ABSTRACT
Mycoplasma pneumoniae is an important cause of respiratory disease, especially in school-age children and young adults. We employed normal human bronchial epithelial (NHBE) cells in air-liquid interface culture to study the interaction of M. pneumoniae with differentiated airway epithelium. These airway cells, when grown in air-liquid interface culture, polarize, form tight junctions, produce mucus, and develop ciliary function. We examined both qualitatively and quantitatively the role of mycoplasma gliding motility in the colonization pattern of developing airway cells, comparing wild-type M. pneumoniae and mutants thereof with moderate to severe defects in gliding motility. Adherence assays with radiolabeled mycoplasmas demonstrated a dramatic reduction in binding for all strains with airway cell polarization, independent of acquisition of mucociliary function. Adherence levels dropped further once NHBE cells achieved terminal differentiation, with mucociliary activity strongly selecting for full gliding competence. Analysis over time by confocal microscopy demonstrated a distinct colonization pattern that appeared to originate primarily with ciliated cells, but lateral spread from the base of the cilia was slower than expected. The data support a model in which the mucociliary apparatus impairs colonization yet cilia provide a conduit for mycoplasma access to the host cell surface and suggest acquisition of a barrier function, perhaps associated with tethered mucin levels, with NHBE cell polarization.
INTRODUCTION
Mycoplasma pneumoniae is a human respiratory tract pathogen primarily associated with tracheobronchitis and pneumonia. Infections are typically not life threatening but can be life altering due to the long-term lung damage that can result, including asthma and chronic obstructive pulmonary disease (1). M. pneumoniae initiates colonization of the airway mucosal epithelium via its terminal organelle (2–4). This highly differentiated polar structure functions in adhesion to host cell receptors, gliding motility, and cell division (5–8). Adhesin proteins P1 and P30 localize to the terminal organelle surface, where they participate directly in adherence to host cells and gliding motility (5, 6, 9, 10).
Colonization of the human airways requires circumvention of mucociliary defenses, which effectively obstruct, capture, and remove inhaled substances, limiting access to the epithelium (11–13). Previous M. pneumoniae colonization models employed submerged organ and tissue culture systems and have contributed to our current understanding of pathogen-host cell interactions, but they are limited in their ability to accurately reflect the environment of the airway mucosa (3, 4, 14–17). Mycoplasma-host interactions in vivo typically begin at mucosal barriers (11–13), which we define here as including ciliary motion, mucus production, and tight-junction formation (11, 18). Gliding motility is required for lung colonization in experimentally infected hamsters and mice (19, 20), and we speculate that this requirement begins with the need to cross the gel layer mucus and gain access to ciliated airway cells.
We previously described the use of normal human bronchial epithelial (NHBE) cells in air-liquid interface (ALI) culture to model M. pneumoniae interactions with the human airway (21) and noted then that impaired gliding motility was correlated with reduced colonization (22). Here, we extend that analysis further in three important ways. First, we assessed mycoplasma colonization of NHBE cells at different developmental stages. The airway epithelium is a pseudostratified population of cells from which underlying basal cells replace their differentiated counterparts in response to turnover or injury (23), and M. pneumoniae is likely to encounter basal cells, in addition to fully differentiated cells, during the course of infection. These analyses also allowed the correlation of colonization patterns specifically with host cell polarization, acquisition of mucus production, and cilium formation and activity. Second, we quantified mycoplasma colonization of fully differentiated NHBE cells spatially and temporally in order to define the steps in that process. Finally, we expanded the analysis of M. pneumoniae gliding and adherence mutants in this model. We observed a sharp decline in colonization efficiency very early, as NHBE cells polarized, followed by a second decline that coincided with gain of full mucociliary function. As expected, colonization was initiated by mycoplasma adherence to the tips of the cilia, with localization patterns suggesting downward movement from there to the base of the cilia. Lateral spread to nonciliated areas was less than expected, raising the possibility of a secondary physical or chemical barrier on the epithelial surface.
MATERIALS AND METHODS
Mycoplasma strains.Wild-type M. pneumoniae (strain M129, 17th broth passage) (15); P30 mutants II-3, II-7, and II-3R (7, 24); the P200 mutant (22); and the prpC mutant (25) were included in the current study. P30 is a terminal organelle protein required for adherence to host cells and gliding motility (5, 6, 26). II-3 has a frameshift mutation in the P30 gene (MPN453) (6). II-7 produces a C-terminally truncated P30 protein (27). II-3R resulted from a second-site mutation in II-3 that restores the correct reading frame for all but 17 residues of P30 (5). The P200 mutant resulted from an IS256 insertion in the P200 gene, MPN567 (22). The prpC protein phosphatase mutant resulted from transposon insertion in MPN247 (28). Table 1 summarizes the adherence and gliding phenotypes of these strains.
Hemadsorption and gliding motility phenotypes of M. pneumoniae strains used in the study
Gliding measurement.M. pneumoniae gliding phenotypes were confirmed as described previously (5) but with modifications. Briefly, mycoplasmas were grown overnight at 37°C in chambered slides (Nunc Nalgene, Naperville, IL) in modified SP-4 medium (30) (without phenol red but with 3% gelatin; pH 7.2). The spent medium was then removed and replaced with fresh, prewarmed, modified SP-4 medium. Mycoplasmas were viewed by using a Leica DM IL inverted microscope (Leica Microsystems, Buffalo Grove, IL) with a digital charge-coupled-device (CCD) camera (Hamamastsu Photonics K.K., Hamamatsu City, Japan) and analyzed using Openlab version 5.5.0 (PerkinElmer, MA). A minimum of 20 uninterrupted frames at a constant time interval were analyzed along a collision-free path for ≥100 individual cells per strain for gliding velocity. The software allows movement tracking with time stamp capability for calculating the distance traveled over time. Gliding frequency was determined by counting the motile cells per field and dividing by the total number of cells in the same field (n = approximately 2,100 cells per strain). The mutant II-3 is nonmotile (5) and was not measured but confirmed visually.
NHBE cell culture.NHBE cells are primary airway cells from the bronchial epithelium and were cultured as described previously (21, 29). Briefly, expanded NHBE cells (Lonza, USA) were seeded onto collagen-coated transwell inserts (Costar, Cambridge, MA) and grown to near confluence in submerged culture in 1:1 bronchial epithelial cell growth medium (Dulbecco's modified Eagle's medium with high glucose [BEGM:DMEM-H; Lonza] supplemented with Single-Quot [Lonza] components: 0.5 ng/ml human recombinant epidermal growth factor, 0.5 μg/ml hydrocortisone, 5 μg/ml insulin, 10 μg/ml transferrin, 0.5 U/ml epinephrine, 6.5 ng/ml triiodothyronine, 50 μg/ml gentamicin, 50 ng/ml amphotericin B, and 65 ng/ml bovine pituitary extract, as well as 5 × 10−8 M all-trans retinoic acid [Sigma-Aldrich, St. Louis, MO]). Cultures were maintained in a humidified atmosphere with 5% CO2 at 37°C. The apical medium was then removed, and the epithelium fed only basally from this point forward. The number of days in development was designated relative to initiation of ALI culture, corresponding to day 0.
Mycoplasma growth and radiolabeling.Mycoplasmas were grown at 37°C in SP-4 medium (30) until the phenol red pH indicator turned orange. Antibiotic selection was included for the prpC mutant (25). For radiolabeling studies, mycoplasmas were cultured in the presence of 100 μCi [methyl-3H]thymidine (6.7 Ci/mmol; PerkinElmer), harvested and collected by centrifugation, and washed 3 times in fresh SP-4 medium as described previously (31). The mycoplasmas were suspended in 1.8 ml SP-4 medium, incubated at 37°C for 10 min, syringe passaged 10 times with a 25-gauge needle, and centrifuged at 123 × g for 5 min at 25°C to remove large clumps. The resulting supernatant was used for NHBE cell infection. Infections were initiated with 1010 CFU of radiolabeled mycoplasmas in 150 μl SP-4 medium (7.6 × 105 to 3.2 × 106 CFU/cpm). Mycoplasma CFU were measured as described previously (32). Cytadherence assays were carried out over a 3-day period for each time point due to the time required for sample processing. Assays were repeated twice, and the two highest and lowest results from 12 data points per sample were omitted before calculating means and confidence intervals.
Mycoplasma infection and microscopic analysis.Antibiotics were omitted from the BEGM beginning four days prior to infection. NHBE cells were washed apically with Hank's balanced salt solution (HBSS; Sigma-Aldrich) 4 h prior to infection and incubated at 37°C in a humidified atmosphere with 5% CO2 to allow mucus accumulation to a depth approximating that in the human airway (13). Unlabeled mycoplasmas (109 CFU) in 150 μl SP-4 broth were added to the apical surface, and at 2 min, 30 min, and 240 min postinfection, the inocula were removed by gentle suction and the cells were washed with HBSS, fixed with phosphate-buffered saline (PBS) plus 7.4% (vol/vol) formalin for 1 h at room temperature, and blocked with 10% goat serum (Sigma-Aldrich) in PBS. Blocked specimens were probed for M. pneumoniae by using rabbit antisera raised against whole organisms (1:250) (33), followed by goat anti-rabbit Cy3 (1:250; Invitrogen, Eugene, OR). Actin localized to tight junctions was identified by using phalloidin-647 (1:40; Invitrogen). Beta-tubulin for cilia was detected with anti-tubulin antibody (1:250; Abcam Inc., Cambridge, MA), and nuclei were stained using DAPI (4′,6-diamidino-2-phenylindole) (Sigma-Aldrich). Periodic acid-Schiff (PAS) staining was conducted as described previously (29). Images were captured, and z-stacks were constructed using either a Zeiss LSM 510 Meta (Carl Zeiss, Thornwood, NY) or a Nikon A1R (Nikon Instruments Inc., Melville, NY) confocal microscope, with subsequent image analysis using ImageJ (National Institutes of Health, Bethesda, MD) (34). To quantify mycoplasma colonization levels with ciliated NHBE cells, regions of interest (ROI) having approximately 200 NHBE cells were identified and categorized based on the density of ciliated cells: <20%, 21 to 80%, and 81 to 100%. The M. pneumoniae levels associated with each ROI were measured by fluorescence intensity above the background control. Confidence intervals or standard deviations were reported as indicated, and the data are representative of a minimum of three independent experiments.
Airway barrier function.Transepithelial electrical resistance (TEER) was measured as an indicator of epithelial polarization and barrier function using the Epithelial Voltohmmeter (World Precision Instruments, Inc., Sarasota, FL) (35). Briefly, the epithelial apical surface was washed once with HBSS to remove secreted mucus, a 500-μl volume of fresh HBSS was added to the apical surface, and electrical resistance was measured according to the manufacturer's instructions, with the baseline defined as the TEER of uncoated transwell inserts without NHBE cells (≤200 Ω cm2).
RESULTS
Airway cell development.We monitored the development of normal human bronchial epithelium visually, by immunohistochemical staining and by TEER (Fig. 1). At days 2 to 4 ALI, epithelial cells were squamous-like morphologically, with cells as large as 30 μm in diameter. The cell layer appeared fully confluent, with no evidence of culture medium from the basal compartment in the apical chamber. Actin labeling demonstrated organized tight junctions over >90% of the transwell surface, and analysis by confocal microscopy indicated that the monolayer was as thin as 10 μm in some regions (data not shown). These observations were consistent with the near-baseline TEER measured at this point (Fig. 1). PAS staining was positive, but secreted mucus was not observed on the apical surface. In summary, day 2 to 4 airway cells characteristically presented as underdeveloped epithelium with no detectable cilia or mucus secretion; we refer to these early developmental cells collectively as day 2 airway cells.
NHBE cell development. (Top) Representative images of actin labeling as described in the text, illustrating epithelium in the process of organizing into a pseudostratified monolayer at the indicated time points in ALI culture. Bars, 15 μm. (Bottom) Features of differentiating NHBE cells. n = 12 for TEER and n = 20 for cell diameters, with standard deviations indicated. −, absent; + to +++, rare to prominent; NT, not tested.
Day 10 to 12 airway cells exhibited a cobblestone morphology, with large squamous-like cells less numerous. Actin labeling was localized to cell borders in all areas of the transwells. PAS staining was positive, but mucus secretion was not typically observed, and ciliated cells accounted for less than 1% of the apically exposed cells (data not shown). In summary, day 10 to 12 airway cells were more polarized, as indicated by an increase in TEER, but lacked mucus production, and ciliation was rare; we refer to these intermediate developmental cells collectively as day 10 airway cells.
Day 15 to 17 airway cells demonstrated cobblestone morphology similar to day 10 airway cells but with fewer squamous-like cells. Fluorescence microscopy indicated the presence of more ciliated cells than with day 10 airway cells, but their presence accounted for <5% of the surface area. PAS staining was positive, and mucus production was clearly visible. Secreted mucus at the apical surface exhibited no evidence of the phenol red indicator from the basal compartment, and TEER was higher at this stage, indicating further cell polarization and adequate barrier function. We refer to these mucus-producing, intermediate-development cells collectively as day 15 airway cells.
Day 43 to 45 airway cells exhibited robust mucus production and ciliated-cell presence, although ciliated cells were unevenly distributed in the transwells. In some regions, populations of ciliated cells were observed in large areas of near confluence, while other areas had only sparse ciliation. Ciliary beating and directional movement of mucus were observed by using fluorescent microspheres (0.05, 2.0, and 9.9 μm) (data not shown). The epithelium had a largely cobblestone appearance, but some ciliated cells appeared to be narrower or larger than the 10- to 15-μm range observed for most cells. In summary, day 43 to 45 airway cells characteristically exhibited abundant mucus secretion and substantial ciliation; we refer to these airway cells as terminally differentiated day 43 airway cells.
Primary airway cell development and M. pneumoniae adherence.We measured attachment of radiolabeled mycoplasmas to the airway epithelium at different stages of development (Fig. 2). The highest adherence for all strains was observed with day 2 airway cells and dropped 3- to 5-fold with day 10 airway cells, correlating with epithelial cell polarization, but not with the appearance of mucociliary function. Wild-type adherence levels were comparable for day 10 and 15 airway cells but then dropped again significantly with day 43 airway cells. As expected, the nonadherent mutant II-3 exhibited the lowest levels of adherence at all stages of NHBE cell development. Mutants II-7 and II-3R exhibited adherence to erythrocytes at 35% and 60% of the wild-type level, respectively, but very low gliding velocities (5) (Table 1) and colonized NHBE cells at day 10 and beyond at levels comparable to the baseline levels for mutant II-3. The P200 mutant adhered at a wild-type level with day 2 airway cells but at levels below wild type for airway cells at later points in development (Fig. 2). We previously showed that complementation of this mutant with the recombinant wild-type allele for P200 in trans restores NHBE cell colonization to wild-type levels (22), but we also cannot rule out the possibility that revertants in the mutant population contribute to the adherence observed both here and previously (22). In contrast to the P200 mutant, the hypermotile prpC mutant attached at lower than wild-type levels with day 2, 10, and 15 cells but at wild-type levels with day 43 cells (Fig. 2).
Colonization of NHBE cells at the indicated stages in development by wild-type (WT) (left) and mutant (right) M. pneumoniae strains. For WT attachment, statistical comparisons were made to day 2 airway cells. For mutant M. pneumoniae strains, statistical comparisons were relative to the WT for that time point. NHBE cells were infected with radiolabeled mycoplasmas on the luminal surface and incubated for 240 min as described in the text. Each bar represents the mean and 95% confidence interval for 8 data points from 3 independent experiments. *, P < 0.05; #, P < 0.10; NS, not significant.
Spatial and temporal colonization patterns.Confocal analysis of day 2 cells infected with wild-type M. pneumoniae revealed mycoplasmas in numerous near-confluent consolidations separated by areas of much lower colonization density but otherwise with no apparent pattern to the mycoplasma distribution (Fig. 3, top). Mycoplasma colonization of day 10 cells was less consolidated and more diffuse, consistent with the reduced adherence observed in binding assays with radiolabeled mycoplasmas. Mycoplasmas appeared randomly localized in areas of diffuse colonization, except that the highest mycoplasma densities were often noted along NHBE cell perimeters, colocalized with actin (Fig. 3, bottom, arrows, and data not shown). The colonization pattern for day 15 cells was similar to that for day 10 cells, with consistent mycoplasma colocalization with actin staining at cellular junctions (data not shown). Thus, for day 2, 10, and 15 airway cells, the data suggested a relationship between reduction in attachment levels and emergence of a localization pattern.
Luminal view of wild-type M. pneumoniae-infected day 2 (top) and day 10 (bottom) NHBE cells at 240 min postinfection. Representative images show Cy3-labeled wild-type mycoplasmas (red) illustrating near-confluent (day 2) and less dense and more diffuse (day 10) localization patterns. The uninfected-control images show nuclear DAPI (blue) and Cy3 background staining (red). The arrows indicate mycoplasma localization to cell perimeters. Bars, 20 μm.
We followed M. pneumoniae localization on day 43 cells at 2, 30, and 240 min postinfection. Figure 4A illustrates the typical interaction of M. pneumoniae with ciliated cells after 30 min. The organisms resided in the same plane as the cilia, and the DAPI stain corresponded to mycoplasmas on the apical surface, while NHBE cell nuclei were beneath the scanning area and not in view. Detection was inconsistent at 2 min and appeared to be very sparse. Colonization patterns became more detectable but changed little between 30 min and 4 h postinfection. At all time points, wild-type M. pneumoniae largely localized to ciliated cells, and colonization favored regions of the epithelium with dense clusters of ciliated cells versus sparsely ciliated regions (Fig. 4B and 5). The overall density of mycoplasma fluorescence associated with ciliated cells increased with incubation time. This pattern was likewise observed with the intermediate-gliding P200 mutant but with reduced overall mycoplasma fluorescence density. As expected, mycoplasma fluorescence was low and often barely detected for mutants II-3R and II-3 and largely failed to correlate with the cilium density (data not shown and Fig. 4, respectively).
(A) Wild-type M. pneumoniae M129 colocalized with cilia on the apical surface of day 43 NHBE cells after 30 min. Blue, DAPI; green, cilia; red, mycoplasmas; purple, mycoplasmas costaining with DAPI. Bars, 10 μm. The DAPI staining corresponds to mycoplasmas on the NHBE cell surface. (B) Day 43 NHBE cells at 2, 30, or 240 min postinfection with the indicated M. pneumoniae strains. Random fields with high ciliated-cell density were selected. Green, cilia; red, mycoplasmas; yellow, mycoplasmas colocalized with ciliated cells. Bars, 20 μm. (C) Quantification of wild-type and mutant mycoplasma colonization in arbitrary fluorescence units at the indicated time points for 6,700 to 9,200 NHBE cells per strain per time point, presented with 95% confidence intervals. For each time point, a background control was subtracted, and comparisons were made to II-3 (negative control). *, P < 0.05; #, P < 0.10; NS, not significant relative to the negative control.
(A) Luminal view of mycoplasma localization and spreading patterns on day 43 NHBE cells as three-dimensional (3D) projections of regions from Fig. 4 at the indicated times postinfection. Likely mycoplasma spreading (yellow arrows) was observed for the wild type but was more limited for the P200 mutant, which correlates with localization to ciliated cells (white arrows). Bar, 10 μm. Purple, mycoplasmas; green, cilia; blue, DAPI. (B) Cross-sectional view of mycoplasmas interacting with ciliated (white circles) and nonciliated (yellow circle) cells. Bars, 10 μm. Red, mycoplasmas; green, cilia; yellow, mycoplasmas colocalized with cilia; purple, actin; pink, mycoplasmas colocalized with actin; blue, DAPI.
In order to further evaluate the relationship between mycoplasma colonization and ciliated-cell density, we analyzed the data from Fig. 4B for each strain and time point according to the extent of ciliation, which we categorized as low, medium, or high (<20%, 20 to 80%, and >80% ciliated cells, respectively) (Fig. 6). Colonization was not consistently quantifiable at 2 min, but at 30 min, mycoplasma fluorescence increased with the ciliated-cell density, and at 240 min, mycoplasma fluorescence associated with medium and high ciliated-cell densities had increased approximately 2.5-fold. Significant association of wild-type mycoplasmas with areas of low ciliated-cell density was not noted until 240 min and was >3-fold higher than that for the P200 mutant. Interestingly, while spreading from ciliated-cell foci may account for colonization of nonciliated cells, at no time did we observe equal distribution of M. pneumoniae on ciliated and nonciliated cells (Fig. 5). Furthermore, removal of the gel layer mucus by washing day 43 airway cells with Hank's balanced salt solution or dithiothreitol (10 mM) did not appear to enhance colonization of nonciliated cells by wild-type M. pneumoniae or negatively impact localization to cilia (data not shown).
Quantification of mycoplasma fluorescence associated with ciliated-cell density for the data set in Fig. 4C; 6,700 to 9,200 NHBE cells per strain per time point were analyzed. The data are reported as means, and the error bars reflect the standard deviations. II-3 was omitted, since it failed to colonize NHBE cells.
DISCUSSION
Our data indicate that changes associated with NHBE cell development impact mycoplasma colonization patterns. The most striking effect was a significant drop in adherence and a corresponding change in the localization pattern with NHBE cell polarization. The development of NHBE cells in our system paralleled lung injury repair models, recently reviewed by Crosby and Waters (23). Similar to repairing an airway, NHBE cells spread, migrate, proliferate, polarize, and differentiate, forming a functional mucosal barrier (36). In an in vivo injury repair model, Dupuit and coworkers reconstituted human nasal epithelial cells in rats in much the same way that NHBE cells differentiated here on transwell inserts (37). Of interest is the observation that, while the markers used for epithelial development differed between studies, the morphological changes and the time frame for development were consistent. Epithelial turnover is normal and ongoing in a healthy airway and elevated in the context of infection or inflammation. The importance of airway epithelial repair to host-pathogen interactions is illustrated in colonization by the opportunistic pathogen Pseudomonas aeruginosa, where attachment is dependent on injured and actively repairing cells (38). Interestingly, polarization of the epithelium also has an impact on P. aeruginosa internalization (39). Although we did not assess mycoplasma internalization by NHBE cells in the current study, our findings nevertheless suggest that basal cells may have a high infection burden, which could be particularly significant with concurrent coinfection or inflammation.
The striking drop in M. pneumoniae adherence and change in localization pattern for day 10 cells relative to day 2 cells correlated with cell polarization, but not gel layer mucus production, and may reflect acquisition of an airway barrier function with NHBE cell development. Tethered mucins are cell surface-associated markers known to form part of the periciliary barrier at mucosal surfaces (40–42). Significantly, production of the tethered mucin Muc-1 in mice is spatially and temporally controlled with the onset of epithelial cell polarization (43). The human tethered mucins MUC1 and MUC4 reportedly develop very early gestationally in relation to gel-forming mucins, suggesting that they have a complex role in development, in addition to limiting pathogenic interactions (44), which has recently been reviewed (42).
Acquisition of gel layer mucus production in the presence of low ciliation for day 15 airway cells had little impact on M. pneumoniae colonization relative to that seen with day 10 cells. Mycoplasma gliding motility was not essential for colonization of day 2 undifferentiated epithelium (Fig. 2), but poorly gliding mutants were severely limited in their ability to overcome the mucus barrier, even in the absence of cilia. Gel layer mucus is a complex matrix that is designed to trap and transport (11–13). While functional trapping likely accounts for the barrier to colonization by poorly gliding mutants, the disulfide-linked mucin glycoprotein network of gel layer mucus (13) may actually support mycoplasma gliding across the layer to allow access to the cilia (45). In that regard, it is noteworthy that human gel layer mucus contains sialic acid in a variety of linkages, including α2,3 N-acetylgalactosamine (46), which can also be found on ciliated cells (47). This is significant, because M. pneumoniae binds (48) and glides (49) on such moieties.
Colonization levels for all strains dropped further, as expected (22), with the development of extensive ciliation (Fig. 2), consistent with a fully developed barrier function. Furthermore, differences in colonization by motility-deficient mycoplasmas become more evident in the presence of this barrier. Confocal microscopy demonstrated that gliding-capable mycoplasmas readily localized to cilia and could be found there at a high level compared to gliding-defective mycoplasmas. Localization to ciliated cells appeared to correlate with ciliated-cell density, while the colonization pattern was consistent with a stepwise process of attachment at the cilium tip, followed by rapid and efficient movement to the cilium base, where at a high multiplicity of infection they appeared to accumulate (Fig. 5). In contrast to the rapid colonization of ciliated epithelium, colonization of nonciliated cells was limited, and lateral spread from ciliated cells was less than expected based on the M. pneumoniae gliding velocity on an inert surface (Table 1). Thus, there was only limited lateral movement by wild-type M. pneumoniae from heavily infected ciliated cells to nonciliated cells over 4 h (Fig. 4 and 5). Furthermore, mycoplasmas associated with nonciliated cells were often seen colocalized with actin at the intercellular junctions.
At least two factors may account for limited colonization of nonciliated cells, both initially and as a result of lateral spread from ciliated cells. The first is a potential difference in the availability of suitable binding sites along the cilia, at their bases, and on nonciliated cells. Two broad receptor types have been identified for M. pneumoniae, glycoproteins having terminal sialic acids α2-3 linked to galactose (48) and sulfated glycolipids (50), but the density and distribution of each on airway cells remains poorly defined, and we cannot rule out the possibility of additional receptor moieties. Alternatively, or in addition, this outcome might reflect the presence of a physical or chemical barrier associated with the periciliary layer rather than the absence of suitable receptors. The barrier model is consistent with the reduced ability of gliding-defective mutants to penetrate the periciliary layer that underlies the gel layer mucus in the absence of cilia, even after removal of the gel layer mucus by gentle washing. Tethered mucins of the periciliary layer (MUC1, MUC4, and MUC16) have been shown to limit the penetration of microspheres to the apical surfaces of primary human airway cultures (41). In preliminary studies, we observed a pattern in which mycoplasmas localized to regions of day 2 airway cultures where MUC4+ cells were absent and on day 10 and day 15 airway cultures to intercellular junctions of MUC4+ cells (data not shown). These preliminary observations are consistent with a model in which the decrease in mycoplasma attachment from day 2 to day 10 is associated with cell surface barrier development. We speculate that periciliary components may impact the colonization patterns of the differentiated airway. Further studies are required to more clearly define the role(s) that tethered mucins may play.
To summarize, our current findings suggest that in vivo, mycoplasmas most likely utilize gliding motility to span the gel layer mucus barrier and access the underlying epithelium via the tips of cilia. Gliding motility and/or forces associated with the ciliary recovery stroke (51) may facilitate localization to the base of the cilia, although we cannot rule out the possibility that mycoplasmas glide on the periciliary layer that supports the overlying gel layer mucus, a possibility raised by the recently proposed “gel-on-brush” model for airway mucus and the periciliary layer (41). The brush layer, consisting of tethered mucins (41), supports the gel layer mucus and may support mycoplasma gliding at the brush-gel interface while limiting lateral spread at the base of the brush layer. Finally, the appearance of tethered mucins is developmentally regulated, forming a rudimentary barrier to colonization that is absent or incomplete on underlying basal cells of the airways, rendering them susceptible to a higher mycoplasma colonization burden than is typically seen with terminally differentiated cells.
ACKNOWLEDGMENTS
This work was supported by Public Health Service research grant AI49194 from the National Institute of Allergy and Infectious Diseases to D.C.K. and the National Institutes of Health Supplement to Promote Diversity in Health-Related Research.
We thank the University of Georgia College of Veterinary Medicine Cytometry Core Facility.
FOOTNOTES
- Received 20 August 2013.
- Returned for modification 20 September 2013.
- Accepted 12 November 2013.
- Accepted manuscript posted online 18 November 2013.
- Copyright © 2014, American Society for Microbiology. All Rights Reserved.
