Mycoplasma bovis mbfN Encodes a Novel LRR Lipoprotein That Undergoes Proteolytic Processing and Binds Host Extracellular Matrix Components

Experimental validation of the putative functions of genes in M. bovis will advance our understanding of the basic biology of this economically important pathogen and is crucial in developing prevention strategies. This study demonstrated the extracellular matrix binding ability of a novel immunogenic lipoprotein of M. bovis, and the role of this protein in adhesion by M. bovis suggests that it could play a role in virulence.

ABSTRACT

Mycoplasma bovis causes serious infections in ruminants, leading to huge economic losses. Lipoproteins are key components of the mycoplasma membrane and are believed to function in nutrient acquisition, adherence, enzymatic interactions with the host, and induction of the host’s immune response to infection. Many genes of M. bovis have not been assigned functions, in part because of their low sequence similarity with other bacteria, making it difficult to extrapolate gene functions. This study examined functions of a surface-localized leucine-rich repeat (LRR) lipoprotein encoded by mbfN of M. bovis PG45. Homologs of MbfN were detected as 48-kDa peptides by Western blotting in all the strains of M. bovis included in this study, with the predicted 70-kDa full-length polypeptide detected in some strains. Sequence analysis of the gene revealed the absence in some strains of a region encoding the carboxyl-terminal 147 amino acids found in strain PG45, which could account for the variation detected by immunoblotting. In silico analysis of MbfN suggested that it may have an adhesion-related function. In vitro binding assays confirmed MbfN to be a fibronectin and heparin-binding protein. Disruption of mbfN in M. bovis PG45 significantly reduced (P = 0.033) the adherence of M. bovis PG45 to MDBK cells in vitro, demonstrating the role of MbfN as an adhesin.

IMPORTANCE Experimental validation of the putative functions of genes in M. bovis will advance our understanding of the basic biology of this economically important pathogen and is crucial in developing prevention strategies. This study demonstrated the extracellular matrix binding ability of a novel immunogenic lipoprotein of M. bovis, and the role of this protein in adhesion by M. bovis suggests that it could play a role in virulence.

INTRODUCTION

Mycoplasma bovis is an economically important pathogen of cattle and other ruminant species, causing a range of diseases including mastitis, pneumonia, keratoconjunctivitis, and arthritis, leading to significant losses (1–3). M. bovis has a worldwide distribution, and infection is detected across the globe (4–6), with the latest incursion reported in New Zealand, which was previously free of M. bovis (https://www.biosecurity.govt.nz/protection-and-response/mycoplasma-bovis).

Mycoplasmas lack a cell wall; they are bounded by a single cytoplasmic membrane, and those elements that are associated with the mycoplasma cell membrane play a crucial role in colonization and survival within the host. Although mycoplasmas are believed to have evolved through genome reduction (7), they are also able to acquire foreign genetic material through horizontal gene transfer (HGT) (8). Gene products that display motifs associated with surface or membrane localization are believed to serve as the primary interface between mycoplasmas and their host (9, 10). Notwithstanding the increased availability of sequence data over the past decade, many M. bovis proteins do not have any functions assigned to them. The low sequence similarity between mycoplasmas and other bacterial species also limits the deduction of their functions by simple amino acid sequence comparison. In order to advance our understanding of the pathogenesis of M. bovis and, indeed, other mycoplasmas, there is a need to characterize the function of these proteins that appear to be specific to mycoplasmas (3). One of the critical gaps in understanding the pathogenesis of M. bovis has been lack of experimental validation of the functions of hypothetical M. bovis genes (11). It is suggested that some of these proteins may play multiple roles during infection because of the evolutionary forces that have driven the reduction in the mycoplasma genome (12).

Mycoplasmas cannot synthesize their own nucleic acid bases or amino acids, so host cell surfaces provide a rich nutrient environment that is conducive for the survival of these organisms (13). Membrane-associated proteins play significant roles in interacting with the host, and many of these proteins, especially the lipoproteins, are believed to be exposed to the extracellular milieu and to have functions akin to those of the periplasmic proteins in Gram-negative bacteria. These functions may include host cell adhesion and invasion and stimulation of the immune response, and these proteins may be targets of growth inhibitory antibody (14–20). The ability of mycoplasmas to adhere to the host’s epithelial surfaces and colonize affected tissue is considered a key virulence factor in the pathogenesis of mycoplasmosis (21). Extracellular matrix (ECM) components from the host are targeted by microorganisms for adhesion and colonization (22). Mycoplasma gene products have been shown to bind to ECM components such as collagen, laminin, fibronectin, plasminogen, and the glycosaminoglycan heparin, thereby mediating cytoadherence, assisting host cell invasion (3, 23, 24). Mycoplasmas possess a large number of virulence factors, but the repertoire of cell surface proteins of M. bovis that interact with host serum or ECM components remains to be determined. While some of the surface-localized proteins have primary adhesin functions, many have moonlighting functions involved in adhesive interactions with ECM components. It has been proposed that this additional adhesive function plays a significant role in host-microbe interactions (25).

Previous studies in our laboratory have focused on defining the biochemical roles of genes of unknown function in mycoplasmas, focusing on those predicted to encode lipoproteins (3, 26–28). Recently, we used in silico three-dimensional protein structural analysis, comparing the resultant structures with those found in the Protein Data Bank (PDB), as an initial step to predicting molecular functions, followed by in vitro biochemical assays of recombinant hypothetical proteins to assign functions (3, 29). In some instances, the assignation of molecular functions has relied on serendipitous discoveries (30). Given the roles played by mycoplasma surface-associated proteins in pathogenesis, elucidating the function of proteins encoded by mycoplasma genes, including validating predictions of multiple functions, is likely to contribute to our understanding of the organisms’ capacities to cause disease. This could also lead to discovery of novel antibacterial targets, identification of basic elements for effective vaccines, and development of novel and highly specific immunodiagnostics.

Ab initio protein structural predictions for MbfN using the Phyre² (31) and I-TASSER (32) online tools revealed moderate structural similarity with EUBVEN_01088 of Eubacterium ventriosum and BACOVA_01565 of Bacteroides ovatus, which are leucine-rich repeat (LRR) motif-containing proteins. Further homology searches with MbfN using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database identified, among other hits, the cytoadherence-associated protein Hlp3 of Mycoplasma gallisepticum and extracellular matrix-binding proteins in many bacterial species. With these in mind, we hypothesized that MbfN may possess some adherence-related function. Here, we show that MbfN is a novel immunogenic LRR lipoprotein of M. bovis PG45 that is recognized by the humoral immune system of the host. MbfN bound both fibronectin and heparin in vitro, a function attributable to adhesins. We also found that disruption of the gene encoding MbfN decreased the capacity of M. bovis PG45 to adhere to eukaryotic cells, suggesting that MbfN is a multifunctional lipoprotein that can be considered a potential virulence factor in M. bovis. These findings demonstrate a role for MbfN in host-pathogen interactions and highlight the likely contribution of this surface-localized lipoprotein in the pathogenesis of the economically important ruminant pathogen M. bovis.

RESULTS

Genomic features, bioinformatics analyses, and structural modeling.

The mbfN gene is located between nucleotides 654047 and 655885 of M. bovis PG45 (NCBI Protein ID no. ADR25360). The region upstream of mbfN includes a lipoprotein gene (MBOVPG45_0564), while the region downstream encodes a putative lipoprotein (MBOVPG45_0566) predicted to be a member of the S41B family of peptidases. ProtParam (ExPASy) predicted MbfN to have a molecular weight of 70 kDa and an isoelectric point of 8.98, while its closest relative, the product of MAGa2670 of Mycoplasma agalactiae, had a predicted molecular weight of 52.3 kDa and an isoelectric point of 8.96. MbfN contains 115 positively charged amino acids and 102 negatively charged amino acids, of which lysine and glutamic acid make up 17.8 and 10.8% of the total amino acids, respectively. The average hydrophobicity of MbfN was −0.722, with an instability index of 35.90, classifying it as stable according to ProtParam analysis. The protein was predicted to have multiple localization sites by PSORTb, with some parts being cytoplasmic (2.5/10.0) and the rest being membrane associated. A putative signal peptide cleavage site was identified between residues 20 and 21 using SignalP.

Several LRR motifs were identified in the MbfN protein sequence, including NGELIIGDDYKYILSDAFADNKK, VETITFNENLKSLEGFNNTK, VKELTLPKKLEKFAAFNSTK, ITKLLLPKTLKSFALSSTF, LEELEFEPDFKFDKVSPDSY, and YESIAINQQLLPSLNKIYVSDG. KEGG searches also revealed the presence of a G protein-coupled receptor kinase-interacting protein 1 (GIT1)-like domain between residues 226 and 284 of MbfN. GIT1 is a multidomain protein that plays an important role in cell adhesion, motility, cytoskeletal remodeling, and membrane trafficking in eukaryotes (33). The ab initio three-dimensional structural model of MbfN, as predicted by I-TASSER and Phyre², revealed structural similarity with EUBVEN_01088 of Eubacterium ventriosum and BACOVA_01565 of Bacteroides ovatus, which are LRR motif-containing proteins. The predicted structure of MbfN and the crystal structure of EUBVEN_01088 (PDB ID 4H09) had the characteristic α/β solenoid horseshoe fold seen in most proteins with LRR regions (see Fig. S1 in the supplemental material). Database searches for homologs of MbfN identified proteins that may be associated with cytoadherence and/or interactions with extracellular matrix components in many bacterial species. The putative LRR motif region of MbfN had moderate similarities with the BspA-like protein of Trichomonas vaginalis and the LRR cell surface protein of Bacteroides species (34).

Cloning and expression of recombinant MbfN.

In order to achieve full-length expression of recombinant MbfN in Escherichia coli, the TGA codon at amino acid position 174 in MbfN was mutagenized to TGG by overlap extension PCR (see Fig. S2A in the supplemental material). The predicted molecular mass of the recombinant MbfN was 96 kDa (including the 26-kDa glutathione S-transferase [GST] polypeptide), as confirmed by polyacrylamide gel electrophoresis, but the protein underwent cleavage during the purification process, resulting in products of 70, 65, and 60 kDa (Fig. S2B). Analysis of MbfN for intrinsically disordered protein (IDP) regions detected areas that were intrinsically disordered, as revealed by D²P² (http://d2p2.pro/search) (see Fig. S3 in the supplemental material), which may contribute to the cleavage observed.

MbfN is an immunogenic lipoprotein.

Lipoproteins are key factors in activation of the host’s humoral immune system. We therefore tested the ability of MbfN to react with antibody produced in calves following experimental infection with M. bovis. The immunogenicity of MbfN was confirmed by probing an immunoblot of recombinant MbfN with sera from the infected calves. The cleaved fragments of MbfN reacted with the sera from M. bovis-infected calves, but recombinant GST protein did not react (Fig. 1A). Western immunoblotting of 8 strains of M. bovis using monospecific anti-MbfN sera detected immunoreactive bands at 48 kDa in all strains and proteins of the predicted molecular mass of the full-length MbfN (70 kDa) in the PG45, 9760EU, and 193731 (appearing faintly on the blot) strains (Fig. 1B). A search for paralogous genes in M. bovis PG45 using the KEGG and NCBI databases did not identify any paralogs.

FIG 1.

Immunogenicity of recombinant MbfN. (A) Decreasing concentrations of recombinant MbfN protein were separated by SDS–12.5% PAGE and transferred onto a PVDF membrane. The membrane was incubated in a 1:100 dilution of pooled sera from calves experimentally infected with M. bovis and then probed with a 1:2,000 dilution of HRP-conjugated sheep anti-bovine antibody. Binding was detected using the Clarity Western ECL blotting substrate (Bio-Rad). M, PageRuler prestained protein ladder in kilodaltons (Thermo Scientific). (B) Immunogenicity of MbfN in different strains of M. bovis. Whole-cell proteins of PG45 and eight other strains of M. bovis were separated by SDS–12% PAGE, transferred onto a PVDF membrane, incubated in a 1:500 dilution of anti-MbfN sera, and then probed with a 1:2,000 dilution of HRP-conjugated rabbit anti-rat antibody.

MbfN is surface exposed and undergoes proteolytic processing.

Freshly cultured M. bovis PG45 cells were mixed with increasing concentrations of trypsin, from 0.5 to 32 μg/ml, and incubated for 20 min at 37°C prior to Western immunoblotting of the whole-cell proteins and probing the blot with anti-MbfN sera. Trypsin cleaved MbfN into smaller fragments, as can be seen in Fig. 2B, demonstrating that the protein was surface exposed. To further confirm that MbfN was an integral membrane protein, M. bovis whole-cell proteins were fractionated into aqueous and detergent phases using Triton X-114, and the fractions were subjected to SDS-PAGE and Western immunoblotting. The 48- and 70-kDa immunoreactive polypeptides were detected in the detergent phase and not in the aqueous phase (Fig. 2C). This concurred with the predicted localization of MbfN by PSORTb analysis as being mainly on the cell membrane.

FIG 2.

Cellular localization of MbfN. (A) Proteins of untreated M. bovis PG45 cells (NT) or cells treated with increasing concentrations of trypsin (0.5, 1, 4, 8, 16, or 32 μg/ml) were separated by SDS–4 to 20% gradient PAGE and stained with Coomassie brilliant blue. (B) A duplicate of the gel shown in panel A was subjected to Western immunoblotting and probed with anti-MbfN sera. Trypsin was able to digest MbfN, suggesting that it was surface exposed. (C) TX-114 partitioned M. bovis PG45 cellular proteins were subjected to Western immunoblotting and probed with anti-MbfN sera. The cellular localization of MbfN was demonstrated by its presence in the hydrophobic fraction. M, PageRuler prestained protein ladder in kilodaltons (Thermo Scientific).

MbfN interacts with ECM components.

On the basis of our in silico analyses of the protein coding regions of MbfN, we decided to test whether MbfN could interact with ECM components. Protein dot blot and microtiter plate assays showed that recombinant MbfN was able to bind fibronectin in a similar manner to the fibronectin-binding protein B of Staphylococcus aureus (FnBPB; used as a positive control), while the negative controls (GST and bovine serum albumin [BSA] proteins) did not bind fibronectin (P < 0.0001) (Fig. 3A and B). MilA was among other M. bovis proteins that were overexpressed in a similar manner to MbfN, and these proteins were later batch tested for fibronectin binding activity. The binding of MilA to fibronectin was therefore an accidental finding. As this was the first proteolytically processed M. bovis lipoprotein shown to interact with fibronectin, we named it MbfN (M. bovis fibronectin-binding lipoprotein). Fibronectin binding by MbfN using the microtiter plate method also showed that the fibronectin binding was dose dependent and saturable (Fig. 3C). To further confirm the specificity of the binding, a binding competition assay was performed, and the results showed that binding of recombinant MbfN to fibronectin could be diminished by the addition of monospecific anti-MbfN sera (Fig. 3D).

FIG 3.

Interactions of MbfN with extracellular matrix components. (A) Serial 2-fold dilutions of fibronectin were dot blotted onto a nitrocellulose membrane and then blocked in 5% BSA in PBS. The blot was incubated with 20 pmol MbfN and probed with goat anti-GST antibody and then HRP-conjugated rabbit anti-goat antibody. Binding was detected using the Clarity Western ECL blotting substrate (Bio-Rad). FnBPB from S. aureus and GST proteins were included as positive and negative controls, respectively. (B) Fibronectin (5 μg) was immobilized on each well of a microtiter plate and blocked with 5% BSA in PBS. Proteins were then added and incubated for 2 h at 37°C, followed by probing with goat anti-GST antibody and then HRP-conjugated rabbit anti-goat antibody. Absorbance readings were taken from triplicate reactions of at least two independent experiments. Significant difference: ****, P < 0.0001; **, P = 0.0023. (C) Recombinant MbfN was incubated with increasing amounts of fibronectin immobilized on microtiter plate wells. (D) A fixed amount of fibronectin was immobilized on microtiter plate wells and incubated with recombinant MbfN that had been incubated with increasing dilutions of the monospecific anti-MbfN sera. Plates were probed with goat anti-GST antibody and HRP-conjugated rabbit anti-goat antibody. ABTS substrate was then added for color development. Absorbance readings were taken from triplicate reactions of at least two independent experiments.

Bioinformatic analysis further revealed the presence of putative heparin binding consensus sequences in MbfN (NKKIKKV between residues 300 and 306, SKKIKL between residues 442 and 447, DKKKKI between residues 473 and 478, and TKKTKS between residues 492 and 497), and about 20% of the amino acid sequence was composed of basic residues. Recombinant MbfN was able to bind to heparin in a dot blot assay (Fig. 4), demonstrating that the heparin binding motifs detected in the sequence of MbfN were functional.

FIG 4.

Glycosaminoglycan binding assay. Using the Bio-Dot apparatus, Hybond-C Extra nitrocellulose membrane was blotted with 15, 7.5, and 3.75 pmol of each of recombinant MbfN, GST, MilA, and BSA and stained with Coomassie brilliant blue (A) or probed with biotinylated heparin (100 μg/ml) (B). Binding was detected with HRP-conjugated goat antibiotin antibody (Cell Signaling) and the Clarity Western ECL blotting substrate (Bio-Rad).

Disruption of the lipoprotein gene mbfN greatly reduces the ability of M. bovis PG45 to adhere to eukaryotic cells in vitro.

A transposon insertion mutant with a disruption of the mbfN gene, from a library of mutants of M. bovis PG45 that had been characterised by direct sequencing of each mutant, was used to examine the contribution of MbfN to cytoadhesion. Gene-specific PCR using the primers 565 ampR and IR inverse and Sanger sequencing confirmed insertion of the transposon into mbfN (Δ565) at a position 47.9% of the full coding sequence downstream of the start codon of mbfN (between nucleotides 880 and 881) (see Fig. S4A in the supplemental material). This transposon insertion was also demonstrated by Western immunoblotting by probing with anti-MbfN sera (Fig. S4B). Our initial results showed that MbfN was surface localized and able to interact with ECM components. In order to assess whether MbfN has a role in cytoadherence, we further examined the ability of MbfN to adhere to MDBK cells in vitro by comparing the adhesion between the Δ565 (ΔMbfN) mutant strain and the wild-type PG45 parental strain (Fig. 5A). Adhesion of the Δ565 mutant to MDBK cells was significantly reduced compared to that of the parental strain (P = 0.0329), indicating that mbfN had a role in adhesion of M. bovis to the MDBK cells. The specificity of the adhesion was further assessed by an indirect method using monospecific anti-MbfN sera. When M. bovis PG45 was incubated with anti-MbfN sera prior to inoculation onto MDBK cells, the adherence of M. bovis was also significantly lowered (Fig. 5B). Inhibition of adhesion by MbfN antisera was dose dependent, with incubation in dilutions of 1/10 resulting in an increased level of inhibition compared to 1/100 dilutions of the antisera (P = 0.0021). Dilutions of anti-MbfN sera at 1/100 significantly inhibited adherence compared to the negative control (P = 0.0364). Serum obtained from rats prior to immunization had no effect on the adhesion of M bovis PG45 to MDBK cells (P = 0.863). To assess whether M. bovis was able to adhere to inert surfaces, we initially incubated suspensions of M. bovis cells in cell culture wells without MDBK cells, and M. bovis did not adhere to the cell culture plate.

FIG 5.

Adherence and inhibition assays with M. bovis PG45. (A) MDBK cell monolayers were inoculated with the Δ565 mutant or the parental M. bovis PG45 strain, and adherent mycoplasmas were enumerated after 2 h of incubation. *, significantly different at P = 0.0329 by Student's t test. (B) M. bovis PG45 cells were incubated with dilutions of anti-MbfN sera before inoculation onto MDBK cells. Adherent mycoplasmas were enumerated after 2 h of incubation. The data presented here show the mean ± standard deviation (SD) from three independent experiments. Significant difference by ANOVA: **, P = 0.0021; *, P = 0.0364; ns, no significant difference (P = 0.8631).

Heterogeneity in sequences of mbfN homologs.

Differences in the immunogenicity profiles of MbfN homologs in several strains of M. bovis prompted the analysis of their nucleotide and amino acid sequences. Analysis of four of the sequences that were available in GenBank revealed differences in the lengths of the homologous genes. On the basis of this, a pair of PCR primers were designed based on the PG45 strain, and a second pair was designed based on MMB_0269 (Hubei strain) in order to amplify homologous genes in the strains of M. bovis that had not been sequenced before and compare their sequences with those that were previously available in GenBank. The PG45 primers (565 ampF and 565 ampR) amplified a product from two strains (1355 and 9760EU), while the MMB_0269 primers (565 F-ch and 565 R-ch) amplified a product from five other strains originating from Australia. Neither primer pair amplified a product from M. bovis strain 48832, despite the detection of an immunoreactive band on the Western blot. The mbfN gene and its homologs in M. bovis 1355 and M. bovis 9760EU clustered more closely in the nucleotide sequence analysis of their PCR products, while the M. bovis 108244, 106202, 149040, 193731, and 3683 strains clustered together with the two published Chinese isolates (Fig. S4A). Nucleotide sequence alignments of all strains included in this study by Clustal Omega revealed the absence of 135 residues (about 15 kDa) towards the terminal end of the C terminus in all strains except 1355, 9670EU, and PG45, with the last 12 amino acids retained in all strains. The absence or deletion of this region in the Chinese isolates and some Australian strains was not associated with the leucine-rich repeat regions. The nucleotide sequence of the retained 12 amino acids also matches the first 36 nucleotides of the gene downstream of mbfN (Fig. S4B).

DISCUSSION

The ability of mycoplasmas to adhere to and colonize host surfaces has long been recognized, but the elements used by M. bovis to attach, colonize, and subsequently cause disease are not fully understood. One of the constraints militating against successful functional annotation of mycoplasma genes is their low sequence similarity with those of other bacteria, making it difficult to predict their functions. This study provides insights into the possible roles played by a lipoprotein encoded by mbfN in the pathogenesis of M. bovis infection. Recombinant MbfN was initially expressed as a single polypeptide following induction with isopropyl-β-d-1-thiogalactopyranoside (IPTG), but the protein was cleaved into fragments during the protein purification process. This was not seen when we expressed and purified several other recombinant membrane proteins of M. bovis PG45 in E. coli JM109, using similar procedures. The other recombinant proteins retained their integrity during the purification steps (results not shown). We trialled a matrix of different expression conditions and IPTG concentrations and also tried optimizing our primers and recloning, but achieved similar results. The reason for the cleavage of MbfN during purification is not clear and persisted even after the addition of a cocktail of protease inhibitors during the purification process. It is possible that MbfN is inherently prone to proteolysis. Our bioinformatic analysis of MbfN using D²P² revealed the presence of several IDPs along the length of the polypeptide. Although many proteins with IDPs may not have stable secondary and tertiary structures under physiological conditions (35), we do not know what the significance of the IDPs was in the cleavage of MbfN during protein purification. Cleavage of recombinant MbfN did not, however, affect analyses of its biological activity in vitro.

Triton X-114 (TX-114) partitioning of M. bovis PG45 cells and Western immunoblotting revealed MbfN to be an integral membrane protein. Trypsin treatment of intact cells confirmed that MbfN was exposed on the surface of M. bovis PG45 cells, which could allow MbfN to interact with host surfaces. Two polypeptides (at 48 and 70 kDa) were observed on the Western blot probed with the monospecific antisera against MbfN in the PG45, 9760EU, and 193731 strains. Only the 48-kDa polypeptide was observed in the other seven strains tested. Differences in the protein profiles observed on the Western blots may be attributable to differences in nucleotide sequence observed in the homologous genes of other M. bovis strains or to posttranslational modifications of the protein. Nucleotide sequence analysis indicated that strains 1355 (isolated in Australia) and 9760EU (isolated in Spain) shared close identity with PG45 (type strain isolated in the United States), while the other Australian strains clustered closely with the Chinese strains that have been sequenced. Recent genotypic characterizations of M. bovis isolates using multilocus sequence typing schemes and pulsed-field gel electrophoresis revealed the presence of a single dominant sequence type (ST-10) clone circulating in China and Israel, with links to calves exported from Australia (36). Variations in mbfN in different strains of M. bovis may result in variations in protein expression and may affect the outcomes of infection (37). Differences in amino acid composition could also lead to both structural and functional changes, such as changes in enzyme activity, aggregation propensity, structural stability, binding, and dissociation (38). Although differences in the pathogenicity of some of these M. bovis strains have not been determined, at least two strains, 3683 and HB0801 (39, 40), and the type strain PG45 have been shown to be pathogenic. The major difference between the nucleotide sequences of the genes in the strains sequenced in this study and those already published was the absence of a region encoding ∼15-kDa at the carboxylic terminal end in all strains except 1355, 9760EU, and PG45, with only the last 12 codons at the C terminus retained (QWENIVQVKPKK, present in PG45) in all strains. Nucleotide sequence alignment revealed 100% identity between the last 37 bp from mbfN and the first 37 bp in MBOVPG45_0566, a gene downstream of mbfN. Similar overlapping arrangements of these genes were seen in other strains of M. bovis and in homologs in its closest relative, M. agalactiae (see Fig. S4B in the supplemental material). The retention of this terminal sequence may be necessary for the expression of the N-terminal region of the downstream open reading frame. The reason for the absence of the region encoding the remaining 15-kDa of the carboxyl terminus of the protein in some strains of M. bovis is still unclear, but the absence/deletion was not within the leucine-rich repeat region. It is reasonable to suggest that the absence or presence of this polypeptide in the M. bovis genomes may be a consequence of genome reorganization due to reductive evolution or the acquisition of a foreign piece of chromosomal DNA that has then been transferred to subsequent generations.

Surface-localized proteins of mycoplasmas have been shown to undergo posttranslational modifications generating smaller polypeptides (3, 41), which could result in alterations to protein function or even excision of newly formed proteins as a result of proteolysis, thereby regulating a wide range of biological processes (42). In the studies reported here MbfN was shown to be surface exposed, and two polypeptides were recognized in the PG45, 9760EU, and 193731 strains when their whole cells were probed with anti-MbfN sera on the Western blot, indicating that this protein is probably undergoing some form of posttranslational cleavage. MbfN is the second protein of M. bovis PG45 shown to undergo proteolytic processing. We have earlier reported on the first M. bovis protein (MilA) to be found to undergo proteolytic processing (3). Although the cleavage events in MilA and MbfN were discovered serendipitously and the consequences of these events remain to be elucidated, proteolytic processing in M. bovis may be widespread, especially in those surface-associated proteins that are predicted to have very high molecular masses. Analysis of the global processing events in other mycoplasma species, such as Mycoplasma hyopneumoniae, have identified a large proportion of surface-localized proteins to be targets of endoproteolytic processing events (41). Proteolytic processing may serve as a fundamental mechanism generating a variety of proteoforms on the surface of mycoplasmas that may be important in separating functionally important binding motifs (43).

Host-pathogen interactions can be influenced by associations between pathogenic elements and ECM components of the host, leading to adhesion and/or evasion of immune surveillance by the host (44). MbfN was found to contain LRR motifs, and in silico analysis of its amino acid sequence revealed functional similarity with fibronectin-binding proteins Hlp3 of Mycoplasma gallisepticum (45) and BspA from Bacteroides forsythus (34). Recombinant MbfN was shown to be able to bind fibronectin using the dot blot and microtiter plate methods. The fibronectin binding was dose dependent, and this interaction could be competitively inhibited by addition of monospecific anti-MbfN sera. Although a capacity to bind fibronectin is widespread among surface proteins of many bacterial species, not much is known about the proteins of M. bovis that interact with fibronectin. The binding of MilA to fibronectin was a serendipitous discovery, adding to the list of M. bovis proteins that interact with fibronectin. This concurs with our earlier hypothesis that MilA is a multifunctional membrane protein of M. bovis (3). Fibronectin production is upregulated during pneumonia, and this dramatic increase in fibronectin levels provides scaffolding for microbial attachment and subsequent colonization of the host (46, 47). In addition to its ability to bind fibronectin, four heparin-binding consensus sequences were identified in MbfN. The ability to bind to heparin was confirmed using a dot blot assay, demonstrating that the heparin-binding motifs in MbfN were functional. MbfN is the second heparin-binding protein identified in M. bovis PG45 (3). The capacity of pathogenic microorganisms to bind heparin and related glycosaminoglycans as a means of interacting with host mucosal surfaces is well documented. These organisms recruit ECM components to their surfaces, impacting key aspects of microbial pathogenicity (48).

The results of our in silico modeling of MbfN and functional similarities detected in homology searches suggested that MbfN may have cytoadherence-related ECM binding functions. In addition to demonstrating that MbfN was a fibronectin and heparin-binding protein, we hypothesized that MbfN may have an adhesion-related function. We assessed the differential adherence of the parental M. bovis PG45 strain and its mutant (Δ565) derivative to MDBK cells. Disruption of mbfN by transposon mutagenesis led to a 3-fold decline in adherence to MDBK cells compared with the parental M. bovis PG45 strain. The specificity of the decrease in adherence of the Δ565 mutant to MDBK cells was further assessed by preincubating M. bovis PG45 cells with monospecific anti-MbfN sera before addition of the mycoplasma cells to the MDBK monolayer. Treatment of M. bovis PG45 with monospecific anti-MbfN sera resulted in a significant decrease in adherence to MDBK cells. Previous studies have demonstrated the contribution of surface-exposed proteins to adherence by mycoplasmas to eukaryotic cell lines using trypsin-shaving experiments (49–51). Current genetic tools have enabled studies examining the effect of gene disruptions on adherence of mycoplasmas to eukaryotic cell lines (52, 53). The identification of surface-localized adhesion-related proteins in M. bovis could highlight potential targets for the development of drugs, vaccines, and/or immunodiagnostic tests for this economically important ruminant pathogen.

Conclusion.

This study identified a novel fibronectin-binding protein, MbfN, in M. bovis PG45 and demonstrated its role in adherence in vitro. Adherence to host cells is considered an initial and decisive step in infection (54). MbfN was also able to elicit an antibody response in calves experimentally infected with M. bovis. These roles played by MbfN have, therefore, improved our understanding of the pathogenesis of M. bovis infection.

MATERIALS AND METHODS

Bacterial strains, plasmids, and culture conditions.

M. bovis PG45 (ATCC 25523) and its mutant derivative were cultivated to the mid-logarithmic growth phase in M. bovis broth (modified Frey’s broth containing 21 g pleuropneumonia-like organism [PPLO], 37 ml yeast extract, 100 ml inactivated swine serum, 4 ml 1.6% phenol red solution, 300 mg penicillin G, 859 ml distilled water, pH adjusted to 7.8) at 37°C for approximately 18 to 20 h, with 50 μg gentamicin/ml included when required to maintain selection pressure for carriage of the Tn4001 transposon. The wild-type and mutant strains were cultured routinely at a 1/10 dilution in fresh medium. The concentrations of M. bovis PG45 and the mutant were determined as color-changing units (CCU) per milliliter as previously described (55). The expression vector pGEX-4T-1 and a derivative of it were maintained in Escherichia coli strain JM109 in Luria-Bertani (LB) broth at 37°C. Broth cultures of E. coli were incubated on a rotary shaker at 200 rpm. For selection of transformants, LB agar was supplemented with 100 μg ampicillin/ml. Staphylococcus aureus CC4 was grown on Columbia sheep blood agar at 37°C for 24 h. The origins and growth conditions of other strains used in this study were reported earlier (3).

DNA cloning and protein expression.

Genomic DNAs from S. aureus CC4 and M. bovis strains were extracted using the High Pure PCR Template Preparation kit (Roche). The DNA sequence of mbfN from M. bovis PG45 (NCBI ID no. CP002188.1) was used to design primer pairs incorporating a BamHI cleavage site at the 5′ end (Table 1). The mycoplasma TGA tryptophan codon in the sequence of mbfN was altered to TGG using overlap extension PCR before heterologous expression in E. coli JM109 (Fig. 6B). A SalI site followed the protein-coding sequence at the 3′ end. All PCRs were carried out in a 25-μl volume containing approximately 50 ng of M. bovis strain PG45 genomic DNA, 1× Green GoTaq Flexi buffer (Promega), 2 mM MgCl₂, 200 μM each deoxynucleoside triphosphate (dNTP; Bioline), 0.1 μM each primer, and 1.25 U of GoTaq Flexi DNA polymerase (Promega). To amplify the different lengths of mbfN, touchdown PCRs were performed in an iCycler under the following conditions: 95°C for 5 min, then 18 cycles of 95°C for 1 min, with the annealing temperature (T_a) + 10°C lowered by 1.3°C every two cycles for 1 min until it reached T_a, and 68°C for 2 min, followed by 25 cycles of 95°C for 30 s, T_a for 30 s, and 68°C for 2 min, and a final extension of 68°C for 5 min. PCR products were analyzed in 1% agarose gels, cleaned using the QIAquick gel extraction kit (Qiagen), and subsequently used as the template (combined molar ratio of each amplicon) in the overlap extension PCR (see Fig. S2A in the supplemental material). The fnbB gene from S. aureus CC4 was amplified using primer pairs FnBPB-fwd and FnBPB-rev under the following conditions: 95°C for 5 min followed by 30 cycles of 95°C for 30 s, 58°C for 30 s, and 72°C for 2 min, and a final extension of 72°C for 4 min. PCR products were ligated into the expression vector pGEX-4T-1 (GE Healthcare). E. coli JM109 cells were transformed with the ligation mixtures, and clones containing the amplified regions were selected on LB agar plates supplemented with 100 μg ampicillin/ml. The DNA sequences of the recombinant mbfN and fnbB genes in pGEX-4T-1 were confirmed by Sanger sequencing using a BigDye Terminator cycle sequencing reaction kit (Applied Biosystems) and the oligonucleotides pGEXfwd and pGEXrev (Table 1), with capillary electrophoresis performed at the Centre for Translational Pathology, The University of Melbourne, Australia. The expression of the recombinant glutathione S-transferase (GST)–MbfN and FnBPB (used as positive control) fusion proteins was induced in a logarithmic-phase culture of the transformants by the addition of isopropyl-β-d-1-thiogalactopyranoside (IPTG) (Sigma) to a final concentration of 0.5 mM. Expression of the fusion proteins was assessed by sodium dodecyl sulfate–12% polyacrylamide gel electrophoresis (SDS–12% PAGE) and Western immunoblotting of whole-cell lysates.

TABLE 1.

Oligonucleotides used in this study

Primer no.	Primer name	Binding sitea	Sequence (5′→3′)b	Ta (°C)c
1	565-BamH1F	55–75	ATAggatccGCTTGGGATAATAAGGAT	51
2	565-1WR	534–505	AGATTTATAAGTCCATTTTAGCAGGTTTGT
3	565-1WF	505–534	ACAAACCTGCTAAAATGGACTTATAAATCT	52
4	565-Sal1R	1806–1786	gtcgacTCATTGGTTTGAATCACTTCC
5	565 ampF	1–29	ATGAAAAAATTTTCTTTTGTACTTTTACC	50
6	565 ampR	1839–1813	TTATTTTTTGGTTTTTACTTGAACGAT
7	565 F-ch	26–47	TACCAATTTTAACATTCCCGGC	50
8	565 R-ch	1355–1333	TCTGTGTTATCACTTCCTTCTG
9	FnBPB-fwd	481–502	TAggatccGAAGCTAAAGCGACAGGTACAG	58
10	FnBPB-rev	2080–2100	TAgtcgacGAATGACTGATTACCGCTATT
11	pGEXfwd		GGGCTGGCAAGCCACGTTTGGTG	55
12	pGEXrev		CCGGGAGCTGCATGTGTCAGAGG
13	IR inverse		TGGCCTTTTTACTTTTACACAAT

^aThe binding site refers to the location on the gene sequence. Primers 1 to 6 are based on MBOVPG45_0565 sequence (NC_014760; 654047 to 655885 complement), while primers 7 and 8 are based on the MMB_0269 sequence (CP002513; 320775 to 322166 complement) of M. bovis. Primers 9 and 10 are based on the fnbB gene sequence of Staphylococcus aureus N315 (NC_002745.2; 2568323 to 2571208 complement).

^bBoldface letters indicate nucleotide differences from the published gene sequence. These changes were made to mutate the cysteine codon at the lipoprotein cleavage site or to change the TGA tryptophan codon to TGG. Lowercase letters indicate restriction endonuclease cleavage sites (BamHI, ggatcc; SalI, gtcgac).

^cT_a, annealing temperature.

FIG 6.

Physical map of MBOVPG45_0565 from M. bovis PG45 and fnbB from Staphylococcus aureus SA2290. (A) Schematic diagram to show the relative positions of the leucine-rich repeats (black rectangles) in MBOVPG45_0565 (orange rectangle with numbers above showing the amino acid position. (B) Diagram illustrating the mutagenesis of the TGA tryptophan codon in MBOVPG45_0565 to TGG by overlap extension PCR and the introduction of restriction endonuclease cleavage sites to facilitate cloning into pGEX-4T-1. The portion of the gene upstream of the prokaryotic cleavage site is shaded. (C) Structural organization of the fibronectin-binding protein (FnBPB) from S. aureus SA2290 and cloning of the fragment from residues 161 to 700. FnBPB contains a signal sequence (S) followed by fibrinogen and an elastin binding domain consisting of subdomains N1, N2, and N3. The fibronectin-binding domain has tandemly repeated fibronectin-binding motifs. At the C terminus are proline-rich repeats (PRR) and cell wall (W)- and membrane (M)-spanning domains.

Construction of the plasmid-carrying transposon and generation of the mbfN mutant.

The construction of the Tn4001-based transposon plasmid (pTn4001complete) used for the random transposon mutagenesis has been described in detail previously (56). The plasmid construct was used to transform M. bovis PG45 to generate mutant libraries. Transposon insertion into mbfN was confirmed by PCR using primers 565 ampR and IR inverse and by Sanger sequencing. The insertion site was confirmed by searching the DNA sequence against the M. bovis PG45 genome using the BLAST algorithm.

Purification of recombinant fusion proteins.

The induced E. coli JM109 cells containing the recombinant MbfN and FnBPB were pelleted and resuspended in phosphate-buffered saline (PBS: 137 mM NaCl, 2.7 mM KCl, 8 mM Na₂HPO₄, 1.46 mM KH₂PO₄, pH 7.4). The cells were then lysed by incubation in 2 mg lysozyme/ml for 2 h in lysis buffer (PBS, 1 μM EDTA, 1 μM phenylmethylsulfonyl fluoride). The sample was then sonicated, followed by incubation on ice for 30 min in 1% Triton X-100. The recombinant proteins were purified from the soluble total protein fraction by affinity chromatography with glutathione-Sepharose 4B beads (GE Healthcare) and elution with free glutathione. The purified proteins were analyzed by SDS-PAGE and then dialyzed in PBS at 4°C with three changes every 8 h within a 24-h period. GST, used as a negative control, was expressed from the pGEX-4T-1 vector in E. coli JM109 cells and purified in the same manner as the other recombinant proteins. Protein quantitation was carried out using the Pierce bicinchoninic acid (BCA) protein assay kit (Thermo Fisher Scientific).

Polyclonal antibody production.

Two healthy Sprague-Dawley rats were used to raise antisera against the recombinant MbfN. Blood samples were taken via the tail vein before being immunized by subcutaneous injection with 50 μg of recombinant protein in solution emulsified in Freund’s complete adjuvant (Sigma-Aldrich). The rats were given booster inoculations with 50 μg of the protein in Freund’s incomplete adjuvant (Sigma-Aldrich) at day 21. A second booster in Freund’s incomplete adjuvant was given 2 weeks after the first booster. Blood was finally collected by cardiac puncture under anesthesia with isoflurane at day 45. Serum samples were stored at –20°C until further use.

Trypsin treatment of intact M. bovis PG45 cells.

A 20-ml culture of M. bovis PG45 was grown to mid-logarithmic phase (1.2 × 10⁶ CCU/ml) and pelleted by centrifugation at 13,000 × g for 5 min at 4°C. The cell pellet was washed in PBS and suspended in 400 μl PBS. Protein was quantitated, and 20 μg of cells was diluted with PBS to give a final volume of 200 μl for each trypsin treatment. To partially digest the cell surface proteins of the intact M. bovis, trypsin (Sigma-Aldrich) was added to separate aliquots of the cell suspensions to a final concentration of 0.5, 1, 4, 8, 16, or 32 μg/ml. Samples were incubated at 37°C for 30 min and centrifuged at 13,000 × g for 5 min, and the supernatant was discarded. The cell pellets were suspended in SDS-PAGE lysis buffer (4% [wt/vol] SDS, 20% [vol/vol] glycerol, 125 mM Tris-HCl [pH 6.8], bromophenol blue, and 10% [vol/vol] β-mercaptoethanol) and heated at 95°C for 5 min. The cell lysate was then analyzed by SDS–12% PAGE and stained with Coomassie brilliant blue or transferred onto polyvinylidene difluoride (PVDF) membrane for Western immunoblotting using the monospecific anti-MbfN sera.

Triton X-114 phase partitioning of M. bovis PG45.

Triton X-114 (TX-114; Sigma-Aldrich) detergent-soluble M. bovis PG45 proteins were prepared according to the method described by Duffy et al. (57), with minor modifications. M. bovis PG45 cells were pelleted from 45 ml of a late-logarithmic-phase (1.6 × 10⁸ CCU/ml) broth culture by centrifugation at 12,000 × g for 15 min at 4°C. The cell pellet was washed three times with cold PBS and suspended in 0.5 ml of 0.5% (vol/vol) TX-114 in PBS, and the mixture was incubated on ice for 90 min, with mixing every 15 min. The solution was centrifuged at 12,000 × g for 30 min at 4°C, and the supernatant was carefully loaded on top of a 1-ml sucrose cushion (6% [wt/vol] sucrose and 0.06% [vol/vol] TX-114 in PBS) and incubated at 37°C for 9 min to achieve phase separation. The sample was then centrifuged at 500 × g for 6 min at 37°C. The supernatant, containing mainly the hydrophilic proteins, was carefully aspirated, leaving an oily pellet containing the hydrophobic fraction. The pellet was suspended in 0.5 ml of ice-cold PBS, and the proteins were precipitated by adding 0.5 ml methanol-chloroform (4:1), followed by centrifugation at 12,000 × g for 1 min at 4°C. The hydrophilic proteins in the supernatant were also precipitated. The proteins were partially air dried, dissolved in 100 μl of 8 M urea in PBS, and stored at –20°C until further use. The insoluble proteins and hydrophilic and hydrophobic fractions were analyzed by SDS–12% PAGE gel and Western immunoblotting.

SDS-PAGE and Western immunoblotting.

Ten micrograms of whole-cell proteins of Mycoplasma bovis strains or 2 μg of the purified recombinant protein was separated by SDS-PAGE. Gels were stained with Coomassie brilliant blue or the proteins were transferred onto polyvinylidene difluoride (PVDF) membranes (Amersham Biosciences). The PageRuler Prestained Protein Ladder (Thermo Scientific) was used as the molecular weight standards. Following transfer, membranes were incubated overnight at 4°C in 5% (wt/vol) skim milk in PBS. The membranes were washed three times for 5 min each with PBS-T (PBS containing 0.05% Tween 20 [vol/vol]) and then incubated with gentle rocking for 1 h at room temperature in a 1:100 dilution of pooled sera from calves experimentally infected with M. bovis (58) or a 1:500 dilution of monospecific anti-MbfN sera. Membranes were washed and incubated at room temperature for 1 h in a 1:2,000 dilution of horseradish peroxidase (HRP)-conjugated sheep anti-bovine antibody (Dako) or 1:2,000 dilution of HRP-conjugated rabbit anti-rat antibody (Dako), respectively, in PBS-T containing 0.05% skim milk. Following a final washing step, the bound antibody was detected using the Clarity Western ECL (enhanced chemiluminescence) blotting substrate (Bio-Rad). Images were captured by chemiluminescence using the ChemiDoc XRS+ system (Bio-Rad Laboratories).

Fibronectin binding studies. (i) Dot blot assay.

Serial 2-fold dilutions of fibronectin from human plasma (Sigma-F2006), from 2.5 to 0.078 μg, were generated, and 5 μl of each dilution was applied as spots onto Hybond-C Extra nitrocellulose membrane (Amersham) and allowed to dry for 5 min at room temperature. The membrane was blocked in blocking solution (1% bovine serum albumin [BSA] in Tris-buffered saline [TBS; 10 mM Tris, pH 7.4, and 150 mM NaCl]) for 1 h at room temperature and washed three times in TBS-T (0.05% Tween 20 in TBS). The membrane was then incubated for 2 h at 4°C with gentle mixing on a rocker in a solution containing 20 nM recombinant MbfN in TBS-T. The membrane was washed and then incubated for 1 h at room temperature in a 1:2,000 dilution of goat anti-GST antibody (Dako) in TBS-T containing 0.05% (wt/vol) BSA. Following a washing step, the membrane was incubated at room temperature in a 1:2,000 dilution of HRP-conjugated rabbit anti-goat antibody (Dako) in TBS-T. Detection of fibronectin-bound protein was achieved using the ECL chemiluminescence substrate (GE Healthcare). FnBPB and GST proteins were included as positive and negative controls, respectively. Images were captured using the ChemiDoc XRS+ system. The assay was performed twice to ensure reproducibility.

(ii) Microplate binding method.

Microplate binding assays were performed using 96-well microtiter plates (Maxisorp Nunc-Immunoplate; Thermo Scientific). Serial 2-fold dilutions of fibronectin from 10 to 0.078 μg per well in 100 μl PBS were used to coat the wells of the microtiter plate by incubating the plate overnight at 4°C. The plate was then washed three times with 300 μl of PBS-T per well and blocked with 5% BSA in PBS for 2 h at 37°C. Twenty picomoles of recombinant MbfN in 100 μl of PBS was added to each of the wells and incubated for 2 h at 37°C. The wells were washed, 100 μl of goat anti-GST antibody diluted 1:2,000 in 0.05% BSA in PBS-T was added to each well, and the plate was incubated for 1 h at room temperature. The wells were washed, 100 μl of HRP-conjugated rabbit anti-goat antibody diluted 1:1,000 in 0.05% BSA in PBS-T was added to each well, and the plates were incubated for another 1 h at room temperature. After a final wash, 100 μl of 0.3 g 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS; KPL)/liter was added to the wells, and the plate was incubated for 3 min at room temperature for color development. The absorbance of each well was measured at 405 nm using a Hybrid Multimode microplate reader (BioTeK). Readings from triplicate reactions were taken from two independent experiments.

For the competition binding assay, wells of the microtiter plate were coated with 5 μg of fibronectin. Decreasing concentrations of the monospecific anti-MbfN sera were incubated with 20 pmol of recombinant MbfN in separate tubes for 1 h at 37°C, before addition to the microtiter wells coated with fibronectin and incubation for 2 h at 37°C. After washing, MbfN bound to fibronectin was detected by probing with goat anti-GST antibody and HRP-conjugated rabbit anti-goat antibody as described above.

Heparin binding assay.

Using the Bio-Dot apparatus (Bio-Rad), Hybond-C Extra nitrocellulose membrane (Amersham) was blotted with different amounts of recombinant MbfN (30, 15, and 7.5 pmol), and the membrane was allowed to dry for 5 min at room temperature. MilA was used as a positive control, while GST and BSA proteins served as negative controls. The membrane was blocked in 2% BSA in TBS for 1 h at room temperature and then washed three times for 5 min each with TBS-T. The membrane was then incubated for 1 h at room temperature in 100 μg biotinylated heparin (Sigma-Aldrich)/ml in 1% BSA in TBS-T with gentle rocking to allow for binding of the heparin to the immobilized proteins. The membrane was then washed and then incubated for 1 h in a 1:2,000 dilution of HRP-conjugated goat antibiotin antibody (Cell Signaling) in 1% BSA in TBS-T. The blot was washed, bound heparin was detected using the Clarity Western ECL substrate, and the image was captured using the ChemiDoc XRS+ system. To ensure that equivalent amounts of protein had been blotted at each position, a duplicate blot was stained with Coomassie brilliant blue. The assay was carried out in two independent experiments.

Adherence and inhibition assay of M. bovis PG45.

MDBK cells were grown in Dulbecco’s modified Eagle’s medium (DMEM; D2902 [Sigma-Aldrich]) with 0.42 g sodium hydrogen carbonate/liter, 0.0032% HCl, and 0.5% phenol red, supplemented with 5% (vol/vol) fetal bovine serum (FBS), 2 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES; pH 7.4), and 50 mg ampicillin/ml. All cells were maintained at 37°C in a humidified 5% CO₂ atmosphere. Cells of M. bovis PG45 and its mutant (Δ565) derivative were grown to logarithmic phase (1.8 × 10⁷ CCU/ml) in M. bovis broth, and 20 ml of each culture was centrifuged to remove the mycoplasma culture medium and resuspended in DMEM. The concentration of the Mycoplasma bovis cells was adjusted with DMEM so as to infect MDBK cells at a multiplicity of infection (MOI) of 100 in 6-well flat-bottom tissue culture plates (Costar). Plates were incubated for 2 h at 37°C in 5% CO₂ (vol/vol) in air, and nonadherent mycoplasmas were removed by washing with PBS. The MDBK cells and adherent mycoplasmas were scraped off each well using a cell scraper, resuspended in 1 ml M. bovis broth, and passed through a 20-gauge needle to mechanically lyse the MDBK cells before titration in 96-well microtiter plates. The concentrations of viable mycoplasma cells in each well were determined using the most probable number technique and expressed as CCU per milliliter (55). For the adherence inhibition assay, M. bovis PG45 cells were preincubated with different dilutions (1:10 and 1:100) of heat-inactivated monospecific anti-MbfN sera for 2 h at 37°C and then inoculated onto the MDBK cells. The percentage of adherent M. bovis cells relative to the input amount of M. bovis was calculated as follows: % adherence = (no. of M. bovis cells recovered/no. of M. bovis cells inoculated) × 100. Serum from an unimmunized rat was used as a negative control. Assays were carried out in triplicate in three independent experiments.

Sequencing of mbfN and its homologs.

PCR primers were designed to target the mbfN gene (565 ampF and 565 ampR) from M. bovis PG45 and the MMB_0269 gene (565 F-ch and 565 R-ch) from the M. bovis Hubei-1 strain. These primer pairs were used to amplify corresponding homologs from other strains of M. bovis. The PCRs were carried out in a 25-μl volume containing approximately 50 ng of genomic DNA, 1× Green GoTaq Flexi buffer, 2 mM MgCl₂, 200 μM each dNTP, 0.1 μM each primer, and 1.25 U of GoTaq Flexi DNA polymerase. Thermal cycling was performed with the following settings: 95°C for 5 min, followed by 30 cycles of 95°C for 30 s, 50°C for 30 s, and 72°C for 2 min. This was followed by a final extension at 72°C for 5 min. PCR products were analyzed in a 1% agarose gel. The PCR products were purified with an UltraClean GelSpin DNA extraction kit (MoBio Laboratories). DNA sequencing of PCR products was carried out at the Centre for Translational Pathology, The University of Melbourne, Australia. PCR products were sequenced in both directions with the same primers used for the PCR amplification. DNA sequences were assembled and trimmed using Geneious version 7.1.5 (59). BLAST searches were performed against M. bovis genomes in the NCBI database, and Clustal Omega was used to align the sequences. Phylogenetic analysis was performed using the neighbor-joining method and global alignment with free end gaps in Geneious version 7.1.5.

In silico analyses and structural modeling of mbfN gene.

In silico analysis of the coding sequence of MBOVPG45_0565 (mbfN) was performed using the Kyoto Encyclopedia of Genes and Genome (KEGG) and National Center for Biotechnology Information (NCBI) databases. The coding domain sequence of mbfN from the M. bovis PG45 genome (positions 654047 to 655885) was retrieved from the NCBI database. Analyses of codon usage, codon optimization, and hydropathy were carried out using DNA Strider version 2.0 f1.3 (60). For ab initio structural modeling and ligand binding site predictions, Phyre² (31), I-TASSER (32), and RaptorX (61) were used. Cellular localization and signal peptide predictions were performed using PSORTb v. 3 (62) and SignalP v. 4.0 (63), respectively. LRR regions within the predicted amino acid sequence were identified based on motif searches using the NCBI database and a previously described method (64). Analysis for intrinsically disordered (IDP) motifs in MbfN was performed using the Database of Disordered Protein Prediction tool D²P² (http://d2p2.pro/search) (65).

Statistical analysis.

Statistical analyses were performed using GraphPad Prism version 8.0e for Mac OSX (GraphPad Software, Inc., San Diego, CA). A one-way analysis of variance (ANOVA) was used to compare three or more means followed by post hoc analysis using Tukey’s multiple-comparison tests. Student's t test was used to compare the difference in adhesion between the M. bovis PG45 and Δ565 strains to MDBK cells.

Ethical statement.

Animal experiments were performed according to the International Guiding Principles for Biomedical Research Involving Animals with approval of The University of Melbourne Animal Ethics Committee (approval no. 1413074).

Accession numbers(s).

Sequences have been deposited in GenBank under accession no. MF136693 to MF136699.