|Home | About | Journals | Submit | Contact Us | Français|
Moraxella catarrhalis is a human pathogen causing otitis media in infants and respiratory infections in adults, particularly patients with chronic obstructive pulmonary disease. The surface protein Hag (also designated MID) has previously been shown to be a key adherence factor for several epithelial cell lines relevant to pathogenesis by M. catarrhalis, including NCIH292 lung cells, middle ear cells, and A549 type II pneumocytes. In this study, we demonstrate that Hag mediates adherence to air-liquid interface cultures of normal human bronchial epithelium (NHBE) exhibiting mucociliary activity. Immunofluorescent staining and laser scanning confocal microscopy experiments demonstrated that the M. catarrhalis wild-type isolates O35E, O12E, TTA37, V1171, and McGHS1 bind principally to ciliated NHBE cells and that their corresponding hag mutant strains no longer associate with cilia. The hag gene product of M. catarrhalis isolate O35E was expressed in the heterologous genetic background of a nonadherent Haemophilus influenzae strain, and quantitative assays revealed that the adherence of these recombinant bacteria to NHBE cultures was increased 27-fold. These experiments conclusively demonstrate that the hag gene product is responsible for the previously unidentified tropism of M. catarrhalis for ciliated NHBE cells.
Moraxella catarrhalis is a gram-negative pathogen of the middle ear and lower respiratory tract (29, 40, 51, 52, 69, 78). The organism is responsible for ~15% of bacterial otitis media cases in children and up to 10% of infectious exacerbations in patients with chronic obstructive pulmonary disease (COPD). The cost of treating these ailments places a large financial burden on the health care system, adding up to well over $10 billion per annum in the United States alone (29, 40, 52, 95, 97). In recent years, M. catarrhalis has also been increasingly associated with infections such as bronchitis, conjunctivitis, sinusitis, bacteremia, pneumonia, meningitis, pericarditis, and endocarditis (3, 12, 13, 17-19, 24, 25, 27, 51, 67, 70, 72, 92, 99, 102-104). Therefore, the organism is emerging as an important health problem.
M. catarrhalis infections are a matter of concern due to high carriage rates in children, the lack of a preventative vaccine, and the rapid emergence of antibiotic resistance in clinical isolates. Virtually all M. catarrhalis strains are resistant to β-lactams (34, 47, 48, 50, 53, 65, 81, 84). The genes specifying this resistance appear to be gram positive in origin (14, 15), suggesting that the organism could acquire genes conferring resistance to other antibiotics via horizontal transfer. Carriage rates as high as 81.6% have been reported for children (39, 104). In one study, Faden and colleagues analyzed the nasopharynx of 120 children over a 2-year period and showed that 77.5% of these patients became colonized by M. catarrhalis (35). These investigators also observed a direct relationship between the development of otitis media and the frequency of colonization. This high carriage rate, coupled with the emergence of antibiotic resistance, suggests that M. catarrhalis infections may become more prevalent and difficult to treat. This emphasizes the need to study pathogenesis by this bacterium in order to identify vaccine candidates and new targets for therapeutic approaches.
One key aspect of pathogenesis by most infectious agents is adherence to mucosal surfaces, because it leads to colonization of the host (11, 16, 83, 93). Crucial to this process are surface proteins termed adhesins, which mediate the binding of microorganisms to human cells and are potential targets for vaccine development. M. catarrhalis has been shown to express several adhesins, namely UspA1 (20, 21, 59, 60, 77, 98), UspA2H (59, 75), Hag (also designated MID) (22, 23, 37, 42, 66), OMPCD (4, 41), McaP (61, 100), and a type 4 pilus (63, 64), as well as the filamentous hemagglutinin-like proteins MhaB1, MhaB2, MchA1, and MchA2 (7, 79). Each of these adhesins was characterized by demonstrating a decrease in the adherence of mutant strains to a variety of human-derived epithelial cell lines, including A549 type II pneumocytes and Chang conjunctival, NCIH292 lung mucoepidermoid, HEp2 laryngeal, and 16HBE14o-polarized bronchial cells. Although all of these cell types are relevant to the diseases caused by M. catarrhalis, they lack important aspects of the pathogen-targeted mucosa, such as the features of cilia and mucociliary activity. The ciliated cells of the respiratory tract and other mucosal membranes keep secretions moving out of the body so as to assist in preventing colonization by invading microbial pathogens (10, 26, 71, 91). Given this critical role in host defense, it is interesting to note that a few bacterial pathogens target ciliated cells for adherence, including Actinobacillus pleuropneumoniae (32), Pseudomonas aeruginosa (38, 108), Mycoplasma pneumoniae (58), Mycoplasma hyopneumoniae (44, 45), and Bordetella species (5, 62, 85, 101).
In the present study, M. catarrhalis is shown to specifically bind to ciliated cells of a normal human bronchial epithelium (NHBE) culture exhibiting mucociliary activity. This tropism was found to be conserved among isolates, and analysis of mutants revealed a direct role for the adhesin Hag in binding to ciliated airway cells.
Strains and plasmids are described in Table Table1.1. M. catarrhalis and Escherichia coli were cultured as previously reported by our laboratory (7, 22, 23, 41, 42, 100). Haemophilus influenzae was grown at 37°C using brain heart infusion medium (BD Diagnostic Systems) supplemented with 50 μg/ml hemin chloride and 10 μg/ml NAD (Sigma). Antimicrobial supplementation for M. catarrhalis involved kanamycin (20 μg/ml), spectinomycin (15 μg/ml), or Zeocin (5 μg/ml). Recombinant strains of Escherichia coli were selected with chloramphenicol (15 μg/ml). Recombinant H. influenzae bacteria were selected with 50 μg/ml of spectinomycin.
The human epithelial cell lines A549 (type II alveolar lung epithelium; ATCC CCL85) and NCIH292 (lung mucoepidermoid; ATCC CRL-1848) were cultured as reported by Timpe and colleagues (100). The methods described by Krunkosky et al. (57, 58) were used to expand, cryopreserve, and culture NHBE cells (LONZA) in an air-liquid interface system. By means of these methods, the NHBE cultures were grown on Transwell permeable inserts (Corning), and their apical surfaces were exposed to air for at least 3 weeks prior to use in biological assays to ensure the proper development of cilia.
Standard molecular biology techniques were performed as described elsewhere (82), using E. coli strain EPI300 (Epicentre Biotechnologies), H. influenzae strain DB117 (86), or M. catarrhalis strain O35E.TN2 (42) as the host for recombinant DNA manipulations. The procedure used for electroporating M. catarrhalis has been described in detail by others (42) and was also utilized to electroporate H. influenzae DB117. Plasmid DNA was purified with a QIAprep Spin miniprep system (Qiagen).
The previously described plasmid pELO35.Hag (23) was digested with BamHI, and a 6.2-kb fragment corresponding to the O35E hag open reading frame was excised from an agarose gel, purified with the High Pure PCR product purification kit (Roche), and ligated into the BamHI site of pWW115 (106). This ligation mixture was introduced into the M. catarrhalis hag mutant strain O35E.TN2 (42) by electroporation. Spectinomycin-resistant colonies were then screened in colony blot immunoassays to identify recombinant clones expressing Hag. This approach yielded a strain containing the plasmid pRB.Hag. A similar strategy was used to subclone the 3.5-kb insert of the previously characterized plasmid pBBHS10.9 (22) in pWW115. This insert specifies the expression of O35E-Hag in which residues 71 to 384, 746 to 1193, and 1546 to 1707 have been removed (22). The resulting pWW115-based plasmid construct was designated pLF.10.9.
The method used to prepare whole-cell lysates is described elsewhere (28, 74). These preparations were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using 7.5% polyacrylamide gels and transferred to polyvinylidene difluoride (PVDF) membranes (Millipore). Western blot experiments were performed as reported by Bullard and colleagues (22, 23). Protein bands were visualized by chemiluminescence, using the SuperSignal West Dura extended duration substrate (Thermo Scientific) and a Foto/Analyst Luminary/FX imaging system (Fotodyne Inc.). The polyclonal murine antibodies against the C terminus of O35E-Hag (22) and residues 51 to 650 of O35E-McaP (61) that were used in the Western blot experiments have been described previously.
For colony blot immunoassays, patched colonies were transferred to a PVDF membrane and incubated at 60°C for 1 h. The membrane was then incubated for 1 h at room temperature in phosphate-buffered saline (PBS) supplemented with 0.05% Tween 20 (PBST) and 3% skim milk (blocking buffer). Detection of protein was accomplished with the aforementioned mouse polyclonal antibodies against the C terminus of Hag, followed by incubation with a secondary goat anti-mouse antibody conjugated to horseradish peroxidase (Southern Biotech). Primary and secondary antibodies were diluted in blocking buffer. Antibody binding was detected by chemiluminescence as described above.
Quantitative adherence assays were performed as reported by Balder et al. (7). Briefly, plate-grown bacteria were suspended in sterile PBS supplemented with 0.15% gelatin. Portions of these suspensions (107 bacteria) were used to inoculate 105 epithelial cells. After a brief centrifugation (165 × g, 5 min), the infected cells were incubated at 37°C, washed to remove unbound bacteria, treated with a solution containing saponin and EDTA, serially diluted, and spread onto agar plates. After overnight incubation at 37°C, CFU were counted, and these numbers were used to determine the binding efficacy of bacterial strains. The results are expressed as the percentage (± standard error) of bound bacteria relative to the inoculum used to infect epithelial cells. Duplicate assays were repeated on at least three separate occasions for each strain.
Following infection with M. catarrhalis or H. influenzae strains, air-liquid interface cultures of NHBE cells were washed to remove unbound bacteria, and the epithelium-containing inserts were fixed with 4% paraformaldehyde in PBS. The epithelium-containing inserts were then treated with PBS supplemented with 0.5% Triton X-100, washed with PBST, and blocked overnight at 4°C with 10% normal goat serum diluted in PBST (blocking buffer). Next, the epithelium-containing inserts were probed at room temperature for 1 h with a rabbit antibody against β-tubulin (Abcam) and murine polyclonal antibodies against either residues 51 to 650 of O35E-McaP (for NHBE cells infected with M. catarrhalis) or the C terminus of O35E-Hag (for NHBE cells infected with recombinant H. influenzae bacteria). These antibodies were used at a dilution of 1:200 in blocking buffer. The epithelium-containing inserts were subsequently washed with PBST, and incubated for 1 h at room temperature with a goat anti-rabbit antibody labeled with Alexa Fluor 488 (Molecular Probes) and a goat anti-mouse antibody labeled with Alexa Fluor 546 (Molecular Probes) diluted 1:400 in blocking buffer. Following this incubation, the epithelium-containing inserts were washed with PBST, and F-actin was labeled with Alexa Fluor 647 phalloidin (Molecular Probes). Lastly, the epithelium-containing inserts were stained with DAPI (4′,6-diamidino-2-phenylindole) (Molecular Probes), washed with PBST, removed from their Transwell supports, mounted onto glass slides with the SlowFade antifade reagent (Molecular Probes), and examined by laser scanning confocal microscopy (LSCM) using a Zeiss LSM 510 Meta confocal system. Immunofluorescent staining and LSCM (IFS-LSCM) were performed on at least two separate occasions for each strain.
Statistical analyses were performed with the GraphPad Prism software using the Mann-Whitney test. P values of <0.05 were considered to be statistically significant.
To date, most published reports examining M. catarrhalis adherence have utilized cell lines of undifferentiated epithelial cells in submerged cultures. In an effort to use a model system that would mimic the structure and function of the respiratory epithelium more accurately, we studied the interaction of M. catarrhalis with NHBE cells grown in an air-liquid interface culture system previously described for examining M. pneumoniae cytadherence (58). Under the growth conditions used, NHBE cultures form a pseudostratified epithelium with tight junctions containing ciliated as well as nonciliated cells. This airway epithelium also exhibits transepithelial resistance, mucus secretion, mucociliary activity, and an apical surface not submerged in medium, therefore representing an environment similar to that of the airway lumen in vivo (58).
We first tested whether M. catarrhalis binds to these differentiated NHBE cultures by the use of IFS-LSCM. Figures 1A to E depict a representative confocal image of an NHBE culture infected for 3 h with the M. catarrhalis wild-type (WT) strain O35E as viewed from the apical surface of the epithelium. To visualize the cellular architecture of the culture, the infected cells were labeled with phalloidin (binds F-actin and outlines the cells; Fig. Fig.1A),1A), an antibody against β-tubulin (highlights tufts of cilia on the surface of cells; Fig. Fig.1C),1C), and DAPI (stains nucleic acids; Fig. Fig.1D).1D). Bacteria were visualized using an antibody against the M. catarrhalis adhesin McaP (Fig. (Fig.1B)1B) and the DAPI stain (Fig. (Fig.1D).1D). Figure Figure1E1E shows a merged image of Fig. 1A to D. This merged image, along with Fig. Fig.3A3A and the staining patterns observed in Fig. 1B and C, clearly demonstrates an association between M. catarrhalis O35E and the ciliated cells of the airway epithelium. Moreover, a cross-sectional view (i.e., z-plane stack of LSCM images) of the infected culture revealed that M. catarrhalis binds to the apical portion of cilia (Fig. (Fig.1F1F and and22 ).
Previously characterized isogenic mutant strains of M. catarrhalis O35E were tested for their ability to bind to NHBE cultures. This collection of mutants included those lacking expression of the adhesins UspA1 (1), McaP (100), outer membrane protein CD (41), MhaB1 (7), MhaB2 (7), and Hag (42). These strains were first examined by IFS-LSCM as described above, and mutants that appeared to be reduced in their attachment were then tested using quantitative adherence assays (data not shown). Using this approach, we discovered that expression of the adhesin Hag is necessary for the binding of M. catarrhalis strain O35E to ciliated NHBE cells. Figure Figure4A4A shows the adherence of O35E in comparison to that of its hag mutant strain O35E.TN2 after three different infection periods. These quantitative results demonstrate that lack of Hag expression considerably reduces adherence to the airway epithelium at all three time points (85 to 95% reduction). Infected cultures were also analyzed by IFS-LSCM, and Fig. 4B and C depict representative images of NHBE cultures incubated with strains O35E and O35E.TN2 for 15 min, respectively. Figure Figure4B4B shows that the WT isolate associates exclusively with ciliated cells, which is consistent with the results presented in Fig. Fig.11 after a 3-h infection. As illustrated in Fig. Fig.4C,4C, the hag mutant O35E.TN2 no longer associates with ciliated cells. This pattern of binding for O35E and O35E.TN2 was also observed at the 3 h and 6 h infection time points (data not shown).
To determine whether the attachment of M. catarrhalis O35E to ciliated NHBE cells and the involvement of Hag in this binding are strain-specific phenotypes, previously characterized hag mutants of M. catarrhalis isolates O12E, V1171, McGHS1, and TTA37 (23) were examined for their ability to adhere to NHBE cultures. Results of quantitative assays are presented in Fig. 5A and D and show that O12E binds to NHBE cells very efficiently (mean adherence of 110%), while strain V1171 attached at lower levels (mean adherence of 5%). Despite this difference in the overall binding capabilities of O12E and V1171, disruption of the hag gene in these isolates resulted in significantly lower adherence (96% reduction for O12E.Hag and 64% reduction for V1171.Hag). IFS-LSCM confirmed that strains O12E and V1171 associate with ciliated cells (Fig. 5B and E, respectively). Figures 5C and F illustrate the greatly reduced adherence of the hag mutants O12E.Hag and V1171.Hag to NHBE cultures, respectively. Similar patterns of binding with the WT isolate and hag mutants were also observed by IFS-LSCM for strains McGHS1, McGHS1.Hag, TTA37, and TTA37.Hag (data not shown), and quantitative assays revealed adherence values of 31.6% ± 10.3%, 3.6% ± 1.4%, 19.6% ± 6.9%, and 1.5% ± 0.6%, respectively. Taken together, the results of these experiments demonstrate that the tropism of M. catarrhalis for ciliated NHBE cells and the role of the adhesin Hag in this interaction are conserved among all strains tested.
Wang et al. (106, 107) as well as Hoopman and colleagues (43) recently demonstrated the utility of the plasmid pWW115 for trans-complementing mutations in the genome of M. catarrhalis. Based on these investigators' successes, we sought to complement the adherence-negative phenotype of the hag mutant O35E.TN2 using this vector. Complementation was accomplished by subcloning the insert of the previously described plasmid pELO35.Hag, which specifies expression of the O35E-Hag protein in E. coli (23), into the BamHI site of pWW115. The resulting construct, designated pRB.Hag, restored the expression of Hag in the mutant O35E.TN2 (Fig. (Fig.6A,6A, lane 3). Of note, the binding of polyclonal antibodies against the adhesin McaP was used to demonstrate that equivalent amounts of cell lysates were analyzed in these Western blot experiments (Fig. (Fig.6B).6B). Figure Figure6C6C shows that the complemented mutant bound to NHBE cultures at WT levels, while O35E.TN2 harboring the vector pWW115 showed no restoration of adherence. IFS-LSCM demonstrated that the plasmid pRB.Hag also restored O35E.TN2's ability to attach to ciliated NHBE cells (data not shown).
Additional experiments were performed to confirm a direct role for Hag in M. catarrhalis adherence to ciliated NHBE cells. To address the possibility that the absence of Hag in the outer membrane of the aforementioned mutants affected proper surface display of other adhesins, which themselves mediate attachment to ciliated cells, the plasmid pRB.Hag was introduced into the H. influenzae strain DB117. This organism does not normally adhere well to human epithelial cells (59, 94) and therefore provides a heterologous genetic background that is appropriate to examine the adhesive properties of Hag. Importantly, this strain has also been shown to maintain the plasmid pWW115 (106). Figure Figure77 shows that DB117 harboring pRB.Hag expresses Hag (Fig. (Fig.7A,7A, lane 2) and gains the ability to associate apically with ciliated NHBE cells (Fig. (Fig.3B3B and and7B).7B). Moreover, quantitative assays revealed that the adherence of H. influenzae expressing O35E-Hag to NHBE cultures was 27 orders of magnitude greater than that of recombinant bacteria carrying the vector pWW115 (Fig. (Fig.7C).7C). These experiments conclusively demonstrate that the hag gene product is responsible for the tropism of M. catarrhalis for ciliated NHBE cells.
Previous work by our laboratory indicated that amino acids 385 to 745 of O35E-Hag specify adherence to the human lung cell line NCIH292 (22). Other investigators also reported that a purified recombinant protein corresponding to amino acids 715 to 863 binds to A549 type II pneumocytes (37). To gain more information pertaining to the region of O35E-Hag important for attachment to ciliated NHBE cells, we subcloned the insert of the previously characterized plasmid pBBHS10.9 (22) into the BamHI site of pWW115. This construct (i.e., pBBHS10.9) specifies the expression of O35E-Hag in which residues 71 to 384, 746 to 1193 (i.e., most of the proposed A549 binding region), and 1546 to 1707 have been removed. Importantly, Hag was found to confer E. coli with the ability to bind to NCIH292 cells despite these mutations (22). The resulting pWW115-based construct, designated pLF.10.9, was introduced into the hag mutant strain O35E.TN2 in order to complement its adherence-negative phenotype.
As shown in lane 4 of Fig. Fig.6A,6A, O35E.TN2 harboring pLF.10.9 expresses a Hag protein migrating with a lower molecular mass than that of the WT molecule (Fig. (Fig.6A,6A, lanes 1 and 3), which is consistent with the aforementioned deletions introduced in O35E-Hag. As expected, pLF.10.9 restored the adherence of O35E.TN2 to NCIH292 cells to WT levels (Fig. (Fig.6D).6D). However, the mutated Hag protein did not complement the adherence-negative phenotype of O35E.TN2 for NHBE cultures (Fig. (Fig.6C).6C). Consistent with these results, we discovered that pLF.10.9 increased the binding of H. influenzae DB117 to NCIH292 cells by 12 orders of magnitude (Fig. (Fig.7D)7D) but did not confer these recombinant bacteria with the ability to attach to NHBE cultures (Fig. (Fig.7C).7C). These experiments indicate that the cell binding domain of O35E-Hag for NHBE cultures differs from that necessary for adherence to NCIH292 cells.
Earlier reports by our laboratory (23, 42) and others (37, 66) demonstrated that Hag expression is important for attachment of M. catarrhalis to A549 human lung cells. These reports were published prior to the availability of plasmid pWW115 (106), and therefore, the mutant strains described in these studies were not complemented due to a lack of a suitable M. catarrhalis vector. For these reasons, we measured the adherence of O35E.TN2 containing plasmids pRB.Hag and pLF.10.9 to A549 monolayers. Figure Figure6E6E shows that WT O35E-Hag (i.e., pRB.Hag) fully complemented the adherence-negative phenotype of O35E.TN2, while the protein specified by pLF.10.9 did not restore binding. These findings were supported by experiments with recombinant H. influenzae bacteria in which full-length O35E-Hag increased binding to A549 cells 18-fold (pRB.Hag in Fig. Fig.7E),7E), whereas the mutated adhesin did not (pLF.10.9 in Fig. Fig.7E).7E). Taken together, the data demonstrate that Hag directly mediates adherence to A549 pneumocytes and that the region of O35E-Hag specifying this attachment differs from that necessary for binding to NCIH292 cells.
The first goal of this study was to determine whether M. catarrhalis adhered to air-liquid interface cultures of NHBE cells. These cultures represent an environment resembling the airway lumen in vivo in that they are fully differentiated, form a pseudostratified epithelium with tight junctions, and contain ciliated and mucus-producing goblet cells (58). A key aspect of this culture system consisted of exposing the apical surface of NHBE cells to air for at least 3 weeks prior to performing biological assays to ensure the proper development and function of cilia. Using a combination of visual (Fig. (Fig.1)1) and quantitative (Fig. (Fig.4A)4A) assays, we demonstrated that the M. catarrhalis strain O35E adheres efficiently to these cultures. Moreover, these experiments revealed that the organism binds principally to ciliated cells and that this association appears to take place at the apical portion of cilia (Fig. (Fig.1,1, ,2,2, and and3A).3A). Tropism for ciliated cells has been reported for other pathogens of the respiratory tract, including Bordetella pertussis (62, 80, 89, 101), Bordetella bronchiseptica (5, 33, 85), M. hyopneumoniae (44, 45), M. pneumoniae (58), P. aeruginosa (6, 31, 38, 46, 108), and Klebsiella pneumoniae (36). B. pertussis and M. pneumoniae have been shown to associate primarily with the base of cilia, whereas P. aeruginosa preferentially locates at the tip. Interestingly, though not the focus of the current study, we did not detect M. catarrhalis inside NHBE cells at any of the infection periods tested (up to 24 h; data not shown). This observation is in contrast with reports that the organism invades immortalized cell lines in submerged cultures, including Chang, A549, and BEAS-2B cells (87, 88, 90). M. catarrhalis was also reported to invade normal small airway epithelial cells (87, 88), but it is not clear whether these cells were grown in submerged cultures and/or under conditions promoting their differentiation.
The second goal of this study was to determine which M. catarrhalis adhesin(s) might be important for attachment to ciliated NHBE cells. To accomplish this, we utilized IFS-LSCM to examine the adherence of previously characterized mutants of strain O35E. These experiments identified Hag as a potential adhesin for ciliated cells (Fig. 4B and C), and quantitative assays verified that the hag mutant O35E.TN2 is impaired in its attachment to NHBE cultures (Fig. (Fig.4A).4A). Importantly, the adherence to ciliated NHBE cells and the involvement of Hag in this binding were found to be conserved traits among M. catarrhalis isolates of various clinical and geographical origins (Fig. (Fig.5).5). Differences in the adherence values of WT strains were noted after a 3-h infection, with almost 50% of O35E (Fig. (Fig.4A),4A), only 5% of V1171 (Fig. (Fig.5D),5D), 19% of TTA37 (data not shown), more than 100% of O12E (Fig. (Fig.5A),5A), and 31% of McGHS1 (data not shown) cells interacting with NHBE cultures. Such variability in the overall adherence of M. catarrhalis WT isolates is not uncommon and has been previously reported (7, 22, 23, 37, 59, 66). For example, strain O12E consistently binds more avidly than does O35E to various human airway epithelial cell lines grown in submerged cultures (7, 22, 23, 59). The epithelial cell type used to perform assays also appears to influence the overall adherence levels of a specific isolate. Strain V1171, for instance, has been shown to attach relatively well to NCIH292 and Chang cells (42 to 46%) (22) but poorly to A549 pneumocytes (<1%) (23). These differences are likely caused by observed variability in expression levels of adhesins (7, 23, 59) and autoagglutination properties of strains (23, 42, 76, 96), as well as differing cell binding specificities of individual adhesins (20, 21).
Expression of O35E-Hag in the heterologous genetic background of H. influenzae conclusively demonstrated that this protein mediates adherence to the apical portion of cilia on NHBE cells (Fig. (Fig.3B3B and 7B and C). Previous research has shown that O35E-Hag forms elongated structures that protrude from the surface of M. catarrhalis (76). This appears to be a conserved feature of molecules mediating adherence to ciliated cells, as the type 1 fimbriae of K. pneumoniae (36), the pilus of P. aeruginosa (31, 46), the filamentous hemagglutinin FHA of Bordetella species (33, 62, 80), and the terminal organelle of M. pneumoniae (8, 9, 30, 49, 54-56, 58) have been implicated in such interactions. Hag (also designated MID) is one of the best studied adhesins of M. catarrhalis, and several binding domains of this multifunctional adherence factor have been identified, including those for NCIH292 cells (22), immunoglobulin D (73), type IV collagen (22), A549 cells (37), and primary cultures of human middle ear epithelium (22). Our experiments with recombinant bacteria carrying the plasmid pLF.10.9 indicate that the region of Hag important for adherence to NCIH292 cells differs from that necessary for attaching to A549 cells or NHBE cultures (panels C to E in both Fig. Fig.66 and and7).7). The hypothesis that Hag's cell binding domain for A549 pneumocytes is the same as that for NHBE cultures is currently being investigated. These structure-function studies will not only improve our understanding of the molecular basis by which Hag contributes to host-pathogen interactions by M. catarrhalis, but may also identify amino acid motifs shared by bacterial adhesins that specify adherence to ciliated airway cells.
Our ability to clone hag in the vector pWW115 allowed us to revisit some of the previously published phenotypes associated with Hag expression, specifically adherence to A549 and NCIH292 cells (22, 23, 42). In these previous studies, it was demonstrated that lack of Hag expression by strain O35E results in decreased adherence to A549 and NCIH292 monolayers. Additionally, recombinant E. coli cells expressing O35E-Hag were shown to have gained the ability to bind to NCIH292 cells, but not to A549 pneumocytes. These results suggested that Hag alone is not sufficient for mediating attachment to A549 cells and led us to hypothesize that Hag requires expression of a coadhesin in order to perform its role in adherence to this particular cell line (22, 23). As expected, the plasmid pRB.Hag restored adherence of the hag mutant O35E.TN2 to NCIH292 and A549 cells to WT levels (Fig. 6D and E, respectively) and increased the binding of H. influenzae to NCIH292 monolayers by 10 orders of magnitude (Fig. (Fig.7D).7D). Surprisingly, we discovered that expression of O35E-Hag by H. influenzae also increased adherence to A549 cells 18-fold (Fig. (Fig.7E).7E). These last results contradict our previous finding that E. coli cells expressing O35E-Hag do not gain the ability to attach to A549 monolayers (23). One potential explanation for this discrepancy is that Hag's cell binding domain for A549 cells is properly displayed on the surface of H. influenzae but not that of E. coli. Alternatively, adherence to A549 pneumocytes may require posttranslational modification of Hag, which is achieved in the heterologous genetic background of H. influenzae but not that of E. coli. Nonetheless, the results of adherence assays with recombinant H. influenzae strains conclusively demonstrate that Hag alone is sufficient to mediate binding to A549 cells. These data are consistent with a previous report by Forsgren and colleagues (37) in which a purified recombinant protein corresponding to residues 715 to 863 of O35E-Hag was shown to bind to A549 monolayers by the use of a cellular enzyme-linked immunosorbent assay.
In summary, we demonstrated that M. catarrhalis has tropism for ciliated NHBE cells and that the adhesin Hag mediates this interaction. Most isolates tested to date contain a hag gene product (23, 37, 66, 76, 77, 105), and studies have shown that Hag is a major target of new immunoglobulin A antibodies purified from the sputum of COPD patients who have successfully cleared M. catarrhalis infections (68). Taken together, these data support a key role for the Hag protein in the pathogenesis of M. catarrhalis in the human airway. Luke et al. (64) recently used a chinchilla model to study the contribution of the M. catarrhalis type IV pilus to nasopharyngeal colonization. These investigators were able to show that the organism can stably colonize the nasopharynx and nasoturbinates of chinchillas for 2 weeks. These results are noteworthy in light of our findings with NHBE cultures, because the mucosa of the chinchilla nasopharynx consists primarily of ciliated columnar epithelium with some goblet cells, and these cell types are also present in the nasoturbinates (64). For these reasons, we believe that the air-liquid interface cultures of NHBE cells and the chinchilla model of nasopharyngeal colonization provide powerful, relevant, and complementary tools to study host-pathogen interactions by M. catarrhalis. Determining whether hag mutants are impaired in their ability to colonize the nasopharynx of chinchillas, together with testing the vaccinogenic potential of Hag in this model, represents a key area for future study.
This study was supported by a grant from NIH/NIAID (AI051477) to E.R.L. as well as start-up funds provided by the College of Veterinary Medicine at the University of Georgia.
We thank Eric Hansen at the University of Texas Southwestern Medical Center in Dallas, TX, for technical assistance and for providing the plasmid pWW115.
Editor: J. N. Weiser
Published ahead of print on 10 August 2009.
§Supplemental material for this article may be found at http://iai.asm.org/.