The Membrane Bound LRR Lipoprotein Slr, and the Cell Wall-Anchored M1 Protein from Streptococcus pyogenes Both Interact with Type I Collagen

Streptococcus pyogenes is an important human pathogen and surface structures allow it to adhere to, colonize and invade the human host. Proteins containing leucine rich repeats (LRR) have been indentified in mammals, viruses, archaea and several bacterial species. The LRRs are often involved in protein-protein interaction, are typically 20–30 amino acids long and the defining feature of the LRR motif is an 11-residue sequence LxxLxLxxNxL (x being any amino acid). The streptococcal leucine rich (Slr) protein is a hypothetical lipoprotein that has been shown to be involved in virulence, but at present no ligands for Slr have been identified. We could establish that Slr is a membrane attached horseshoe shaped lipoprotein by homology modeling, signal peptidase II inhibition, electron microscopy (of bacteria and purified protein) and immunoblotting. Based on our previous knowledge of LRR proteins we hypothesized that Slr could mediate binding to collagen. We could show by surface plasmon resonance that recombinant Slr and purified M1 protein bind with high affinity to collagen I. Isogenic slr mutant strain (MB1) and emm1 mutant strain (MC25) had reduced binding to collagen type I as shown by slot blot and surface plasmon resonance. Electron microscopy using gold labeled Slr showed multiple binding sites to collagen I, both to the monomeric and the fibrillar structure, and most binding occurred in the overlap region of the collagen I fibril. In conclusion, we show that Slr is an abundant membrane bound lipoprotein that is co-expressed on the surface with M1, and that both these proteins are involved in recruiting collagen type I to the bacterial surface. This underlines the importance of S. pyogenes interaction with extracellular matrix molecules, especially since both Slr and M1 have been shown to be virulence factors.


Introduction
Streptococcus pyogenes is an important human pathogen. It most commonly causes throat and skin infections, such as pharyngitis and impetigo [1]. It can also cause severe invasive diseases such as sepsis, necrotizing fasciitis and toxic shock syndrome, as well as complications post infection such as rheumatic fever and glomerulonephritis [2]. Surface structures of S. pyogenes allow the bacteria to adhere to, colonize and invade mucus membranes and human skin. Some of these structures are M protein, M-like proteins, collagen type I-binding protein (Cpa) and streptococcal fibronectin-binding protein I (SfbI) [3,4].
Mammalian cells have a variety of ways to detect invading pathogens and to alert the immune system. The best known are the Toll-like receptors (TLR), which are transmembrane receptors containing an extracellular leucine rich repeat (LRR) domain. It is the LRR domain in TLRs that is responsible for recognizing diverse microbial components [5]. More recently indentified receptors involved in bacterial recognition are receptors with a carboxyl-terminal LRR domain, a central nucleotide binding and oligomerization domain (NOD). They are implicated in the cytosolic detection of bacterial components, mediated through the LRR domain [6,7]. LRR proteins are also involved in proteinprotein interactions such as signal transduction, cell adhesion and apoptosis [8,9]. LRR proteoglycan decorin has been shown to bind to fibrillar collagens. Other examples of LRR proteins are proline arginine-rich end leucine-rich repeat protein (PRELP), chondroadherin (CHAD) and biglycan [10,11].
LRRs are typically 20-30 amino acids long and the defining feature of the LRR motif is an 11-residue sequence LxxLxLxxNxL (x being any amino acid). The number of LRR repeats ranges from 2 to 45 and they are divided into a conserved and variable segment [12,13]. The conserved amino-terminal stretch of 9-12 amino acids forms the b-strand and the variable segment is a carboxy-terminal stretch of 10-19 amino acids that varies in length, sequence and structure. The arrangement of the repeats results in a horseshoe-shaped structure with the b-sheet on the concave side and the variable stretches on the convex side [14,15].
Several pathogenic bacteria, both Gram positive and Gram negative express surface proteins with LRR regions. LRR proteins have also been identified in viruses, archaea and eukaryotes [15,16]. Internalins of Listeria monocytogenes are the most studied bacterial proteins with LRR domains. There are at least 9 proteins in this family and all have been implicated in the invasion of the human cell [17]. The streptococcal leucine rich (Slr) protein is predicted to be a lipoprotein attached to the cell membrane and has been shown to be camouflaged for antibody recognition by the M6 protein [18,19]. Antibodies against Slr are developed during a S. pyogenes infection in humans and an S. pyogenes strain lacking Slr had lower resistance against neutrophil phagocytosis in vitro and was less virulent in a mouse model of infection [18,19]. Slr lacks a LPXTG motif, which links Gram-positive proteins to the cell wall; instead a TLIA lipobox is present and acts as a membrane anchoring motif. The Slr protein is similar to virulence proteins of the internalin family of L. monocytogenes but there are differences between the proteins. The LRR region in Slr is located in the Cterminal half of the molecule unlike the members of the internalin family in which the LRR region is located in the N-terminal part of the protein. At the end of the N-terminal region there are four histidine triad motifs that are not present in the internalin family of proteins.
Adherence of human pathogens to certain tissue components might reflect preferences for specific sites of infection. Bacterial adhesins that interact with extracellular matrix components such as collagen, fibronectin, fibrinogen and laminin-related polysaccharides have been identified for both Gram-negative and Grampositive bacteria. Groups A, B, C, D and G streptococci have all been shown to exhibit collagen binding ability [20]. Protein FOG is one of the proteins that recruit collagen type I in group G streptococci [21]. For S. pyogenes, adhesion to collagen type IV by surface protein M3 has been reported as well as binding to Mnegative strains [2,22]. Since Slr is a putative surface located lipoprotein containing LRR motifs, we hypothesized that Slr could mediate binding to collagen. In this study we present experimental evidence that Slr is a surface attached lipoprotein containing LRR regions. Furthermore, using electron microscopy of immunogold labeled Slr and other experimental techniques, we were able to show that Slr is a collagen I binding protein.

Reagents, bacteria and culture conditions
Bacteria and plasmids used in this study are described in Table 1. The S. pyogenes strain of serotype M1 (AP1) and mutants MB1 and MC25 were cultured in Todd-Hewitt broth (TH) (Difco, Detroit, DI, USA) or TH broth with 0.2% yeast extract (THY) (Oxoid LTD, Hampshire, UK) at 37uC in 5% CO 2 atmosphere, including 150 mg/ml kanamycin for growth of the mutants. Collagen I was purchased from Sigma (St.Louis, MO, USA). Antibodies used in this study were goat anti-rabbit IgG-peroxidase labeled antibody (Bio-Rad, CA, USA) and rabbit anti-collagen I/ III antibody (Calbiochem, Darmstadt, Germany). The S. pyogenes mutant strain MC25 used in this study was constructed as previously described by Collin et al [23].

Generation of Slr mutant strain
A PCR product was generated from AP1 chromosomal DNA using oligonucleotide primers 59- ATT-ATT-CTC-GAG-GGT-CTA-TTG-TTA-TCA-T-39 and 59-TTG-ACT-AGA-TCT-TAT-AAT-GGA-AGT-G-39 with XhoI and BglII sites. The XhoI and BglII digested PCR fragment was ligated into XhoI and BglII digested pFW13 and transformed into TOP10 chemically competent cells (Invitrogen, Carlsbad, CA). Purified plasmids were electroporated into competent AP1 cells as previously described [24]. For preparation of competent AP1 cells, the bacteria were cultured over night in THY broth, washed with water and 0.5 M sucrose solution. The electroporator settings were 2.3 kV, 3 mF, 800 Ohms. The electroporated cells underwent phenotypic expression for 2 h in SOGAS solution at 37uC in 5% CO 2 . The AP1 recombinants were selected on THY-plates containing 150 mg/ml kanamycin. All the S. pyogenes isolates described in Table 1 were tested for the presence of the slr gene using standard PCR. DNA was prepared by boiling single colonies in 100 ml of distilled water followed by centrifugation at 10,0006g for 5 min. One ml of supernatant was used as template for PCR with an annealing temperature of 53uC using the oligonucleotide primers 59-GCT-CCA-ACC-CCA-TTC-CC-ATC-39 and 59-TCG-TGA-GCA-TGA-CCT-TCT-TCA-TC-39 generating an 1169 bp product.

Characterization of mutants by Western Blot
The S. pyogenes wild type strain AP1 and mutants MB1 and MC25 were cultured in THY broth, including 150 mg/ml kanamycin for growth of the mutants, at 37uC in 5% CO 2 atmosphere. The bacterial pellets were resuspended in SDS-PAGE sample buffer and separated by 10% SDS-PAGE and blotted to an Immobilion-P TM PVDF-membrane (Millipore, Bedford, MA) according to Towbin et al [25]. After blotting, the membrane was blocked in 15 ml PBS (140 mM NaCl, 30 mM KCl, 8 mM Na 2 HPO 4 , 2 mM KH 2 PO 4 ) containing 0.1% Tween 20 (PBST) and 5% skim milk for 30 min at room temperature. After addition of primary antibody from Slr and M1 whole rabbit antiserum at dilution 1:1,000 and 1:20,000, the membrane was incubated at 37uC for 30 min in PBST and 5% skim milk. The filter was washed three times for 5 min in PBST and incubated with goat anti-rabbit IgG-peroxidase labeled antibody diluted 1:3,000 in PBST and 5% skim milk. After incubation the filter was washed three times in PBST and developed with SupersignalH West Pico Chemiluminescent substrate (Thermo Scientific, Rockford, IL) followed by visualization using a Chemidoc XRS imaging system and Quantity One image analysis software (Bio-Rad).
Analysis of bacterial mutants and Slr by electron microscopy F(ab9) 2 fragments of rabbit anti-Slr and anti-M1 IgG were generated using the IgG specific streptococcal protease IdeS [26]. IgG from 2 ml Slr or M1 antiserum was adsorbed to 2 ml Protein G Sepharose followed by washing with 865 ml PBS. 1 mg of GST-IdeS in 1 ml PBS was added and incubated for 5 h in 20uC. Filtrate was passed 5 times through Glutathione Sepharose (GE Healthcare, Uppsala, Sweden) to remove GST-IdeS. Integrity and purity of F(ab9) 2 fragments were verified by SDS-PAGE and protein concentration was determined using Advanced Protein Assay Reagent (Cytoskeleton, Denver, CO) (data not shown). The presence and location of individual Slr and M1 molecules on wild type and mutant GAS was analyzed by negative staining and transmission electron microscopy as described previously [27]. F(ab9) 2 fragment samples were conjugated with, 15 nm for Slr and 10 nm for M1, colloidal thiocyanate gold [28]. Bacteria were grown to Log phase (OD 620 = 0.5), washed in 3610 ml TBS (50 mM Tris-HCl, 150 mM NaCl, pH 7.4) and adjusted to 2610 9 colony forming units (cfu) per ml. The bacteria were mixed with the F(ab9) 2 -Au conjugates and incubated for 1 h at room temperature at a 1:2 molar ratio. 5 ml aliquots were adsorbed onto carbon-coated grids for 1 min, washed with two drops of water, and stained on two drops of 0.75% uranyl formate. The grids were rendered hydrophilic by glow discharge at low pressure in air. Specimens were observed in a JEOL JEM 1230 electron

Expression pattern of Slr and M1 protein
The S. pyogenes wild type strain AP1 was cultured in THY broth, at 37uC in 5% CO 2 atmosphere. Growth was recorded hourly by measurement of OD 620 and plotted in KaleidaGraph (Synergy Software, Reading, PA). Samples were collected hourly; Slr and M1 Western blot on cell extracts was performed as described above.

Recombinant expression of Slr and purification of M1 protein
Slr was expressed in Escherichia coli using the GST Gene Fusion System (GE Healthcare). A 2337 base pair slr PCR product was amplified from S. pyogenes AP1 genomic DNA using primers 59-ACT-TTG-GGA-TCC-TGT-CAA-TCA-CGA-39 with a BamHI site (underlined) and 59-TCT-TAG-CTC-GAG-TTA-GTC-AGC-ATG-39 with an XhoI site (underlined). This fragment was digested with BamHI and XhoI, ligated into pGEX-6P-1 generating plasmid pGEXslr, that was used to transform E. coli BL21(DE3)-pLys. pGEXslr/BL21(DE3)pLys was induced with 0.1 mM isopropyl b-D-thiogalactopyranoside and GST-Slr was purified using Glutathione-Sepharose (GE Healthcare) according to manufacturer's instructions. Purified GST-Slr was used for immunization of rabbits according to standard protocols [29]. The GST tag was removed from GST-Slr by on-column cleavage using Precission protease according to manufacturer's instructions (GE Healthcare). M1 protein was purified from culture supernatants from MC25 using affinity chromatography and anti-M1 rabbit antiserum was generated as previously described [23,29].
Lipoprotein analysis using globomycin AP1 bacteria were cultured in TH over night supplemented with 0, 12.5, 25 and 50 mg/ml of the signal peptidase II inhibitor globomycin similarly to what has been described for S. equisimilis, S. equi and S. zooepidemicus [34,35]. Bacteria from 5 ml cultures were pelleted by centrifugation and resuspended in 200 ml PBS. Bacteria were lysed by addition of 100 U of the bacteriophage lysin PlyC [36] for 10 min at room temperature followed by addition of a final concentration of 25 mg/ml of ribonuclease A and deoxiribonuclease (Sigma) and incubation for 10 min at room temperature. Samples were centrifuged for 10 min at 14,0006g at 4uC. Pellets were resuspended in SDS-PAGE sample buffer and separated on 10% SDS-PAGE followed by transfer to PVDF membranes and anti-Slr Western blot as described above.

Absorption of collagen type I to S. pyogenes strains
The strains AP1, MB1 and MC25 were grown in THY to OD 620 = 0.8, with 150 mg/ml kanamycin for growth of the mutants, at 37uC in 5% CO 2 atmosphere. The bacteria were washed with PBST and diluted to 2610 9 cfu/ml solution. 200 ml of the bacterial solution were incubated with 100 mg collagen I for one hour at room temperature. Aliquots of the solution were taken for electron microscopy visualization as previously described in Bober et al [27]. The bacterial solution was washed three times with PBST and the bound collagen was eluted from bacterial surface by incubation with 0.1 M glycine, pH 2.0 for 20 min. The bacteria were removed by centrifugation, 15006g for 10 minutes, and the supernatant was neutralized by addition of 1 M Tris base. 15, 10 and 5 ml of the samples were immobilized on a PVDF membrane using the Milliblot-D system (Millipore). The membrane was blocked in 5 ml PBST and 5% skim milk for 30 min at room temperature. After addition of rabbit anti-collagen I/III antibody (Calbiochem, Darmstadt, Germany) diluted 1:60 in PBST and 5% skim milk, the membrane was incubated at 37uC for 30 min. The filter was washed three times for 5 min in PBST and incubated with goat anti-rabbit IgG-peroxidase labeled antibody (Bio-Rad) diluted 1:3,000 in PBST and 5% skim milk. After incubation, the filter was washed three times in PBST and visualized as above. The samples were also analyzed by separation on a reducing 3-12% SDS-PAGE gel and stained with Coomassie Brilliant Blue.

Radiolabeling of collagen type I and Slr
Radioactive iodine ( 125 I) labeling was performed using IODO-BEADS (Pierce, Rockford, IL, USA) and done according to the manufacturer's instructions. Unbound iodine was removed on a PD-10 column (Amersham Biosciences, Uppsala, Sweden) in PBS buffer containing 0.05% Tween 20 and 0.002% sodium acide (PBSAT). Fractions of 500 ml were collected and radioactivity was measured using a Wallac Wizard 1470 automatic gamma counter. The fraction containing the radiolabeled protein was stored at +4uC until further usage.

In vitro binding studies
The strains AP1, MB1 and MC25 were grown in THY to OD 620 = 0.8, with 150 mg/ml kanamycin for growth of the mutants, at 37uC in 5% CO 2 atmosphere. The bacteria were washed with PBSAT and diluted to 2610 9 cfu/ml solution. 250 ml of the bacterial solution were incubated with approximately 100 0000 counts per minute (cpm) 125 I labeled collagen I for 30 min at room temperature. The unbound collagen was removed by centrifugation, 15006g for 10 minutes, and the pellet was measured for radioactivity using a Wallac Wizard 1470 automatic gamma counter. Percent of binding was calculated with Kaleida-Graph (Synergy Software) and values were illustrated in a dot plot. Experiments were performed in quadruples at three independent time points.
Collagen I, non-denatured and denatured with guanidine hydrochloride was applied directly onto a PVDF membrane using the Milliblot-D system (Millipore). The membrane was washed twice for 30 min in PBS containing 0.25% Tween 20 and 2% bovine serum albumin, incubated with 125 I-Slr overnight at 4uC, and washed three times for 20 min in blocking buffer as described above. Signals of bound ligand were analyzed by phosphoimaging in a Fujix BAS 2000 Bioimaging analyzer (Fujifilm Sverige AB, Stockholm, Sweden).
The presence and location of individual Slr molecules on monomeric and fibrillar collagen I was analyzed by gold labeling, negative staining and transmission electron microscopy as described above.

Surface plasmon resonance (SPR) interaction analysis
Collagen I, diluted to 10 mg/liter in 10 mM sodium acetate (pH 4), was immobilized via amine coupling to CM3 sensorchip flow chambers (GE Healthcare) at moderate response levels (e.g. 1500 RU). Briefly, Collagen was mixed with freshly prepared 100 mM N-hydroxysuccinimide and 400 mM N-ethyl-N'-(dimethylaminopropyl)carbodiimide in equal volumes, and capping of unreacted carboxymethyl sites was achieved by a 1 M ethanolamine (pH 8) injection. A flow chamber subjected to the immobilization protocol but without any addition of protein was used as control (blank) for each experiment. Slr and M1 proteins respectively, were sequentially diluted 2-fold in running buffer (10 mM HEPES, 150 mM NaCl, and 0.005% Surfactant P20 (GE Healthcare) pH 7.5) and injected over the collagen surface starting at 500 nM and 35 ml/min. Bacterial cell suspensions were diluted 2-fold into PBS for injections in PBS in sequence starting from 2610 9 cfu/ml solution. Binding was monitored in a BIAcore 2000 instrument. Between experiments, the collagen surfaces were strictly regenerated with multiple pulses of 2 M NaCl and 1.5 M glycine-HCl pH 2.0 followed by an extensive wash procedure using running buffer or PBS, respectively. After x-axis and y-axis normalization of obtained data, the blank bulk refraction curves from the control flow chamber of each injected concentration or experiment were subtracted. Binding curves were displayed and where applicable, the association (K a ) and dissociation (K d ) rate constants were determined using the BIA Evaluation 4.1 software and its equation for 1:1 Langmuir binding.

Slr is a horseshoe shaped LRR lipoprotein
The slr gene encoding the S. pyogenes LRR protein Slr is present in 32/32 strains of 24 different M serotypes as determined by PCR ( Table 1) and is also present in all currently sequenced S. pyogenes genomes. The slr gene in strain AP1 was cloned and sequenced using oligonucleotide primers based on the genome sequence from the M1 strain SF370 (Accession no. NC_002737) [37]. Sequenc- ing of this gene revealed an ORF encoding a 793 amino acid protein that is 99% identical to the predicted Slr protein from SF370. The sequences of the Slr from AP1 and the corresponding gene slr, have been deposited in Genbank under the accession no. HQ908654. Slr contains a 21 amino acid putative N-terminal signal sequence ending with the amino acid sequence TLIA. This is the recognition sequence for signal peptidase II that allows for lipid modification of the resulting amino-terminal cysteine and insertion into the cellular membrane. In between amino acid 22 and 421 are four histidine triad motifs, that are not present in InlA of L. monocytogenes. There are 13 LRRs in Slr spanning over amino acid numbers 421-705 forming the b-sheets compared to 15 LRRs in InlA. The carboxyl terminal end of Slr contains histidine rich repeat sequences, but lacks a cell wall anchoring motif, while InlA carries a classical LPtTG cell wall anchoring motif ( Figure 1A). The LRR in Slr resembles the consensus motif of the LRR region in InlA ( Figure 1B). To confirm the horseshoe shape of the protein, Slr was visualized by electron microscopy. The Slr is pointed out by arrows and the shape is well visible supporting a horseshoe like conformation of Slr (Figure 2A). Homology modeling of Slr using InlA as the template further supported the protein's structure. The modeled LRRs forming the b-sheets (blue) are suggested to be the protein-protein interaction sites ( Figure 2B).

Slr is a surface lipoprotein
The Slr protein was expressed in E. coli using the GST Gene Fusion system and was used in several in vitro experiments (See Material and Methods). GST-Slr and Slr were separated on a SDS-PAGE for size and purity control. The GST-Slr has an apparent size of 110 kDa and Slr of approximately 85 kDa ( Figure 2C). To confirm that Slr is a lipoprotein, a globomycin inhibition assay was performed. The maturation of lipoproteins is governed by signal peptidase II that removes signal peptides in the NH 2 -terminal signal sequence. This step is inhibited by globomycin, resulting in a surplus of signal peptides and thus an increased molecular size (Figure 2D), confirming that Slr is a lipoprotein. To investigate the role of Slr on the surface of S. pyogenes, we constructed a Slr mutant by single crossover mutagenesis of the AP1 strain of M1 serotype. A 505 bp promotorless fragment of the slr gene was cloned into multiple cloning site I of pFW13, resulting in vector pMB01 ( Figure 3A). One mutant strain denoted MB1 was chosen for further analysis. Correct insertion of pMB01 was confirmed by PCR using gene-and plasmid specific primers ( Figure 3A, primer 1-4, data not shown). To confirm slr disruption in MB1 and the Slr and M1 expression in AP1 and MC25, protein extracts from AP1, MB1 and MC25 cell pellets were analyzed by immunoblotting using anti-Slr and anti-M1 antibodies. A strong signal was detected in the AP1 pellet for both Slr and M1 protein.
The only detectable signal in MB1 was for the M1 protein while in MC25 only Slr expression was detected ( Figure 3B). Further mutant strain analysis and surface protein localization was done using gold labeled F(ab') 2 fragments of anti-Slr (small gold) and anti-M1 (large gold) IgG. The F(ab') 2 fragments were incubated with wild type and mutant strains and analyzed by negative staining and transmission electron microscopy. The AP1 strain, expressing both Slr and M1 protein, bound both M1 and Slr antibodies simultaneously, ( Figure 3C) whereas the mutant MB1 strain only showed binding of M1 antibodies verifying that the Slr protein is not expressed on the surface ( Figure 3C). The mutant MC25 strain only exhibited binding of Slr antibodies, showing that the M1 protein is not present on the bacterial surface ( Figure 3C). The expression pattern of Slr and M1 was investigated in the wild type strain AP1. The growth was plotted as OD 620 versus the incubation time ( Figure 4A), and samples were taken for Western blot analysis. Slr protein can be detected from hour 6 (approx. OD 620 = 0.4) and declines at hour 9 (approx. OD 620 = 0.8), while M1 can be detected from hour 2 (approx. OD 620 = 0.05) ( Figure 4B) and starts to be degraded at hour 7 (data not shown). Taken together, these results indicate that Slr is an abundant membrane bound lipoprotein that is co-expressed on the surface with the M1 protein in early stationary growth phase.

Slr and M1 binds to collagen type I
Previous results implicated that M proteins can interact with collagens (data not shown). Since it has been stipulated that LRRs have collagen binding capacity and since Slr is co-expressed with M protein on the bacterial surface there are reasons to investigate whether Slr can contribute to the collagen binding capacity of S. pyogenes. Therefore, we analyzed a possible Slr-collagen interaction. Wild type AP1 (expressing both M protein and Slr), mutant strain MB1 (lacking Slr protein), and MC25 (lacking M1 protein) bacteria were incubated with collagen I and the bound collagen was eluted with low pH glycine. The eluted proteins were immobilized on a PDVF membrane and anti-collagen Western blot was performed. The wild type AP1 strains exhibited strongest binding ability towards collagen I, while MB1 and MC25 strain showed diminished binding capacity ( Figure 5A). The eluted collagen I from the AP1, MB1 and MC25 strains was also separated on 3-12% SDS-PAGE with collagen I as a standard and similar results could be observed as for anti-collagen Western blot, with AP1 showing a higher binding of collagen I than both MC25 and MB1 ( Figure 5B). In order to further establish the collagen binding capacity of strains AP1, MB1 and MC25, a binding assay with radiolabeled collagen I was performed. The strains were incubated with 125 I labeled collagen I, unbound collagen was removed by centrifugation and the radioactivity value of bound collagen was measured. All strains exhibited binding toward collagen I with the binding capacity of 36.8% for AP1, 36.1% for MB1 and 33.2% for MC25 ( Figure 5C). The binding of collagen to the bacterial strains was also visualized by electron microscopy, where association of collagen I could be observed for all three strains with a tendency of less accumulation in MB1 and MC25 ( Figure 5D). To determine the influence of collagen denaturation during SDS-PAGE on the Slr binding, non-denatured and guanidine hydrochloride-denatured collagen I was examined in a slot blot experiment. The same procedure was used to determine that the binding was not caused by the GST-tag on Slr. The slot blot experiment showed no interaction between collagen I and the GST-tag (data not shown). The radiolabeled Slr bound to both collagen preparations in a similar manner ( Figure 6A). Further in vitro interaction between Slr and both monomer and fibrillar collagen I was investigated. Gold labeled Slr (the dots) and monomer collagen I show four binding sites. The dots are bound to C-and N terminal as well as 70 respectively 100 nm further in on the collagen ( Figure 6B, panel I). More detailed analysis is required to identify the specific site of interaction. The collagen I fibril is composed of triple helices overlapping each other, resulting in a variation of triple helical thickness in the molecule. The short span between light strands represents the overlap region where the collagen triple helices overlap each other, while the dark span represents the gap region where the gap between collagen helices occurs. Binding can be observed in both regions, but is mostly concentrated to the overlap region ( Figure 6B, panel II). These experiments were also executed with gold-labeled GST. There was no interactions between GST and monomeric or fibrillar collagen I (data not shown). The binding of Slr and M1 protein as well as the AP1, MB1 and MC25 strains to collagen I was investigated using surface plasmon resonance. Slr and M1 showed a dosedependent binding to collagen I with a K D of 12 nM and 54 nM, respectively ( Figure 7A and B). The bacterial strains AP1, MB1 and MC25 also showed a dose-dependent binding to collagen I ( Figure 7C-E). Taken together, these results suggest that both Slr and M1 protein can bind collagen I, both when in purified form and in a bacterial surface context. Slr has a somewhat higher affinity towards collagen I than M1.

Discussion
S. pyogenes is a human pathogen responsible for a variety of diseases, from common infections to life-threatening conditions. One of the most studied virulence factors of S. pyogenes is the cell wall anchored M protein. It is involved in adherence to human epithelial cells, such as buccal cells and epidermal keratinocytes [38,39]. The M protein is also intimately linked to antiphagocytic properties and intracellular survival in phagocytic cells [40,41]. Of particular interest for the present study is that the M6 protein present on the surface of S. pyogenes of M6 serotype has been suggested to camouflage the LRR lipoprotein Slr from Slr antibody recognition [19].
Lipoproteins from Gram-positive bacteria are not as well studied as cell wall anchored proteins, but several recent studies have suggested that lipoproteins also are important for immune evasion and adherence during colonization and infection. Furthermore, several lipoproteins have shown promise as vaccine candidates [42]. Previously described lipoproteins of S. pyogenes are linked to metal transport as acquisition, most of them belong to ABC transport systems and some have been shown to be virulence factors in animal models of infection [43,44,45,46].
Proteins containing LRRs vary widely in functions and are found in both prokaryotes and eukaryotes, but the known bacterial LRR proteins, especially from Listeria species, are important virulence factors that mediate adhesion and cellular invasion through protein-protein interactions [17]. The slr gene encoding Slr can be found in all 32 strains of 23 different M serotypes S. pyogenes used in this study, as well as in all currently sequenced S. pyogenes genomes. The apparent conservation of the slr gene, the previous finding that patients with S. pyogenes infections seroconvert to the Slr protein and the fact that Slr negative mutants are attenuated in animal model and contribute to phagocytosis resistance [18], warrant further investigation of the function of Slr and possibly also exploration of Slr as a vaccine candidate.
Slr and M1 protein expression on the bacterial surface reach their peaks at different time points during the logarithmic growth phase and M1 starts to be degraded before Slr. However, during a substantial amount of time in late logarithmic growth phase, both proteins are present at the bacterial surface. This expression pattern could be similar in an in vivo situation, where Slr and M1 surface exposure results in optimal adherence to host tissue. Direct binding experiments indicate that Slr binds strongly to collagen I, both in native and denatured form. This could suggest a fairly limited site of interaction and/or a protein-protein interaction that Figure 6. Binding of Slr to immobilized collagen I. A: Collagen I denatured with 4 M guanidine hydrochloride (+) and non-denatured collagen I (2) were applied to a PVDF membrane and incubated with radiolabeled Slr. The amount of collagen is indicated to the right. B: Binding of gold labeled Slr (the dots) and collagen I using electron microscopy. The dots are bound to C-and N terminal as well as 70 respectively 100 nm further in on the monomeric collagen I (panel I). The short span between light strands represents the overlap region, whiles the dark span represents the gap region on a collagen I fibril. Binding can be observed in both regions, but is mostly concentrated to the overlap region (panel II). Scale bar = 100 nm. doi:10.1371/journal.pone.0020345.g006 is not dependent on the three-dimensional structure of collagen I. In vitro binding assay with the wild type and mutant strains showed that elimination of either Slr or the M1 protein on the bacterial surface did not have a major affect on the strains' total capacity to bind collagen I. A similar binding pattern of Slr and M1 toward collagen could also be observed in direct binding assays, leading to a conclusion that both proteins can use collagen as a ligand.
Interestingly, we did not observe that M1 protein inhibited the binding of gold-labeled Slr antibodies to Slr on the bacterial surface as previously described for the M6 protein [19]. In fact, gold-labeled anti-M1 and anti-Slr antibodies could bind simultaneously to the bacterial surface (Fig. 3C). This observation, in combination with observed collagen binding to both proteins on the surface of the bacteria, suggest that under the conditions studied here, M1 and Slr are simultaneously accessible for some protein-protein interactions on the bacterial surface. On the other hand, elimination of either Slr or the M1 protein from the bacterial surface might reveal additional binding sites toward collagen I on the remaining protein in concordance with the camouflaging described by Waldemarsson et al [19]. This might Figure 7. Surface plasmon resonance interaction analysis. The binding of Slr and M1 protein, the wild type strain AP1, and mutant strains MB1 and MC25 strains to immobilized collagen I was investigated using surface plasmon resonance. Binding curves are displayed and, the association (K a ) and dissociation (K d ) rate constants were determined. A-B: Slr and M1 proteins were diluted 2-fold in running buffer and injected over the collagen surface starting at 500 nM and 35 ml/min. The proteins displayed a dose-dependent binding to collagen I with a K D of 12 nM for Slr and 54 nM for M1 protein. C-E: Bacterial cell suspensions were diluted 2-fold into PBS for injections sequences starting from 2610 9 cfu/ml solution. doi:10.1371/journal.pone.0020345.g007 explain why there was not a major reduction in total collagen I binding capacity of the Slr and M1 mutants compared to the wild type strain.
Another feature of Slr are the four histidine triad motifs in the N-terminal part of the molecule, which are absent in InlA of L. monocytogenes. Such motifs have been identified in four proteins of Streptococccus pneumoniae and might be involved in metal or nucleoside binding [47]. Furthermore, such histidine triad proteins in S. pneumoniae have been shown to alter complement deposition of the bacterial surface [48]. Therefore, enzymatic or other functions than collagen binding of Slr, cannot be excluded.
In conclusion, we have established that Slr is an abundant membrane bound horseshoe shaped LRR lipoprotein from S. pyogenes that, both alone and in concert with M1 protein, interacts strongly with collagen I. The multiple interactions with collagen I through Slr and M protein, indicate that collagen binding is an important feature of S. pyogenes. Therefore, further studies elucidating the regulation of, and possible inhibition/co-operation between Slr and M proteins will be important in understanding how S. pyogenes interacts with the extracellular matrix during colonization and infection.