A Processed Multidomain Mycoplasma hyopneumoniae Adhesin Binds Fibronectin, Plasminogen, and Swine Respiratory Cilia

Abstract

Porcine enzootic pneumonia is a chronic respiratory disease that affects swine. The etiological agent of the disease, Mycoplasma hyopneumoniae, is a bacterium that adheres to cilia of the swine respiratory tract, resulting in loss of cilia and epithelial cell damage. A M. hyopneumoniae protein P116, encoded by mhp108, was investigated as a potential adhesin. Examination of P116 expression using proteomic analyses observed P116 as a full-length protein and also as fragments, ranging from 17 to 70 kDa in size. A variety of pathogenic bacterial species have been shown to bind the extracellular matrix component fibronectin as an adherence mechanism. M. hyopneumoniae cells were found to bind fibronectin in a dose-dependent and saturable manner. Surface plasmon resonance was used to show that a recombinant C-terminal domain of P116 bound fibronectin at physiologically relevant concentrations (K_D 24 ± 6 nm). Plasmin(ogen)-binding proteins are also expressed by many bacterial pathogens, facilitating extracellular matrix degradation. M. hyopneumoniae cells were found to also bind plasminogen in a dose-dependent and saturable manner; the C-terminal domain of P116 binds to plasminogen (K_D 44 ± 5 nm). Plasminogen binding was abolished when the C-terminal lysine of P116 was deleted, implicating this residue as part of the plasminogen binding site. P116 fragments adhere to the PK15 porcine kidney epithelial-like cell line and swine respiratory cilia. Collectively these data suggest that P116 is an important adhesin and virulence factor of M. hyopneumoniae.

Introduction

Porcine enzootic pneumonia is a chronic respiratory disease affecting swine populations worldwide. The etiological agent, Mycoplasma hyopneumoniae, adheres to cilia in the respiratory tract of pigs, resulting in loss of cilia and epithelial cell damage (1). The host becomes more susceptible to secondary infections by Pasteurella multocida (2), swine influenza virus (3), and Actinobacillus pleuropneumoniae (4). M. hyopneumoniae infections result in a significant economic burden to the swine industry (5).

The initial association between M. hyopneumoniae and respiratory cilia is considered an essential process in colonization of the host (6). Commercial vaccines restrict the development and severity of disease progression, but are unable to prevent colonization (7–9). Prevention of the interactions between the host and M. hyopneumoniae prior to colonization would result in a more efficacious vaccine (8, 10). Although research into this area is focusing on proteins specific to these interactions (11–15), M. hyopneumoniae pathogenicity and adherence mechanisms remain poorly characterized (7, 12).

P97, is one of the major cilium adhesins of M. hyopneumoniae and has been well characterized. Despite the role of P97 in adherence, additional surface proteins also contribute to this process including P159 and P216 (5, 11–14). A 126 kDa pre-protein is proteolytically processed to form P97 (5, 16); the gene encoding P97 (mhp183) is located in a two gene operon with that encoding P102 (mhp182) (5). The function of P102 has not been established, but it is processed generating two fragments of 72 and 42 kDa and secreted from the cell during growth in vitro (16). mhp182 is known to be expressed during disease and the genetic linkage of P102 and P97 suggests it may play an accessory role to the P97 adhesin (17).

There are six paralogs of both the P97 adhesin and P102 throughout the M. hyopneumoniae genome. These proteins share >30% identity over 70% of the amino acid sequence (18). The presence of multiple adhesins in other mycoplasma pathogens has been well described. Such Mycoplasma species contain adhesins located in operons (19), with multiple copies of adhesin genes present in the genome (20, 21). The presence of multiple adhesins increases antigenic variation which is advantageous for avoiding recognition by the immune system (5). The majority of the M. hyopneumoniae paralogs are found in two gene operons containing genes encoding both a P97 and a P102 paralog (18). All of these paralogs, with the exception of mhp280, are transcribed in vivo. The gene partners are transcribed as a single operon unit, except mhp107 and mhp108, which are likely to be independently transcribed (17). The presence of a signal peptide sequence is common across the majority of the paralogs (17), suggesting cell wall and/or surface location of these proteins.

These observations stimulated our interest in characterizing the above mentioned M. hyopneumoniae putative adhesins, the P97 and P102 paralogs. The characterization of P116 has been undertaken as part of a larger study to investigate all of the P97 and P102 paralogs. The putative 116 kDa protein (P116) is encoded by mhp108. This study used proteomic analyses to examine the expression and post-translational cleavage of P116. Surface plasmon resonance (SPR)² was employed to demonstrate the ability of recombinant P116 fragments to bind to fibronectin and plasminogen. Recombinant P116 fragments also bound the porcine epithelial-like cell line (PK15) and swine cilia, suggesting a role for P116 in colonization and virulence of M. hyopneumoniae.

EXPERIMENTAL PROCEDURES

Bacterial Strains and Culture Conditions

M. hyopneumoniae strain 232 (11) was grown in modified Friis medium and harvested by centrifugation as described previously (22). Escherichia coli BL21 star (DE3) (Invitrogen) was grown in Luria Bertani medium at 37 °C with shaking at 200 rpm. When appropriate, ampicillin was added to 100 μg/ml.

Plasmid Constructs, Expression, and Purification

Polyhistidine fusion proteins (F1_P116-F4_P116) were generated using the pET151/d-TOPO® cloning kit (Invitrogen), according to the manufacturers' instructions. F1_P116 and F3_P116 were amplified from M. hyopneumoniae strain 232 chromosomal DNA. F2_P116 and F4_P116 were amplified from a pMHP108 construct containing mhp108, with TGA codons converted to TGG to encode tryptophan; TGA encodes tryptophan in M. hyopneumoniae. Primer sequences used in this study can be viewed in supplemental Table S1. Polymerase chain reactions used PfuUltra^TM High-Fidelity II DNA Polymerase (Stratagene) and a Cooled Gradient Palm Cycler GC1-96 (Corbett Research). Site-directed mutagenesis using the QuikChange Site-directed mutagenesis kit from Stratagene, was performed to create a C-terminal lysine deletion construct of F4_P116. All recombinant constructs were sequenced to confirm fidelity of the mhp108 constructs. Expression plasmids were transformed into E. coli BL21 star (DE3) cells, and protein expression was induced by the addition of 1 mm isopropyl-1-thio-β-d-galactopyranoside when cell growth reached an A₆₀₀ of 0.8. Proteins were purified using Ni-NTA-agarose in accordance with the manufacturer's instructions for denaturing conditions (Qiagen). Proteins were dialyzed into 0.01% SDS in phosphate-buffered saline (PBS; 10 mm sodium phosphate, 137 mm sodium chloride, 2 mm potassium phosphate, 2.7 mm potassium chloride; pH 7.4).

Polyclonal Antisera Production and Western Blot Analysis

Rabbit serum was raised against recombinant P116 protein fragments as described previously (12) with three injections of antigen at 2-week intervals. Serum was collected from clotted blood by centrifugation at 2,500 × g for 10 min. The resulting serum was centrifuged at 16,000 × g for 5 min to pellet remaining red blood cells. To affinity purify antisera, 400 μg of antigen was applied to 4 cm² nitrocellulose membrane. Unbound protein was removed by washing with 100 mm glycine/HCl, pH 2.5 for 5 min with gentle shaking. 3% BSA in PBS was used to block the membrane for 1 h at room temperature, with shaking. The membrane was washed twice with PBS. Then, 1 ml of polyclonal antiserum was diluted into 4 ml of PBS and incubated with the membrane for 3 h at room temperature with gentle shaking to allow specific antibodies to bind to the antigen. Following the incubation, the membrane was washed 4 × 5 min with PBS, to remove unbound antibodies. Bound antibodies were eluted with 0.5 ml of 100 mm glycine/HCl, pH 2.5 during a 10 min incubation with shaking. This elution step was repeated. Eluates were pooled, 100 μl of 1 m Tris, pH 8.0 was added to raise the pH and BSA to 1 mg/ml was added to stabilize the purified antisera. Western blot analysis of the purified, depleted and polyclonal antisera against the antigens and M. hyopneumoniae whole cell extracts was used to verify antisera purification. Western blots were performed according to Cole et al. (23).

Proteomic Analyses

Trypsin treatment of intact M. hyopneumoniae, as previously described (24), was used to investigate the subcellular location of P116. Trypsin was added at 0–300 μg/ml. Cell lysates were then analyzed using immunoblots probed with the purified polyclonal antisera described above, negative control blots were probed with affinity purified pre-immune sera. MALDI-TOF and ESI-MS/MS mass spectrometry was undertaken to investigate the expression of P116. Proteins for MALDI-TOF MS analysis were extracted from M. hyopneumoniae cells using Triton X-114 (TX-114) as described by Wise and Kim (25). Initial cell lysis was performed overnight at 4 °C in 1% TX-114, 10 mm Tris pH 8.0, 150 mm NaCl, and 1 mm EDTA. Insoluble material was removed by centrifugation at 16,000 × g for 15 min. Extracted proteins were acetone precipitated, resuspended in SSS buffer (8 m urea, 100 mm DTT, 4% (w/v) CHAPS, 0.8% (v/v) Bio-Lyte® 3/10 ampholyte (Bio-Rad), 40 mm Tris-HCl) and subjected to sonication for 4 × 30 s using a Microson Ultrasonic sonicator. The sample was centrifuged at 16,000 × g for 15 min and the supernatant retained. Samples of M. hyopneumoniae proteins were prepared for two-dimensional gel electrophoresis (2DGE) with a Ready Prep 2-D Cleanup kit (Bio-Rad). 2DGE was performed and results analyzed as described by Cole et al. (23) using 600 μg of M. hyopneumoniae aqueous phase proteins to re-hydrate a 17 cm, non-linear, pH 3–10 ReadyStrip IPG strip (Bio-Rad). Protein spots were identified by peptide mass fingerprinting using MADLI-TOF MS (26).

Further identification of P116 was undertaken using ESI-MS/MS analysis of the M. hyopneumoniae proteome and proteins extracted from SDS-PAGE of the M. hyopneumoniae aqueous phase from a TX-114 extraction. For analysis of the M. hyopneumoniae proteome, lyophilized cells were resuspended in 1 ml of 8 m urea, 100 mm NH₄HCO₃, pH 9 and sonicated with an ultrasonic probe at 80% power for 6 × 30 s on ice. Reduction and alkylation of cysteine was performed in a single step by adding tributylphosphine to a final concentration of 10 mm and acrylamide monomers to 20 mm. The sample was then incubated for 90 min at room temperature. The alkylated proteins were precipitated by adding five volumes of acetone and incubating for 30 min at room temperature. The precipitated proteins were pelleted by centrifugation and then resuspended in 8 m urea, 100 mm NH₄HCO₃, pH 9, before digestion to peptides by addition of 2.5 μg of endoproteinase LysC (Roche, Switzerland) and incubation for 4–6 h at 37 °C. The sample was then diluted to 1 m urea by adding 100 mm NH₄HCO₃ prior to the addition of 2.5 μg of trypsin (Promega) and incubation at 37 °C for 16 h. Peptides were desalted and concentrated using an OASIS HLB SPE column (Waters, 1 cc) as per the manufacturer's instructions. The peptide mixture was diluted with 20 mm KH₂PO₄, 20% acetonitrile, pH 3 (SCX buffer A), and separated using a PolyLC polysulfoethylA column (2.1 × 100 mm) pre-equilibrated in SCX buffer A while connected to an Agilent 1200 HPLC system. After collection of unbound peptides, retained peptides were fractionated by an increasing gradient (0–50% B in 50 min) of SCX buffer A + 0.5 m KCl (SCX buffer B). Eluting peptides were monitored at 214 nm and collected in a 96-well plate by peak detection mode. Peptide fractions were lyophilized to ∼10 μl by rotary evaporation and desalted using OMIX C18 SPE pipette tips (Varian) per the manufacturer's instructions.

To extract M. hyopneumoniae aqueous phase proteins from a SDS-PAGE gel, gel slices were destained by washing twice with 50% acetonitrile/50 mm NH₄HCO₃ pH 9 for 10 min with vortexing. Washes were discarded, and the gel slices were dehydrated with 100% acetonitrile for 10 min and rehydrated in 30 μl of NH₄HCO₃ pH 9 containing 12.5 ng/μl of trypsin (Promega). After incubating at 40 °C for 30 min, 50 μl of NH₄HCO₃ pH 9 were added, and the samples incubated overnight at 37 °C. The digest solution was removed to a new tube and 50 μl of 50% acetonitrile, 2% formic acid added, and incubated for 10 min in a sonicating water bath at full power. The solution was removed and pooled with the digest solution and lyophilized to 15 μl and transferred to an autosampler vial.

Peptide fractions from the M. hyopneumoniae proteome and gel-extracted M. hyopneumoniae proteins were analyzed with MS/MS. Using an Eksigent AS-1 autosampler connected to a Tempo nanoLC system (Eksigent), 10 μl of sample were loaded at 20 μl/min with MS buffer A onto a C8-reversed phase trap column (Michrom). The peptides were washed off the trap at 300 nl/min onto a New Objective IntegraFrit column (75 μm × 100 mm) packed with ProteoPep II C18 resin. A gradient program consisting of the following phases was used to separate peptides: 5–50% MS buffer B (98% acetonitrile + 0.2% formic acid) over 105 min, 50–80% MS buffer B over 5 min, 80% MS buffer B for 2 min, 80–5% MS buffer B for 3 min. The eluting peptides flowed into a MicroIonSpray II-mounted 75 μm ID emitter tip that tapered to 15 μm, ionized by nanoelectrospray with 2,300 V into the source of the QSTAR, which then performed an Information Dependent Acquisition (IDA) experiment. The MS/MS data files produced by the QSTAR were merged using Mascot Daemon (version 2.2.2, provided by the Australian Proteomics Computational Facility, Perkins, D.N., 1999) and searched against the LudwigNR data base (comprised of the UniProt, plasmoDB, and Ensembl databases (vQ209. 8,785,680 sequences, 3,087,386,706 residues) with the following parameter settings. Fixed modifications: none. Variable modifications: propionamide, oxidized methionine. Enzyme: trypsin. Number of allowed missed cleavages: 3. Peptide mass tolerance: 100 ppm. MS/MS mass tolerance: 0.2 Da. Charge state: 2+ and 3+. The results of the search were then filtered by including only protein hits with at least one unique peptide (bold red, see supplemental Table S2) and excluding peptide hits with a p value greater than 0.05. False positive rate was determined by performing a decoy data base search concurrently where each time a protein sequence from the target data base is tested, a random sequence of the same length and average amino acid composition is automatically generated and tested. Peptides identified by Mascot as belonging to P116 were further validated by manual inspection of the MS/MS spectra for the peptide to ensure the b- and y-ion series were sufficiently extensive for an accurate identification (supplemental Tables S2–S5).

Porcine Plasminogen Purification

Plasminogen was isolated from porcine plasma as described previously (27, 28). MALDI Q-TOF tandem MS, performed according to Henrich et al. (26), confirmed the purified protein as plasminogen. The Asp-plasminogen fraction was further purified by gel filtration with Sephacryl S-200 Superfine (Amersham Biosciences), followed by ultrafiltration to concentrate the protein. Plasminogen activity and absence of plasmin contamination was confirmed using Spectrozyme PL (American Diagnostica). 1 μg of plasminogen was incubated at 37 °C with and without 100 ng of human tPA (Calbiochem, Germany) in 50 mm Tris (pH 7.4) for 15 min. Spectrozyme PL (20 μl of 2.5 mm) was added, and absorbance at 405 nm was measured every 5 min for 1.5 h at 37 °C. Biotinylation of plasminogen for whole cell binding assays was performed using Sulfo-NHS-Biotin (sulfosuccinimidobiotin) (Pierce) according to the manufacturer's instructions. To remove excess biotin the biotinylated plasminogen was dialyzed against PBS at 4 °C.

Fibronectin and Plasminogen Whole Cell Binding Assay

Microtiter plate binding assays were performed using 96-well format plates (Linbro/Titertek; MP Biomedicals Inc.). 125 ml of freshly cultured M. hyopneumoniae cells were harvested, washed extensively with PBS, and then fixed in 1% paraformaldehyde-PBS. Cells were diluted in carbonate-coating buffer (18 mm NaHCO₃, 27 mm Na₂CO₃; pH 9.5) and aliquoted into microtiter plate wells at an A₆₀₀ of 0.04. Microtiter plates were then centrifuged at 2,000 rpm for 10 min using a CS-6 benchtop centrifuge (Beckman Coulter). For fibronectin whole cell binding assays, cells were incubated with varying concentrations (0–50 μg/ml) of fibronectin from human plasma (Calbiochem) for 1 h at room temperature. Bound fibronectin was then detected with a 1:3,000 dilution of anti-fibronectin followed by 1:1,500 HRP-labeled sheep anti-rabbit conjugate, both for 1 h at room temperature. Negative control wells received no cells, no fibronectin, no anti-fibronectin or no anti-rabbit conjugate. Blocking, washing and color detection of wells were performed as previously described (12). Optical density readings were taken at 7, 15, 25, and 45 min at 414 nm using a Multiskan Ascent plate reader (Thermo Labsystems). Background reactivity in wells without cells or with no fibronectin added was negligible and has been subtracted from the curve. The plasminogen whole cell microtiter plate binding assays were performed using the protocol described above with the following modifications. The cells were blocked for 1 h at room temperature with 2% BSA in PBS. 2-fold serial dilution of biotinylated porcine plasminogen diluted in 1% BSA in PBS was added (100 μl per well) and incubated at room temperature for 1.5 h. Bound plasminogen was detected with streptavidin peroxidise diluted 1:3,000 in 1% BSA in PBS at room temperature for 1 h. Optical density reading was taken after 10 min using an xMark Microplate Absorbance Spectrophotometer (Bio-Rad).

Surface Plasmon Resonance Analyses

Binding to fibronectin and plasminogen was examined using a Biacore T100 (Biacore AB, Sweden) at 20 °C. The running buffer used for immobilization and kinetics assays was 10 mm HEPES, 150 mm NaCl, 3 mm EDTA, 0.05% v/v Surfactant P20 (HBS-EP⁺; Biacore AB); recombinant P116 fragments were transferred into HBS-EP⁺using Vivaspin 500 columns with a 10 kDa MWCO (Sartorius). Fibronectin or porcine plasminogen was immobilized onto a CM4 series S sensor chip via primary amino acids using an amine coupling kit (Biacore AB). The immobilized levels were ∼2,900 RU of fibronectin and ∼500 RU of plasminogen. A blank immobilized CM4 cell was used as a reference cell. F1_P116-F4_P116 binding kinetics were analyzed in separate experiments. Kinetic assays were performed by injecting recombinant P116 protein fragments at varying concentrations (0–200 nm) for 600 s with a flow rate of 10 μl/min. Dissociation occurred over a 1200 s period. Regeneration was undertaken with 6 m urea in HBS-EP⁺ for 45 s at 15 μl/min for immobilized fibronectin or 2 × 15 s injections of 4 m MgCl₂ at 15 μl/min for immobilized plasminogen. Analysis of the plasminogen binding data used the 1:1 Langmuir binding model in BiaEvaluation software 3.1 (Biacore AB, Sweden). Data for binding of F4_P116 to fibronectin were analyzed manually using a two-component heterogeneous surface model. One component accounting for ∼40% of the total response showed relatively fast on and off rates with k_on showing a linear dependence on [F4_P116], while the other, showed slow on and off rates independent of [F4_P116]; we chose to ignore this second nonspecific component and determined K_D from the ratio of k_d and k_a for the specific binding component.

Microsphere Preparation and PK15 Cell Culture

Fluoresbrite® polychromatic red polystyrene microspheres with a diameter of 0.5 μm were obtained from Polysciences. 2.73 × 10¹⁰ microspheres were pelleted via centrifugation at 16,000 × g for 8 min and washed twice with distilled water. Microspheres were resuspended in 150 μg of protein (F1_P116-F4_P116 or BSA) in 0.5 ml ≤ 0.05% SDS in PBS, with pH one unit greater than the pI of the protein. Proteins were incubated with the microspheres overnight at room temperature with gentle rocking. Microspheres were pelleted and resuspended in 0.5 ml of 10 mg/ml BSA in PBS and left to block for 1 h with gentle rocking. Microspheres were washed twice with 1 ml of filtered PBS, then resuspended in 1 ml of 10 mg/ml BSA in PBS containing 5% glycerol and stored at 4 °C for a maximum of 1 month. To check the coating efficiency of each protein, 2.5 μl of the microsphere suspensions were dotted onto nitrocellulose and allowed to dry. The membrane was probed by Western blotting using polyclonal rabbit antiserum raised against the coating protein. The adherence assay of protein coated microspheres to PK15 cells was then performed according to Burnett et al. (14) with the following modifications. PK15 cells were grown in 5% fetal calf serum with cells seeded at 5 × 10⁴ on 13 mm coverslips. 3.4 × 10⁸ protein-coated microspheres were incubated with PK15 cells in 2% fetal calf serum for 1, 2, 4, and 7 h. Following fixing of washed cells, PK15 cells were stained with a 1:20 dilution of modified Giemsa stain (Sigma-Aldrich).

Cilia Binding Assay

Cilia binding was performed with recombinant P116 proteins, as described by Wilton et al. (13) based on the assay developed by Zhang et al. (29). Cilium adhesin protein P97 (F2_P97) (12) was used as the positive control, and the non cilia-binding N-terminal 64-kDa fragment of P120 (a M. hyopneumoniae protein encoded by mhp107), another P97 paralog, was used as a negative control. Briefly, microtiter plates were coated with cilia isolated from swine trachea, blocked and incubated with recombinant proteins. Plates were washed and bound proteins were detected via hydrolysis of p-nitrophenyl phosphate following incubation with polyclonal antisera and a secondary antibody.

RESULTS

Proteomic and Molecular Analyses of P116

P116 is a 116 kDa P102 paralog encoded by mhp108. Adams et al. (17) reported that mhp108 is transcribed in vivo. Analysis using the SignalP 3.0 Server suggests a 0.993 probability of a signal peptide between amino acids 1–33, whereas TMpred identifies the region corresponding to amino acids 17–36 as a transmembrane domain (score 2,285) (Fig. 1, A and B). Bioinformatic analysis indicates that P116 is transported to the cell surface via a leader sequence with a transmembrane sequence embedded in the cell membrane.

FIGURE 1.

Proteomic identification of P116. A, schematic representation of P116 of M. hyopneumoniae, a 116 kDa protein. The unfilled region depicts the P116 transmembrane and signal peptide sequence. Shown are the amino acid positions of the recombinant P116 protein fragments constructed for this study. The molecular mass of the recombinant protein fragments includes a 3.8 kDa polyhistidine region encoded by the expression vector. B, P116 expression detected using mass spectrometry. Peptide mass fingerprint of M. hyopneumoniae strain 232 identified the underlined amino acids, which are contained within F3_P116. ESI-MS/MS mass spectrometry of the M. hyopneumoniae proteome identified the amino acids shown in bold, while ESI-MS/MS mass spectrometry of M. hyopneumoniae proteins extracted from a SDS-PAGE gel identified the amino acids shown in bold italics. Amino acids in italics and underlined italics indicate the predicted signal peptide sequence and transmembrane sequence respectively. C, Coomassie-stained 12% SDS-PAGE gel of purified P116 fragments; F1_P116 (lane 1), F2_P116 (lane 2), F3_P116 (lane 3), and F4_P116 (lane 4). D, immunoblot analyses of M. hyopneumoniae strain 232 whole cell lysates. Blots were probed with affinity purified anti-F1_P116 (lane 1), anti-F2_P116 (lane 2), anti-F3_P116 (lane 3), and anti-F4_P116 (lane 4) sera. E and F, P116 is cell surface located. M. hyopneumoniae strain 232 was incubated with 0 (lane 1), 5 (lane 2), 10 (lane 3), 50 (lane 4), 150 (lane 5), or 300 (lane 6) μg/ml trypsin prior to preparation as whole cell lysates. Whole cell lysates were transferred from 12% SDS-PAGE gels to nitrocellulose for immunoblotting. Blots were probed with pooled affinity-purified anti-F1_P116, anti-F2_P116, anti-F3_P116, and anti-F4_P116 sera (E) or anti-L7/L12, a cytosolic ribosomal protein (F). G, diagrammatic representation of a theoretical native cleavage pattern of P116. The unfilled region depicts the P116 transmembrane and signal peptide sequence. Evidence from this study to support the hypothesized P116 cleavage fragments is indicated by a +. MALDI refers to peptide mass fingerprinting results from two-dimensional gel electrophoresis, antisera refers to fragments detected in immunoblot analyses of M. hyopneumoniae whole cell lysates, and MS/MS refers to peptides identified during MS/MS analysis of M. hyopneumoniae proteins extracted from a SDS-PAGE gel. Relative positions of recombinant P116 fragments are depicted, amino acid positions are indicated at the bottom of the figure and shown to scale.

Peptide mass fingerprinting of M. hyopneumoniae strain 232 proteins separated by 2DGE identified a region of ∼108 amino acids (Fig. 1B) within the recombinant F3_P116 fragment of P116 (Fig. 1A), with an observed pI of ∼6.3 and an observed mass of 17 kDa. This 17 kDa peptide was also identified during ESI-MS/MS analysis of M. hyopneumoniae strain 232 proteins separated using SDS-PAGE, along with peptides that indicate the presence of two distinct C-terminal P116 fragments with observed sizes within the ranges of 45–52 kDa and 52–68 kDa. Four additional peptides of P116 were detected by ESI-MS/MS analysis of the M. hyopneumoniae proteome. The tryptic peptides ranged in size from 13–23 amino acids and spanned P116. The presence of these peptides establishes that P116 is expressed in vitro, while the peptides detected from a SDS-PAGE gel and 2DGE indicates P116 is proteolytically processed and provides insight into the processing pattern. Four adjoining polyhistidine-tagged fragments of P116 (Fig. 1A) were expressed and purified (Fig. 1C), when SDS-PAGE of these fragments was performed abnormal migration was displayed. Polyclonal rabbit antisera were generated against the recombinant P116 fragments and affinity purified. Immunoblots of whole cell lysates (Fig. 1D) detected a high molecular weight protein ∼120 kDa in size, equivalent in size to full length P116, which was recognized by F1_P116 and F2_P116 antisera. A 70 kDa C-terminal protein fragment was detected by F3_P116 and F4_P116 antisera. An additional 47 kDa protein was recognized only by the C-terminal F4_P116 antiserum. An N-terminal protein fragment of 55 kDa was recognized by F1_P116 and F2_P116 antisera along with a 39 kDa protein fragment recognized only by F1_P116 antiserum (Fig. 1D). Treatment of M. hyopneumoniae cells with concentrations of trypsin above 150 μg/ml resulted in digestion of fragments recognized by P116 antisera, demonstrating surface location (Fig. 1E). The cytosolic L7/L12 protein remained intact (Fig. 1F) (14) and P116 fragments were not recognized by pre-immune sera (supplemental Fig. S6). In conjunction with mass spectrometry analysis, these observations suggest that P116 is a surface located proteolytically processed protein. Based upon observations from peptide mass fingerprinting analysis, ESI-MS/MS analysis, and immunoblots of whole cell M. hyopneumoniae lysates, a model describing proteolytic processing of P116 is hypothesized (Fig. 1G).

F4_P116 Binds Fibronectin

Previous studies describe Mycoplasma species that utilize fibronectin as a ligand for adherence (30–32); fibronectin receptors of M. hyopneumoniae have not previously been identified. M. hyopneumoniae binds soluble fibronectin in a dose-dependent and saturable manner (Fig. 2A) and preliminary investigation using ligand blots suggested fibronectin binds P116 fragments (data not shown). Analysis of the real-time interaction of P116 fragments with immobilized fibronectin using SPR was carried out over a range of concentrations (Fig. 2B). F4_P116 clearly bound specifically to fibronectin in a dose-dependent manner with a physiologically relevant K_D estimated (as described under “Experimental Procedures”) to be 24 ± 6 nm; k_a = (1.3 ± 0.2) × 10⁵ m⁻¹ s⁻¹. F2_P116, and F3_P116 showed only weak, nonspecific binding, while no response was detected with F1_P116 (data not shown).

FIGURE 2.

F4_P116 is a fibronectin-binding protein. A, binding of fibronectin by M. hyopneumoniae. M. hyopneumoniae cells were coated onto microtiter plate wells and increasing concentrations of fibronectin were added until saturation was reached. The data represent 8 replicates from three separate experiments. B, F4_P116 binds immobilized fibronectin. Surface plasmon resonance sensorgram of F4_P116. Protein was injected over fibronectin at concentrations of 10, 50, 100, and 200 nm for 600 s at 10 μl/min. The arrow indicates the end of protein injection, at which stage dissociation of bound protein from the immobilized fibronectin can be observed.

C-terminal Lysine of P116 Facilitates Binding to Porcine Plasminogen

The presence of a C-terminal lysine residue in proteins of other bacterial species have been observed to mediate plasminogen binding ability (33, 34). The C-terminal lysine in P116, isolation of M. hyopneumoniae in deep tissues (35) and previous studies describing the interaction of the broad spectrum protease plasminogen with Mycoplasma species (36, 37) led us to investigate the capacity of M. hyopneumoniae and P116 to bind plasminogen. M. hyopneumoniae was found to bind porcine plasminogen in a saturable and dose-dependent manner (Fig. 3A). SPR analysis was therefore employed to investigate interactions between P116 fragments and porcine plasminogen (Fig. 3, B and C). The C-terminal fragment of P116 (F4_P116) bound to plasminogen with a K_D of 44 ± 5 nm and k_a of (9.2 ± 1.4) × 10⁵ m⁻¹ s⁻¹ (Fig. 3B). The C-terminal amino acid of F4_P116 is a lysine residue. Site-directed mutagenesis was performed to delete the C-terminal lysine residue of F4_P116. SPR analysis indicates that deletion of the C-terminal lysine residue from F4_P116 abolished plasminogen binding ability (Fig. 3C).

FIGURE 3.

F4_P116 C-terminal lysine promotes binding to immobilized porcine plasminogen. A, binding of porcine plasminogen by M. hyopneumoniae. M. hyopneumoniae cells were coated onto microtiter plate wells and increasing concentrations of plasminogen were added. The data represent three replicates from three separate experiments. B and C, surface plasmon resonance sensorgrams of F4_P116 injected over porcine plasminogen. B, F4_P116 was injected at concentrations of 10, 50, 100, and 200 nm over plasminogen with a flow rate of 10 μl/min for 600 s. The arrow indicates the end of the injection at which stage any dissociation of F4_P116 from immobilized plasminogen can be observed. C, association of 200 nm F4_P116 full-length (top curve) and F4_P116 with deletion of the C-terminal lysine (bottom curve). Proteins were injected with a flow rate of 10 μl/min for 600 s for comparison of binding ability.

Adherence of P116 Fragments to PK15 Cells and Swine Cilia

M. hyopneumoniae was shown by Zielinski et al. (38) to adhere to PK15 cells. Given that PK15 cells have been used as a model for M. hyopneumoniae binding to eukaryotic cells (13, 14), we explored the interaction of P116 fragments with PK15 cells. Adherence to PK15 cells was determined using fluorescent microspheres coated with purified P116 fragments. Microspheres coated with F1_P116 were able to significantly adhere to PK15 cells after 2 h, while F3_P116-coated microspheres required a 7 h incubation to adhere to PK15 cells (Fig. 4).

FIGURE 4.

P116 fragments promote PK15 cell adhesion. A–D, confocal microscopy images of protein coated fluorescent microspheres. Bound microspheres coated with BSA (A) or F1_P116 (C) after a 4-h incubation period with PK15 cells, arrows indicate microspheres. Corresponding transmission images of PK15 cells, to which microspheres coated with BSA (B) and F1_P116 (D) are bound. Scale bars, 50 μm. E, graphical summary of protein-coated microspheres bound to PK15 cells following different incubation periods. * denotes results differing significantly to the BSA negative control when a one-way ANOVA is performed.

During disease M. hyopneumoniae adheres specifically and directly to respiratory cilia (1). An in vitro assay of M. hyopneumoniae binding to cilia was developed by Zhang et al. (29) to investigate potential receptors of the bacterium. A similar assay has been used to determine the cilial-binding region of P97 (39) and demonstrate that P216 is a cilium adhesin (13). As this assay has previously been successfully employed to identify cilium adhesins, it was utilized to investigate the ability of P116 fragments to bind purified swine respiratory cilia. The cilium adhesin (F2_P97) was used as a positive control and a N-terminal 64-kDa fragment of P120 as an example of a protein unable to bind cilia. Each of the P116 fragments were found to significantly adhere to porcine respiratory cilia (Fig. 5).

FIGURE 5.

P116 fragments bind swine respiratory cilia. P116 fragments were incubated with swine cilia in a microtiter plate based binding assay. The M. hyopneumoniae cilial adhesion protein F2_P97 is a positive control cilial binding protein, while the P120 N-terminal protein lacks the capability to bind swine cilia. Data represent the response above background plus the upper bound of the 95% confidence interval. Absorbance is measured at 405 nm. * denotes results differing significantly from the background.

DISCUSSION

During disease, the M. hyopneumoniae protein P102 is expressed, whereupon it is processed and secreted from the cell (16, 17). Beyond these findings, the role of P102 and P102 paralogs has not previously been elucidated. In this study we examined the surface expression and proteolytic processing of the P102 paralog P116 in M. hyopneumoniae strain 232, corroborating previous reports of the transcription of mhp108 in vivo (17). P116 is extensively processed with multiple P116 fragments observed. Extensive proteolytic processing has been observed in other M. hyopneumoniae adhesins, including P97 (16). The advantage of adhesin processing is undetermined, although it may represent a mechanism for modification of these proteins to generate functional domains (13, 14). The C-terminal region of P116 binds fibronectin and plasminogen; to our knowledge, this is the first report identifying the capacity of a M. hyopneumoniae protein to bind these proteins. Fibronectin is a key component of the host extracellular matrix and fibronectin-binding proteins have been described in many other bacterial pathogens including Mycoplasma pneumoniae (40), Mycoplasma gallisepticum (30), and Mycoplasma penetrans (41). In this study we report M. hyopneumoniae as a bacterial pathogen capable of binding fibronectin. The ability of bacteria to bind fibronectin can provide a mechanism for host cell adhesion or cytoskeletal rearrangements, which may facilitate bacterial internalization (41–45). SPR analysis demonstrated that F4_P116-bound fibronectin in a dose-dependent manner with a physiologically relevant K_D value. Comparison of the F4_P116 amino acid sequences with known fibronectin binding domains did not reveal a conserved fibronectin binding motif, suggesting that F4_P116 contains a novel fibronectin binding motif. The fibronectin binding ability of M. hyopneumoniae may mediate the adherence of the bacterium to swine respiratory cilia, once epithelial cell damage has occurred.

P116 contains a C-terminal lysine residue which mediates binding to porcine plasminogen. Plasminogen is a proenzyme found in plasma and extracellular fluids, which is converted to the broad spectrum serine protease plasmin upon activation (46). Active plasmin is capable of degrading fibronectin, fibrin, vitronectin, and laminin, with considerable evidence for the ability of plasmin to facilitate extracellular matrix degradation and invasion (34). In this study, plasminogen binding was observed for M. hyopneumoniae. Plasminogen binding ability has been reported widely among bacterial pathogens, including Mycoplasma fermentans and Mycoplasma sp. bovine group 7 (34, 36, 37). P116 is the first M. hyopneumoniae plasminogen-binding protein to be described. The dissociation constant of the C-terminal fragment F4_P116 for porcine plasminogen is 44 ± 5 nm, which is a physiologically relevant value. The capacity of M. hyopneumoniae to capture host plasminogen at the bacterial surface may facilitate the penetration of this pathogen into host organs via the circulatory system, with M. hyopneumoniae recently being isolated from the liver, kidneys and spleen of swine (35).

P97 and the P97 paralog P216 are known to bind swine respiratory cilia. Additionally, adherence to the porcine epithelial-like PK15 cell line is promoted by P159 and P216 (13, 14, 39). P116 is the first P102 paralog member demonstrated to exhibit the capacity to bind PK15 cells and swine respiratory cilia. Intimate association between M. hyopneumoniae and swine cilia have been observed in vivo (1, 17). The observations that P97 and P102 paralogs mediate bacterial adherence strongly suggest a central and unified role for these protein families in the process of colonization of the host respiratory tract. The differential binding of P116 fragments to fibronectin, plasminogen, PK15 cells, and swine respiratory cilia suggests that this protein contains multiple functional domains with divergent roles. Multiple multifunctional proteins have been identified as a central theme in bacterial pathogenesis. NhhA is a surface-located multi-functional protein expressed by Neisseria meningitides: NhhA is able to bind human epithelial cells, laminin, and heparin sulfate (47). The M protein virulence determinant of Streptococcus pyogenes is known to mediate binding to a range of extracellular matrix proteins and serum proteins including plasminogen (48), while α-enolase, a virulence factor of Streptococcus suis mediates binding to fibronectin, plasminogen and endothelial cells (49). Collectively, our findings suggest that P116 is a multifunctional virulence determinant involved in the M. hyopneumoniae disease process.