Abstract
Although mycoplasmas possess a very limited genome, little is known about their virulence mechanisms and methods of persistence in the host. Examination of a wide range of mycoplasma species found considerable variation in their ability to form a biofilm. Mycoplasma putrefaciens, M. cottewii, M. yeatsii, M. agalactiae and M. bovis produced prolific biofilms. Conversely, the highly pathogenic mycoplasma and causative agent of contagious bovine pleuropneumonia, Mycoplasma mycoides subsp. mycoides SC, was unable to produce a biofilm. Biofilms were found to be considerably more resistant to stress, including heat and desiccation, than planktonic cells. A link between the biofilm phenotype and genotype as determined by molecular typing was found for M. bovis. Analysis of biofilms using fluorescent staining combined with confocal microscopy demonstrated that mycoplasma biofilms formed a highly differentiated structure with stacks and channels. Biofilm formation may indicate that mycoplasmas are capable of surviving in the environment.
INTRODUCTION
Mycoplasmas are characterized by their small genome size and are thought to have undergone reductive evolution, losing many genes possessed by more complex bacteria in the process. They lack many genes, including those for cell wall synthesis and for the production of all 20 amino acids, as well as genes encoding enzymes of the citric acid cycle and the majority of all other biosynthetic genes (reviewed by Razin et al., 1998). Presumably they can survive with a reduced genome as they have evolved in such a way as to acquire these products from their host in vivo and as such they are considered the model organisms for the study of essential functions in living cells. Despite their small genome size, mycoplasmas cause a wide range of disease in both humans and animals. They are associated with pneumonia, arthritis and reproductive disorders, and often cause disease of a chronic and persistent nature. However, very little is known about their pathogenicity factors or mechanisms of persistence in the host. There are only a few characterized virulence factors of mycoplasmas; these include, in certain mycoplasma species, the production of hydrogen peroxide (Brenann & Feinstein, 1969; Miles et al., 1991), the carbohydrate capsule (Tajima et al., 1982; Almeida & Rosenbusch, 1991), the ability to scavenge arginine from host cells (Sasaki et al., 1984) and T-cell mitogens (Tu et al., 2005). It is difficult to explain how mycoplasmas manage to cause such severe and chronic infection given their paucity of virulence factors and lack of cell wall.
Bacteria are thought to persist in the host by the formation of an adherent biofilm. Biofilms can be defined as sessile bacteria attached to a substratum, or each other, often surrounded by an extracellular polysaccharide matrix. Biofilms often exhibit different phenotypes and physiology compared with planktonic cells, particularly with regard to gene transcription and growth rate (Donlan & Costerton, 2002). Perhaps the most important and widely studied property of biofilms is their vastly increased recalcitrance to host defences and resistance to stress, including that of antimicrobials: compared with planktonic cells, biofilms are commonly 10–1000 times more resistant (Mah & O'Toole, 2001; Costerton et al., 1999). Unattached bacteria can be cleared by antibodies and phagocytes and are normally susceptible to antibiotics. However, adherent biofilm cells are resistant to antibiotics, antibodies and phagocytes. In addition, biofilms can cause host damage as phagocytes are attracted but phagocytosis is frustrated, and phagocytic enzymes are released which damage surrounding tissue and exacerbate infection. As well as enabling chronic infection of hosts, biofilms may cause bouts of acute infection when planktonic cells are periodically released from the biofilm. Despite their discovery in many other bacterial species studied to date, biofilms have not been studied in mycoplasmas. We will investigate whether the phenomenon of biofilm formation is ubiquitous among bacterial genera and occurs even in organisms with a very limited genome. We propose that biofilm formation facilitates the survival of mycoplasma species in the environment and speculate whether it contributes to the persistence of mycoplasmas in the host.
METHODS
Strains and growth conditions.
The bacterial strains used in this study are listed in Table 1⇓. In addition, 56 Mycoplasma bovis strains as described in our previous study (McAuliffe et al., 2004) were used. All strains were stored at −70 °C. Planktonic cells were grown in Eaton's broth at 37 °C with 5 % CO2 without aeration, as previously described (Nicholas & Baker, 1998). Biofilms were grown either on 11 mm2 glass coverslips (Sigma) or in multiwell plates (Corning). Sterile coverslips were positioned vertically in a 50 ml conical tube (Corning) with 7·5 ml pre-warmed Eaton's medium. Coverslips were inoculated with a 1 : 100 dilution of a 20 h planktonic culture and left at 37 °C with 5 % CO2 without aeration for 48 h. Biofilms were grown in multiwell plates by placing 180 μl medium in each well and inoculating with 20 μl of 20 h planktonic culture. Biofilms were left to grow for 48 h at 37 °C with 5 % CO2 without aeration.
Mycoplasma strains used and their origins
Semi-quantitative analysis of biofilm growth by crystal violet staining.
Crystal violet staining was performed as described previously with minor modifications (O'Toole & Kolter, 1998). Biofilms grown on glass coverslips and in microtitre plates were rinsed briefly in PBS to remove non-adherent cells and stained with 0·5 % crystal violet solution for 30 min. Biofilms were then washed profusely (five times) in distilled water before being left to dry at room temperature for at least 30 min. Coverslips were broken into smaller pieces using sterile forceps and 1 ml 100 % ethanol was added to solubilize the crystal violet. Crystal violet in stained biofilms in microtitre plates was solubilized by the addition of 200 μl 100 % ethanol. Biofilm production was quantified by measuring the absorbance (560 nm) of 100 μl of the solubilized crystal violet in a microtitre plate. At least eight wells were analysed for each biofilm strain and each experiment was replicated in duplicate.
Assay of sensitivity of mycoplasma to stress of heat and drying.
M. bovis biofilms were grown on glass coverslips for 48 h and subjected to stress of heat or drying. For the heat assay, biofilm and planktonic bacteria were exposed to a temperature of 50 °C for 40 min. At 10 min intervals viable counts were made to enumerate the number of surviving bacteria, and triplicate samples were taken of both biofilm and planktonic cells. Coverslips were rinsed in PBS to remove non-adherent cells and were then placed into 2 ml Eaton's broth, broken into smaller pieces using sterile forceps, and vortexed for 1 min to remove biofilm cells from the coverslip. Viable bacteria were enumerated according to the methods of Miles & Misra (1938). Serial dilutions (20 μl) were performed in microtitre plates using 180 μl pre-warmed Eaton's medium as diluent. Triplicate aliquots of diluted cells (10 μl) were dropped on to overdried Eaton's agar plates (without phenol red) from a height of approximately 1 cm. Plates were incubated at 37 °C for 72 h and the resulting colonies counted.
For the drying assay, planktonic cells aliquoted on to 2 cm2 pieces of sterile filter paper and biofilms on glass coverslips were placed in sterile Petri dishes and exposed to desiccation at room temperature for 31 h. Prior to the beginning of the assay, a piece of filter paper was placed briefly on top of the biofilms on coverslips to remove excess moisture. At intervals of 0, 3, 6, 17, 20, 24 and 31 h, triplicate biofilm and planktonic samples were taken and placed in 2 ml Eaton's medium, vortexed for 1 min and serial dilutions performed to enable the enumeration of surviving mycoplasmas as described above.
One-way analysis of variance was used to demonstrate significant differences between the performance of biofilm and planktonic cells in each assay.
Determination of effect of biofilm formation on the susceptibility to antibiotics for M. bovis and M. putrefaciens.
M. bovis and M. putrefaciens biofilms and planktonic cells were subjected to various concentrations of the antibiotics danofloxacin, enrofloxacin and oxytetracycline. MICs were determined for planktonic cultures by diluting stationary-phase cultures to a standard OD600 of 0·005 in 10 ml Eaton's medium with appropriate concentrations of antibiotics. Dilutions were made of the antibiotics to give a range of concentrations between 0·001 and 50 μg ml−1. Planktonic cells were incubated with the antibiotic for 48 h and the tubes were examined for growth and colour change of the phenol red indicator. The MIC was determined as the lowest concentration of antibiotic that inhibited growth. Biofilms were grown on glass coverslips in 7·5 ml Eaton's medium for 24 h and then transferred to a fresh tube containing 10 ml Eaton's medium with the appropriate concentration of antibiotics. Biofilms were left in contact with antibiotic for 48 h and tubes were monitored for growth and colour change.
Confocal scanning laser microscopy.
To examine the architecture of biofilm formation, M. bovis and M. putrefaciens were analysed by confocal microscopy. Biofilms were grown on glass coverslips and stained using the BacLight bacterial viability assay kit (Molecular Probes) according to the manufacturer's instructions. Essentially, live cells stained green with a SYT09 stain whereas dead cells with damaged membranes stained red with propidium iodide. Coverslips were mounted on to glass slides and examined using Perkin Elmer UltraView ERS using an excitation wavelength of 488 nm for SYTO9 and an emission filter of 500–550 nm; for propidium iodide an excitation of 568 nm and an emission filter of 580–650 nm were used. The presence of extracellular polysaccharide was also examined using Calcofluor White (Sigma) staining, with an excitation of 405 nm and an emission filter of 580–650 nm.
RESULTS
Biofilm formation by mycoplasma species
Eleven mycoplasma species were assessed for their ability to form a biofilm. At least two strains were examined for each species to ensure that there was no intraspecific variability in the ability to form a biofilm. Crystal violet staining of biofilms on coverslips revealed a striking variability in the ability of mycoplasma species to form a biofilm (Table 2⇓). Some species, notably M. putrefaciens, M. cottewii, M. yeatsii, M. bovis and M. agalactiae, all formed prolific biofilms. A clearly defined line of biofilm growth approximately 1–2 mm wide was seen on the coverslip at the air/liquid interface, and below this line the coverslip was heavily stained, indicating biofilm growth. M. capricolum subsp. capricolum, Mycoplasma sp. bovine group 7 and M. ovipneumoniae all showed very sparse biofilm formation with little staining evident. Interestingly, one species, M. mycoides subsp. mycoides SC, the causative agent of contagious bovine pleuropneumonia (CBPP), exhibited no biofilm formation at all. To confirm these results, an additional 24 strains of M. mycoides subsp. mycoides SC were analysed for their ability to form a biofilm and none was able to do so. Other mycoplasma species such as M. mycoides subsp. capri and M. mycoides subsp. mycoides LC showed an intermediate level of biofilm formation and the lower half of the coverslip below the air/liquid interface showed staining. Microscopic analysis of crystal-violet-stained M. bovis biofilms on coverslips indicated that individual cells were aggregating together to form microcolonies (Fig. 1⇓). Biofilm growth by M. bovis was followed over a time-course of 48 h. By 20 h post-inoculation, cells had begun to adhere to the coverslips and a thin covering of individual cells could clearly be seen (Fig. 1a⇓), by 36 h the cells had began to form small groups (Fig. 1b⇓), and by 48 h larger groups of microcolonies were evident (Fig. 1c⇓). Scanning electron microscopic (SEM) analysis of coverslips also indicated that whereas M. bovis formed a prolific biofilm, M. mycoides subsp. mycoides SC did not adhere very well. Coverslips were densely populated by M. bovis (Fig. 2a⇓); conversely, only a few fields of view showed the presence of any M. mycoides subsp. mycoides SC at all (Fig. 2b⇓).
Microscopic analysis of a biofilm formation on glass coverslips by M. bovis at (a) 20 h, (b) 36 h and (c) 48 h post-inoculation using crystal violet staining. Bar, 100 μm.
SEM analysis of biofilm formation by M. bovis (a) and M. mycoides subsp. mycoides SC (b) cultures grown on glass coverslips.
Ability of mycoplasma species to form a biofilm
Mycoplasmas were also analysed quantitatively for biofilm formation and the stationary-phase cell counts compared for both biofilm and planktonic cells. Cell counts for biofilms showed a good correlation with results obtained using crystal violet staining (Table 2⇑). M. putrefaciens, M. cottewii, M. yeatsii, M. bovis and M. agalactiae demonstrated the most prolific biofilm formation, with cell counts in the range of 106–107 cells per coverslip, M. mycoides subsp. mycoides SC demonstrated minimal adherence (101 cells per coverslip) and the other species tested showed intermediate adherence (103–105 cells per coverslip).
M. mycoides subsp. mycoides SC strains were also tested on a variety of different surfaces to confirm that they were unable to adhere. Tissue-culture-treated and untreated polystyrene multiwell plates (Corning), Nunclon tissue-culture-treated (Sigma) and polypropylene plates (Corning) were all tested. Tissue-culture-treated flasks, Nunc tissue-culture-treated 25 cm2 and Corning 25 cm2 cell culture flasks were also tested for SC adherence. However, M. mycoides subsp. mycoides SC was unable to adhere to any surface tested (results not shown).
Intraspecific variation in biofilm formation by M. bovis and correlation with molecular typing
M. bovis strains were found to vary widely in their ability to form a biofilm as determined by semi-quantitative crystal violet staining of biofilms grown on PVC microtitre plates (Fig. 3⇓). Strains showing poor adhesion and biofilm formation gave A560 values as low as 0·08 whereas strains that formed prolific biofilms gave values up to 0·2. Interestingly, the type strain PG45 showed a low level of biofilm formation. There was some correlation between the ability to form a biofilm and the groups generated using molecular typing techniques (McAuliffe et al., 2004), as shown in Table 3⇓. Isolates that gave higher A560 values (>0·15) and formed more prolific biofilms, such as 158B01, 76B98, 9B99, 164B02, 226B02, 152B01, 148B02 and 232B02, all fell into subgroup B as determined by amplified fragment length polymorphism (AFLP) and random amplified polymorphic DNA (RAPD). Similarly, isolates that were unable to form an abundant biofilm (with A560 values <0·100), such as 147B02, 162B02, 68B98, PG45, 217B02, 86B96, 125B01, 54B98 and 2B01, fell into typing group A using AFLP and RAPD.
Semi-quantitative analysis of M. bovis biofilm formation on microtitre plates using crystal violet staining. Isolates of M. bovis are derived from those described by McAuliffe et al. (2004).
Influence of molecular typing profile and Vsp type on adherence of M. bovis
When biofilm ability was compared with the Vsp profiles of the isolates (as determined in McAuliffe et al., 2004), it was found that there was a strong correlation between the expression of certain Vsps and biofilm formation (Table 3⇑). The majority of isolates (70 %) that formed a poor biofilm (with A560 values less than 0·100) showed expression of VspF by Western blotting. Conversely, expression of VspF was not seen in any isolates that formed prolific biofilms (with A560 values in excess of 0·15) and expression of Vsps B and O was much more common, with VspB expressed in 86 % of isolates and VspO expressed in 63 % of isolates.
Contribution of biofilm formation to resistance of M. bovis to stress of heat and drying
Biofilm-grown M. bovis were more resistant to heat at 50 °C than planktonic M. bovis (Fig. 4a⇓). For both strains analysed there was at least one log cycle more biofilm cells alive after 40 min compared with planktonic-grown cells. For strain 4, the 0·3 % biofilm cells alive was significantly greater (P<0·0001) than the 0·008 % planktonic cells alive after 40 min exposure to heat at 50 °C. Similarly, for strain 43, the 0·09 % biofilm cells alive was significantly greater (P=0·022) than the 0·002 % planktonic cells alive after 40 min at 50 °C.
Effect of heat at 50 °C (a), or desiccation (b), on survival of biofilm or planktonic cells of M. bovis. Biofilm grown strain 43 (⧫), biofilm-grown strain 4 (□), planktonic strain 43 (▴), planktonic strain 4 (×). Bars indicate standard error (n=3).
Biofilm M. bovis were also more resistant to desiccation than planktonic-grown cells (Fig. 4b⇑). There were no detectable surviving planktonic cells after 24 h drying whereas significantly greater proportions of strain 43 (0·0013 %, P=0·009) and strain 4 (0·01 %, P=0·026) biofilm cells were still viable after 32 h.
Effect of biofilm formation on the susceptibility of M. bovis and M. putrefaciens to antibiotics
Biofilm formation did not affect the MIC value of any of the antibiotics tested for either M. bovis or M. putrefaciens. Both planktonic and biofilm M. bovis strains 4 and 43 had an MIC of 0·75 μg ml−1 for danofloxacin, 25 μg ml−1 for oxytetracycline and 0·4 μg ml−1 for enrofloxacin. When M. putrefaciens was tested, it was found to have an MIC of 0·75 μg ml−1 for danofloxacin, 0·8 μg ml−1 for oxytetracycline and 0·4 μg ml−1 for enrofloxacin. Although the planktonic and biofilm MICs did not differ significantly, biofilm formation did have some phenotypic effects on M. bovis. It was found that exposure to oxytetracycline inhibited polysaccharide film formation in M. bovis at much lower concentrations of antibiotic for planktonic compared with biofilm cultures. Polysaccharide film production was prevented by 0·75 μg ml−1 in planktonic cultures but by 7·5 μg ml−1 in biofilm cultures.
Confocal microscopy of biofilms
Confocal imaging of 24, 48 and 72 h M. bovis and M. putrefaciens biofilms was undertaken. When biofilms were stained using a viability stain (BacLight, Molecular Probes) the majority of cells stained green at 24 and 48 h, indicating that cells were live or had intact membranes. By 72 h the number of dead or membrane-compromised cells had increased to approximately 70 % of total cells, and live cells were seen predominantly in the centre of the biofilm (Fig. 5⇓). Biofilms were seen to initiate from a layer of single, adherent cells at 24 h; by 48 h the biofilms had developed a more heterogeneous structure with stacks of cells approximately 20 μm high (Fig. 6⇓). Channels were also seen between stacks. It was found that the SYTO9 stain stained both live and dead cells but as propidium iodide only stained dead cells these cells were easily differentiated by superimposing the images produced using the two wavelengths of excitation. Calcofluor white staining demonstrated a profuse layer of polysaccharide covering the biofilm (Fig. 7⇓).
Confocal image of M. putrefaciens biofilm with BacLight staining, showing persistence of live cells within microcolonies. Bar, 20 μm.
Three-dimensional confocal image of M. putrefaciens biofilm stained using BacLight (Molecular Probes). After 24 h a layer of cells has formed on the coverslip (a); after 48 h stacks of cells can be seen within the biofilm (b). Bar, 20 μm.
Three-dimensional confocal image of calcofluor white-stained M. putrefaciens biofilms. A thick layer of polysaccharide (stained blue) can be seen covering stacks of cells within the biofilm Bar, 20 μm.
DISCUSSION
The mycoplasma species examined in this study demonstrated a remarkable diversity in their ability to form a biofilm. Species such as M. putrefaciens and M. bovis formed a prolific biofilm that resembled that of other, more complex bacterial species. Biofilms were seen to form a highly differentiated structure with channels and stacks like those described in Pseudomonas aeruginosa, Staphylococcus epidermidis and others (Dunne, 2002). Although many species were capable of forming a prolific biofilm, there was one notable exception, M. mycoides subsp. mycoides SC. Paradoxically, M. mycoides subsp. mycoides SC causes the most severe disease of all the mycoplasma species, with high mortality and morbidity, but it was unable to adhere to any surface tested in this study or to form a biofilm. Interestingly, the closely related small ruminant species M. mycoides subsp. mycoides LC did form a biofilm. Previous studies have also indicated that these two pathogens exhibit differing adherence abilities and reported that whereas LC adhered to erythrocytes and caprine endothelial cells, SC did not (Valdivieso-Garcia et al., 1989). Intriguing studies by Minga (1981) described two colony types for SC. A smooth colony type was found that was thought to be more virulent, causing more widespread disease, with arthritis and weight loss in addition to respiratory symptoms. However, rough colony types caused more localized disease symptoms with only respiratory involvement. Rough colony types were seen to form ‘filaments and comets' in liquid media whereas smooth colony types gave a uniform turbidity. We speculate that the rough colony type may have been capable of biofilm formation and that the filaments and comets described by Minga were the first account of biofilm formation by a mycoplasma species. None of the 24 M. mycoides subsp. mycoides SC strains we tested showed rough colony morphology. It seems plausible that the inability of SC to adhere to the substrata tested in this study may be a result of the polysaccharide capsule produced by SC. The capsule of M. mycoides subsp. mycoides SC has been used as a diagnostic tool in a latex agglutination test (LAT) (March et al., 2003). In the LAT, latex microspheres are coated with M. mycoides subsp. mycoides SC polyclonal immunoglobulin G antiserum and used to detect M. mycoides subsp. mycoides SC antigen in the serum of cattle infected with CBPP and in growth medium containing the organism (March et al., 2003). It was found that there was cross-reactivity between M. mycoides subsp. mycoides SC, M. mycoides subsp. mycoides LC and M. mycoides subsp. mycoides capri but not between other members of the mycoides cluster (March et al., 2003). It is feasible that this is because these three mycoplasma species produce the galactan-based capsule and that other mycoides clusters produce a different capsule. We speculate that the galactan-based capsule of M. mycoides subsp. mycoides SC may inhibit adherence in vitro. Similarly, M. ovipneumoniae, which is also thought to produce a galactan-based capsule similar to SC (Niang et al., 1998), was also unable to form a biofilm in this study. However, it has previously been suggested that the capsule may actually facilitate the adhesion of M. ovipneumoniae to the epithelium (Niang et al., 1998). The idea that the polysaccharide capsule may actually inhibit adhesion has been proposed previously. In Neisseria meningitidis, an inverse correlation between capsule and biofilm formation has been described (Yi et al., 2004). Similarly, in Vibrio vulnificus, capsule production has also been shown to prevent biofilm formation (Joseph & Wright, 2004). However, as many mycoplasma species such as M. agalactiae and M. bovis produce polysaccharide films, it will be essential to also examine the role of these macromolecules in biofilm formation.
This study is one of the first to address some fundamental questions regarding the survival of mycoplasma species when exposed to environmental stress. Previously it has been assumed that, given their paucity of resistance mechanisms and lack of cell wall, mycoplasmas would survive for very little time outside of laboratory culture or the animal host. Surprisingly, all the M. bovis isolates tested were much more resistant to heat than would be expected for an organism that lacks many of the heat-shock proteins commonly seen in other bacterial species. Mycoplasma species sequenced to date vary in the heat-shock-associated genes they possess. Most mycoplasma species do not possess GroE/L but M. penetrans, M. pneumoniae and M. genitalium do. The ClpB chaperonin is present in most mycoplasma species, with the exception of M. gallisepticum and M. pulmonis, but the ClpP/X proteases are not found in any sequenced mycoplasma species. Lon protease is also found in most mycoplasma species, as are the DnaK and DnaJ chaperonins.
Similarly, M. bovis was also surprisingly resistant to desiccation, with biofilm cells surviving over 30 h in the environment. As with heat shock, biofilm cells were significantly more resistant to drying than planktonic cells, providing further support for the notion that biofilm growth may help mycoplasmas to persist in the environment. Previously it has been shown in other bacterial species that biofilm cells are typically 10–1000 times more resistant to stress than their planktonic counterparts (Mah & O'Toole, 2001). The resistance of M. bovis to drying may have important implications for outbreak-handling and disease control, i.e. premises must be thoroughly disinfected following outbreaks and before restocking with cattle as it is plausible that M. bovis may persist in the environment. There is further evidence of the contribution of biofilm growth to the persistence of some mycoplasma species. When late-stationary-phase cultures that were starved of nutrients were observed using confocal microscopy and Live/Dead staining, it was found that live cells persisted in the centre of biofilm microcolonies even after all other cells around the edge of the biofilm had died. It seems likely that mycoplasmas in the centre of the biofilm are protected from environmental stress compared with free-living counterparts or those at the edge of the biofilm.
One of the most commonly studied properties of biofilms is their increased resistance to the effects of antibiotics and disinfectants. It is thought that the properties of biofilms that may cause recalcitrance to antibiotics include slow growth, phenotypic heterogeneity, the presence of persister cells, prevention of antibiotic penetration due to the exopolysaccharide matrix and inactivation of antibiotics within the biofilm matrix (Stewart, 2002). In this study, biofilm M. bovis was not found to be significantly more resistant to the actions of any of the antibiotics tested, although some antibiotics did cause phenotypic effects at concentrations below the MIC. The antibiotic oxytetracycline caused the inhibition of polysaccharide ‘film’ production in M. bovis planktonic cells at concentrations far below the MIC levels. However, film production was only inhibited by much higher concentrations of antibiotic in biofilm cells. Previous studies have also indicated that subinhibitory concentrations of antibiotics may have phenotypic effects, particularly on adhesion and polysaccharide production (Rachid et al., 2000; Fonseca et al., 2004).
This study has raised further questions regarding how mycoplasma species adhere to surfaces and initiate biofilm formation, as they lack all of the genes commonly associated with biofilm formation by other bacterial species. Mycoplasmas are thought to have evolved from the branch of low G+C Gram-positive species that includes the Bacillus and Lactobacillus species (Woese et al., 2000). They have undergone degenerative evolution, losing a large proportion of genomic DNA in the process (Woese et al., 1980). It is therefore not surprising that they lack biofilm-associated gene homologues, such as sigB, that are normally present in Gram-positive organisms, or the staphylococcal genes icaABC, bap or its homologues esp and bhp, arl or sarA.
We also examined the influence of variable surface proteins on biofilm formation by M. bovis. Membrane lipoproteins and proteins are recognized virulence factors in mycoplasmas as they have been shown to induce blastogenesis and secretion of proinflammatory cytokines by a mechanism distinct from that of lipopolysaccharides (Brenner et al., 1997; Herbelin et al., 1994). In addition, membrane lipoproteins have been linked to biofilm formation and adhesion in Escherichia coli (Otto & Silhavy, 2002). There was found to be some correlation between the Vsp expressed by the strains and the ability to form a biofilm for M. bovis. Strains that expressed VspF showed poor biofilm formation whilst those expressing VspO or VspB formed prolific biofilms. These results are contradictory to the findings obtained using a bovine bronchial epithelial cell line, where the use of monoclonal antibodies to VspF significantly reduced the adherence of M. bovis (Thomas et al., 2003). However, it is important to note that the monoclonal used in the study may show cross-reactivity against other surface lipoproteins and that it was not tested if the M. bovis strain used was expressing VspF (Thomas et al., 2003). Previous studies have also shown a link between the expression of specific lipoproteins and adhesion for M. bovis. Sachse et al. (1996, 2000) demonstrated that the Vsp lipoproteins were involved in a variety of attachment and binding processes.
It is important to remember that this study has examined biofilm formation under in vitro laboratory conditions. Further studies utilizing animal models will be crucial to determine if biofilms form in vivo and whether they contribute to mycoplasma persistence in the host.
It has previously been suggested that as the inclination for bacteria to become surface bound is so ubiquitous in diverse ecosystems it suggests a strong survival and/or selective advantage for surface dwellers over their free-ranging counterparts (Dunne, 2002). We propose that the biofilm mode of growth offers a selective advantage in mycoplasmas, as it contributes to the persistence of many mycoplasma species. As mycoplasmas possess only a small genome, it seems likely that biofilm growth is beneficial, if not essential, for many mycoplasmas; otherwise it is likely that the genes necessary for this mode of growth would have been lost during degenerative evolution.
Acknowledgments
We wish to thank DEFRA for their continuing support and Bill Cooley for SEM analysis.