|Home | About | Journals | Submit | Contact Us | Français|
Biofilms are communities of microorganisms that are encased in polymeric matrixes and grow attached to biotic or abiotic surfaces. Despite their enhanced ability to resist antimicrobials and components of the immune system in vitro, few studies have addressed the interactions of biofilms with the host at the organ level. Although mycoplasmas have been shown to form biofilms on glass and plastic surfaces, it has not been determined whether they form biofilms on tracheal epithelium. We developed a tracheal organ mounting system that allowed the entire surface of the tracheal lumen to be scanned by using fluorescence microscopy. We observed the biofilms formed by the murine respiratory pathogen Mycoplasma pulmonis on the epithelium of trachea in tracheal organ culture and in experimentally-infected mice and found similar structure and biological characteristics as biofilms formed in vitro. This tracheal organ mounting system can be used to study interactions between biofilms formed by respiratory pathogens and the host epithelium and to identify the factors that contribute to biofilm formation in vivo.
Mycoplasmas are significant and chronic pathogens of man and animals that primarily colonize the mucosal surfaces of the respiratory and genitourinary tracts (Cassell, et al., 1985). Although they lack a cell wall and would appear to be sensitive to the host’s immune system, the mycoplasmas persist despite an intense inflammatory response. A key to the survival of Mycoplasma pulmonis is the V-1 variable surface antigen (Vsa protein). The length of the tandem repeat region of the Vsa protein modulates biofilm formation and the susceptibility of individual mycoplasma cells to killing by complement (Simmons & Dybvig, 2003; Simmons, et al., 2004; Simmons, et al., 2007). Mycoplasmas producing a short Vsa protein containing few tandem repeats are sensitive to complement when dispersed but form complement-resistant biofilms on glass and plastic surfaces. Mycoplasmas producing a long Vsa protein containing as many as 40 to 60 tandem repeats do not form biofilms but are nevertheless resistant to complement.
Biofilms formed in vitro protect mycoplasmas from the lytic effects of not only complement but also the small antimicrobial peptide gramicidin (Simmons & Dybvig, 2007). The biofilm has honeycombed regions consisting of thin layers of mycoplasmas pocketed with cavities. Interspersed in and arising out of the honeycombed regions are towers (Simmons, et al., 2007). The mycoplasmas in the honeycombed region are sensitive to complement and gramicidin while the mycoplasmas within the towers are protected (Simmons & Dybvig, 2007). The ability to survive in the host may be attributed to the ability to form biofilms, or perhaps protective tower structures, on the tracheal epithelium. It remains to be determined whether mycoplasmas form biofilms or tower structures in vivo and whether factors that modulate biofilm formation in vitro function similarly on tracheal epithelium. In the current study mice were intranasally infected with mycoplasmas to determine whether biofilms form on tracheal epithelium in vivo. Whether the Vsa protein influences biofilm formation on tracheal epithelium was examined by infecting tracheal organ cultures (TOC) with mycoplasma strains that differ in the length of the Vsa protein produced.
For observation of biofilms in TOC, whole trachea were aseptically excised from anesthetized female C57BL/6 or BALB/c mice (15 to 20 grams) and placed into 1 mL of mycoplasma broth (pH 7.8) containing 20% horse serum, 100 μg mL−1 ampicillin, and phenol red as a pH indicator (Gumulak-Smith, et al., 2001). After acclimation to this medium at 37°C overnight, the trachea were incubated in 1 mL of mycoplasma broth containing 105 colony-forming units (CFU) of M. pulmonis strains producing either a short VsaA protein (strain CT182-R3), a long VsaA protein (strain CT182-R40), a short VsaG protein (strain CTG-R5) (Simmons & Dybvig, 2003; Simmons, et al., 2004) or in mycoplasma broth only. Strains CT182-R3 and CTG-R5 form biofilms in vitro and infect mice (Simmons, et al., 2007). Acid production was monitored though the use of phenol red as a pH indicator to assess the growth of the mycoplasmas. The mycoplasma broth was changed daily in all of the cultures as the pH neared 7.0.
Two days after infection, the explants were washed 3 times for 30 minutes in 1 mL of phosphate-buffered saline (PBS) with gentle shaking. This was followed by incubation in 1 mL of 10% neutral buffered formalin (NBF; 4.0% formaldehyde, 0.4% sodium dihydrogen orthophosphate, 0.65% disodium hydrogen orthophosphate [pH 7.0]) for four hours at 4°C followed by incubation in a 10-fold dilution of NBF in PBS overnight at 4°C. The explants were then washed 3 times for 30 minutes in PBS at room temperature.
To detect biofilms or towers, the explants were incubated with PBS containing 10% normal goat serum for 30 minutes and incubated for 2 hours in a 200-fold dilution of rabbit immune serum containing antibodies that recognize the VsaA protein (isotype A) (Bhugra, et al., 1995; Simmons, et al., 2007). After washing the explants 3 times for 20 minutes in PBS, the explants were incubated with goat anti-rabbit antibodies conjugated to the fluorochrome AF-594 (Molecular Probes; A11012; at 1 μg mL−1) for 2 hours and washed 3 times again for 20 minutes in PBS.
Microscope slides were washed in 100% ethanol. Four pedestals were cut from plastic that was 0.51 mm thick and fixed to the glass slides with ethyl-2-cyanoacrylate (Superglue™, Loctite Corporation). One explant per slide was placed between the pedestals onto a 40-μL drop of a 50:50 mixture of glycerol and Prolong Gold mounting solution (Molecular Probes), observed with a 40x power binocular dissecting microscope and trimmed of excess tissue with a single-edged razor blade. The trachea were cut longitudinally with a single thrust of the blade and the tracheal sections aligned so that the lumen were positioned upward. The explants were sealed onto the glass slide by fixing a glass cover slip to the pedestals with ethyl-2-cyanoacrylate. The slides were allowed to settle overnight with a pressure of 70 g cm−2 applied evenly to the glass cover slip.
The trachea were observed with a Leica HC microscope using a Chroma 86012v2 Texas Red and DAPI filter set to detect the VsaA epitopes and DNA, respectively. Stacks of images were acquired at 320x (Leica 20x HC PL FLUOTAR objective, numerical aperture 0.50; through a 1.6x tube lens) or 1600x (Leica 100x HCX PL FLUOTAR oil immersion objective, numerical aperture 1.30; through a 1.6x tube lens) magnification using the MetaMorph software package (Version 6.3r2, Molecular Devices, Inc.). Images were thresholded to remove background, deconvolved using the iterative 3D deconvolution ImageJ plugin and three-dimensional reconstructions of the resulting image stacks were performed with the software application Image J (National Institutes of Health, version 1.38x). In some images the DAPI and Texas Red channels were merged to give two-color micrographs where red represents the VsaA epitopes and blue represents the DNA of the nuclei of the epithelial cells and the DNA of the mycoplasmas (shown in supplementary material). The majority of the images of the mycoplasmas were acquired in the single channel mode to detect the Vsa epitopes. To enhance the visibility of the mycoplasmas over that of grayscale images, these single-channel images were pseudocolored with the HotRed indexed color lookup table provided with ImageJ.
Ten tracheal explants were inoculated with 105 CFU of either strains CT182-R3 or CT182-R40 in 1 ml of mycoplasma broth for 2 days. To determine the level of non-specific binding by the VsaA-specific antibodies, two explants were incubated with mycoplasma broth only and 2 explants were inoculated with 105 CFU of the VsaG-producing strain CTG-R5. The medium was changed daily. The explants were processed for indirect immunofluorescence microscopy as described above. Thirty stacks of images in the Z-plane were acquired at various x, y coordinates on the tracheal surface from explants that were incubated with CT182-R40. Twenty-eight stacks of images were acquired from the explants that were incubated with CT182-R3. The images were acquired at 320x magnification. Z-projections of these image stacks produced images were identically thresholded to remove background. We defined a tower as an object with a surface area on the image equal or greater than 20 μm2, which corresponds to the size of a small mycoplasma tower. Using the particle analysis function of ImageJ, the total number of objects identified as either towers or non-towers was determined. For each strain, the number of objects that were identified as towers relative to the number of objects that were not identified as towers were analyzed by the Chi Square method.
Experiments utilizing NOD/SCID or C57BL/6 mice (15 to 20 grams; Jackson Laboratories) were performed separately. In total, 5 NOD/SCID and 4 C57BL/6 mice were intranasally inoculated with 108 CFU of mycoplasma strains CT182-R3, and 5 NOD/SCID and 2 C57BL/6 mice were inoculated with strain CTG-R5 as described (Denison, et al., 2005). After 3 days the trachea were excised and prepared for immunofluorescence as described above. One trachea was mounted per slide and replicate fields of view were analyzed as described above.
Biofilms of M. pulmonis strain CT182-R3 were grown in mycoplasma broth on 22 mm by 22 mm glass cover slips for 2 days. The biofilms were prepared to detect VsaA epitopes by immunofluorescence microscopy as previously described (Simmons, et al., 2007). The cover slips were mounted directly to the microscope slides without using plastic pedestals.
We compared the biofilms and towers formed by mycoplasmas on tracheal epithelium in organ culture and in vivo to the biofilms they formed in vitro. The mycoplasma towers formed on the epithelium of TOC, in vivo and in vitro are shown in Figure 1, Figure 2 and Figure S1, respectively. As mycoplasmas form biofilms on glass, they progress from individual cells attaching to the glass and forming small towers, to tower structures and to continuous films (Simmons, et al., 2007). We observed all of these structures on tracheal epithelium. Fluorescence microscopy performed with the DAPI and Texas Red filter sets revealed the DNA of the epithelial and mycoplasma cells, and the VsaA epitopes of the mycoplasmas, respectively. At 320x magnification we observed individual mycoplasmas and many larger structures formed by the mycoplasmas (Figure 1, panel A). High-resolution microscopy revealed these larger structures formed on the epithelium of the tracheal explants (Figure 1, panels C to H) and the epithelium of trachea in vivo (Figure 2, panels A to D) and that they resembled the towers of mycoplasmas grown on glass cover slips (Figure S1, panels A to E). The diameters of the towers ranged from about 5 μm (Figure 1, panel G; Figure 2 panels C and D) to over 20 μm (Figure 1, panels C to F; Figure 2, panels A to D), and many of the towers were over 10 um in height. This is comparable to what was observed in vitro (Figure S1). Films of mycoplasmas were observed on the epithelium of trachea from mice that were intranasally inoculated with strain CT182-R3 (Figure 2, panel E) but not on the epithelium of the explants. This could have resulted from the mice being infected for three days while the explants were incubated with the mycoplasmas for only two days. Because the antibody used was specific for VsaA, no mycoplasmas were observable on tracheal epithelium of mice or on the explants that were infected with the strain of mycoplasma that produced VsaG (Figure 1, panel B; Figure 2, panel F). Also the antibody did not bind to the explants that were incubated with mycoplasma broth only (not shown). Therefore the VsaA-specific antibody had no significant background binding to the epithelium or the VsaG-producing mycoplasmas as previously reported (Simmons, et al., 2007). No differences were detected in the ability of the mycoplasmas to form towers on the trachea of the BALB/c, NOD/SCID or C57BL/6 mice.
By mounting the tracheal explants longitudinally on the slides, we were able to examine the entire surface of the lumen for rapid analysis of a large continuous section of tracheal surface. An examination of the tracheal lumen indicated that strains of mycoplasmas that have been shown previously to form biofilms in vitro similarly formed biofilms on the tracheal epithelium. The mycoplasmas that produced a short Vsa protein formed significantly more towers than did the mycoplasmas that produced a long Vsa protein (P < 0.001; Figure S2, panels A and B respectively; representative towers are noted by yellow arrows). 6.1 % of the objects identified on trachea that were infected with strain CT182-R3 were towers. In contrast only 0.6 % of the objects identified on the trachea that were infected with strain CT182-R40 were towers. This was likely not due to differential growth of the mycoplasmas in the cultures as about equal numbers of CFU were recovered from the cultures. This is consistent with the differential ability of these strains to form biofilms in vitro (Simmons & Dybvig, 2003; Simmons, et al., 2004; Simmons, et al., 2007). No mycoplasmas were detectable on explants that were incubated with strain CTG-R5 (Figure S2, panel C) because of the specificity of the antibody for VsaA epitopes.
A feature observed with the tracheal explant system is that three-dimensional reconstructions of the epithelium revealed that not all mycoplasmas or towers lie in the same Z-plane (Figure 1, panels H; Figure 2, panels B and D). The points at which the mycoplasmas attached to the glass lied uniformly in the same plane (Figure S1, panel E). This likely reflects the mycoplasmas attaching at the surface of the epithelium which itself may lie in different planes of the micrographs. Alternately, this could result from mycoplasmas penetrating into the epithelium to varying degrees. As nuclei of the epithelial cells are positioned basally near the basement membrane in the trachea, co-localizing the mycoplasmas to the epithelial nuclei would not provide an accurate measurement of how deep the mycoplasmas penetrate into the epithelial layers. These measurements would be best determined utilizing markers that define the luminal surfaces of the epithelial cells, perhaps markers that recognize tubulin or actin.
The depth to which the Vsa epitopes can be detected in the tower structures, an indicator of the depth to which antibodies penetrate the mycoplasmal towers, is similar on the towers formed in vitro and on the tracheal epithelium. Consistent with our previous results in vitro, the Vsa epitopes can only be detected on the outer layer of the towers (Figure S1, panel C; Figure 1, panel F) (Simmons, et al., 2007). This need not be the result of reduced VsaA expression in the tower interior because complement and the small antimicrobial peptide gramicidin cannot penetrate into the towers (Simmons & Dybvig, 2007). However, mycoplasma cells were efficiently killed by complement and gramicidin once the biofilms were dispersed. These results suggest that towers formed on tracheal epithelium protect the mycoplasmas from host immunity.
These results indicate the biofilms formed on tracheal epithelium of three different strains of mice are similar to the biofilms that are formed in vitro. The structure of the towers formed on the tracheal epithelium is similar to the structure of the towers in mycoplasma biofilms formed in vitro, with similar resistance to penetration by antibodies and dependence on a short Vsa protein. The towers are densely packed with mycoplasma cells that have a typical diameter of 500 nm. Towers with a 20 um2 cross section and a height of 10 μm might contain over twenty thousand cells and represent a substantial reservoir of mycoplasmas that are resistant to host immunity and from which chronic infections could be maintained.
As the ability of the mycoplasmas to form biofilms in vivo may parallel their ability to form biofilms in TOC, an examination of the effects of virulence factors on biofilm formation in TOC may provide an indicator of how those factors affect virulence in the animal. Additionally, the tracheal organ mounting system described here can be applied to biofilms formed by other bacterial species to test the validity of the current models for biofilm formation that have been developed in vitro. Rather than inferring that biofilms form on tracheal epithelium in vivo by extrapolating from in vitro data, direct observations of biofilm formation can be made and it can be determined whether factors that contribute to in vitro spatial and temporal organization of biofilms perform similar functions ex vivo or in vivo.
Figure S1. Immunofluorescence microscopy of tower formed on glass. All images were acquired at 1600x magnification in a single-channel mode and pseudocolored to enhance visibility. Overhead view (A), view at 40 degree angles (B), and a cross-sectional view (C) of towers formed on glass cover slips. Panels D and E are overhead views or side views, respectively, of small mycoplasma towers. Scale bars represent 10 μm. Yellow arrows denote examples of towers and green arrows denote individual mycoplasma cells.
Figure S2. The length of the Vsa protein modulates tower formation on epithelium in TOC. Low power magnification of towers formed on TOC by strains CT182-R3 (A), CT182-R40 (B) or CTG-R5 (C; two-color micrograph, DNA of the tracheal epithelial cells in blue and Vsa epitopes in red). All images are pseudocolored to enhance visibility. The yellow arrows point to examples of towers. Scale bars represent 100 μm.
This work was supported by the National Institutes of Health grant AI64848. We thank Paula B. Simmons for her assistance in developing the TOC slide mounting system.