2.5 µm. (B) Upper part: 25 µm, inset: 12.5 µm. (C) 3.4 µm. (D) 83 µm.

Interestingly, although it was totally obstructed, the bronchus shown in Figures 2A and 2C revealed totally intact cilia after preparation of the specimen for SEM. Since glutaraldehyde poorly links the polysaccharide components of the obstructing sputum to the epithelium, the whole sputum material, together with the bacterial load, is lost during SEM preparation and the nature of the epithelium becomes visible. Using a special glue, the sputum material is preserved, revealing intensive obstruction (Figure 2D).

In summary, the results suggest that S. aureus is predominantly mucus-bound in CF airways. In order to investigate this hypothesis in more detail in vitro, a complex cell culture system was used.

Adherence of S. aureus to Mucus-secreting and Mucus-depleted Cell Balls from CF Patients and Healthy Individuals

Cell balls (Figure 3A) from CF patients and normal healthy individuals did not differ in the ratio of goblet cells to ciliated cells and nonciliated cells after 3 mo of culture. Thus, this material was appropriate for our adherence experiments. When cell balls from CF patients and healthy individuals were used for the adhesion assays without prior washing, mucicarmine staining demonstrated large amounts of mucus on top of ciliated cells as well as on nonciliated cells. S. aureus was found by SEM to adhere to the mucus overlaying the CF and control cell balls (Figures 3B and 3C) at a distance between the ciliated cell membrane and the bacteria of approximately 10-20 µm. Bacteria:cell ratios of 21.9 ± 1.5 were counted on CF cell balls (Figure 4). Similar bacteria:cell ratios were found on epithelial cells from healthy individuals, resulting in a nonsignificant difference between CF cells and non-CF cells (P = 0.27). These results suggest that S. aureus adheres predominantly to mucus components secreted from cells of the respiratory epithelium and that CF mucus does not contain a higher number of receptors for S. aureus adhesins or receptors which bind S. aureus more avidly compared with mucus from healthy individuals.

View larger version (9K): [in this window] [in a new window]   Figure 4.   Adherence of S. aureus to primary nasal epithelial-cell cell balls of five patients with CF (open columns) and five normal healthy individuals (closed columns). Unwashed (with mucus) (1) and washed (without mucus) (2) cell balls were incubated with S. aureus for 2 h, nonadherent bacteria were removed using a cell strainer and centrifugation, and adherent bacteria were quantified by SEM. (1) Unwashed cell balls; (2) washed, mucus- depleted cell balls; (3) cell balls treated with 1 µg/ml neutrophil elastase before bacterial incubation; (4) cell balls treated with 1 µg/ ml neutrophil elastase after bacterial incubation. For quantification of adherent bacteria, about 500 cells from 10 cell balls were examined. Values represent means ± SD of five independent experiments of each of the individuals. *CF patients: 1 and 2: P < 0.001. Normal individuals: 1 and 2: P < 0.007; Student's t test.

In order to investigate whether S. aureus adheres to the cilia or the membrane components of the nasal epithelial cells, cell balls were mucus depleted by extensive washing with PBS. Thereafter, mucus was visible by mucicarmine staining only in the intercellular spaces of cell balls. Adherence of S. aureus to cell membranes and cilia of 3- to 5-mo-old cell balls from CF patients (Figure 3D) or healthy individuals was not visible. However, S. aureus was found to adhere to the intercellular spaces of cell balls (bacteria/ cell ± SD: CF: 4.2 ± 0.3; non-CF: 5.0 ± 1.5; Figure 4). The presence of adhering bacteria and positive mucicarmine stain correlated significantly, suggesting that S. aureus adhered to the mucus which was still present after washing. There was no significant difference in the adherence of S. aureus to mucus-depleted cell balls of CF patients or healthy individuals (CF:non-CF: P = 0.38). The difference in binding of S. aureus to unwashed and mucus-depleted cell balls was significant both for CF cell balls (P < 0.001) and cell balls derived from normal healthy individuals (P < 0.007) (Figure 4). The results suggest that S. aureus adheres predominantly to mucus and that the mucus depleted respiratory epithelial cell membrane or the cilia of CF patients do not express receptors for S. aureus adhesins.

Effect of Hypersecretion on the Adherence of S. aureus to Mucus Components on Cell Balls of CF Patients and Healthy Individuals

In order to investigate the effect of hypersecretion on the adherence of S. aureus, cell balls from CF patients and healthy individuals were treated with 1 µg/ml NE for 30 min. This treatment did not induce cytotoxic effects, as shown by an unchanged structure of the epithelium and cilia compared by SEM with untreated controls (Figure 5A). Likewise, the function of the cilia was also not impaired, as the movement of the cell balls did not differ significantly before (8-12 rpm) and after (8-12 rpm) proteinase treatment (data not shown). NE induced hypersecretion was shown by the significantly increased numbers of granules located outside the apical region of the cells after treatment (Figure 5A) which were not detectable before proteinase treatment. Hypersecretion was also visible by a more intensive mucicarmine staining of the cells after treatment (not shown).

View larger version (107K): [in this window] [in a new window]   Figure 5.   (A) Scanning electron micrograph of cell balls of a CF patient, and (B) a transmission electron micrograph of cell balls of the same patient. Unwashed cell balls were treated with 1 µg/ml neutrophil elastase prior to incubation with S. aureus for 2 h, nonadherent bacteria were removed using a cell strainer, and adherent bacteria were quantified by SEM. In A, note membranes and cilia free of S. aureus as well as secretion of mucus granula (arrow). In B, S. aureus attached and embedded in a mucus granulum secreted from a cell ball. Magnifications: (A) ×8,100. (B) ×11,500. Bars: (A) 0.93 µm. (B) 0.8 µm.

When S. aureus was incubated with NE pretreated cell balls, no bacteria adherent to the mucus layer on top of the cilia (10-20 µm from the apical membrane) were detectable by SEM (Figure 4). As demonstrated by TEM, S. aureus was associated with the large granules distant from the cells (Figure 5B). Cells from CF patients and healthy individuals did not show different results (Figure 4). NE treatment was also effective in clearing S. aureus from the cell-associated mucus, when cell balls were incubated first with the bacteria and NE treatment was initiated 30 min thereafter (not shown).

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

S. aureus is provided with a variety of membrane-bound molecules which may act as adhesins for different surface structures. In general, the nature of these adhesins is thought to be protein because treatment of S. aureus with proteinases significantly reduces adherence (14, 17, 19, 20). Receptors for such adhesins on epithelial surfaces include membrane bound glycolipids (13, 15); connective tissue glycoproteins, such as fibronectin (30); collagens (31); laminin (32); vitronectin (33); and proteoglycans, such as heparansulfate (14). From these data, therefore, it is conceivable that S. aureus chronically infecting the lower airways of CF patients adheres to respiratory epithelial cell membranes and to components of the connective tissue.

In the present study we present evidence, however, that adherence of S. aureus to membrane components, including cilia of human respiratory epithelial cells, is negligible compared with adherence to components of the secreted mucus layer. This notion is supported by the following observations. First, SEM and TEM examinations of S. aureus-infected bronchi and bronchioli of three CF patients after lobectomy, as well as the examination of one infected polyp by immunofluorescence, clearly showed that S. aureus was embedded in the obstructing mucus but not in the vicinity (about 10 µm) of the respiratory epithelium. Second, mucus-depleted cell balls, when examined by SEM after incubation with S. aureus, revealed only low numbers of S. aureus adhering to cellular interspaces where mucus was still present after washing. Third, when cell balls were not mucus-depleted, large numbers of S. aureus adhered to mucus on top of the cilia and overlaying nonciliated cells. Fourth, the induction of mucus hypersecretion significantly enlarged the distance of S. aureus from the cells.

This body of information confirms results from previous in vitro and animal studies. Sanford and associates (18) demonstrated that S. aureus was associated with the mucus gel coating the upper respiratory tract of ferrets and that S. aureus bound to purified mucins in vitro (18, 19). Also, Shuter and colleagues (20) showed mucin binding of S. aureus in vitro and proved that S. aureus adherence to mucus-coated cells was greater than to nonmucus-coated cells.

The molecular mechanism of the binding of S. aureus to components of secreted mucus is not clear. Calcium ions may act as a bridge to crosslink the surface of S. aureus to negatively charged mucin receptors (19), and secretory IgA may also be involved in the binding of S. aureus to mucins (34). Based on the studies of Liang and coworkers (14) and Thomas and associates (19), mucin binding may be due to ionic interactions between the S. aureus cell surface amino groups of several mucin-binding S. aureus peptides (19, 20) and sulfate and carboxyl groups of mucins or proteoglycans, including chondroitin sulfate or heparin sulfate. Interestingly, mucins secreted from CF respiratory epithelial cells showed a higher level of sulfation (11, 12, 21). The present results suggest that increased sulfation of CF mucins does not result in an increased binding capacity to mucins secreted from CF cell balls when compared with controls.

Although the CF patients were chronically infected with S. aureus for several years, the percentage of ciliated cells, a marker for an intact respiratory epithelium, was surprisingly high in some infected bronchi and bronchioli as shown by SEM. Thus, binding of S. aureus to the mucus layer may prevent tissue damage due to S. aureus toxins (35), invasion, and systemic infection with S. aureus in CF. Interestingly, S. aureus sepsis has never been described in CF patients.

Mucus hypersecretion induced by NE and cathepsin G (23) may additionally be active in preventing the direct contact of S. aureus with the respiratory epithelial membrane. NE is present very early in CF airways even before bacterial infection (36). By cleaving secreted mucins (37) (as has been shown for pronase [20]) and proteoglycans such as chondroitin sulfate (23), NE may further increase the distance of S. aureus to the epithelial cells, as demonstrated in the present study.

Similarly to S. aureus, Pseudomonas aeruginosa, the other major pathogen in CF patients, may also adhere to mucins of the CF respiratory tract rather than to the CF epithelial membrane in vivo. Such a mechanism has been proposed particularly by Ramphal (38), based on the observation that until now studies have been lacking which unequivocally prove direct adhesion of these pathogens to the membrane of respiratory epithelial cells in CF patients. Rather, the organism has been found in the mucus (39).

Since smokers also reveal increased mucin production (41) but do not show chronic S. aureus lung infection, it seems likely that the cause for the chronic respiratory infections with S. aureus and other bacterial pathogens is due to the impaired mucociliary clearance in CF patients. This view is supported by a recent study in CF mice (42) where the authors show that decreased airway mucociliary clearance is present before any airway infection. The nature of the factors which are responsible for impaired mucociliary clearance in CF are still a matter of debate and may include a reduced water content of the mucus (1), an increased ion concentration (43), and an abnormal chemical composition of the mucin components (11, 12, 44).