Interactions of surface-displayed glycolytic enzymes of Mycoplasma pneumoniae with components of the human extracellular matrix

G Model

ARTICLE IN PRESS

IJMM-51080; No. of Pages 11

International Journal of Medical Microbiology xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

International Journal of Medical Microbiology journal homepage: www.elsevier.com/locate/ijmm

Interactions of surface-displayed glycolytic enzymes of Mycoplasma pneumoniae with components of the human extracellular matrix Anne Gründel, Enno Jacobs, Roger Dumke ∗ Technische Universität Dresden, Medizinische Fakultät Carl Gustav Carus, Institut für Medizinische Mikrobiologie und Hygiene, Fetscherstrasse 74, 01307 Dresden, Germany

a r t i c l e

i n f o

Article history: Received 2 June 2016 Received in revised form 30 August 2016 Accepted 1 September 2016 Keywords: Mycoplasma pneumoniae Host-pathogen interaction Surface-displayed glycolytic enzyme Human extracellular matrix protein

a b s t r a c t Mycoplasma pneumoniae is a major cause of community-acquired respiratory infections worldwide. Due to the strongly reduced genome, the number of virulence factors expressed by this cell wall-less pathogen is limited. To further understand the processes during host colonization, we investigated the interactions of the previously confirmed surface-located glycolytic enzymes of M. pneumoniae (pyruvate dehydrogenase A-C [PdhA-C], glyceraldehyde-3-phosphate dehydrogenase [GapA], lactate dehydrogenase [Ldh], phosphoglycerate mutase [Pgm], pyruvate kinase [Pyk] and transketolase [Tkt]) to the human extracellular matrix (ECM) proteins fibrinogen (Fn), fibronectin (Fc), lactoferrin (Lf), laminin (Ln) and vitronectin (Vc), respectively. Concentration-dependent interactions between Fn and Vc and all eight recombinant proteins derived from glycolytic enzymes, between Ln and PdhB-C, GapA, Ldh, Pgm, Pyk and Tkt, between Lf and PdhA-C, GapA and Pyk, and between Fc and PdhC and GapA were demonstrated. In most cases, these associations are significantly influenced by ionic forces and by polyclonal sera against recombinant proteins. In immunoblotting, the complex of human plasminogen, activator (tissue-type or urokinase plasminogen activator) and glycolytic enzyme was not able to degrade Fc, Lf and Ln, respectively. In contrast, degradation of Vc was confirmed in the presence of all eight enzymes tested. Our data suggest that the multifaceted associations of surface-localized glycolytic enzymes play a potential role in the adhesion and invasion processes during infection of human respiratory mucosa by M. pneumoniae. © 2016 Elsevier GmbH. All rights reserved.

1. Introduction Mycoplasma pneumoniae is a common cause of a broad spectrum of infections of the human respiratory tract ranging from mild, often undiagnosed forms of tracheobronchitis to severe interstitial pneumonia. Infections due to M. pneumoniae are transmitted by aerosols and have been described in all age groups but older children and young adults are more frequently affected (Atkinson et al., 2008). Besides small-scale endemic transmission in populations with close person-to-person contact, nation- or even world-wide incidence peaks every 5–7 years have been reported, in which up to 40% of all cases of community-acquired pneumonia are attributed to this pathogen (Dumke et al., 2015). Extra-pulmonary manifestations, mainly of the central nervous system and of the skin, further complicate the clinical signs of infections (Narita, 2016). Members of the class Mollicutes are cell wall-less bacteria characterized by a strong reduction of genetic resources during

∗ Corresponding author. E-mail address: [email protected] (R. Dumke).

evolutionary interaction with their hosts. In consequence, M. pneumoniae is not only limited in its metabolic capabilities (Kühner et al., 2009), requiring uptake of many metabolites from the environment, but also in the repertoire of factors determining pathogenicity and the clinical symptoms of infected humans as the only known natural host. To date, the complex adhesion apparatus of the bacteria (Hasselbring et al., 2006), the expression of the CARDS toxin (Kannan and Baseman, 2006) and the release of cell-damaging substances such as hydrogen peroxide (Hames et al., 2009) and hydrogen sulfide (Großhennig et al., 2016) are confirmed or potential virulence factors. First described in streptococci (Pancholi and Fischetti, 1992), bacterial proteins with a primary function in cytosol-localized metabolic processes can be transported by an unknown mechanism to the surface of the cells. These proteins also include glycolytic enzymes essential for glycolysis. Surface-displayed glycolytic enzymes have now been demonstrated in many different bacterial species (Henderson and Martin, 2011; Wang et al., 2013; Kainulainen and Korhonen, 2014) indicating a common and multifaceted mechanism among prokaryotes. In addition to the typical representatives of this class of proteins, such as enolase (Eno) and

http://dx.doi.org/10.1016/j.ijmm.2016.09.001 1438-4221/© 2016 Elsevier GmbH. All rights reserved.

Please cite this article in press as: Gründel, A., et al., Interactions of surface-displayed glycolytic enzymes of Mycoplasma pneumoniae with components of the human extracellular matrix. Int. J. Med. Microbiol. (2016), http://dx.doi.org/10.1016/j.ijmm.2016.09.001

G Model

ARTICLE IN PRESS

IJMM-51080; No. of Pages 11

A. Gründel et al. / International Journal of Medical Microbiology xxx (2016) xxx–xxx

2

glyceraldehyde-3-phosphate dehydrogenase (GapA), the number of further enzymes with a dual role in the cytosol and at the cell surface has increased in recent years (Henderson, 2014). Because of interaction(s) with host factors, mainly with components of the extracellular matrix (ECM), it has been suggested that surfacedisplayed proteins increase the efficiency of micro-organisms in colonizing the host. Especially in mycoplasmas, multi-function of proteins might be a way to compensate for the limited genetic resources. In this context, further investigation of moonlighting proteins is also important for a general understanding of the organisation of the successful parasitic lifestyle of many members of the Mollicutes group. Regarding the pathogenic species M. pneumoniae, the characterization of glycolytic enzymes that associate with human factors is an aspect of host-pathogen interaction that also may contribute to pathogenesis. Recently, we confirmed the surface localization of the following eight glycolytic enzymes of M. pneumoniae: GapA, lactate dehydrogenase (Ldh), phosphoglycerate mutase (Pgm), pyruvate kinase (Pyk), pyruvate dehydrogenase A to C (PdhA-C) and transketolase (TkT) and demonstrated that all these proteins bind to human plasminogen (Plg; Gründel et al., 2016), an important factor of the fibrinolysis system (Sanderson-Smith et al., 2012). In addition, proteolytically active plasmin was generated in the presence of all glycolytic enzymes, Plg and the urokinase plasminogen activator (uPA). However, only the complexes between Plg and the enzymes PdhB as well as Pgm are able to induce the in vitro degradation of human fibrinogen (Fn). With the recombinant proteins derived from genes coding for the confirmed surface-localized enzymes and their corresponding monospecific antisera, experimental tools to investigate possible interactions with other factors of the human ECM are now available. These host factors include Fn, vitronectin (Vc), fibronectin (Fc), lactoferrin (Lf), laminin (Ln), collagen and elastin which are regarded as important targets of bacterial proteins to promote the ability of microorganisms to escape the host response, to adhere effectively and/or to invade tissues (Singh et al., 2010, 2012). The complex of ECM proteins comprises diverse components mediating and influencing many essential processes in the host. For example, via binding to different factors multidomain Fn and Fc are involved in matrix physiology, cell migration and cell signalling (Henderson et al., 2011; Yamaguchi et al., 2013; Halper and Kjaer, 2014) whereas Vc is a complement regulator and promotes many processes such as cell adhesion and differentiation (Singh et al., 2010; Preissner and Reuning, 2011). Laminins play an important role in structuring the ECM matrix by interactions with other proteins not least to protect tissues from pathogens (Chagnot et al., 2012; Singh et al., 2012). The iron-binding glycoprotein Lf contributes to iron homeostasis in the host and is a protective component of the innate immune system (Gonzalez-Chavez et al.,

2009). Binding of Lf to bacterial proteins is discussed as important for iron supply of micro-organisms (Morgenthau et al., 2013). Binding between individual glycolytic enzymes of M. pneumoniae and selected host components such as Fc or Fn has been reported previously (Dallo et al., 2002; Dumke et al., 2011). Nevertheless, overall knowledge of the interactions of surface-localized glycolytic enzymes with ECM factors is far from well-documented and not only for M. pneumoniae. In this study, we investigated these associations to better understand the network of interactions between host ECM factors and a class of primary metabolic bacterial proteins in a common pathogen. 2. Materials and methods 2.1. Bacteria and human cells The M. pneumoniae strains M129 (ATCC 29342), FH (ATCC 15531), the Escherichia coli strain BL21(DE3) (Novagen, Darmstadt, Germany) and human A549 cells (human lung carcinoma cell line ATCC CCL-185) were grown as described previously (Schurwanz et al., 2009). Protein concentration was measured using the BCA protein assay kit (Pierce, Rockford, USA) as recommended by the manufacturer. 2.2. Recombinant production of surface-localized glycolytic enzymes of M. pneumoniae and polyclonal antibodies The E. coli strains containing the pET vectors with the cloned full-length genes coding for the glycolytic enzymes of M. pneumoniae (Table 1) were produced as described (Gründel et al., 2016). Expression of the N-terminal 6 x His-tagged proteins was induced with 1 mM isopropylthiogalactoside (Sigma, St. Louis, USA) for 5 h at 37 ◦ C. The proteins were purified under denaturing conditions by using Ni-nitrilotriacetic acid columns (Qiagen, Hilden, Germany) as recommended by the manufacturer. The eluates were concentrated with ultrafiltration spin columns (Vivascience, Hannover, Germany) and the protein concentration was measured as described. The quality of the expression results was checked after separation of recombinant proteins by SDS-PAGE, using Coomassie blue staining and standard immunoblotting procedures as reported (Gründel et al., 2016). Polyclonal antisera were produced in guinea pigs (Charles River, Sulzfeld, Germany). The animal experiments were conducted in accordance with the recommendations of the Federation of Laboratory Animal Science Associations (FELASA) and approved by the ethical board of Landesdirektion Sachsen, Dresden, Germany (permit no. 24-9168.25-1/2011/1). Primary subcutaneous immunization of guinea pigs with total proteins of M. pneumoniae (positive control serum) or the recombinant proteins, booster

Table 1 Summary of binding characteristics of surface-located glycolytic enzymes of M. pneumoniae to selected components of human ECM. Glycolytic enzyme ECM component Lactoferrin a

PdhA PdhB PdhC GapA Pgm Pyk Tkt Ldh a b c

Laminin b

c

Vitronectin

Fibrinogen

Fibronectin

Binding

NaCl

Anti-serum

Binding NaCl Anti-serum Binding NaCl Anti-serum Binding NaCl Anti-serum Binding NaCl Anti-serum

± + + + − + − −

+ + + +

± ± − ±

+

±

− + ± + + + + +

+ + + + + + +

± ± + + + + ±

± + + + + + + +

± + − ± − ± ± ±

+ + ± + + + ± +

± ± ± + ± + + +

− − ± ± − ± ± −

+ + ± + + + + ±

− − ± + − − − −

− ±

+ +

Strength of binding (+: OD values after comparative binding of highest concentration of ECM component >0.4; ±: OD values ≥ 0.2 − ≤ 0.4; -: OD values < 0.2). Influence of ionic interactions on binding (+: reduction of 50–100%; ±: 10–50%; -: <10% in comparison with negative control). Influence of antiserum to glycolytic enzyme on binding (+: reduction of 50–100%; ±: 10 − ≤ 50%; -: <10% in comparison with negative control).

Please cite this article in press as: Gründel, A., et al., Interactions of surface-displayed glycolytic enzymes of Mycoplasma pneumoniae with components of the human extracellular matrix. Int. J. Med. Microbiol. (2016), http://dx.doi.org/10.1016/j.ijmm.2016.09.001

G Model IJMM-51080; No. of Pages 11

ARTICLE IN PRESS A. Gründel et al. / International Journal of Medical Microbiology xxx (2016) xxx–xxx

immunizations and serum collection were done as described (Thomas et al., 2013).

2.3. Binding of recombinant proteins to components of the human ECM and influence of ionic interactions and polyclonal antisera on the interaction with human A549 cells The interaction of M. pneumoniae whole cells and of glycolytic enzymes with components of human ECM was analysed as described previously (Gründel et al., 2016). To test the general ability of M. pneumoniae whole cells to interact with different host ECM proteins, binding of increasing concentrations (0, 5, 25, 50 ␮g/ml each) of human elastin (polymer, Millipore, Billerica, USA), GluPlg, collagen type 1, Lf, Ln (trimer), plasma-Fc, Fn (hexamer) and plasma-Vc (mixture of 65 and 75 kDa polypeptides) (Sigma) to mycoplasma cells was investigated in ELISA experiments. Freshly harvested M. pneumoniae whole cells (100 ␮g/ml) were used to coat cavities of 96-well ELISA plates (Greiner, Frickenhausen, Germany) for 2 h at 37 ◦ C. After washing and blocking, ECM proteins were added and incubated for 2 h at 37 ◦ C. Bound proteins were detected by specific polyclonal antisera (Sigma; anti-elastin 1:500; anticollagen 1:500; anti-Lf 1:2000; anti-Ln 1:750; anti-Vc 1:1500; anti-Fn 1:1000; and anti-Fc 1:1000, respectively) and peroxidaselabelled anti-rabbit IgG or anti-goat IgG (Dako, Hamburg, Germany; 1:1000 each). The substrate (tetramethylbenzidine; Sigma) was added, the reaction was stopped with 1 M HCl and absorbance was measured at 450 nm with a reference wavelength of 620 nm. Recombinant proteins were bound to human A549 cells as reported recently (Gründel et al., 2016). Briefly, freshly grown cells were harvested, the protein concentration was adjusted (100 ␮g/ml) and the cell suspension was used to coat the wells of an ELISA plate for 2 h at 37 ◦ C. In parallel, wells were coated with recombinant proteins (10 ␮g/ml) as controls for maximum binding of antigens. As positive and negative controls we used whole M. pneumoniae cells, recombinant proteins rP12 (C-terminal part of the main P1 adhesin protein which has been confirmed as involved in cytoadherence; Schurwanz et al., 2009) and rP8 (part of the P1 protein which is not able to bind to human cells; Schurwanz et al., 2009). After blocking, the proteins (10 ␮g/ml) were added to the cavities containing human cells for 2 h at 37 ◦ C. During this incubation, wells coated with recombinant proteins were treated in parallel with PBS. Bound proteins were detected using polyclonal antisera to whole M. pneumoniae antigen, PdhA-C, Pyk, Pgm, Ldh, Tkt, GapA, rP8 and rP12 (each 1:750) and anti-guinea pig IgG (Dako; 1:1000). Viability of A549 cells in the presence of recombinant proteins was confirmed microscopically after staining with fluorescein diacetate (Sigma; 5 mg/ml) and propidium iodide (Sigma; 1 mg/ml). To test the specificity of the binding, freshly harvested A549 cells (100 ␮g/ml) were immobilized in wells of ELISA plates for 2 h at 37 ◦ C. In parallel, recombinant proteins were pre-incubated with corresponding guinea pig antisera or a pre-immune serum as control (1:50 each) for 1.5 h at room temperature. After washing and blocking, the mixture was added to the cavities of the plate. Bound proteins were detected by a rabbit antiserum to the Triton X100soluble fraction of M. pneumoniae total proteins and anti-rabbit IgG (1:750). The presence of ECM proteins on the surface of A549 cells as potential binding partners of the glycolytic enzymes of M. pneumoniae was tested in ELISA experiments. For this, cells (100 ␮g/ml) were used to coat cavities of a 96-well plate for 2 h at 37 ◦ C. After washing and blocking, antisera to the different ECM proteins (antiLf 1:2000; anti-Ln 1:750; anti-Vc 1:2000; anti-Fn 1:1000; and anti-Fc 1:1000, respectively) were added and incubated for 1.5 h at 37 ◦ C. Detection was performed using anti-goat IgG or anti-rabbit IgG (both 1:1000).

3

To confirm the specific interaction of glycolytic enzymes of M. pneumoniae with human ECM proteins, the wells of 96-well plates were incubated with recombinant proteins and BSA as negative control (each 15 ␮g/ml) overnight at 4 ◦ C. Preliminary experiments were carried out to determine the concentration range of ECM proteins resulting in evaluable results. Different concentrations of Lf (0.5, 1.0, 2.5 ␮g/ml PBS), Ln (5.0, 10.0, 25.00 ␮g/ml), Fn (1.0, 5.0, 10.0 ␮g/ml), Fc (1.0, 5.0, 10.0 ␮g/ml) and Vc (0.5, 1.0, 2.0 ␮g/ml) and PBS were added for 2 h at 37 ◦ C. Bound ECM proteins were detected using the corresponding antisera (anti-Lf 1:2000; anti-Ln 1:750; anti-Vc 1:1500; anti-Fn 1:1000; anti-Fc 1:1000) and anti-rabbit IgG or anti-goat IgG (both 1:1000). Dissociation constants (KD ) for binding affinity of surface-displayed glycolytic enzymes to ECM proteins were determined as described (Lin et al., 2009; Fernandes et al., 2012). Briefly, ECM proteins (10 ␮g/ml) were immobilized in ELISA plate wells, blocked and incubated with increasing concentrations (c) of recombinant protein (0, 0.1, 0.25, 0.5, 1, 2, 3, 4, 5 ␮M). Bound proteins were detected with the corresponding guinea pig antisera (1:1000) and HRP-conjugated anti-guinea pig IgG (1:1000). Using the ELISA data (means of three replicates per concentration) the KD values were calculated based on the equation OD450 = ODmax *[cProtein ]/KD + [cProtein ]. Furthermore, the influence of ionic interactions and of the corresponding monospecific antisera to surface-exposed glycolytic enzymes of M. pneumoniae on the binding of recombinant proteins to host factors was analysed. Confirmed protein–protein interactions (OD values > 0.2) were tested. The 96-well plates were coated with recombinant proteins as described above. The different ECM components (Lf: 2.5 ␮g/ml, Ln: 25.0 ␮g/ml, Fc: 10.0 ␮g/ml, Fn: 10.0 ␮g/ml, Vc: 2 ␮g/ml) were diluted in PBS with high concentrations of either NaBr or NaCl (both 650 mM; Roth, Karlsruhe, Germany), added to the cavities with recombinant proteins and incubated for 2 h at 37 ◦ C. As controls, wells coated with recombinant proteins were incubated with the same concentrations of ECM factors in PBS without NaBr or NaCl. Bound ECM proteins were detected as described above. To study the influence of the corresponding antisera to glycolytic enzymes on binding to ECM components, we immobilized recombinant proteins for 2 h at 37 ◦ C as described. In parallel, the ECM proteins (in the same concentration as used for analysis of ionic interactions) were pre-incubated with antisera and guinea pig pre-immune serum as control (1:50 each) for 1.5 h at room temperature with overhead rotation. The suspension was incubated with the recombinant proteins (2 h, 37 ◦ C) and the bound ECM components were detected as described above.

2.4. Degradation of human ECM proteins in the presence of Plg and surface-localized glycolytic enzymes of M. pneumoniae The proteolytic potential of generated plasmin to cleave ECM components was tested in an assay as reported recently (Koenigs et al., 2013). Briefly, recombinant proteins (10 ␮g/ml) and BSA as negative control (10 ␮g/ml) were immobilized in wells of an ELISA plate. After washing and blocking, proteins were incubated with human Plg (10 ␮g/ml) for 2 h at 37 ◦ C. Subsequently, different ECM components (Lf, Ln, Vc, Fc; 10 ␮g/ml each) with tissue (-type) plasminogen activator (tPA) or uPA (4 ng/well; both Millipore) were added and plates were stored at room temperature. Samples were taken after different time points, protein sample buffer was added and the reaction was stopped by heating at 96 ◦ C for 10 min. Total proteins in the samples were separated by SDS-PAGE and blotted on a nitrocellulose membrane as reported previously (Gründel et al., 2016). Different ECM proteins were detected with the specific primary antibodies (same dilutions as in binding experiments) and HRP-conjugated secondary antibodies (1:2000).

Please cite this article in press as: Gründel, A., et al., Interactions of surface-displayed glycolytic enzymes of Mycoplasma pneumoniae with components of the human extracellular matrix. Int. J. Med. Microbiol. (2016), http://dx.doi.org/10.1016/j.ijmm.2016.09.001

G Model IJMM-51080; No. of Pages 11

ARTICLE IN PRESS A. Gründel et al. / International Journal of Medical Microbiology xxx (2016) xxx–xxx

4

Fig. 1. Concentration-dependent binding of human extracellular matrix proteins to M. pneumoniae. Mycoplasma cells (100 ␮g/ml) were immobilized in wells of ELISA plates and incubated with ECM proteins. Bound proteins were detected with rabbit antibodies to human proteins and HRP-conjugated anti-rabbit IgG secondary antibodies. Mean and standard deviation of eight parallels.

3. Results 3.1. Binding of human ECM components to M. pneumoniae cells From the broad range of host factors that might be interaction partners for surface-localized M. pneumoniae-specific proteins, pre-testing of freshly grown mycoplasma cells to bind selected components of the human ECM was used to narrow down potential associations. Concentration-dependent binding of Fn, Fc, Ln, Lf, Vc and Plg was confirmed for type 1 strain M129 (Fig. 1). In addition, the binding rates of type 2 strain FH to these host proteins demonstrated no significant differences (data not shown). In contrast, interaction of M. pneumoniae cells to human collagen and elastin was found very low. These proteins were not further tested in subsequent experiments. Compared to Plg, Fc, Lf and Vc, detection of low concentrations (5 ␮g/ml) of bound Fn and Ln was difficult. However, addition of 50 ␮g/ml host protein resulted in all cases in strong binding (OD values of ≥ 0.85) and in relatively slight differences between the reactivity of the different ECM components with mycoplasma cells. 3.2. Binding of surface-displayed glycolytic enzymes of M. pneumoniae to human lung epithelial cells and to components of the human ECM Recombinant proteins derived from eight glycolytic enzymes with surface-exposed regions were tested for adherence to human A549 cells. Wells coated with recombinant protein and incubated with PBS were compared with wells coated with A549 cells pre-incubated with the same recombinant protein under comparable conditions. Binding was confirmed with the corresponding anti-recombinant protein serum. Viability of human cells during incubation was verified by fluorescence microscopy (data not shown). Using protein rP12 (C-terminal part of the main P1 adhesin) as positive control, nearly the same OD values were obtained for wells with immobilized recombinant protein in comparison with cell-coated wells and addition of recombinant proteins, confirming strong interaction of rP12 with A549 cells (Fig. 2A). As further positive control, binding of M. pneumoniae whole cells to human cells resulted in an OD value of 0.45. In contrast, cell-binding of recombinant protein rP8 as part of P1 protein not involved in adhesion remains low (OD value: 0.14). Using the surface-located glycolytic enzymes of M. pneumoniae, the reactivities of recombinant proteins bound to human cells are reduced to OD values between 25% (PdhC) and 76% (GapA) compared to those

of wells coated with recombinant proteins. Nevertheless, with the exception of PdhC, OD values of glycolytic enzymes bound to cells reached at least 0.5 confirming an interaction with surface components of A549 cells. Analyzing the specificity of these interactions, pre-incubation of recombinant proteins with corresponding guinea pig antisera resulted in all cases in a significant reduction of binding to human cells in comparison with a pre-immune serum (Fig. 2B). However, the quantitative influence of antibodies on interactions was variable, ranging from 24% (Ldh) to 96% (PdhC). To obtain preliminary information regarding possible interaction partners on the host site, A549 cells were analysed by ELISA for the presence of selected ECM proteins (Fig. 2C). The results demonstrated high OD values for human Vc and Plg whereas the other ECM components showed low reactivity (0.23–0.35). To prove the specific interaction between surface-located glycolytic enzymes of M. pneumoniae and the particular human ECM components, binding studies were carried out by ELISA (Fig. 3). In all experiments, the negative control BSA was not able to bind increasing concentrations of human Lf, Ln, Vc, Fn, and Fc, respectively (OD values ≤ 0.1). Regarding the glycolytic enzymes, a complex binding pattern was found. This included a range of concentrations of ECM proteins from up to 2.5 mg/ml (Lf and Vc) to 25 mg/ml (Ln) which are needed to demonstrate an increasing interaction. Furthermore, protein-specific binding of the different host and bacterial partners was confirmed. For example, the concentration-dependent interaction of all glycolytic enzymes of M. pneumoniae with human Vc was measured (Fig. 3C). In contrast, definite binding to Fc (OD ≥ 0.2) was demonstrated for GapA and, with significantly lower intensity, for PdhC. The highest binding affinity of human Lf was confirmed after incubation with PdhC, PdhB, Pyk, and GapA (Fig. 3A), of human Ln with GapA, PdhB, Ldh, Pyk, and Tkt (Fig. 3B), and of human Fn with GapA, Pyk, Tkt, and Ldh (Fig. 3D), respectively. As indicated in Table 2, KD values for confirmed interactions of glycolytic enzymes with Lf ranged from 81.5 (PdhA) to 2,361.3 nM (PdhC), with Ln from 104.7 (Pyk) to 1,328.1 nM (PdhC), with Vc from 64.6 (Pyk) to 1,358.4 nM (PdhC), with Fn from 106.2 (Pyk) to 1,106.7 nM (PdhC), and with Fc from 122.5 (GapA) to 2,174.7 nM (PdhC), respectively. 3.3. Influence of ionic interactions and of polyclonal antisera on binding of glycolytic enzymes of M. pneumoniae to human ECM proteins The importance of ionic interactions for the binding of recombinant proteins with different components of ECM was analysed by adding 650 mM NaCl or 650 mM NaBr to host factors before incubation with recombinant proteins. Furthermore, the influence of corresponding guinea pig antiserum to the glycolytic enzyme on this association was tested by pre-treatment of ECM proteins with antibodies. Experiments were limited to confirmed interactions between proteins (OD values ≥ 0.2, Fig. 3). When both salts were added, binding of Lf to glycolytic enzymes was significantly reduced to ≤ 20% of the mean OD values of controls using PBS (set to 100%), demonstrating the strong influence of ionic forces on all interactions with this ECM protein (Fig. 4A). Pre-incubation of Lf with guinea pig antibodies to glycolytic enzymes resulted in a mean decrease of bound Lf to PdhA, PdhB, GapA and Pyk between 44 and 28%. Interestingly, the strong interaction of PdhC with Lf was significantly reduced by ionic forces but not by the anti-PdhC serum. Interaction of Ln with recombinant proteins (Fig. 4B) was also weakened by the influence of both salts to rates between 67% (NaBr and PdhB) and 99% (NaBr and PdhA). In the presence of the specific antiserum, bound Ln to all glycolytic enzymes decreased significantly with rates between 12% (Pyk) and 82% (PdhB), respectively. The interaction of Vc with most recombinant proteins was reduced by adding both salts (reduction: 23–66%) but with greater

Please cite this article in press as: Gründel, A., et al., Interactions of surface-displayed glycolytic enzymes of Mycoplasma pneumoniae with components of the human extracellular matrix. Int. J. Med. Microbiol. (2016), http://dx.doi.org/10.1016/j.ijmm.2016.09.001

G Model

ARTICLE IN PRESS

IJMM-51080; No. of Pages 11

A. Gründel et al. / International Journal of Medical Microbiology xxx (2016) xxx–xxx

5

Fig. 2. Comparative binding of surface-displayed glycolytic enzymes of M. pneumoniae to human lung epithelial cells. (A) Freshly grown A549 cells were immobilized in wells of ELISA plates and incubated with recombinant surface-displayed enzymes. In parallel, wells were coated with the same concentration of recombinant proteins and incubated with PBS. The reactivity of guinea pig antisera to glycolytic enzymes with immobilized and with cell-bound recombinant proteins was measured. Whole M. pneumoniae cells (M129), recombinant proteins P12 (C-terminal part of main P1 adhesin with confirmed association with human cells) and P8 (part of P1 protein not involved in adhesion) were used as positive and negative controls, respectively. Bars represent means and standard deviations of eight parallels from a single experiment representative of three independent experiments. (B) Influence of preincubation of surface-localized glycolytic enzymes of M. pneumoniae with their corresponding antisera on binding to immobilized A549 cells. Parallel incubation of each recombinant protein with a guinea pig pre-immune serum (PIS) was used as control. Means and standard deviations of eight replicates from a single experiment representative of two independent experiments. t-test: **P ≤ 0.01, ***P ≤ 0.001. (C) Reactivity of immobilized A549 cells with antisera to human ECM proteins with confirmed binding to M. pneumoniae cells. Bars represent means and standard deviations of eight replicates from a single experiment representative of three independent experiments.

differences between glycolytic enzymes (Fig. 4C). The binding of Vc to PdhC in the presence of NaCl and NaBr was not different from the control PBS and interaction with Pgm was reduced by NaCl only. Adding the specific antiserum resulted in significant inhibition of binding with rates between 43% (PdhC) and 86% (Pyk). Interestingly, the interaction of Fn with recombinant proteins is not influenced by NaCl and significant decreases of association in the presence of NaBr was found only in three (GapA, Pyk, and Tkt) of eight cases (Fig. 4D). In contrast, pre-incubation of ECM proteins with specific antibodies to glycolytic enzymes resulted in a reduc-

tion of bound host component in the range of 16% (Tkt) to 54% (PdhC) of the OD values with PBS. Finally, Fc-binding to all five recombinant proteins was reduced in the presence of NaCl and to three glycolytic enzymes after pre-incubation with NaBr (Fig. 4E). Binding to the bacterial proteins decreased significantly to between 12% (GapA) and 63% (PdhB) after pre-treatment with specific antiserum. Taking the results together, binding of most human ECM proteins to glycolytic enzymes of M. pneumoniae was attributed to ionic interactions. Exceptions, like binding of Fn in general or PdhC

Table 2 Binding affinities (KD values ± standard deviation, nM) of recombinant surface-displayed glycolytic enzymes of M. pneumoniae to human ECM proteins. Glycolytic enzyme

PdhA PdhB PdhC GapA Pgm Pyk Tkt Ldh a

ECM component Lactoferrin

Laminin

Vitronectin

Fibrinogen

Fibronectin

81.5 ± 50.0 206.0 ± 71.8 2,361.3 ± 1,194.3 174.4 ± 112.6 n.b. 223.8 ± 117.0 n.b. n.b.

n.b.a 181.8 ± 89.9 1,328.1 ± 542.9 238.6 ± 123.2 199.5 ± 59.3 104.7 ± 68.4 230.8 ± 156.5 654.8 ± 297.8

145.0 ± 63.9 124.3 ± 46.3 1,358.4 ± 770.3 136.9 ± 114.1 89.9 ± 37.8 64.6 ± 17.7 114.4 ± 49.0 3,137.8 ± 827.8

120.3 ± 59.4 168.8 ± 79.4 1,106.7 ± 509.4 135.5 ± 31.4 218.6 ± 158.2 106.2 ± 69.7 140.2 ± 72.5 607.7 ± 720.7

n.b. n.b. 2,174.7 ± 680.6 122.5 ± 65.1 n.b. n.b. n.b. n.b.

n.b. − no binding (see Table 1).

Please cite this article in press as: Gründel, A., et al., Interactions of surface-displayed glycolytic enzymes of Mycoplasma pneumoniae with components of the human extracellular matrix. Int. J. Med. Microbiol. (2016), http://dx.doi.org/10.1016/j.ijmm.2016.09.001

G Model IJMM-51080; No. of Pages 11 6

ARTICLE IN PRESS A. Gründel et al. / International Journal of Medical Microbiology xxx (2016) xxx–xxx

Fig. 3. Concentration-dependent binding of human ECM proteins to immobilized surface-displayed glycolytic enzymes of M. pneumoniae. Bound ECM proteins were detected with rabbit antisera to lactoferrin (A), laminin (B), vitronectin (C), fibrinogen (D) and fibronectin (E), respectively. BSA was used as negative control in all experiments. Bars represent means and standard deviations of eight replicates from a single experiment representative of three independent experiments.

to Vc, show that further molecular mechanisms for these interactions can be suggested. After pre-incubation of host proteins with monospecific polyclonal antisera, a quantitatively different but in nearly all cases significant reduction of binding was found, confirming the specificity of the interaction and the occurrence of antigenic regions of glycolytic enzymes that play a role in association with the different host factors.

3.4. Degradation of human ECM proteins in the presence of Plg and glycolytic enzymes of M. pneumoniae The degradation of human Fn by the protease plasmin generated after activation of Plg with uPA in the presence of PdhB and Pgm of M. pneumoniae was reported previously (Gründel et al., 2016). Furthermore, after addition of tPA, plasmin generation increased in

Please cite this article in press as: Gründel, A., et al., Interactions of surface-displayed glycolytic enzymes of Mycoplasma pneumoniae with components of the human extracellular matrix. Int. J. Med. Microbiol. (2016), http://dx.doi.org/10.1016/j.ijmm.2016.09.001

G Model IJMM-51080; No. of Pages 11

ARTICLE IN PRESS A. Gründel et al. / International Journal of Medical Microbiology xxx (2016) xxx–xxx

7

Fig. 4. Influence of ionic interactions and of polyclonal guinea pig antisera on binding of surface-displayed glycolytic enzymes of M. pneumoniae to human ECM components. ECM proteins were preincubated with PBS, NaCl, NaBr or antisera to glycolytic enzymes and added to immobilized recombinant proteins. Bound ECM proteins were detected with rabbit antisera to lactoferrin (A), laminin (B), vitronectin (C), fibrinogen (D) and fibronectin (E), respectively. Binding of recombinant proteins in wells incubated with ECM proteins and PBS was considered to be 100%. Means and standard deviations of eight replicates from a single experiment representative of three independent experiments. t-test: *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001.

Please cite this article in press as: Gründel, A., et al., Interactions of surface-displayed glycolytic enzymes of Mycoplasma pneumoniae with components of the human extracellular matrix. Int. J. Med. Microbiol. (2016), http://dx.doi.org/10.1016/j.ijmm.2016.09.001

G Model IJMM-51080; No. of Pages 11

ARTICLE IN PRESS A. Gründel et al. / International Journal of Medical Microbiology xxx (2016) xxx–xxx

8

Fig. 5. Degradation of human vitronectin (Vc) in the presence of surface-localized glycolytic enzymes of M. pneumoniae bound to human plasminogen (Plg). (A) Plg was activated by uPA. Samples were taken after 0 and 24 h, proteins were separated by SDS PAGE and blotted. Vc was detected by rabbit antiserum. Experiments with BSA were used as negative controls. (B) Time-dependent degradation of Vc by Pgm- and PdhB-bound Plg activated with the urokinase-type plasminogen activator (uPA) or tissue plasminogen activator (tPA).

the presence of all M. pneumoniae-specific proteins in comparison with the results of parallel experiments using uPA and tPA demonstrated a 4-fold higher mean catalytic activity (data not shown). To investigate the influence of activated Plg on further components of the ECM, Plg bound to the eight surface-located glycolytic enzymes was incubated with Lf, Ln, Vc, and Fc together with the activators tPA or uPA, respectively. The reaction was stopped at different time points and total proteins were analysed by SDS-PAGE and immunoblotting with antisera to ECM factors. Degradation of Lf, Ln and Fc within 24 h was not found (data not shown). In contrast, the 75 kDa protein part of the Vc monomer disappeared in the presence of all recombinant proteins and uPA (Fig. 5A ) as well as tPA (data not shown). The 65 and 75 kDa bands of Vc were detected at the start (0 h) of each experiment, in the presence of immobilized BSA throughout the incubation time, and in controls lacking Plg. As examples, time-dependent analysis of degradation of Vc in the presence of recombinant proteins Pgm and PdhB of M. pneumoniae is summarized in Fig. 5B. With both proteins, degradation of the 75 kDa part of Vc was confirmed after an incubation time of 8 h without major differences between uPA and tPA used as Plg activators. 4. Discussion Associations of surface-localized bacterial proteins with host factors have been described in many species and seem to be a common aspect of host-pathogen interactions (Henderson, 2014). These bacterial components include not only metabolic enzymes but also proteins with other functions such as chaperones or elongation factors (Kainulainen and Korhonen, 2014). However, microbial enzymes involved in the cytosol-localized process of glycolysis are the best-investigated class of these moonlighting proteins. In recent years, surface-displayed glycolytic enzymes have also been reported in different mycoplasma species with clinical importance for animal and human health (Dallo et al., 2002; Hoelzle et al., 2007; Chen et al., 2011; Dumke et al., 2011; Schreiner et al., 2012; Song et al., 2012; Thomas et al., 2013; Bao et al., 2014, 2015; He et al., 2015; Gründel et al., 2016). With genome sizes between 801 and 858 Mbp (http://www.ncbi.nlm.nih.gov/ genome), M. pneumoniae is one of the smallest self-replicating bacteria with pathogenic potential for humans, resulting in a greatly

reduced repertoire of metabolic proteins and virulence factors. Colonization of hosts by M. pneumoniae is limited to the respiratory epithelium and pathogenesis is initiated by adhesion of bacteria to target cells via a highly specialized adherence complex (Hasselbring et al., 2006). After release of substances such as hydrogen peroxide (Hames et al., 2009), hydrogen sulfide (Großhennig et al., 2016) and the CARDS toxin (Kannan and Baseman, 2006), damage to or at least disturbance of the integrity of the epithelial layer of the respiratory mucosa of infected persons can be suggested. Knowledge about further processes that explain the long-term carriage of M. pneumoniae (Spuesens et al., 2013) is limited despite a strong immune response of the host and appropriate antibiotic treatment. Interactions of microbial proteins with different host factors such as components of the human ECM or Plg have been reported as contributing to adhesion, immune evasion and tissue invasion (Wang et al., 2013). Based on the data from M. pneumoniae, surface-displayed glycolytic enzymes seem to be involved in these processes generally. In addition, especially in mycoplasmas, this dual function of proteins can be a way to use the limited genomic resources effectively. Previous reports confirmed the localization of selected glycolytic enzymes such as PdhB (Dallo et al., 2002) and GapA (Dumke et al., 2011) on the surface of the M. pneumoniae cell. In a recent paper, the complete set of 19 glycolytic enzymes of M. pneumoniae was investigated and eight enzymes (GapA, Ldh, Pgm, Pyk, Tkt and PdhA to C) were found to be surface-displayed (Gründel et al., 2016). In contrast to other bacterial species (Nogueira et al., 2013; Götz et al., 2015), none of the proteins was detected in supernatants of liquid cultures of M. pneumoniae indicating that interactions with host factors are linked to intact surface structures of the bacteria. All surface-localized glycolytic enzymes of M. pneumoniae were able to bind human Plg suggesting the presence of corresponding binding sites in many bacterial proteins (Gründel et al., 2016). The central role of Plg associated with and activated in the presence of bacterial factors for degradation of host components such as Fn has been well-investigated (Bhattacharya et al., 2012; Sanderson-Smith et al., 2012). Regarding M. pneumoniae it has been demonstrated that human Fn was degraded after activation of Plg to the protease plasmin in the presence of the enzymes Pgm and PdhB, (Gründel et al., 2016). In contrast to human Plg, less is known about the interactions of surface-localized glycolytic enzymes with components of the ECM.

Please cite this article in press as: Gründel, A., et al., Interactions of surface-displayed glycolytic enzymes of Mycoplasma pneumoniae with components of the human extracellular matrix. Int. J. Med. Microbiol. (2016), http://dx.doi.org/10.1016/j.ijmm.2016.09.001

G Model IJMM-51080; No. of Pages 11

ARTICLE IN PRESS A. Gründel et al. / International Journal of Medical Microbiology xxx (2016) xxx–xxx

Fig. 6. Simplified overview of confirmed interactions of surface-displayed glycolytic enzymes of M. pneumoniae with human plasminogen (Plg) and ECM proteins lactoferrin (Lf), laminin (Ln), vitronectin (Vc), fibrinogen (Fn) and fibronectin (Fc), respectively. Solid lines: in vitro binding; dotted lines: activation/degradation; thin lines: known interactions (Dumke et al., 2011; Gründel et al., 2016); thick lines: interactions described in the present study. Suggested role of interactions in colonization processes was summarized according to data described in Morgenthau et al. (2013), Sanderson-Smith et al. (2012), Singh et al. (2010), Singh et al. (2012) and Verhamme et al. (2015), respectively.

Investigation of human A549 cells in the present study confirmed the expression of different factors suggesting possible associations with mycoplasmal proteins. Basically, ECM is a complex of proteins secreted by endothelial and epithelial cells not only involved in structural support but also in aspects of signalling and maintenance of the integrity of tissues (Hynes, 2009). Among many other components, the factors Fc, Fn, Ln, Vc and Lf investigated here contribute to processes such as functional scaffolding, adhesion, migration, differentiation of the cell network and iron homeostasis (Chagnot et al., 2012; Singh et al., 2010, 2012; Morgenthau et al., 2013; Halper and Kjaer, 2014). Proteins of different classes of many pathogens are able to bind these host factors as described for Fc (Henderson et al., 2011; Yamaguchi et al., 2013) and Vc (Kohler et al., 2015; Hallström et al., 2015, 2016) suggested to play a role in colonization and/or immune evasion. Interaction of surfacelocalized glycolytic enzymes such as GapA, Pgm or Tkt with Fc, Vc and Ln was demonstrated in Candida species (Gozalbo et al., 1998; Lopez et al., 2014; Kozik et al., 2015). Among bacteria, binding to different ECM proteins was mainly reported for GapA and Eno (Kainulainen and Korhonen, 2014). Recently, interaction of Pgm with Fc and of Ldh with Ln and Fc in Streptococcus suis (Zhang et al., 2014; Li et al., 2015), and of GapA and PdhB with Fc as well as with Fc in Lactobacillus plantarum (Glenting et al., 2013; Vastano et al., 2014) have been found. Regarding M. pneumoniae, binding of GapA to Fn (Dumke et al., 2011) and of PdhB to Fc (Dallo et al., 2002) was demonstrated. Furthermore, Tryon and Baseman (1987) confirmed the acquisition of Lf by M. pneumoniae without identification of specific bacterial proteins as binding partners. In addition, Eno (not surface-localized in M. pneumoniae; Thomas et al., 2013) was found as interaction partner in different bacterial species including Mycoplasma synoviae (Bao et al., 2014). The results of the present study have extended the spectrum of binding between glycolytic enzymes and human ECM components broadly (Fig. 6). Of the eight proteins investigated, at least one (to Fc) up to seven enzymes (to Vc) showed strong interaction(s) in ELISA experiments (Table 1). As expected, differences in affinity between the defined partners

9

were observed (Table 2), making conclusions difficult. However, a strong association with at least two ECM factors was confirmed for most of the bacterial proteins. GapA, PdhB and Pyk in particular can be characterized as multi-binding enzymes showing distinct interactions with three to all host factors analyzed. The results of this study support the findings of previous reports confirming the occurrence of different microbial factors that are able to interact with a defined host protein as well the ability of a glycolytic enzyme to bind different host factors (Henderson and Martin, 2011; Copley, 2012; Henderson, 2014). The tested ECM components can be found on the surface of A549 cells to which M. pneumoniae cells are able to bind but with lower affinity in comparison to many recombinant proteins tested. Nearly all surface-displayed glycolytic enzymes (except PdhC) interact with the human cell line A549, indicating their involvement in the adhesion process. Further studies are needed regarding the anchoring of surfacelocalized proteins in the mycoplasma membrane and for deeper analysis of protein regions involved in association with host factors. This is of special importance for M. pneumoniae proteins that show strong binding affinity for different host components such as GapA and PdhB. In comparison with BSA, pre-incubation of Plg with recombinant proteins to mask binding sites demonstrated a complex picture of influences on subsequent binding of the pre-treated Plg to the different glycolytic enzymes (Gründel et al., 2016) requiring further investigations for generalized conclusions. Two studies narrowed the regions of PdhB and elongation factor Tu involved in association with Plg (Thomas et al., 2013) and Fc (Balasubramanian et al., 2008) resulting in the characterization of small as well as multiple regions responsible for binding. The confirmed in vitro interactions of surface-localized glycolytic enzymes of M. pneumoniae with components of the human ECM are a precondition for a possible influence on in vivo processes such as adhesion. However, further aspects with importance for infection, such as degradation of host factors in the presence of microbial proteins bound to Plg, are crucial for pathogenesis. Beside the primary function of activated Plg in hosts, like degradation of fibrin clots and activation of metalloproteinases, activation by bacterial factors can support the processes of pathogen invasion (Verhamme et al., 2015). In addition to Fn, it has been documented that activated Plg bound by bacteria can degrade Ln, Fc and other ECM proteins (Lähteenmäki et al., 2000; Vieira et al., 2009). In contrast, the degradation of Ln, Fc and Lf was not confirmed in the present study. The role of different factors complicates the performance of experimental investigations. As an example, the confirmed influence of plasminogen activator on plasmin production was also found in a previous report (Fulde et al., 2014). It is further important to note that A549 cells are able to activate human Plg without addition of an activator and the expression of uPA on the surface of the cells can be confirmed by using an anti-PLAU antibody (data not shown). Here, Plg activation in the presence of nearly all tested glycolytic enzymes of M. pneumoniae (except PdhC) was not reduced in comparison to the BSA control and the degradation of human Vc was confirmed with both activators. Despite the fact that plasmin is important for the degradation of Vc (Gechtman et al., 1997), to our knowledge we have been the first to demonstrate Vc degradation by plasmin in the presence of bacterial proteins. In conclusion, the study demonstrated not only multiple associations of surface-located glycolytic enzymes of M. pneumoniae with different components of the human ECM but also the influence of bound enzymes on Plg activation and degradation of Vc. It can be suggested that this network of interactions plays a role in the complex processes that occur during the colonization of human respiratory tract by M. pneumoniae.

Please cite this article in press as: Gründel, A., et al., Interactions of surface-displayed glycolytic enzymes of Mycoplasma pneumoniae with components of the human extracellular matrix. Int. J. Med. Microbiol. (2016), http://dx.doi.org/10.1016/j.ijmm.2016.09.001

G Model IJMM-51080; No. of Pages 11 10

ARTICLE IN PRESS A. Gründel et al. / International Journal of Medical Microbiology xxx (2016) xxx–xxx

Acknowledgement The study was supported by a grant from the Deutsche Forschungsgemeinschaft (DU 1280/1-1). The authors thank S. Gäbler for technical assistance. References Atkinson, T.P., Balish, M.F., Waites, K.B., 2008. Epidemiology, clinical manifestations, pathogenesis and laboratory detection of Mycoplasma pneumoniae infections. FEMS Microbiol. Rev. 32, 956–973. Balasubramanian, S., Kannan, T.R., Baseman, J.B., 2008. The surface-exposed carboxyl region of Mycoplasma pneumoniae elongation factor Tu interacts with fibronectin. Infect. Immun. 76, 3116–3123. Bao, S., Guo, X., Yu, S., Ding, J., Tan, L., Zhang, F., Sun, Y., Qiu, X., Chen, G., Ding, C., 2014. Mycoplasma synoviae enolase is a plasminogen/fibronectin binding protein. BMC Vet. Res. 10, 223. Bao, S., Chen, D., Yu, S., Chen, H., Tan, L., Hu, M., Qiu, X., Song, C., Ding, C., 2015. Characterization of triosephosphate isomerase from Mycoplasma gallisepticum. FEMS Microbiol. Lett. 362, fnv140. Bhattacharya, S., Ploplis, V.A., Castellino, F.J., 2012. Bacterial plasminogen receptors utilize host plasminogen system for effective invasion and dissemination. J. Biomed. Biotechnol. 2012, 482096. Chagnot, C., Listrat, A., Astruc, T., Desvaux, M., 2012. Bacterial adhesion to animal tissues: protein determinants for recognition of extracellular matrix components. Cell. Microbiol. 14, 1687–1696. Chen, H., Yu, S., Shen, X., Chen, D., Qiu, X., Song, C., Ding, C., 2011. The Mycoplasma gallisepticum ␣-enolase is cell surface-exposed and mediates adherence by binding to chicken plasminogen. Microb. Pathog. 51, 285–290. Copley, S.D., 2012. Moonlighting is mainstream: paradigm adjustment required. Bioessays 34, 578–588. Dallo, S.F., Kannan, T.R., Blaylock, M.W., Baseman, J.B., 2002. Elongation factor Tu and E1 ␤ subunit of pyruvate dehydrogenase complex act as fibronectin binding proteins in Mycoplasma pneumoniae. Mol. Microbiol. 46, 1041–1051. Dumke, R., Hausner, M., Jacobs, E., 2011. Role of Mycoplasma pneumoniae glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in mediating interactions with the human extracellular matrix. Microbiology 157, 2328–2338. Dumke, R., Schnee, C., Pletz, M.W., Rupp, J., Jacobs, E., Sachse, K., Rohde, G., Capnetz study group, 2015. Mycoplasma pneumoniae and Chlamydia spp. infection in community-acquired pneumonia, Germany, 2011–2012. Emerg. Infect. Dis. 21, 426–434. Fernandes, L.G., Vieira, M.L., Kirchgatter, K., Alves, I.J., de Morais, Z.M., Vasconcellos, S.A., Romero, E.C., Nascimento, A.L., 2012. OmpL1 is an extracellular matrix- and plasminogen-interacting protein of Leptospira spp. Infect. Immun. 80, 3679–3692. Fulde, M., Bernardo-García, N., Rohde, M., Nachtigall, N., Frank, R., Preissner, K.T., Klett, J., Morreale, A., Chhatwal, G.S., Hermoso, J.A., Bergmann, S., 2014. Pneumococcal phosphoglycerate kinase interacts with plasminogen and its tissue activator. Thromb. Haemost. 111, 401–416. Götz, F., Yu, W., Dube, L., Prax, M., Ebner, P., 2015. Excretion of cytosolic proteins (ECP) in bacteria. Int. J. Med. Microbiol. 305, 230–237. Gechtman, Z., Belleli, A., Lechpammer, S., Shaltiel, S., 1997. The cluster of basic amino acids in vitronectin contributes to its binding of plasminogen activator inhibitor-1: evidence from thrombin-, elastase- and plasmin-cleaved vitronectins and anti-peptide antibodies. Biochem. J. 325, 339–349. Glenting, J., Beck, H.C., Vrang, A., Riemann, H., Ravn, P., Hansen, A.M., Antonsson, M., Ahrné, S., Israelsen, H., Madsen, S., 2013. Anchorless surface associated glycolytic enzymes from Lactobacillus plantarum 299 v bind to epithelial cells and extracellular matrix proteins. Microbiol. Res. 168, 245–253. Gonzalez-Chavez, S.A., Arévalo-Gallegos, S., Rascón-Cruz, Q., 2009. Lactoferrin: structure, function and applications. Int. J. Antimicrob. Agents 33 (301), e1–e8. Gozalbo, D., Gil-Navarro, I., Azorin, I., Renau-Piqueras, J., Martinez, J.P., Gil, M.L., 1998. The cell wall-associated glyceraldehyde-3-phosphate dehydrogenase of Candida albicans is also a fibrinogen and laminin binding protein. Infect. Immun. 66, 2052–2059. Gründel, A., Pfeiffer, M., Jacobs, E., Dumke, R., 2016. Network of surface-displayed glycolytic enzymes in Mycoplasma pneumoniae and their interactions with human plasminogen. Infect. Immun. 84, 666–676. Großhennig, S., Ischebeck, T., Gibhardt, J., Busse, J., Feussner, I., Stülke, J., 2016. Hydrogen sulfide is a novel potential virulence factor of Mycoplasma pneumoniae: characterization of the unusual cysteine desulfurase/desulfhydrase HapE. Mol. Microbiol. 100, 42–54. Hallström, T., Uhde, M., Singh, B., Skerka, C., Riesbeck, K., Zipfel, P.F., 2015. Pseudomonas aeruginosa uses dihydrolipoamide dehydrogenase (Lpd) to bind to the human terminal pathway regulators vitronectin and clusterin to inhibit terminal pathway complement attack. PLoS One 10, e0137630. Hallström, T., Singh, B., Kraiczy, P., Hammerschmidt, S., Skerka, C., Zipfel, P.F., Riesbeck, K., 2016. Conserved patterns of microbial immune escape: pathogenic microbes of diverse origin target the human terminal complement inhibitor vitronectin via a single common motif. PLoS One 11, e0147709. Halper, J., Kjaer, M., 2014. Basic components of connective tissues and extracellular matrix: elastin, fibrillin, fibulins, fibrinogen, fibronectin, laminin, tenascins and thrombospondins. Adv. Exp. Med. Biol. 802, 31–47.

Hames, C., Halbedel, S., Hoppert, M., Frey, J., Stülke, J., 2009. Glycerol metabolism is important for cytotoxicity of Mycoplasma pneumoniae. J. Bacteriol. 191, 747–753. Hasselbring, B.M., Jordan, J.L., Krause, R.W., Krause, D.C., 2006. 2006: Terminal organelle development in the cell wall-less bacterium Mycoplasma pneumoniae. Proc. Natl. Acad. Sci. USA 103, 16478–16483. He, S., Qi, J., Yu, S., Yin, Y., Tan, L., Bao, S., Qiu, X., Wang, X., Fei, R., Ding, C., 2015. Expression and immunological characteristics of the surface-localized pyruvate kinase in Mycoplasma gallisepticum. Microb. Pathog. 89, 161–168. Henderson, B., Martin, A., 2011. Bacterial virulence in the moonlight: multitasking bacterial moonlighting proteins are virulence determinants in infectious disease. Infect. Immun. 79, 3476–3491. Henderson, B., Nair, S., Pallas, J., Williams, M.A., 2011. Fibronectin: a multidomain host adhesin targeted by bacterial fibronectin-binding proteins. FEMS Microbiol. Rev. 35, 147–200. Henderson, B., 2014. An overview of protein moonlighting in bacterial infection. Biochem. Soc. Trans. 42, 1720–1727. Hoelzle, L.E., Hoelzle, K., Helbig, M., Aupperle, H., Schoon, H.A., Ritzmann, M., Heinritzi, K., Felder, K.M., Wittenbrink, M.M., 2007. MSG1, a surface-localised protein of Mycoplasma suis is involved in the adhesion to erythrocytes. Microbes Infect. 9, 466–474. Hynes, R.O., 2009. The extracellular matrix: not just pretty fibrils. Science 326, 1216–1219. Kühner, S., van Noort, V., Betts, M.J., Leo-Macias, A., Batisse, C., Rode, M., Yamada, ˜ D., Chen, W.H., Devos, T., Maier, T., Bader, S., Beltran-Alvarez, P., Castano-Diez, D., Güell, M., Norambuena, T., Racke, I., Rybin, V., Schmidt, A., Yus, E., Aebersold, R., Herrmann, R., Böttcher, B., Frangakis, A.S., Russell, R.B., Serrano, L., Bork, P., Gavin, A.C., 2009. Proteome organization in a genome-reduced bacterium. Science 326, 1235–1240. Kainulainen, V., Korhonen, T.K., 2014. Dancing to another tune −adhesive moonlighting proteins in bacteria. Biology 3, 178–204. Kannan, T.R., Baseman, J.B., 2006. ADP-ribosylating and vacuolating cytotoxin of Mycoplasma pneumoniae represents unique virulence determinant among bacterial pathogens. Proc. Natl. Acad. Sci. U. S. A. 103, 6724–6729. Koenigs, A., Hammerschmidt, C., Jutras, B.L., Pogoryelov, D., Barthel, D., Skerka, C., Kugelstadt, D., Wallich, R., Stevenson, B., Zipfel, P.F., Kraiczy, P., 2013. BBA70 of Borrelia burgdorferi is a novel plasminogen-binding protein. J. Biol. Chem. 288, 25229–12243. Kohler, S., Hallström, T., Singh, B., Riesbeck, K., Sparta, G., Zipfel, P.F., Hammerschmidt, S., 2015. Binding of vitronectin and Factor H to Hic contributes to immune evasion of Streptococcus pneumoniae serotype 3. Thromb. Haemost. 113, 125–142. Kozik, A., Karkowska-Kuleta, J., Zajac, D., Bochenska, O., Kedracka-Krok, S., Jankowska, U., Rapala-Kozik, M., 2015. Fibronectin-, vitronectin- and laminin-binding proteins at the cell walls of Candida parapsilosis and Candida tropicalis pathogenic yeasts. BMC Microbiol. 15, 197. Lähteenmäki, K., Kuusela, P., Korhonen, T.K., 2000. Plasminogen activation in degradation and penetration of extracellular matrices and basement membranes by invasive bacteria. Methods 21, 125–132. Li, Q., Liu, H., Du, D., Yu, Y., Ma, C., Jiao, F., Yao, H., Lu, C., Zhang, W., 2015. Identification of novel laminin- and fibronectin-binding proteins by far-western blot: capturing the adhesins of Streptococcus suis type 2. Front. Cell. Infect. Microbiol. 5, 82. Lin, Y.P., Lee, D.W., McDonough, S.P., Nicholson, L.K., Sharma, Y., Chang, Y.F., 2009. Repeated domains of Leptospira immunoglobulin-like proteins interact with elastin and tropoelastin. J. Biol. Chem. 284, 19380–19391. Lopez, C.M., Wallich, R., Riesbeck, K., Skerka, C., Zipfel, P.F., 2014. Candida albicans uses the surface protein Gpm1 to attach to human endothelial cells and to keratinocytes via the adhesive protein vitronectin. PLoS One 9, e90796. Morgenthau, A., Pogoutse, A., Adamiak, P., Moraes, T.F., Schryvers, A.B., 2013. Bacterial receptors for host transferrin and lactoferrin: molecular mechanisms and role in host-microbe interactions. Future Microbiol. 8, 1575–1585. Narita, M., 2016. Classification of extrapulmonary manifestations due to Mycoplasma pneumoniae infection on the basis of possible pathogenesis. Front. Microbiol. 7, 23. Nogueira, S.V., Backstedt, B.T., Smith, A.A., Qin, J., Wunder, E.A., Ko, A., Pal, U., 2013. Leptospira interrogans enolase is secreted extracellularly and interacts with plasminogen. PLoS One 8, e78150. Pancholi, V., Fischetti, V.A., 1992. A major surface protein on group A streptococci is a glyceraldehyde-3-phosphate dehydrogenase with multiple binding activity. J. Exp. Med. 176, 415–426. Preissner, K.T., Reuning, U., 2011. Vitronectin in vascular context: facets of a multitalented matricellular protein. Semin. Thromb. Hemost. 37, 408–424. Sanderson-Smith, M.L., De Oliveira, D.M.P., Ranson, M., McArthur, J.D., 2012. Bacterial plasminogen receptors: mediators of a multifaceted relationship. J. Biomed. Biotechnol. 2012, 272148. Schreiner, S.A., Sokoli, A., Felder, K.M., Wittenbrink, M.M., Schwarzenbach, S., Guhl, B., Hoelzle, K., Hoelzle, L.E., 2012. The surface-localized ␣-enolase of Mycoplasma suis is an adhesion protein. Vet. Microbiol. 156, 88–95. Schurwanz, N., Jacobs, E., Dumke, R., 2009. A strategy to create chimeric proteins derived from functional adhesin regions of Mycoplasma pneumoniae for vaccine development. Infect. Immun. 77, 5007–5015. Singh, B., Su, Y.C., Riesbeck, K., 2010. Vitronectin in bacterial pathogenesis: a host protein used in complement escape and cellular invasion. Mol. Microbiol. 78, 545–560.

Please cite this article in press as: Gründel, A., et al., Interactions of surface-displayed glycolytic enzymes of Mycoplasma pneumoniae with components of the human extracellular matrix. Int. J. Med. Microbiol. (2016), http://dx.doi.org/10.1016/j.ijmm.2016.09.001

G Model IJMM-51080; No. of Pages 11

ARTICLE IN PRESS A. Gründel et al. / International Journal of Medical Microbiology xxx (2016) xxx–xxx

Singh, B., Fleury, C., Jalalvand, F., Riesbeck, K., 2012. Human pathogens utilize host extracellular matrix proteins laminin and collagen for adhesion and invasion of the host. FEMS Microbiol. Rev. 36, 1122–1180. Song, Z., Li, Y., Liu, Y., Xin, J., Zou, X., Sun, W., 2012. ␣-Enolase, an adhesion-related factor of Mycoplasma bovis. PLoS One 7, e38836. Spuesens, E.B.M., Fraaij, P.L.A., Visser, E.G., Hoogenboezem, T., Hop, W.C.J., van Adrichem, L.N.A., Weber, F., Moll, H.A., Broekman, B., Berger, M.Y., van Rijsoort-Vos, T., van Belkum, A., Schutten, M., Pas, S.D., Osterhaus, A.D., Hartwig, N.G., Vink, C., van Rossum, A.M., 2013. Carriage of Mycoplasma pneumoniae in the upper respiratory tract of symptomatic and asymptomatic children: an observational study. PLoS Med. 10, e1001444. Thomas, C., Jacobs, E., Dumke, R., 2013. Characterization of pyruvate dehydrogenase subunit B and enolase as plasminogen-binding proteins in Mycoplasma pneumoniae. Microbiology 159, 352–365. Tryon, V.V., Baseman, J.B., 1987. The acquisition of human lactoferrin by Mycoplasma pneumoniae. Microb. Pathog. 3, 437–443. Vastano, V., Salzillo, M., Siciliano, R.A., Muscariello, L., Sacco, M., Marasco, R., 2014. The E1 beta-subunit of pyruvate dehydrogenase is surface-expressed in Lactobacillus plantarum and binds fibronectin. Microbiol. Res. 169, 121–127.

11

Verhamme, I.M., Panizzi, P.R., Bock, P.E., 2015. Pathogen activators of plasminogen. J. Thromb. Haemost. 13, S106–S114. Vieira, M.L., Vasconcellos, S.A., Goncales, A.P., de Morais, Z.M., Nascimento, A.L., 2009. Plasminogen acquisition and activation at the surface of leptospira species lead to fibronectin degradation. Infect. Immun. 77, 4092–4101. Wang, G., Xia, Y., Cui, J., Gu, Z., Song, Y., Chen, Y.Q., Chen, H., Zhang, H., Chen, W., 2013. The roles of moonlighting proteins in bacteria. Curr. Issues Mol. Biol. 16, 15–22. Yamaguchi, M., Terao, Y., Kawabata, S., 2013. Pleiotropic virulence factor − Streptococcus pyogenes fibronectin-binding proteins. Cell. Microbiol. 15, 503–511. Zhang, H., Zheng, J., Yi, L., Li, Y., Ma, Z., Fan, H., Lu, C., 2014. The identification of six novel proteins with fibronectin or collagen type I binding activity from Streptococcus suis serotype 2. J. Microbiol. 52, 963–969.

Please cite this article in press as: Gründel, A., et al., Interactions of surface-displayed glycolytic enzymes of Mycoplasma pneumoniae with components of the human extracellular matrix. Int. J. Med. Microbiol. (2016), http://dx.doi.org/10.1016/j.ijmm.2016.09.001