Centre for Veterinary Science, Department of Clinical Veterinary Medicine, University of Cambridge, Madingley Road, Cambridge CB3 0ES, UK
Correspondence
Andrew Preston
apreston{at}uoguelph.ca
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Present address: Department of Microbiology, University of Guelph, Guelph, ON, Canada N1G 2W1.
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Little is known about what determines the host range or the differing pathologies caused by different Bordetella species. The genomes of a representative strain of each of these bacteria have recently been sequenced and analysed (Parkhill et al., 2003; Preston et al., 2004
). This information allows the generation of new hypotheses to explain the molecular basis for Bordetella pathogenesis.
The bordetellae are known to initially infect the cilia of the respiratory epithelium and it is likely that this is the major, and possibly only, site of infection. Our current understanding is that Bordetella pathogenesis involves adherence of bacteria to the cilia of the respiratory tract, induction of ciliostasis, mucus production, damage to the epithelium and induction of an inflammatory response in the respiratory tract (Bemis et al., 1977; Bemis & Kennedy, 1981
; Bemis & Wilson, 1985
; Funnell & Robinson, 1993
; Matsuyama, 1977
; Muse et al., 1977
; Thompson et al., 1976
). However, these observations have been made in separate systems, many of them highly artificial, that have not allowed the temporal relationship between these pathologies to be investigated.
We report here the use of air-interface organ culture from canine tracheal tissue, infected with B. bronchiseptica, to model the interaction between these bacteria and ciliated respiratory tissue. We have identified ciliostasis as a very early event in the host tissuepathogen interaction that occurs before obvious damage to the epithelial surface and mucus production. We thus hypothesize that ciliostasis occurs very early during B. bronchiseptica infection of the respiratory tract and that this prevents clearance of the bacteria and allows further colonization of the tissue. It is also likely to prevent clearance of the mucus and extruded cells that are subsequently induced, and accumulation of this material in the respiratory tract is likely to contribute to the pathology associated with Bordetella infections.
![]() |
METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Construction of air-interface organ culture.
Canine tracheae were obtained from beagles used as naïve controls in other research experimental studies with the cooperation of Glaxo-SmithKline, Welwyn Garden City, UK, and other research groups at the University of Cambridge.
Organ cultures were constructed based on the method developed for human tissues by Jackson et al. (1996). The dogs that were used in this study did not have signs of clinical disease and had not received treatment with a pharmacological agent. Tracheae were taken from both male and female dogs whose age ranged from 9 months to 4 years. The age and sex of the dog was not considered when constructing organ cultures.
The whole trachea, from the larynx to the carina, was removed from the euthanased dog by surgical excision. Connective tissue that was adherent to the trachea was removed by blunt dissection. The trachea was placed, intact, into a 400 ml capped bottle filled with pre-warmed Dulbecco's modified Eagle's medium (DMEM) containing 10 mM glutamine, 2·5 µg amphotericin ml1, 100 U penicillin ml1 and 50 µg streptomycin ml1 and incubated for 2 h at 37 °C. It was then transferred to another 400 ml capped bottle filled with pre-warmed DMEM without antibiotics and incubated for a further 2 h at 37 °C. During this period the medium was changed twice. Used medium was poured away and the bottle refilled with fresh, pre-warmed medium. These washes removed antibiotics from the tracheae.
Following the washes the trachea was placed on a dissecting board. The trachea ligament that holds the C-shaped cartilage in place was removed using sharp-ended scissors. Using a scalpel the trachea was cut lengthways into four equal strips, each approximately 5 mm wide. Each strip was then cut at every second cartilage ring. The resulting pieces were 5 mm2 and consisted of respiratory mucosa on cartilage. Pieces were placed, mucosa side uppermost, directly onto agarose plugs (Fig. 1). Plugs were made by pipetting 1 ml 1 % agarose in distilled water into a well of a 24-well cell-culture dish. After the agarose had set, plugs were removed and placed in a 6-well cell-culture dish; 3 ml 0·5 % agarose in DMEM medium was pipetted around the plugs. The lid of the cell-culture dish was then replaced to prevent contamination. Organ cultures were incubated at 37 °C in a humidified atmosphere containing 5 % CO2. The ciliary function of organ cultures was tested (see below) before use in experiments, and any organ cultures failing to clear beads were discarded.
|
Scanning electron microscopy (SEM).
Tissue pieces were fixed in 5 % glutaraldehyde in PIPES buffer, post-fixed in 1 % osmium tetraoxide, dehydrated in a graded ethanol series and subjected to critical-point drying. Samples were mounted on aluminium stubs and sputter-coated with 15 nm of gold. Samples were examined on a Philips XL30 FEG scanning electron microscope at an accelerating voltage of 5 kV. SEM analysis was performed at the Multi-Imaging Centre, Department of Anatomy, University of Cambridge.
Morphometric analysis.
Organ cultures were analysed as described previously (Jackson et al., 1996). Briefly, the proportion of organ culture surface covered by particular features was estimated. The presence of four features was recorded: cilia, mucus, cellular damage and processing damage. Processing damage consisted of tissue shearing. Cellular damage included the presence of extracellular matrix, extruded ciliated epithelial cells or the presence of denuded basement membrane. Twenty predetermined squares were examined at x2000 magnification. Each field had an edge of 30 µm and typically contained 325 cells. A transparent acetate sheet with 100 equal squares was placed over the video display unit to allow scoring of these fields. The predominant cell type in each square was recorded. Each score represented 1 % of the total area of that field. The occurrence of each feature was expressed as a percentage of the visible epithelium in each image. Percentages were adjusted to account for tissue damaged by processing. To determine the effect of infection on the percentage of tissue covered by mucus, the percentage of mucus present on uninfected organ cultures was subtracted from the percentage of mucus present on the infected samples. Comparisons were made between samples from individual tracheae, which negated variation between individuals. Data represent the mean and standard deviation of six independent experiments obtained from organ cultures prepared from separate animals. The significance of differences between samples was assessed using MannWhitney analysis. Comparisons between uninfected and infected cultures, wild-type and bvg mutant infections, and RB50 and SB796 infections were made. All differences discussed in the text are significant at the 5 % level.
Light microscopy.
Samples were submerged in 10 % formalin in PIPES buffer and fixed for between 24 and 72 h. The majority of tissue cartilage was removed using a scalpel, to leave only a thin section attached to the mucosa. Samples were placed in histology cassettes and processed using a Shandon Citadel 2000 automatic processor. Samples were dehydrated in a series of graded methylated spirits of increasing concentration, submerged in chloroform and then embedded in molten Tissue-TEK III embedding centre wax (Raymond Lamb Laboratory Supplies). Samples were correctly oriented in molten wax before cooling to allow the hardening of wax and setting of tissue. Then 5 µm transverse sections of tissue were cut along a lateral edge of the organ culture on a Leica Ultracut E microtome, dried overnight, taken through a series of graded methylated spirits to remove wax, cleared in xylene and stained.
Staining of sections.
Sections were stained with haematoxylin and eosin by standard methods (Bancroft & Cook, 1994), using a Linistain GLX Staining Machine (Shandon), in methylene blue for 2 min or with PAS/alcian blue using standard methods (Bancroft & Cook, 1994
). Schiff reagent was purchased from Fisher Scientific. Staining reagents were purchased from Surgipath Europe. Sections were mounted using SubX mounting media (Surgipath) and examined using a Nikon Optiphot microscope.
Quantification of mucus on canine organ cultures.
Mucus on organ cultures was quantified as described previously (Thornton et al., 1989). Mucus was gently scraped from the mucosal surface using a plastic loop and placed into 1 ml PBS. Samples were shaken vigorously and stored at 20 °C. For assay, samples were thawed at room temperature and heated for 30 min at 37 °C. Then 50 µl was transferred to a nitrocellulose membrane (0·2 µm pore size) using a dot-blot apparatus. Mucin from bovine submaxillary gland (1 mg ml1; Sigma Aldrich, catalogue no. M3895) was used as a positive control and water as a negative control. Following transfer the membrane was removed and dried in a fume hood. It was then washed in distilled water for 2 min, incubated in a freshly prepared solution of 1 % (v/v) periodic acid in 3 % (v/v) acetic acid for 30 min and then rinsed twice for 2 min in freshly prepared 0·1 % (w/v) sodium metabisulphite in 1 mM HCl. Schiff's reagent was added and the colour allowed to develop for 15 min, after which time the membrane was rinsed in 0·1 % sodium metabisulphite in 1 mM HCl and then dried in a fume hood.
The quantity of mucin contained within each sample was calculated by measurement of the reflectance density of each blot using the Bio-Rad Quantity One Quantification Software program and a Bio-Rad GelDoc 2000 imaging system. Each sample was positioned in a small rectangular region of analysis of 0·5 cm2 and the volume of pixels within this region was measured. The show saturated pixel feature was turned on during measurements to ensure that saturation had not been reached. The amount of mucin in samples correlates to the volume of pixels contained within the analysed region, and was calculated by V=TB, where V is the corrected volume of pixels of a sample, T is the total volume of pixels of the analysed rectangle and B is the volume of background pixels. The relative levels of mucin within samples were compared. Data represent the mean and standard deviation of nine independent experiments obtained from organ cultures prepared from three dogs.
Latex bead clearance assay.
Ten microlitres of 1 µm diameter polybead polystyrene microsphere beads (latex beads; Polyscience) was pipetted onto the ciliated surface of organ cultures such that the beads covered the entire mucosal surface. Cultures were incubated for 30 min at 37 °C and 5 % CO2 and then assessed by visual inspection. Ciliary-mediated clearance was assessed to be intact if the latex beads had been cleared to one edge of the organ culture piece. Preliminary experiments determined that for normal tissue, this occurred within 30 min (data not shown).
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Both post-mortem and organ-culture samples predominantly comprised ciliated tissue, with very few non-ciliated areas, although post-mortem tissue often had mucus that obscured the underlying cilia (Fig. 2). A mean of 65 % (range 3296 %) or 95 % (range 8998 %) of the surface area was ciliated in post-mortem or organ culture samples, respectively. Mucus covered 32 % (range 268 %) or 3 % (range 09 %) of the surface areas of the post-mortem or organ culture samples, respectively, and where present it obscured cilia from view (Fig. 2
). The post-mortem samples, on average, contained a higher level of mucus and showed wider variation in mucus levels than did organ culture samples. It is likely that mucus that was present on the tracheal tissue at the time of post-mortem was removed during the washing steps that were performed during the setting up of the organ cultures, resulting in the lower level of mucus observed in organ culture tissue. These data also suggested that maintenance of the tissue in vitro did not induce large amounts of mucus expression onto the tissue surface. Some splits in the tissue were observed; these were classed as processing damage that occurred during sample preparation for SEM. This affected up to 10 % of the surface of samples in both post-mortem and organ culture samples. In experiments with extended incubation periods the appearance of tissue surface remained unchanged for up to 72 h. Beyond this time, cilia became less organized and some were not upright (data not shown). Therefore organ cultures were maintained for 24 h or less in subsequent experiments.
|
Pathology induced by B. bronchiseptica
To investigate changes in tracheal tissue induced by B. bronchiseptica infection, organ cultures were inoculated with different strains of B. bronchiseptica, incubated for 24 h and then analysed by SEM/morphometric analysis. The surface area of the tissue that comprised cilia, mucus or epithelial cell damage was recorded and compared with uninfected controls. Two wild-type (WT) strains were used: RB50, an isolate from a rabbit and the genome sequence strain, and SB796, a clinical isolate from a case of kennel cough. Strain RB50 was used because it is a well-characterized strain for which a large number of well-defined mutants are available. Strain SB796 was used in order to ensure that the interaction between a clinical isolate and the cognate host tissue was being studied. In addition, organ cultures were infected with isogenic bvg mutants of each strain, RB54 and SB799, respectively. These mutants are locked in the Bvg-minus phase and thus do not express any of the protein adhesins and toxins previously implicated in B. bronchiseptica pathogenesis (Cotter & Miller, 1994). Organ culture pieces were inoculated with 108, 105 or 102 c.f.u. In all experiments, pure cultures of B. bronchiseptica were recovered from infected organ cultures, while uninfected pieces remained sterile.
Organ cultures inoculated with 108 c.f.u. of WT bacteria were covered in sheets of mucus to which numerous rod-shaped bacteria were adhered. Underlying surface structures were obscured from view by the mucus (Fig. 3). Extruded cells were not observed. Uninfected control organ culture pieces were not covered with mucus, consistent with previous results (see above). Both of the WT strains induced a similar increase in the surface area of tissue that was covered by mucus when compared to uninfected controls. Organ culture samples infected with RB50 or SB796 had a mean of 84 % (±20 %) or 91 % (±9 %), respectively, more tissue surface covered in mucus than uninfected controls (Fig. 4
). Infection with 105 or 102 c.f.u. of WT bacteria did not induce any changes in surface morphology when compared to uninfected controls (data not shown). Some mucus was observed on the surface of these samples although the level of mucus was not significantly different from that in uninfected controls (Fig. 4
). These data suggested that infection by WT bacteria induced mucus expression by the organ culture tissue, that a certain number of bacteria were required for this induction and that this threshold level was reached by inoculation with somewhere between 105 and 108 bacteria.
|
|
Induction of ciliostasis
Coordinated, directional ciliary beat is a fundamental property of respiratory tract physiology. Previous reports suggested that B. bronchiseptica induced ciliostasis of tracheal cells cultured as explant outgrowths (Bemis & Kennedy, 1981; Bemis & Wilson, 1985
). To assay the ability of B. bronchiseptica to induce ciliostasis in air-interface organ culture tissue, a simple assay was developed that measured the ability of the organ culture to clear a suspension of latex beads from its surface. When inoculated on to the surface of the organ cultures, a drop of latex bead suspension spread to form a white layer on the surface of the tissues (Fig. 5
). Clearance of the beads by the organ cultures was observed in real time as the beads were moved by ciliary beat action to one edge of the tissue piece (Fig. 5
; see also the movie file available as supplementary data with the online version of this paper at http://mic.sgmjournals.org). Clearance of the beads indicated that the cilia were beating and doing so in a coordinated fashion. By noting the orientation of the tissue piece in relation to its original anatomical orientation it was possible to establish that ciliary-mediated clearance was in a distal to proximal (lungs-to-larynx) direction, and that the organ culture tissue maintained the same direction of ciliary beat that occurred in vivo. Organ culture samples continued to clear beads for 48 h. After this time, the number of organ cultures clearing beads dropped (data not shown). Thus, during the 24 h time period of the experiments reported here, organ culture pieces maintained coordinated and functional ciliary beat.
|
|
Time-course of infection by WT bacteria
The results above demonstrated that WT bacteria, at an inoculum size of 108 c.f.u., induced mucus expression and abrogation of ciliary-mediated clearance by 24 h post-inoculation. In order to define further the time-course of these events, organ cultures were inoculated with 108 c.f.u. of SB796 and analysed at 1, 2, 4, 8 and 12 h post-inoculation. At 1 h all of the organ cultures tested failed to clear beads (and thus were not tested for bead clearance at later time points). In SEM analysis, very few bacteria were visible adherent to cilia at this point; the cilia appeared healthy and there was little mucus present, suggesting that ciliostasis occurred without extensive colonization of or damage to the cilia. At each subsequent time point there was an increase in the number of bacteria observed on the tissue surface. By 12 h, much of the tissue surface was covered by mucus, obscuring the bacteria adherent to cilia. In regions not covered by mucus, large numbers of bacteria were visible. Representative images for each time point are shown in Fig. 7. At 1 and 2 h post-inoculation no mucus was observed. By 4 h, some fibrous mucus was evident on the tissue surface, which increased at 8 h. By 12 h much of the tissue surface was covered by mucus; therefore much of the mucus observed at 24 h (see above) was expressed on to the tissue surface by 12 h. Representative images of this analysis are shown in Fig. 8
. Thus bacteria adhered to the cilia of the organ cultures and proliferated on tissue in organ culture before significant mucus expression. This suggested that mucus expression occurred in response to colonization of the tissue by bacteria. These data also demonstrated that inhibition of bead clearance was an early event following inoculation of the organ culture with bacteria and that it occurred prior to mucus expression on the tissue surface and prior to obvious damage to the ciliated epithelium.
|
|
Incubation of tissue pieces in organ culture induced a significant increase in mucus compared to cultures at the start of the incubation period (i.e. immediately following inoculation) in all cases, including uninfected controls (Fig. 9), demonstrating that incubation of tissue in organ culture induced expression of mucus onto the tissue surface. At 2 h post-inoculation, there was no difference in the quantity of mucus measured in infected and uninfected organ cultures. Thus, during this time period the bacteria did not stimulate mucus expression on to the tissue surface. However, by 8 h post-inoculation significantly more mucus was measured in infected organ cultures than in controls, suggesting that bacteria induced expression of mucus on to the tissue surface between 2 and 8 h post-inoculation. The same trend was also apparent at 24 h post-inoculation, although the quantity of mucus was not significantly different on infected and control organ cultures. There was no significant difference between the amount of mucus produced in response to WT or bvg mutant bacteria at any time point. Thus, the incubation conditions employed induced the expression of some mucus on to the tissue surface in the absence of infection by bacteria, but infection by WT or bvg mutant bacteria increased mucus expression.
|
Organ cultures were inoculated with 108 c.f.u. of WT or bvg mutant bacteria, and incubated for either 0 or 24 h. Sections of tissue samples were cut, stained by PAS/alcian blue and examined by light microscopy. In this procedure neutral mucus was stained magenta, acid mucus was stained blue and goblet cells containing mucus not yet released onto the tissue surface were easily identified.
Organ cultures incubated for 24 h without infection contained numerous goblet cells that stained magenta, indicating that they still contained mucus (Fig. 10a). The number of stained goblet cells was less than at 0 h incubation (data not shown), suggesting that organ culture conditions stimulate some release of mucus from goblet cells. Organ cultures infected with WT bacteria did not contain any stained goblet cells (Fig. 10b
), indicating that these bacteria stimulated the release of mucus from all of the goblet cells in the tissue piece. Organ cultures infected with bvg mutant bacteria contained a small number of stained goblet cells (Fig. 10c
) but not as many as uninfected cultures, suggesting that these bacteria stimulated mucus release, but not to as great an extent as did WT bacteria.
|
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
A previous study investigated interactions between the exclusively human pathogen B. pertussis and human adenoidal tissue or nasal turbinate tissue maintained in organ culture at an air interface (Soane et al., 2000). In this system, B. pertussis infection resulted in damage to the ciliated tissue surface and release of mucus on to the tissue surface. This model was thus proposed as a physiologically relevant one in which to analyse the interaction between B. pertussis and host tissue. However, amounts of human tissue similar to the amounts of canine tissue used in this study, particularly if that tissue is tracheal or bronchial, are rarely available, making it difficult to fully develop the model for humans. We decided to follow on from this study with human tissue by using B. bronchiseptica infection of canine air-interface tracheal organ culture. This choice retains a natural pathogenhost combination but utilizes a tissue that is far more readily available than human tissue. Our data demonstrate that for at least 24 h, the tracheal tissue is maintained in a state that closely resembles that of the tissue in vivo. Beyond this time, some deterioration of the tissue was observed. It is likely that the tissue starts to deteriorate as soon as it is harvested from the host. In some 24 h organ culture samples examined by transmission electron microscopy, some tight-cell junctions had come apart and there was evidence of heterochromatin in the nuclei of some cells (data not shown), indicative of some cell death. However, these changes did not appear to materially affect the tissue, as even these samples retained bead clearance activity and their tissue surface appeared normal by SEM (data not shown). Nevertheless, it is likely that beyond 24 h in organ culture, natural tissue deterioration will make the use of samples less attractive for the study of infection-induced pathology.
Infection of the organ culture tissue with B. bronchiseptica resulted in changes that are consistent with pathology observed in vivo and previous studies: B. bronchiseptica adhered to cilia, induced mucus expression on to the tissue surface, resulted in damage to the epithelial layer and induced ciliostasis. Thus, we propose that this model is suitable for the study of the interaction between B. bronchiseptica and host ciliated respiratory tract tissue. We consider that this interaction is likely to be one of the very early steps during B. bronchiseptica infection and that the outcome of this interaction is likely to influence greatly whether infection is established or the bacteria are cleared from the host.
We consider that the model used here has several advantages over the submerged organ cultures or submerged ciliated cell outgrowths that had been used previously to study Bordetella pathology. Mucus expression on to the tissue surface was not observed in the submerged systems. Our data, and those of Soane et al. (2000), suggest that this response is an important component of the host tissue response to Bordetella infection.
Submerged organ cultures systems did not permit an easy assessment of ciliary activity. Previous studies measured a reduction of ciliary beat frequency (CBF) of individual cells, grown as tracheal explant outgrowths, when infected by B. bronchiseptica (Bemis & Kennedy, 1981; Bemis & Wilson, 1985
). However, ciliary clearance function is dependent not only on ciliary beat, but also on the coordination of ciliary beat across the entire tissue. Ciliary beat of cells is coordinated to produce a wave-like motion (metachronal waves) on the tissue surface that provides clearance of material in a lungs-to-larynx direction. The simple bead-clearance assay used here measures this functional clearance activity of the tissue. Furthermore, the need to use a microscope to focus on a small region of cells for CBF measurement means that simultaneous measurement of multiple samples is not possible. The bead clearance assay provides a high-throughput analysis in which ciliary-mediated clearance can be assessed in many samples simultaneously. It is also possible to perform the assay multiple times on the same organ culture sample with no apparent loss of function.
Our data demonstrated that ciliary-mediated clearance was abrogated very soon after inoculation of the organ cultures with bacteria. This occurred before the expression of mucus on to the tissue surface and before obvious damage to the epithelial layer. It was not possible using the bead clearance assay to determine whether ciliary beat had been completely halted or whether its coordinated nature had been interrupted to prevent bead clearance. However, clumps of ciliated cells were removed by gentle scraping from organ cultures 1 h after inoculation with bacteria and thus at a time when bead clearance had been halted. These clumps were viewed by light microscopy and cilia were observed to be clearly motionless (data not shown). When cells were viewed in this way from uninfected control samples, cilia were observed to be clearly beating. This strongly suggests that infection had induced ciliostasis rather than disorganization of ciliary beat. Thus, we propose that B. bronchiseptica induces ciliostasis very soon after interacting with the ciliated epithelium.
An elegant series of previous studies demonstrated that B. pertussis released a soluble peptidoglycan fragment, termed tracheal cytotoxin (TCT), during normal growth and that TCT exhibited cytotoxicity to submerged hamster ciliated tracheal organ cultures (Heiss et al., 1993). Furthermore, it was demonstrated that TCT-mediated cytotoxicity was due to induction of nitric oxide (NO) synthesis via interleukin-1-induced expression of type II NO synthase (Goldman et al., 1994
). More recently, it was demonstrated that TCT acted in synergy with LPS to cause cytotoxicity, that NO synthesis occurred exclusively in non-ciliated cells of the hamster organ culture and that the NO diffused to neighbouring ciliated cells, causing their death (Flak & Goldman, 1999
; Flak et al., 2000
). Thus, it was proposed that TCT and LPS induction of NO synthesis is a major contributor to loss of ciliary function during B. pertussis infection. In our studies ciliostasis occurred within 1 h of infection, whereas TCT/LPS-induced cytotoxicity occurred at least 36 h after infection of the organ cultures with B. pertussis. Furthermore, addition of TCT (a kind gift from W. E. Goldman, Washington University, MO, USA) in combination with LPS to organ cultures failed to abrogate bead clearance in our model (data not shown). A low level of ciliostasis was observed following infection of organ cultures with bvg mutants (Fig. 6
). TCT secretion and LPS expression occur in the absence of Bvg stimulation and thus it is possible that TCT/LPS contribute, albeit at a low level, to the ciliostasis observed in our studies, although the effect of Bvg-mediated LPS variation (Preston et al., 2003
) on TCT/LPS-induced cytotoxicity is not clear. However, the ciliostasis demonstrated in our studies probably occurs through a different mechanism than TCT/LPS-induced synthesis of NO.
Although bvg mutants did not abrogate bead clearance to a large extent they did not adhere to the cilia; thus it was not possible to determine whether induction of ciliostasis is a Bvg-dependent event, or whether ciliostasis requires prior adherence of the bacteria to the cilia, and the adherence is a Bvg-dependent event. Previously, it was demonstrated that heat-killed or chloramphenicol-inhibited B. bronchiseptica adhered to ciliated canine tracheal cells without inducing ciliostasis, suggesting that adherence and induction of ciliostasis were independent events (Bemis & Wilson, 1985). However, this study also demonstrated that induction of ciliostasis required an intimate association between bacteria and cilia. Recently, it was demonstrated that adherence of B. bronchiseptica to ciliated rabbit tracheal cells, grown as explant outgrowths, induced calcium transients within the cells and that this required a functional Bvg system (Groathouse et al., 2003
). Induction of calcium transients occurred within seconds of adherence of the bacteria to the cilia. It is possible that in this study the Bvg-dependence for induction of calcium transients was due to the need for adherence to the cilia for this phenomenon and that adherence was Bvg-dependent. These data suggest that calcium-mediated signal transduction is stimulated within ciliated cells by adherence of B. bronchiseptica. It is tempting to speculate that these signal transduction events are involved in the induction of ciliostasis.
In the studies reported here, induction of mucus expression on to the tissue surface occurred hours after inoculation of the organ cultures with bacteria. Uninfected controls also released mucus, suggesting that culture conditions stimulated mucus release from goblet cells. However, at 8 h post-inoculation, mucus levels were higher on infected tissue than on uninfected controls, suggesting that the bacteria were also a stimulus to release mucus. Interestingly, bvg mutants stimulated the same level of mucus release as WT bacteria, suggesting that this stimulation is Bvg-independent and that it does not require adherence of the bacteria to host cells. Mucin secretion can be stimulated by a variety of molecules, including bacterial components such as LPS (Kim et al., 2003). The amount of mucus measured in our studies was highly variable between samples. Mucus within granules in goblet cells is highly condensed and expands up to 600-fold following release on to the tissue surface (Rogers, 2003
). Thus, a small variation in the number of goblet cells stimulated to release mucus may result in large differences in measurable mucus volume on the organ cultures. Therefore it is likely that the measurement of the number of goblet cells that contain mucus is a more accurate estimate of mucus release than is direct measurement of mucus volume. In this analysis, infection by B. bronchiseptica resulted in the release of mucus from more goblet cells than did culture conditions alone, but with no significant difference between WT- and bvg-mutant-induced mucus release. Our data thus suggest that induction of mucus expression onto the tissue surface and ciliated cell death and damage to the epithelial layer all occur after abrogation of ciliary-mediated clearance. We thus hypothesize that induction of ciliostasis by B. bronchiseptica is a key early step during infection. Induction of ciliostasis abrogates the clearance activity of the mucociliary escalator and thus disables this important respiratory tract defence mechanism. This in turn renders the respiratory tract susceptible to further colonization by B. bronchiseptica bacteria that proliferate on the surface of the respiratory epithelium. Clearance from the respiratory tract of the mucus, cell debris and inflammatory-mediated influx that is induced after ciliostasis by B. bronchiseptica infection is likely to be impaired and accumulation of this material in the respiratory tract is likely to contribute to the pathology associated with B. bronchiseptica infection.
A limitation of the organ culture system is the lack of immune cells in this model. Thus, a key aspect of the response to B. bronchiseptica infection, induction of an influx of inflammatory cells into the lumen of the respiratory tract, is missing. Studies of this aspect of the host response are still limited to animal models of infection. However, it is very likely that induction of the inflammatory response occurs after the initial interaction between B. bronchiseptica and ciliated tissue, and thus after the induction of ciliostasis described here.
Organ cultures maintained at an air interface more closely resemble the in vivo respiratory tissue environment than do submerged organ cultures, and have been used to study the interaction between respiratory tract tissue and a number of pathogens (Dowling et al., 1998; Middleton et al., 2003
; Wilson et al., 1992
). Here, the use of the air-interface organ culture model allowed the assessment of adherence of B. bronchiseptica to canine tracheal cilia, induction of mucus expression on to the tissue surface, induction of ciliostasis and damage to the ciliated epithelium in a single model system. To our knowledge, this is the first time that the temporal sequence of induction of ciliostasis, induction of mucus release and ciliated epithelial damage in response to Bordetella infection has been conclusively determined.
![]() |
ACKNOWLEDGEMENTS |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Bancroft, J. & Cook, H. (1994). Manual of Histological Techniques and their Diagnostic Application. Edinburgh: Churchill Livingstone.
Bemis, D. A. & Kennedy, J. R. (1981). An improved system for studying the effect of Bordetella bronchiseptica on the ciliary activity of canine tracheal epithelial cells. J Infect Dis 144, 349357.[Medline]
Bemis, D. A., Greisen, H. A. & Appel, M. J. G. (1977). Pathogenesis of canine bordetellosis. J Infect Dis 135, 753762.[Medline]
Bemis, D. A. & Wilson, S. A. (1985). Influence of potential virulence determinants on Bordetella bronchiseptica-induced ciliostasis. Infect Immun 50, 3542.[Medline]
Cherry, J. D. (1996). Historical review of pertussis and the classical vaccine. J Infect Dis 174, S259S263.[Medline]
Cotter, P. A. & Miller, J. F. (1994). BvgAS-mediated signal-transduction analysis of phase-locked regulatory mutants of Bordetella bronchiseptica in a rabbit model. Infect Immun 62, 33813390.[Abstract]
Dowling, R. B., Newton, R., Robichaud, A., Cole, P. J., Barnes, P. J. & Wilson, R. (1998). Effect of inhibition of nitric oxide synthase on Pseudomonas aeruginosa infection of respiratory mucosa in vitro. Am J Respir Cell Mol Biol 19, 950958.
Flak, T. A. & Goldman, W. E. (1999). Signalling and cellular specificity of airway nitric oxide production in pertussis. Cell Microbiol 1, 5160.[CrossRef][Medline]
Flak, T. A., Heiss, L. N., Engle, J. T. & Goldman, W. E. (2000). Synergistic epithelial responses to endotoxin and a naturally occurring muramyl peptide. Infect Immun 68, 12351242.
Funnell, S. G. & Robinson, A. (1993). A novel adherence assay for Bordetella pertussis using tracheal organ cultures. FEMS Microbiol Lett 110, 197203.[CrossRef][Medline]
Goldman, W. E., Luker, K. E., Flak, T. A. & Heiss, L. N. (1994). Tracheal cytotoxin induction of nitric-oxide synthesis role in the pathogenesis of pertussis. J Cell Biochem S18A, 40.
Goodnow, R. A. (1980). Biology of Bordetella bronchiseptica. Microbiol Rev 44, 722738.[Medline]
Groathouse, N. A., Heinzen, R. A. & Boitano, S. (2003). Functional BvgAS virulence control system in Bordetella bronchiseptica is necessary for induction of Ca2+ transients in ciliated tracheal epithelial cells. Infect Immun 71, 72087210.
Heiss, L. N., Flak, T. A., Lancaster, J. R., Jr, McDaniel, M. L. & Goldman, W. E. (1993). Nitric oxide mediates Bordetella pertussis tracheal cytotoxin damage to the respiratory epithelium. Infect Agents Dis 2, 173177.[Medline]
Jackson, A. D., Rayner, C. F., Dewar, A., Cole, P. J. & Wilson, R. (1996). A human respiratory-tissue organ culture incorporating an air interface. Am J Respir Crit Care Med 153, 11301135.[Abstract]
Kim, K. C., Hisatsune, A., Kim do, J. & Miyata, T. (2003). Pharmacology of airway goblet cell mucin release. J Pharmacol Sci 92, 301307.[CrossRef]
Martinez de Tejada, G., Miller, J. F. & Cotter, P. A. (1996). Comparative analysis of the virulence control systems of Bordetella pertussis and Bordetella bronchiseptica. Mol Microbiol 22, 895908.[CrossRef][Medline]
Matsuyama, T. (1977). Resistance of Bordetella pertussis phase I to mucociliary clearance by rabbit tracheal mucous membrane. J Infect Dis 136, 609616.[Medline]
Middleton, A. M., Chadwick, M. V., Nicholson, A. G., Dewar, A., Feldman, C. & Wilson, R. (2003). Investigation of mycobacterial colonisation and invasion of the respiratory mucosa. Thorax 58, 246251.
Muse, K. E., Collier, A. M. & Baseman, J. B. (1977). Scanning electron microscopic study of hamster tracheal organ cultures infected with Bordetella pertussis. J Infect Dis 136, 768777.[Medline]
Nennig, M. E., Shinefield, H. R., Edwards, K. M., Black, S. B. & Fireman, B. H. (1996). Prevalence and incidence of adult pertussis in an urban population. JAMA 275, 16721674.[Abstract]
Parkhill, J., Sebaihia, M., Preston, A. & 50 other authors (2003). Comparative analysis of the genome sequences of Bordetella pertussis, Bordetella parapertussis and Bordetella bronchiseptica. Nat Genet 35, 3240.[CrossRef][Medline]
Porter, J. F., Connor, K. & Donachie, W. (1996). Differentiation between human and ovine isolates of Bordetella parapertussis using pulsed-field gel-electrophoresis. FEMS Microbiol Lett 135, 131135.[CrossRef][Medline]
Preston, A., Maxim, E., Toland, E., Pishko, E. J., Harvill, E. T., Caroff, M. & Maskell, D. J. (2003). Bordetella bronchiseptica PagP is a Bvg-regulated lipid A palmitoyl transferase that is required for persistent colonization of the mouse respiratory tract. Mol Microbiol 48, 725736.[CrossRef][Medline]
Preston, A., Parkhill, J. & Maskell, D. J. (2004). The bordetellae: lessons from genomics. Nat Rev Microbiol 2, 379390.[CrossRef][Medline]
Rogers, D. F. (1994). Airway goblet cells: responsive and adaptable front-line defenders. Eur Respir J 7, 16901706.
Rogers, D. F. (2003). The airway goblet cell. Int J Biochem Cell Biol 35, 16.[CrossRef][Medline]
Soane, M. C., Jackson, A., Maskell, D., Allen, A., Keig, P., Dewar, A., Dougan, G. & Wilson, R. (2000). Interaction of Bordetella pertussis with human respiratory mucosa in vitro. Respir Med 94, 791799.[CrossRef][Medline]
Thompson, H., McCandlish, I. A. P. & Wright, N. G. (1976). Experimental respiratory disease in dogs due to Bordetella bronchiseptica. Res Vet Sci 20, 1623.[Medline]
Thornton, D. J., Holmes, D. F., Sheehan, J. K. & Carlstedt, I. (1989). Quantitation of mucus glycoproteins blotted onto nitrocellulose membranes. Anal Biochem 182, 160164.[Medline]
Wilson, R., Read, R. & Cole, P. (1992). Interaction of Haemophilus influenzae with mucus, cilia, and respiratory epithelium. J Infect Dis 165, S100S102.[Medline]
Received 28 April 2004;
revised 26 May 2004;
accepted 3 June 2004.
HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
INT J SYST EVOL MICROBIOL | MICROBIOLOGY | J GEN VIROL |
J MED MICROBIOL | ALL SGM JOURNALS |