Identification of exopolysaccharide-deficient mutants of Mycoplasma pulmonis

Authors


*E-mail dybvig@uab.edu; Tel. (+1) 205 934 9327; Fax (+1) 205 975 4418.

Summary

The presence of capsular exopolysaccharide (EPS) in Mollicutes has been inferred from electron micrographs for over 50 years without conclusive data to support the production of complex carbohydrates by the organism. Mycoplasma pulmonis binds the lectin Griffonia simplicifolia I (GS-I), which is specific for terminal β-linked galactose residues. Mutants that failed to produce the EPS bound by GS-I were isolated from a transposon library. All of the mutants had the transposon located in open reading frame MYPU_7410 or MYPU_7420. These overlapping genes are predicted to code for a heterodimeric pair of ABC transporter permeases and may code for part of a new pathway for synthesis of EPS. Analysis by lectin-affinity chromatography in conjunction with gas chromatography demonstrated that the wild-type mycoplasma produced an EPS (EPS-I) composed of equimolar amounts of glucose and galactose that was lacking in the mutants. Phenotypic analysis revealed that the mutants had an increased propensity to form a biofilm on glass surfaces, colonized mouse lung and trachea efficiently, but had a decreased association with the A549 lung cell line. Confounding the interpretation of these results is the observation that the mutants missing EPS-I had an eightfold overproduction of an apparent second EPS (EPS-II) containing N-acetylglucosamine.

Introduction

Polysaccharides on the surface of pathogenic bacteria have diverse roles. Although the bacterial glycans vary in size, composition, mode of attachment and structure, the importance of this class of macromolecules in bacteria–host interactions is indisputable. The term exopolysaccharide (EPS) is appropriate for either capsular polysaccharide that is anchored to the bacterial surface or polysaccharide that is secreted into the extracellular matrix (Sutherland, 1972). EPS is required for full virulence for many bacteria including Streptococcus pneumoniae (Avery and Dubos, 1931; Watson and Musher, 1990), Mycobacterium tuberculosis (Stokes et al., 2004) and Staphylococcus aureus (Stone and Saiman, 2007). Bacterial glycans can modulate many properties including antiphagocytosis, antibacteriolytic activity, adhesion, immune responses and evasion, and biofilm production (Comstock and Kasper, 2006).

Mycoplasma pulmonis is a diminutive bacterium and the aetiological agent of murine respiratory mycoplasmosis, a pneumonia-like disease in mice and rats (Watson et al., 1989; Simecka et al., 1992). The class Mollicutes (trivial name, mycoplasmas) is characterized by the lack of a cell wall and a small genome size. These bacteria have evolved by degenerative evolution from a Gram-positive progenitor to exist in a specific niche within a single host (Woese et al., 1980). In 1958 the galactan of Mycoplasma mycoides ssp. mycoides was among the first toxic molecules identified from pathogenic mycoplasmas (Plackett and Buttery, 1958). The research that has been conducted on the capsule of pathogenic mycoplasma species in the intervening 50 years has been incomplete. Capsules have been suggested for several species of Mycoplasma in addition to M. mycoides ssp. mycoides (Gourlay and Thrower, 1968). Examples include Mycoplasma dispar (Almeida and Rosenbusch, 1991), M. gallisepticum (Chu and Horne, 1968), M. hyopneumoniae (Tajima and Yagihashi, 1982), M. penetrans (Neyrolles et al., 1998), M. pneumoniae (Wilson and Collier, 1976) and M. synoviae (Ajufo and Whithear, 1978), based mainly on electron microscopy and cationic staining using ruthenium red (Almeida and Rosenbusch, 1991). Capsular-deficient mutants have not been previously described for any species of mycoplasma, and it has not been shown whether capsular material was synthesized by the mycoplasma or adsorbed from the environment.

We show here that M. pulmonis produces an EPS (EPS-I) that is recognized by the lectin Griffonia simplicifolia I (GS-I) and is composed of galactose and glucose. Evidence is presented that the mycoplasma produces a second EPS (EPS-II) that is recognized by the lectins G. simplicifolia II (GS-II) and wheat germ agglutinin (WGA). From the lectin binding, EPS-II is predicted to have a terminal N-acetylglucosamine (GlcNAc) residue. Mutants from a transposon library were isolated that failed to produce EPS-I. Each of these mutants has the transposon located in one of two overlapping genes, MYPU_7410 and MYPU_7420, that are both annotated as ATP-binding cassette (ABC) transporter and permease proteins. The results indicate that EPS-I is indeed produced by the mycoplasma, and we suggest that the machinery for polysaccharide synthesis is novel.

The EPS-I mutants were examined for their ability to form a biofilm on glass surfaces and associate with epithelial cells. Previous studies have shown that the Vsa proteins of M. pulmonis modulate several properties including the ability of the mycoplasma to attach to glass or plastic and form a biofilm (Simmons and Dybvig, 2003; Simmons et al., 2007). When the Vsa protein is long with many tandem repeats (up to 60) in its carboxyl domain, attachment does not occur and no biofilm is formed. When the Vsa protein is short with few (five or less) tandem repeats, attachment occurs and biofilms are formed. The length of the Vsa protein varies due to slipped-strand mispairing during DNA replication of the tandem repeat region (Dybvig, 1993). In addition to size variation via slipped-strand mispairing, the Vsa family of surface proteins phase-vary as a result of site-specific DNA inversions (Bhugra et al., 1995; Shen et al., 2000). However, biofilm formation is dependent only on the length of the Vsa protein and not on which particular member of the Vsa family is produced. Unlike wild-type cells, we find that the EPS-I mutants formed a biofilm even when a long Vsa protein was produced. The enhanced ability of the mutants to form a biofilm may result from their overproduction of EPS-II. The EPS-I mutants colonized the mouse respiratory tract efficiently but had reduced association with the human A549 lung cell line, suggesting that EPS-I may have a role in cytadherence.

Results

Evidence for polysaccharide production by M. pulmonis

Previous studies have shown that cultures of M. pulmonis react strongly with the lectins GS-I, GS-II and WGA (Simmons et al., 2007) suggesting that the mycoplasma produces at least one and possibly additional glycomoieties. Electron micrographs reveal extracellular material that appears to be capsular extending well beyond the surface of the mycoplasma, suggestive of EPS (Fig. 1).

Figure 1.

Representative electron micrograph of M. pulmonis cell with surrounding capsular material.

A transposon library of M. pulmonis was screened to identify mutants that were deficient in polysaccharide production. Disrupted in the transposon library are about half of the mycoplasma's 782 protein-coding regions (French et al., 2008). At least one mutant for each of the dispensable genes was examined for polysaccharide deficiency by electrophoresis of M. pulmonis lysates on polyacrylamide gels followed by staining with the periodic-Acid Schiff's (PAS) reaction. When the gels were fixed prior to PAS staining, a white precipitant was evident at the top of the wells. The mutants described here were initially identified by the absence of this material. Upon staining with PAS, the mutants exhibited a reduced level of stained material at the top of the wells (Fig. 2). Four mutants were identified as summarized in Table 1. The transposon in transformants CTG2028, CTG1701 and CTG1516 is inserted in the gene MYPU_7410 while the transformant CTG1291 has the transposon inserted in the overlapping gene MYPU_7420 (Fig. 3).

Figure 2.

Polyacrylamide gel electrophoresis of M. pulmonis lysates stained with PAS (A), destained, and restained with Coomassie (B). Numbers on left refer to protein standards in kilodaltons. The EPS does not enter the gel as noted by the arrows.

Table 1.  Mycoplasma strains used in this study.
StrainRelevant propertySource or reference
CTWild-type parent of all other mycoplasma strains used in this studyDavis et al. (1986); French et al. (2008)
CTG-R40Long Vsa protein, non-biofilm formingSimmons et al. (2007)
CTG-R5Short Vsa protein, biofilm formingSimmons et al. (2007)
CTG38Long Vsa protein, non-biofilm forming, indistinguishable from the parental CT strain in all assays used in this study, contains transposon at presumably innocuous site for the purposes of this study, disrupting MYPU_0860 (coding for a conserved hypothetical protein)This study
CTG2028EPS-I mutant, intermediate-length Vsa protein, biofilm forming, transposon disrupts MYPU_7410This study
CTG1516EPS-I mutant, long Vsa protein, biofilm forming, transposon disrupts MYPU_7410This study
CTG1291EPS-I mutant, long Vsa protein, biofilm forming, transposon disrupts MYPU_7420This study
CTG1701EPS-I mutant, long Vsa protein, biofilm forming, transposon disrupts MYPU_7410This study
CTG1701-CCTG1701 complemented with MYPU_7410 and 7420, produces EPS-IThis study
Figure 3.

Schematic of the overlapping MYPU_7410 and 7420 genes and flanking regions. Arrows refer to direction of transcription. The location of the transposon in the mutants are as shown. The nucleotide position of the transposon within the M. pulmonis genome in CTG1701, CTG1516, CTG2028 and CTG1291 is 911295, 911453, 911543 and 913643 respectively. In CTG1701, CTG1516, CTG2028, the 542-amino-acid MYPU_7410 gene product would be truncated to 176, 228, 258 amino acids respectively. In CTG1291, the 538-amino-acid MYPU_7420 gene product would be truncated to 423 amino acids. Small arrows refer to the binding sites of the forward (For) and reverse (Rev) primers used to amplify the MYPU_7410 and 7420 genes for complementation of EPS mutants.

Comparison of polysaccharides from mutant and wild-type mycoplasmas

Fluorescein isothiocyanate (FITC)-tagged lectins were used to examine the glycomoieties of the mutants and wild-type mycoplasmas. The mutants were compared with the wild-type parent strain for the initial experiments, but in later experiments transformant CTG38 was used in place of the wild-type strain. CTG38 was chosen as a control because it has the transposon inserted at an innocuous site and was found to produce glycomoieties identical to those of the parent strain. The lectin GS-I bound strongly to the parent strain and to CTG38 but not to each of the mutants (Fig. 4A–C), suggesting that the mutants had lost a terminal galactose moiety. Surprisingly, the lectin GS-II reacted more strongly with the mutants than with the parent strain or CTG38, suggesting that the mutants had more terminal GlcNAc than did wild-type cells (Fig. 4D–F). Although it is not as specific as GS-II, WGA also binds to terminal GlcNAc and was similarly found to react more strongly with the EPS-I mutants than to the parent strain (data not shown).

Figure 4.

M. pulmonis strains reacted with FITC-conjugated lectins (green) and stained with Hoechst 33342 (blue) to label DNA. (A), (B), (C) and (G) are wild-type CT, the EPS-I mutant CTG2028, the EPS-I mutant CTG1291 and the complemented mutant CTG1701-C, respectively, reacted with GS-I lectin. (D), (E) and (F) are CT, CTG2028 and CTG1291, respectively, reacted with GS-II lectin.

EPS-I was isolated from wild-type M. pulmonis by using a GS-I lectin affinity column. The isolated samples were subjected to methanolysis to generate the methyl glycosides and the trimethylsilyl derivatives were analysed by gas chromatography (GC). Monosaccharide analysis of the polysaccharide (Fig. 5A) showed equimolar levels of galactose and glucose as determined by comparison to a standard derived from the glucose-galactose disaccharide melibiose (Fig. 5B). No sugars other than glucose and galactose were detected in purified EPS-I. These data are consistent with a model for EPS-I as a polymer containing alternating residues of galactose and glucose. When lysates of the polysaccharide-deficient mutants were applied to the affinity column, the eluted material had no EPS-I detected by GC (Fig. 5C and D).

Figure 5.

Gas chromatogram of methyl glycosides of EPS. (A)–(D), wild-type M. pulmonis strain CT, melibiose standard (a disaccharide composed of glucose and galactose), EPS-I mutant CTG2028 and EPS-I mutant CTG1291 respectively.

The GC analysis of lysates from EPS-I mutants and wild-type mycoplasmas indicated that the mutants had about a eightfold increase in GlcNAc as compared with the wild-type mycoplasma (Fig. 6). This increase was calculated by comparing the area under the peaks of the sugars for samples of equilibrated protein concentration. This finding is consistent with the observation that the GS-II lectin reacted more strongly with the mutants than to wild-type cells (Fig. 4). We propose that M. pulmonis produces a second EPS (EPS-II) that contains GlcNAc and is overproduced in the EPS-I mutants.

Figure 6.

Gas chromatogram of methyl glycosides of EPS. (A)–(C), wild-type M. pulmonis strain CT, EPS-I mutant CTG1516 and GlcNAc standard respectively.

MYPU_7410 and MYPU_7420 sequences

The coding regions of the MYPU_7410 and 7420 genes in the mycoplasma genome begin at nucleotide position 910 767 and 912 373 and end at 912 392 and 913 986 respectively. Thus, the coding regions overlap by 20 bp. The 542- and 538-amino-acid products of MYPU_7410 and MYPU_7420, respectively, share 29% amino acid sequence identity and are both annotated as ABC transporter ATP-binding and permease proteins. Both proteins have multiple domains associated with multidrug resistance and also a domain associated with glucan exporters. blast analysis indicated that both proteins exhibit significant sequence similarity with numerous proteins annotated as multidrug transporters from a broad range of bacteria. The top hits by blast were scanned to identify adjacent genes that would code for a pair of proteins highly similar to the MYPU_7410 and 7420 pair. The MYPE2210 and MYPE2200 genes from M. penetrans (Accession No. NP_757608 and NP_757607), and several pairs of genes from various strains of M. hyopneumoniae are candidates for having similar function to MYPU_7410 and 7420. The accession numbers for the M. hyopneumoniae pairs are YP_116194 and YP_116193, YP_288049 and YP_288050, YP_288013 and YP_288012, YP_288050 and YP_288049, YP_279423 and YP_279422, and YP_116157 and 116156.

Complementation of CTG1701

Strain CTG1701-C, strain CTG1701 complemented with the intact MYPU_7410 and 7420 genes, produced wild-type levels of EPS-I as evidence by restoration of binding to GS-I (Fig. 4G) and of wild-type levels of glucose and galactose detected by monosaccharide analysis of lysates (Fig. 7). The amount of EPS-II produced by CTG1701-C was reduced to that of wild-type CTG38. Therefore, overproduction of EPS-II in the mutants appears to be linked to the loss of EPS-I.

Figure 7.

Gas chromatogram of methyl glycosides of EPS. (A)–(C), strains CTG1701, CTG1701-C and CTG38 respectively.

Biofilm formation

To study the influence of EPS on biofilm formation, the experimental plan was to compare mutants with wild-type strains that have the same-length Vsa protein. The length of the Vsa proteins was examined by Western blot analysis using a Vsa-specific antibody. Most of the mutants had a long Vsa protein similar to the parent strain (Fig. 8). One of the mutants (CTG2028) was excluded from biofilm studies because it had a comparably short Vsa protein.

Figure 8.

Western blot of M. pulmonis proteins reacted with Vsa-specific monoclonal antibody. Numbers on left refer to protein standards in kilodaltons. Arrows indicate the upper most Vsa band which corresponds to a Vsa protein with about 40 tandem repeats in CTG38 and a smaller protein with fewer tandem repeats in CTG2028.

Microscopic examination of M. pulmonis grown on glass coverslips revealed that wild-type cells producing a long Vsa protein failed to form a biofilm and wild-type cells producing a short Vsa protein did form a biofilm as reported in previous studies (Simmons and Dybvig, 2003; Simmons et al., 2007). Unexpectedly, the EPS-I mutants producing a long Vsa protein formed a biofilm, nearly as robust as the biofilm formed by wild-type mycoplasmas producing a short Vsa protein (Fig. 9).

Figure 9.

Biofilms formed on glass by M. pulmonis strains. Shown in (A)–(E) are biofilms for strains CTG1701, CTG1516, CTG1291, CTG-R40 and CTG-R5 respectively. Top or overhead view of each biofilm is shown on the left and side view shown on the right.

Colonization of SCID mice

EPS may influence many aspects of the infection process. In the present study, the ability of the EPS-I mutants to colonize mice was examined. Animals with a severe combined immunodeficiency (scid) mutation were chosen to study colonization per se, in the absence of a robust immune response. At 7 days post inoculation, the colony-forming unit (cfu) recovered from the lung and trachea of mice infected with the EPS-I mutant CTG1291 was 2.0 ± 0.2 × 107 (mean ± standard error of the mean) compared with 2.9 ± 0.1 × 106 for animals infected with the control strain CTG38. Although the difference in cfu recovered from mice infected with these two strains was statistically significant, the study is limited in scope and the only strong conclusion we reach is that EPS-I is not required for colonization. In additional experiments, we have shown that EPS-I mutants also colonize efficiently mouse strains DBA2/J and C57BL/6J (J.R. Bolland and K. Dybvig, unpubl. data).

Association with A549 cells

Compared with the wild-type mycoplasmas, statistically significant fewer EPS-I mutants were found in association with A549 cells after co-incubation for 2.5 h (Fig. 10). We propose that the EPS-I mutants adhered less efficiently to the A549 cells, but further studies would be required to rule out the alternative possibility that the EPS-I mutants were simply being killed. Killing seems unlikely because the total number of cfu recovered after incubation with the A549 cells was the same for both the wild-type cells and the mutant mycoplasmas.

Figure 10.

Association of M. pulmonis with A549 cells. A549 cells (1 × 106) were incubated with the indicated strain of M. pulmonis for 2.5 h, washed and assayed for mycoplasma cfu. The data represent a combination of three separate experiments that were performed in triplicate. The data are represented as the mean (± standard error of the mean). Asterisk refers to the statistical difference between CTG38 and the other two groups.

Discussion

In this work we have identified and characterized the first polysaccharide-deficient mutants of a mycoplasma, providing the tools for an extended study on the interactions of bacterial glycans and the host. With its minimal genome and simplistic cellular structure, M. pulmonis provides an excellent murine model for the study of the role of EPS in pulmonary infections. Our data clearly establish the presence of an EPS that is produced by M. pulmonis and composed of an equimolar amount of the monosaccharides glucose and galactose. We predict that EPS-I consists of a linear chain of alternating residues of glucose and galactose. EPS-I should have a terminal β-linked galactose residue based on its binding with GS-I lectin. We are not aware of reports of an EPS with the proposed structure of EPS-I in other bacterial species. However, based on homology to MYPU_7410 and 7420, M. hyopneumoniae and M. penetrans may produce a molecule very similar if not identical to EPS-I.

Many bacteria produce EPS with a simple structure consisting of alternating residues of two sugars. For example, the type III capsule of S. pneumoniae consists of alternating residues of glucose and glucuronic acid. The glycosyltransferase that synthesizes this type of polysaccharide is a processive synthase located in the bacterial membrane (Cartee et al., 2005). In bacteria that have only a single membrane and no periplasmic space, the synthase pumps the polysaccharide through the membrane as it is produced.

One might expect M. pulmonis to have a processive synthase similar to that of S. pneumoniae for the production of EPS-I and perhaps a second synthase to produce EPS-II. However, M. pulmonis has only one annotated glycosyltransferase gene (MYPU_7700). The gene product has a predicted domain for binding amino sugars. MYPU_7700 has not been disrupted in our transposon library (French et al., 2008). Apparent orthologues of this gene have also not been disrupted in transposon libraries of Mycoplasma arthritidis and Mycoplasma genitalium (Glass et al., 2006; Dybvig et al., 2008). Perhaps MYPU_7700 and its orthologues code for an enzyme that is required for synthesis of a glycolipid, such as the glycolipid MfGl-II described for Mycoplasma fermentans that has been shown to induce cytokine production and have a role in attachment to host cells (Brandenburg et al., 2003). Because mycoplasmas have only a single membrane and no cell wall, it would be surprising if the machinery for synthesis does not simultaneously excrete the EPS. Although the MYPU_7410 and 7420 gene products have no obvious similarity to known glycosyltransferases, we propose that these proteins form a heterodimeric synthase that produces EPS-I, with one protein binding glucose and the other galactose.

The overproduction of EPS-II in the EPS-I mutants is interesting. EPS-II production may be upregulated in response to the lack of EPS-I, but we prefer an alternative model in which the hypothesis of gene regulation is not invoked. In this model, more EPS-II is produced because the loss of EPS-I results in an increase in the intracellular concentration of the sugar substrates required for EPS-II as has been described for control of polysaccharide chain length in S. pneumoniae (Ventura et al., 2006). Compared with control strains that produce the same length of Vsa protein, the EPS-I mutants are proficient at biofilm formation. One possibility is that EPS-I hinders biofilm formation, as seen for some soluble polysaccharides (Valle et al., 2006). However, GlcNAc-containing polysaccharides are adhesive with a major role in biofilm formation in many bacterial species (Mack et al., 1996; Wang et al., 2004). We view it likely that the overproduction of EPS-II in the EPS-I mutants is responsible for enhanced biofilm formation and that EPS-I, or the lack of it, has no major role. If EPS-II is adhesive, the decreased association with EPS-I mutants with A549 cells is unexpected. Perhaps A549 cells bind EPS-I, but it is clear that EPS-I is not required for colonization of mice. We have recently shown that M. pulmonis forms biofilm structures in vivo on mouse trachea (Simmons and Dybvig, 2009). Thus, the efficient colonization of SCID mice with the EPS-I mutant CTG1291 may be due at least in part to the ability of EPS-I mutants to form biofilms.

Experimental procedures

Screening for EPS mutants

A total of about 320 genes have been knocked out in a transposon library of M. pulmonis strain CT as described previously (French et al., 2008). For each of these dispensable genes, at least one mutant was screened for a deficiency in EPS. Mycoplasmas were cultured in mycoplasma broth (MB) and assayed for colonies on mycoplasma agar (MA) (Dybvig et al., 2000; Simmons and Dybvig, 2003). Mycoplasmas (5 ml of cultures) were grown to stationary phase, harvested by centrifugation, washed twice in phosphate-buffered saline (PBS), and suspended in 40 μl of 1× running buffer (25 mM Tris, 192 mM glycine, 3.5 mM sodium dodecyl sulphate, pH 8.3) and 10 μl of loading buffer (40% glycerol, 1% bromophenol blue, 10% beta-mercaptoethanol, 49% 1× running buffer). Samples were sonicated at full power at 90% duty cycle on a Branson Sonifier 450 for 60 s and then heated at 100°C to denature. Samples were equilibrated to 130 μg of protein in 50 μl and loaded on an 18% Tris-HCl Ready Gel (Bio-Rad). Gels were electrophoresed at 100 V until completion (1.5 h) and stained with PAS (Zacharius et al., 1969). Gels were photographed, destained overnight in a 7% acetic acid, further destained for 2 h in a solution of 10% acetic acid and 40% methanol, and restained with Coomassie blue (Zacharius et al., 1969).

Genetic complementation

The operon containing MYPU_7410 and 7420 including its promoter was amplified by PCR with the forward primer GGAAGCGGCCGCAAAACATGACATGCCAC and reverse primer GTCTCCGTGGACTAACAGGATCAGTAGC. Incorporated into the forward and reverse primers was a NotI and BtgI restriction site (underlined) respectively. The high-fidelity DNA polymerase iPoof (Bio-Rad) was used for PCR amplification to minimize mutation. The PCR product was cloned into the NotI–BtgI site of transposon Tn4001TF in plasmid pTF85 (Luo et al., 2008). The resulting plasmid was named pTF74. For complementation experiments, a chloramphenicol resistance marker was incorporated in the plasmid. First, the tetM tetracycline resistance determinant gene was removed by digestion with BamHI. The vsa-cat gene was then excised from plasmid pIVC (Dybvig et al., 2000) by BamHI digestion and cloned into BamHI site of pTF74. The resulting plasmid, pTF74c, was transformed into EPS mutants of M. pulmonis, selecting for resistance to chloramphenicol, by using standard polyethylene glycol-mediated transformation methods (Dybvig et al., 1995; 2000). Strain CTG1701-C was a transformant of CTG1701 chosen for further study.

Fluorescence microscopy

Samples (10 μl) of mycoplasma cultures were diluted with 50 μl of PBS, placed on glass slides and allowed to dehydrate. The cells were fixed for 15 min at room temperature in 10% buffered neutral formalin (4.0% formaldehyde, 0.4% sodium dihydrogen orthophosphate, 0.65% disodium hydrogen orthophosphate, pH 7.0). The slides were washed three times in PBS and stained with Hoechst 33342 (Molecular Probes) to detect DNA and simultaneously incubated with FITC-conjugated GS-I or GS-II lectins (EY Laboratories) at a concentration of 20 and 50 μg ml−1, respectively, in 0.01 M phosphate-0.15 M NaCl containing 0.5 mM CaCl2, pH 7.4. After 45 min incubation at room temperature, the slides were washed three times in PBS with a final wash in deionized water. The slides were mounted with ProLong Gold antifade reagent (Molecular Probes) and observed at a magnification of 1600× with a Leica HC fluorescence microscope fitted with the Chroma 86012v2 filter set. Hoechst fluorescence was observed with 350 nm excitation and 475 nm emission filters. The FITC-labelled lectins were observed with 495 nm excitation and 535 nm emission filters. Images were analysed using the MetaMorph Imaging System (version 7.0; Molecular Devices Corporation).

Lectin affinity chromatography

Bacteria were grown to stationary phase, centrifuged and washed twice in PBS. The protein concentration was determined for each sample and the volume was adjusted to contain 300 μg of total protein. The samples underwent Bligh and Dyer lipid extraction as described (Bligh and Dyer, 1959). The aqueous fraction was collected and cooled to 4°C. The precipitate was then suspended in 3 ml of 5 mM CaCl2, 10 mM MgCl2, 25 mM Tris-HCl (pH 8.0) digestion buffer. To this was added 25 μl of a 1 mg ml−1 deoxyribonuclease I solution (Sigma) and 25 μl of a 1 mg ml−1 ribonuclease A solution (Sigma). After overnight incubation at 25°C with continuous rotation, the sample was heat denatured for 10 min at 100°C and allowed to cool to room temperature. Proteinase K (Sigma) was added (25 μl of a 1 mg ml−1 solution), and the sample was incubated overnight at 25°C with constant rotation. The sample was heat denatured again at 100°C for 30 min to deactivate the proteinase K. After cooling the pH was adjusted to 10 using ammonium chloride. This sample was passed over a GS-I lectin affinity column (EY laboratories) using their established protocol. The column was eluted with melibiose. The sample was collected and dialysed against deionized H2O using 2 kDa dialysis cassettes (Pierce) with 15 changes to assure removal of the elutant melibiose.

Monosaccharide analysis

The monosaccharide composition of the M. pulmonis cells was determined by gas chromatographic analysis of the trimethylsilyl derivatives of the sugar methyl glycosides. Samples were dried, suspended in 400 μl of 1.45 N methanolic HCl and heated at 80°C for 12 h. The methanolic HCl was removed by vacuum centrifugation, and the sample was suspended in 200 μl of methanol, followed by the addition of 25 μl of acetic anhydride and 25 μl of pyridine. This mixture was allowed to react for 30 min at room temperature and then evaporated. The samples were trimethylsilylated using 50 μl of Tri-Sil (Pierce), and the vials were sealed under argon. The samples were analysed using a Varian GC-MS in the electron ionization mode. The monosaccharide composition was determined by comparison with known standards from Sigma-Aldrich. For comparison of the amount of EPS produced by wild-type and mutant strains, the samples were quantified before monosaccharide analysis by using the BCATM Protein Assay Kit (Pierce) and equilibrated to 300 μg of total protein.

Electron microscopy

Mycoplasmas were grown in 5 ml of MB with all subsequent manipulations performed at room temperature. The culture was centrifuged at 8000 g for 5 min. The pellet was gently suspended in 500 μl of PBS. The suspension was centrifuged at 10 000 g for 5 min and the supernatant discarded. The pellet was suspended in 40 μl of PBS containing 1 mM ruthenium red and 3.75% sucrose and incubated for 25 min. Four hundred microlitres of PBS containing 3.75% sucrose was added, and the suspension centrifuged at 10 000 g for 2 min. The resulting pellet was suspended in 20 μl of PBS, and 7 μl of the suspension was adsorbed to a 400-mesh formvar-coated copper electron microscopy grid (Electron Microscopy Sciences) for 1 min. The grids were dried and observed at 75 kV with a Hitachi H7000 transmission electron microscope.

Western blot

Mycoplasma proteins were separated by sodium dodecyl sulphate-polyacrylamide gel electrophoresis, transferred to nitrocellulose, and reacted with the monoclonal antibody 7.1-2 which recognizes the constant region conserved in all Vsa proteins as described previously (Denison et al., 2005). The Vsa protein bands were visualized by development of the blot with 5-bromo-4-chloro-3-indolyl phosphate (Sigma) as the substrate.

Informatics

Annotations of the M. pulmonis genome sequence (Accession No. AL445566) are available at http://genolist.pasteur.fr/MypuList/. The MYPU_7410 and 7420 gene products were analysed to identify conserved domains and compare amino acid sequence similarity by protein blast at http://blast.ncbi.nlm.nih.gov/Blast.cgi.

Biofilm formation by M. pulmonis strains

Biofilms were grown on glass coverslips in MB for 2 days as previously described (Simmons and Dybvig, 2007; Simmons et al., 2007). The biofilms were fixed in 10% neutral buffered formalin for 15 min, washed in PBS, and incubated with an aqueous solution of Hoechst 33342 at 10 μg ml−1. Following washing in water, the coverslips were mounted to glass slides using Prolong Gold mounting medium (Invitrogen). Stacks of images were acquired with a Leica HC fluorescence microscope at 320× magnification using a Hoechst/DAPI filter set. Three-dimensional reconstructions were made with the ImageJ software application (version 1.38×, National Institutes of Health).

Animal infection

Female (3- to 5-week-old), NOD.CB17-Prkdcscid mice were acquired from Jackson Laboratories (Bar Harbor, Maine). In two independent experiments, 4- to 6-week-old mice, three mice per strain of mycoplasma per experiment, were infected intranasally with 1 × 108 cfu of M. pulmonis strains CTG1291 and CTG38 in a 30 μl total volume. The animals were sacrificed on day 7. The lung and trachea from each animal were collected together and placed in 1 ml of MB in a glass vial. The tissues were homogenized, sonicated gently to disrupt cell aggregates, and assayed for cfu. Sonication conditions were 30 s at 90% output in a cup-horn sonifier (Branson Ultrasonics model 450). Statistical significance was determined by Student's t-test with a P-value less than 0.01.

Association of M. pulmonis with A549 cells

Monolayers of human lung carcinoma cells A549 (ATCC CCL-185) were grown to confluence in DMEM (Mediatech) supplemented with 10% fetal bovine serum in 12-well tissue culture plates (1 × 106 cells per well). An average of the number of A549 cells per well was determined, and the cells were washed three times with fresh medium. M. pulmonis was added (1:1 ratio of mycoplasma cfu to A549 cells). The cells were incubated for 2.5 h at 37°C in 5% CO2 and a sample of broth removed from each well for cfu determination. The wells were washed three times with PBS and adherent cells were released by addition of trypsin-EDTA (0.25% trypsin, 0.02% EDTA) (Sigma) and incubation at 37°C for 10 min. The mycoplasma cfu associated with the A549 cells was then determined. Treatment of mycoplasmas with trypsin-EDTA alone in the absence of A549 cells had no affect on cfu. Statistical significance was determined by anova with Tukey's post hoc comparison with a P-value less than 0.01.

Acknowledgements

This work was supported by NIH Grants AI63909 and AI64848 and by training Grant T32 AI 07041. We thank S. Dong and R.T. Cartee for their assistance and D.G. Pritchard for sage advice.

Ancillary

Advertisement