Functional Analysis of the Collagen Binding Proteins of Streptococcus parasanguinis FW213

Bacteria generally can utilize multiple adhesins to establish themselves in the host. We found that Streptococcus parasanguinis, a dominant oral commensal and an opportunistic pathogen for subacute endocarditis, possesses at least three collagen-binding proteins that enable S. parasanguinis to successfully colonize damaged heart tissues and escape innate immune clearance. The binding specificities of these three proteins for extracellular matrix molecules differ, although all three proteins participate in biofilm formation by S. parasanguinis. The “multiligand for multisubstrate” feature of these adhesins may explain the high adaptability of this microbe to different tissue sites.

through the activity of collagen binding proteins (CBPs) (5). CBPs in Gram-positive bacteria belong to a group of proteins with a common structure known as MSCRAMMs (microbial surface components recognizing adhesive matrix molecules) (6,7). The general structure of MSCRAMMs includes an N-terminal leader peptide followed by a ligand binding domain (also known as the A region), a variable number of B domain repeats, and a cell wall sorting region (8). Although the function of B domain repeats is not fully defined, it has been suggested that sequences of the B repeats form isopeptides and, thus, can provide structural stability to the protein (7). Some CBPs bind collagen proteins only, whereas others may recognize multiple substrates. For example, both Cna of Staphylococcus aureus (9) and Ace of Enterococcus faecalis (10) bind only to collagen, and although Cna and Ace share a high degree of sequence similarity, these two proteins show different affinities for type I collagen (11)(12)(13). On the other hand, CbpA of Trueperella pyogenes (previously classified as Arcanobacterium pyogenes) binds collagen and fibronectin at different subsites in CbpA (14).
CBPs play important roles in the pathogenesis of streptococci (15). For instance, CpA of Streptococcus pyogenes binds to type I collagen, and CpA mutant strains exhibit decreased attachment to HEp-2 cells (16). In Streptococcus mutans, Cnm is required for adherence to and invasion of human coronary artery endothelial cells (17) and is associated with hemorrhagic stroke, cerebral microbleeds, and IgA nephropathy (18)(19)(20), whereas Cbm is associated with infective endocarditis (IE) (21). Other examples include CbpA of Streptococcus gordonii that can enhance the survival of S. gordonii in vivo (22). Furthermore, Cbp40 of Streptococcus suis serotype 2 is required for optimal adhesion to HEp-2 cells, biofilm formation, and virulence in the zebra fish infection model (23).
Streptococcus parasanguinis is an early colonizer and a common isolate from dental plaque (24). When S. parasanguinis gains access to the bloodstream, it can successfully colonize heart tissue and, thus, is a common cause of bacterial subacute endocarditis (25). Despite the significance of S. parasanguinis in the oral ecosystem and systemic infections, the known virulence factors of this microbe are limited to Fap1, BapA1, and FimA of the FimCBA ATP binding cassette transport system for Mn 2ϩ and Fe 2ϩ . Specifically, both Fap1 (26), the structural subunit of the long fimbriae, and Bap1 (27), a surface protein, are associated with biofilm formation, whereas FimA participates in the development of IE in a rat model, presumably by binding to the fibrin monolayer (28,29). Whether other MSCRAMMs also contribute to the development of IE by S. parasanguinis is yet to be determined. Here, we elucidate the functions of the putative CBPs in the colonization, biofilm formation, and survival of S. parasanguinis against macrophage clearance.

RESULTS
Identification and basic characterization of loci encoding CBPs. Three putative CBP-encoding loci, Spaf_0420, Spaf_1570, and Spaf_1573, were identified from the genome of S. parasanguinis FW213 (30) by a domain search and BLAST using sequences from known CBPs. The basic characteristics of these open reading frames (ORFs) are listed in Table 1. In silico sequence analyses indicated that all three proteins harbor a multidomain architecture similar to those of the known MSCRAMMs (Fig. 1). In addition to the collagen binding A domain and the B domain repeats, all three ORFs harbor an SDR (serine-aspartate repeat-containing protein family)-like immunoglobulin (Ig) domain that is frequently found in bacterial surface adherence proteins, including the Ace collagen binding protein of E. faecalis (12). Furthermore, the SpaA domain, named after the shaft pilin SpaA of Corynebacterium diphtheriae (31), and the FctA domain, known as the FCT (fibronectin-collagen-T antigen) pilus locus of S. pyogenes (32), were observed. Both domains are also found on a variety of bacterial surface proteins, including ECM binding proteins. Similar to the B domain repeats, structures of both SpaA and FctA domains contain intramolecular isopeptides (31,33). All three genes were insertionally inactivated to generate single-, double-, or triple-knockout strains ( Fig. 2A), as detailed in Materials and Methods, to define their functions.
To determine whether mutations in the putative CBPs would affect the growth of S. parasanguinis, the growth kinetics of wild-type FW213 and mutant strains carrying single knockouts were examined (Fig. 2B). Although all strains reached similar final optical density at 600 nm (OD 600 ) values, the inactivation of any of the three ORFs reduced the growth rate of S. parasanguinis. The estimated generation time of wildtype FW213 was 57 Ϯ 5 min, whereas the doubling times of MPH8, MPH11, and MPH19 were 92 Ϯ 8, 82 Ϯ 6, and 80 Ϯ 6 min, respectively.
Adherence of S. parasanguinis FW213 to ECM proteins. A binding curve for type I collagen, type IV collagen, fibronectin, and laminin was initially determined in wildtype S. parasanguinis FW213 by an enzyme-linked immunosorbent assay (ELISA). Maximal binding was observed when wells were coated with 2 to 3 g of the ECM substrates (Fig. 3A). The results also indicated that S. parasanguinis FW213 bound to type I collagen and fibronectin most effectively, with a relatively low affinity for type IV collagen and laminin.
The inactivation of any of the three ORFs (MPH8, MPH11, and MPH19 strains) reduced the binding of S. parasanguinis to type I collagen (Fig. 3B). Although the ZC1 strain (where both Spaf_1570 and Spaf_1573 are inactivated) exhibited a binding level comparable to those of strains carrying a single knockout, the inactivation of all three ORFs (strain ZC2) resulted in the lowest binding level observed. The difference in the binding levels between ZC2 and the other mutant strains was further confirmed by analysis of variance (ANOVA) followed by Tukey's test (P Ͻ 0.01). Thus, it is suggested that all three ORFs participate in binding to type I collagen and that Spaf_0420 may interact with the substrate at a site different from those of the other two ORFs. With the exception of MPH19 (with Spaf_1573 inactivated), all mutant strains showed reduced binding to type IV collagen. All mutant strains, except MPH8 (with Spaf_0420 inactivated), exhibited a significant reduction in binding to fibronectin compared to wildtype FW213. Unexpectedly, enhanced binding to fibronectin was observed in MPH8 (Fig. 3B); the nature of this phenotype is currently unknown. On the other hand, MPH8, MPH19, ZC1, and ZC2 exhibited reduced binding to laminin compared to wild-type a The distance between the ribosome binding site (RBS) and the translational start site is listed in parentheses. nt, nucleotides. b The target to which the ORF shares the highest homology. The length of the alignment over the target size and the percent identity at the deduced amino acid level are listed in parentheses. c GenBank accession number of the locus listed in the identity column. Collagen Binding Proteins in S. parasanguinis FW213, indicating that Spaf_0420 and Spaf_1573 mediated binding to laminin. Finally, as ZC2 remained able to bind to type I collagen and fibronectin, S. parasanguinis FW213 is likely to possess additional ligands for these substrates. Taken together, the data demonstrate that Spaf_0420 participates in binding to type I collagen, type IV collagen, and laminin; Spaf_1570 participates in binding to type I and type IV collagen and fibronectin; and Spaf_1573 participates in binding to type I collagen, fibronectin, and laminin. Thus, all three ORFs were able to bind to multiple substrates, with type I collagen being the common substrate for all CBPs tested.
Role of CBP homologues in S. parasanguinis binding to damaged swine heart tissues. As type I collagen is the major component of normal human mitral valves (34), all three CBPs are likely to contribute to the colonization and development of endocarditis by S. parasanguinis. This hypothesis was tested using trypsin-treated swine heart valves. Confocal laser scanning microscopy (CLSM) examination indicated that all mutant strains bound less efficiently to the damaged valve tissues than wild-type FW213 (Fig. 4), confirming the activity of these proteins in the colonization of host tissues. All S. parasanguinis strains were recognized by the anti-lipoteichoic acid antibody at similar efficiencies using a dot blot assay. Additionally, we found that FW213 failed to bind to intact swine heart tissues (data not shown), indicating that exposure of ECM proteins through trypsin treatment is essential for colonization by S. parasanguinis.
Impact of CBPs on biofilm formation of S. parasanguinis FW213. To analyze whether the deletion of these CBP-encoding genes could affect the growth and structure of the biofilm, the distribution of live and dead cells in a biofilm culture was determined by staining and CLSM examination using both static and flow cell systems. In the static system, MPH8 formed a biofilm containing a majority of dead cells, compared to wild-type FW213 (Fig. 5A). Slightly larger amounts of dead cells were also observed in the biofilms of MPH11 and MPH19 (Fig. 5A). The number of dead cells in each of the biofilms was further analyzed by flow cytometry (Fig. 5B). In agreement with the observations from CLSM analyses, biofilm cultures of MPH8 contained the largest amounts of dead cells among the four strains, whereas the percentages of dead cells in the biofilms of MPH11 and MPH19 were similar to those found in wild-type FW213, indicating that Spaf_0420 was critical for the viability of S. parasanguinis FW213 in a static biofilm.
A gross examination of the biofilm in a flow cell system revealed that biofilms created by MPH8 and MPH19 were thinner than those formed by wild-type FW213, whereas MPH11 was most capable of resisting the shearing force from the medium flow and formed the thickest biofilm among all strains (Fig. 6A). Similar to the static Significant differences between wild-type FW213 and its derivatives were determined by ANOVA followed by Dunnett's test. ***, P Ͻ 0.0001; **, P Ͻ 0.001; *, P Ͻ 0.01. system, the biofilm of MPH8 was comprised mostly of dead cells (Fig. 6B). The views of the x-y plane of the biofilms created by the MPH11 and MPH19 strains suggested that the inactivation of Spaf_1570 or Spaf_1573 promoted cell-cell aggregation, which could result in a less structured biofilm. Taken together, the results suggest that all the CBPs were required for the fitness and subsequent biofilm formation of S. parasanguinis, especially the product of Spaf_0420.
Inactivation of Spaf_0420 and Spaf_1570 results in reduced virulence of S. parasanguinis FW213 in the G. mellonella model. To analyze whether the CBPs contribute to the virulence of S. parasanguinis FW213, all strains were tested in the Galleria mellonella larva model of systemic infection (Fig. 7). The highest mortality rate was observed with larvae infected with the wild-type strain FW213. The mortality of larvae infected with MPH19 was similar to that of larvae infected with FW213, whereas Ͼ50% of the larvae (12 out of 20 larvae) remained viable after infection with MPH8 and MPH11, indicating that both Spaf_0420 and Spaf_1570 are important for the virulence of S. parasanguinis in systemic infections.
CBPs are essential for survival of S. parasanguinis within macrophages but are not involved in macrophage recognition. To ascertain whether the CBPs were recognized by macrophages during phagocytic clearance, the amount of internalized bacterial cells was quantified at 15 min postinfection by flow cytometry analysis (Fig. 8A). At a multiplicity of infection (MOI) of 10, the percentages of the gated population measured from macrophages infected with FW213, MPH11, and MPH19 were comparable (42 to 48%), whereas fewer (20%) macrophages had internalized MPH8 cells. At an MOI of 50, the percent gated population increased among all strains tested, with only minor differences between strains, presumably due to maximal loading of macrophages with bacteria. To analyze whether the differences observed between MPH8 and other strains at an MOI of 10 were significant, the relative internalization efficiency of each strain was calculated, compared to wild-type FW213 (set at 100%), from three independent experiments. Although the average of the percent gated population in MPH8 was lower than that of wild-type FW213, there was no significant difference between strains (Fig. 8B). We next determined the bacterial load per macrophage by examining the geometric mean (Geo mean). The trend of the Geo means between strains at each of the MOIs tested was the same as that seen in the percent gated population. Thus, Spaf_0420, Spaf_1570, and Spaf_1573 of S. parasanguinis FW213 were not the primary targets for macrophage recognition. We next asked whether the CBPs also impact the survival of S. parasanguinis FW213 within macrophages. At an MOI of 10, the survival rates of all three CBP-deficient strains were significantly lower than that of wild-type FW213 (Fig. 8C). At an MOI of 50, the survival rates were approximately 10-fold lower than those observed at an MOI of 10 in all strains. Although the average survival rates of the mutant strains remained lower than that of wild-type FW213, the differences are not statistically significant (all P values are greater than 0.01). The lower bacterial survival rates at an MOI of 50 were likely the result of increased amounts of extracellular bacteria and toxicity from the bacteria compared to those at an MOI of 10. Taken together, the data show that the deletion of the CBP-encoding loci compromises the capacity of S. parasanguinis to survive within macrophages at a low MOI.

DISCUSSION
S. parasanguinis FW213 is not unique in possessing multiple CBPs. For instance, 5 putative CBP-encoding loci (SSA_0227, SSA_0805, SSA_1019, SSA_1663, and SSA_1666) are present in the genome of Streptococcus sanguinis SK36, although the function of these loci has not been characterized. However, it was unexpected that all three CBPs analyzed in this study were specific for other ECM proteins besides collagen. Furthermore, we observed that the triple mutant strain (ZC2) possessed approximately 50% of the wild-type level of binding to all substrates tested, suggesting the presence of additional binding proteins. Presumably, the "multiligand for multisubstrate" feature could facilitate colonization by S. parasanguinis on different host tissues. As seen in S. aureus, the fibronectin binding proteins FnbpA and FnbpB recognize fibronectin, fibrinogen, and elastin, which can facilitate the establishment of an infection by the microbe in a variety of host species and tissue types (35). We were unable to identify additional CBPs in the S. parasanguinis FW213 genome based on sequence analysis thus far; however, a putative fibronectin binding protein, Spaf_1409, was found to share 88% identity with fbpA of S. sanguinis NCTC3168 at the deduced amino acid level. Our recent study shows that this locus is required for the optimal binding of S. parasanguinis FW213 to fibronectin (36), and thus, FW213 possesses at least four ECM binding proteins.
Our analyses on the binding of S. parasanguinis strains to host tissues (Fig. 4) and biofilm formation (Fig. 6) suggest that of the identified CBPs, at least Spaf_1570 and Spaf_1573 are more critical in the initial establishment of IE but not in the growth of the vegetation. Nevertheless, the compromised fitness in the CBP-inactivated mutant strains could also contribute to the reduced colonization of swine heart tissue. To analyze whether the deletion of the CBPs would reduce envelope integrity, we tested the resistance of the CBP deletion strains using an autolysis assay (37). We observed elevated lysis rates in the CBP deletion derivatives (data not shown). Since the N-terminal signal peptides remain intact in the mutant strains, it is possible that the truncated proteins in the recombinant mutant strains cause significant cell envelope stress between the cell membrane and cell wall and, subsequently, reduced viability. On the other hand, we did not observe significant binding of S. parasanguinis to intact swine heart tissues, confirming that a preexisting heart abnormality is a key factor for IE by oral streptococci.
Extracellular polymeric substances, such as extracellular DNA (eDNA), proteins, and polysaccharides, play important roles in the development and stability of biofilm (38). Thus, polymeric substances released from dead cells may potentially enhance biofilm formation. However, the addition of an excess amount of FW213 chromosomal DNA or DNase I in the biofilm culture did not significantly alter the final biofilm mass of wild-type FW213 (see Fig. S1 in the supplemental material), suggesting that eDNA only marginally impacts biofilm formation by S. parasanguinis. This is supported by the fact that larger amounts of eDNA and extracellular proteins were detected routinely in the biofilm cultures derived from MPH8 and MPH19 than in wild-type FW213, yet an elevated biofilm mass was detected only in MPH11 cultures (Fig. S2). We also found that, similar to wild-type FW213, neither excess amounts of eDNA nor DNase I treatment significantly affected the biofilm mass from MPH11 (Fig. S1). On the other hand, as the surface hydrophobicity of oral streptococci could affect their adherence to the salivary pellicle (39) and their interaction with other oral microbes (40), we thought that perhaps truncated Spaf_1570 in MPH11 had enough of an impact on surface hydrophobicity, leading to enhanced cell-cell interactions. As seen in a study by Nakao and colleagues, elevated surface hydrophobicity caused by truncated lipopolysaccharide (LPS) could enhance biofilm formation by Escherichia coli (41). When measuring the surface hydrophobicity of all strains used in this study using a xylene hydrophobicity assay (42), elevated hydrophobicity was observed only in MPH11 (Fig. S3). Thus, it is tempting to suggest that the truncation of Spaf_1570 in MPH11 was responsible for the elevated surface hydrophobicity and biofilm mass.
Insects such as G. mellonella possess a complex, multicomponent innate immune system that is similar to the system used by mammals (43). Thus, the G. mellonella larva model has been widely used to study the virulence of a range of bacterial and fungal pathogens (44). As all three CBPs are not involved in the recognition of macrophages but could impact survival against clearance by macrophages, especially at lower MOIs (MOI ϭ 10), the reduced virulence of MPH8 and MPH11 in larvae suggests that Spaf_0420 and Spaf_1570 are required for resistance to clearance by oxidative burst and other antimicrobial peptides in G. mellonella.
In conclusion, this study demonstrates that the CBPs of S. parasanguinis are essential for binding to the host ECM, the colonization of damaged heart tissues, and the evasion of host innate immune surveillance. The differential impacts of Spaf_0420, Spaf_1570, and Spaf_1573 on biofilm formation and larval killing may reflect distinct roles for these proteins in the oral cavity and the bloodstream.

MATERIALS AND METHODS
Bacterial strains and growth conditions. All strains used in this study are listed in Table 2. S. parasanguinis strains were cultivated in Todd-Hewitt (TH) broth in the presence of 10% CO 2 at 37°C. Erythromycin (Em) at 10 g ml Ϫ1 , kanamycin (Km) at 250 g ml Ϫ1 , and spectinomycin (Sp) at 500 g ml Ϫ1 were used, where indicated, to maintain recombinant S. parasanguinis strains. To evaluate biofilm formation, strains were grown in biofilm medium (BM) (45) supplemented with 20 mM glucose (BMG).
Recombinant E. coli strains were grown at 37°C with aeration in LB medium containing ampicillin (Ap) at 100 g ml Ϫ1 , Em at 200 g ml Ϫ1 , Km at 50 g ml Ϫ1 , Sp at 50 g ml Ϫ1 , or chloramphenicol (Cm) at 25 g ml Ϫ1 , as needed.
Construction of isogenic mutant strains deficient in collagen binding homologues in S. parasanguinis FW213. To generate recombinant strains with deletions in Spaf_0420, Spaf_1570, and Spaf_1573, sequences containing the coding regions or the 5=-flanking regions of the loci were Em r Km r Sp r ; Spaf_0420, Spaf_1570, and Spaf_1573 null Spaf_0420::spe in ZC1 This study generated from S. parasanguinis FW213 by PCR with specific primers. All fragments were cloned into plasmid pSU21 (46) and established in E. coli DH10B. The identities of the recombinant plasmids were verified by plasmid isolation and restriction endonuclease analysis. Inverse PCR was then performed with the correct recombinant plasmids to generate deletions within the loci and the concomitant introduction of restriction sites for the cloning of a polar Em resistance gene (erm) (47). The identity of the erm-containing plasmids was confirmed by sequence analysis, and the correct chimeric plasmids were introduced into S. parasanguinis FW213 by electroporation (48) to generate isogenic mutant strains via double-crossover recombination. The allelic-exchange event in the Em-resistant isolates was confirmed by colony PCR with primers outside the Em gene insertion site as well as sequencing of the interrupted loci. The resulting strains MPH8, MPH11, and MPH19 contain a deletion of amino acids (aa) 311 to 380 in Spaf_420, a deletion of aa 309 to 415 in Spaf_1570, and a deletion of aa 41 to 108 in Spaf_1573, respectively. A similar approach was used to construct strains defective in both Spaf_1570 and Spaf_1573 and in all three loci. Briefly, the sequence encoding aa 308 to 328 of Spaf_1570 was replaced by a polar Km resistance gene (⍀kan) (49) in MPH19. The resulting strain, which bears deletions in both Spaf_1570 and Spaf_1573 and is resistant to both Km and Em, was designated ZC1. Amino acids 311 to 380 of Spaf_0420 were then replaced with a polar Sp resistance gene (spe) (50) in ZC1 to generate a triple-knockout mutant strain, ZC2.
Growth kinetics. The growth kinetics of wild-type FW213 and its derivatives in TH broth was monitored using a Bioscreen C reader (Oy Growth Curve AB Ltd.), as previously described (51). Briefly, bacterial cultures at an OD 600 of 0.3 were diluted 1:50 in fresh TH broth and inoculated in the wells of a 96-well microtiter plate. All wells were overlaid with 50 l sterilized mineral oil for anaerobic growth. Six biological repeats were used for each strain to estimate the generation time. The growth curves were constructed using GraphPad Prism 5.
ECM binding assay. The binding of S. parasanguinis strains to fibronectin, laminin, type I human collagen, or type IV human collagen was analyzed by an ELISA. The 96-well plates (enzyme immunoassay [EIA]/radioimmunoassay [RIA] plate; Costar) were coated with various amounts of ECM proteins as previously described (52). The wells were blocked with 200 l of phosphate-buffered saline (PBS) containing 2% bovine serum albumin (BSA) at room temperature for 1 h prior to the addition of bacterial suspensions. Log-phase cultures (OD 600 ϭ 0.4) were washed once with PBS and resuspended in PBS to an OD 600 of 0.4. An aliquot of 75 l of the bacterial suspension was added to ECM molecule-coated wells and incubated at 37°C with 10% CO 2 for 90 min. Wells with PBS only were used as a background control. At the end of the incubation, the wells were washed three times with 200 l PBST (PBS with 0.1% Tween 20), followed by the addition of a monoclonal antibody specific for Fap1 (E42) of S. parasanguinis FW213 (53). Plates were incubated at room temperature for 1 h. Unbound E42 was removed by three washes with 200 l PBST each. Horseradish peroxidase-conjugated goat anti-mouse IgG was then used to detect bound E42. The color reaction was developed using the ECL system (BD Biosciences). The reactions were stopped by the addition of 2 N H 2 SO 4 , and the final absorbance at 450 nm was recorded in a microplate reader (SoftMax Pro; Molecular Devices). Each assay was performed with triplicate samples.
Examination of binding of S. parasanguinis FW213 to swine heart valves. Swine hearts were purchased from a local wet market. Intact mitral valves were dissected from swine heart and then trypsin treated for 2 h as described previously (54). Intact swine heart valves were used as a negative control. Cultures of S. parasanguinis strains grown to an OD 600 of 0.6 were washed once with PBS and resuspended at 1 ϫ 10 10 CFU ml Ϫ1 in PBS. A piece of trypsin-treated tissue (approximately 1 cm 2 ) was placed in a 15-ml conical tube to which 1 ml of the bacterial suspension was added, and adherence was carried out for 90 min at 37°C in 10% CO 2 . The infected tissues were washed gently with PBS three times and then fixed in 10% formalin overnight. Goat anti-type I collagen affinity-purified polyclonal antibody (Millipore) was used to label exposed type I collagen, and mouse anti-lipoteichoic acid antibody (Novus) was used to localize S. parasanguinis. The locations of these primary antibodies were recognized by rhodamine-conjugated mouse anti-goat IgG and fluorescein isothiocyanate (FITC)-conjugated rabbit anti-mouse IgG, respectively. Samples were examined using confocal laser scanning microscopy (CLSM) (LSM510 Meta microscopy; Zeiss) with a 63ϫ oil immersion lens objective. The images were displayed with LSM Image Browser software.
Preparation of static and flow cell biofilms for CLSM examination. To examine the biofilm structure by CLSM, static biofilms were prepared using an 8-well Nunc chamber slide system (Thermo Scientific). Briefly, early-log-phase cultures (OD 600 ϭ 0.3) were diluted 1:10, and 400 l of the diluted cultures was inoculated into the wells of an 8-well chamber. The biofilm was grown for 24 h at 37°C. Biofilm cultures in the flow cell system were prepared as previously described (51), with minor modifications. Briefly, early-log-phase cultures were adjusted to an OD 600 of 0.1 in fresh BMG. A 250-l aliquot of the diluted cultures was injected into the growth chamber. The chamber was left upside down without medium flow for 4 h at 37°C. The medium pump was set at 5 ml Ϫ1 h Ϫ1 per channel, and the chamber was incubated for 18 h at 37°C.
Both the static and flow cell biofilms were stained with SYTO-9 and propidium iodide (PI) (Live/Dead biofilm viability kit; Invitrogen) according to the manufacturer's recommendations. The stained biofilms were examined by CLSM with a 63ϫ oil immersion lens objective. The biofilm images were displayed using Imaris 7.2.3 (Imaris).
Quantitation of live/dead cells in static biofilms by flow cytometry. Static biofilm cultures were prepared in 24-well plates, stained with SYTO-9 and PI as described above, and then analyzed by flow cytometry (55) to quantify the percentage of dead cells. For each sample, bacterial cells from three wells were used. An equal volume of 1ϫ PBS was added to the stained bacterial suspension prior to flow cytometry analysis using a FACSCalibur instrument (BD Biosciences). The data were analyzed using CELLQuest software (BD Biosciences).
Virulence of S. parasanguinis FW213 in the G. mellonella model. G. mellonella larvae with a body weight of approximately 220 Ϯ 50 mg were used in this study. Cultures of S. parasanguinis strains grown overnight in TH broth were washed twice with PBS and resuspended in PBS at 1.0 ϫ 10 9 CFU ml Ϫ1 . Ten microliters of the bacterial suspension was injected into the proleg of each larva with a 26-gauge needle (Hamilton). PBS alone and heat-inactivated (75°C for 30 min) bacterial suspensions were used as negative controls. The infected larvae were kept at 37°C in a petri dish, without feeding for 14 days, and survival was monitored daily until the death of the negative controls. Larvae were scored as dead when they displayed no movement in response to touch. Twenty larvae were infected with each strain. Survival curves were constructed using the Kaplan-Meier method, and all data were analyzed by a log rank test using GraphPad Prism 5.
Determination of the phagocytosis efficiency in Raw264.7 macrophages by flow cytometry. To determine the impact of CBPs on the phagocytosis efficiency in Raw264.7 macrophages, early-log-phase cultures (OD 600 ϭ 0.3) of S. parasanguinis cells were labeled with FITC (isomer I). Briefly, bacteria were washed twice with PBS and then resuspended in FITC (0.1 mg ml Ϫ1 ) at approximately 5 ϫ 10 9 CFU ml Ϫ1 . Labeling was carried out at room temperature for 1 h. At the end of the incubation, bacteria were washed twice with PBS prior to use. Mouse Raw264.7 macrophages were seeded in 6-well plates at 1.2 ϫ 10 6 cells per well in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum (FBS) and 2 mM L-glutamine and allowed to adhere to plastic plates for 12 h prior to infection with bacteria. The macrophages were infected with labeled bacteria in DMEM without 10% FBS at MOIs of 10 and 50 for 15 min at 37°C in a 10% CO 2 atmosphere. At the end of the infection, the extracellular bacteria were removed by three washes with PBS. The cell monolayers were harvested and then fixed with 2% paraformaldehyde in PBS at 4°C for 15 min. Cells were washed once with PBS, resuspended in 1% formaldehyde, and kept at 4°C overnight. An equal volume of PBS was added to each sample before analyzing the cell suspension using FACSCalibur. The data were analyzed using CELLQuest software.
Macrophage survival assay. Mouse Raw264.7 macrophages were seeded in 6-well plates at 1.2 ϫ 10 6 cells per well in DMEM containing 10% FBS and 2 mM L-glutamine and allowed to adhere to plastic plates for 12 h prior to infection with S. parasanguinis. All bacterial strains were grown to an OD 600 of 0.3, washed twice with PBS, and resuspended in DMEM at approximately 5 ϫ 10 9 CFU ml Ϫ1 . Raw264.7 macrophages were infected with bacteria at a MOIs of 10 and 50 for 30 min. At the end of the infection, the cell monolayer was washed with PBS twice, and the remaining extracellular bacteria were killed by the addition of penicillin at 100 U ml Ϫ1 , streptomycin at 100 g ml Ϫ1 , and gentamicin at 100 g ml Ϫ1 at 37°C for 1 h. The macrophage monolayer was washed with PBS, harvested, and lysed in PBS containing 0.1% Triton X-100 for 10 min. The intracellular viable bacterial counts were determined by serial dilution and plating.
Statistical analysis. Significant differences between wild-type FW213 and its derivatives were analyzed using one-way analysis of variance (ANOVA) followed by Dunnett's test or Tukey's test using GraphPad Prism 5. Differences were considered significant if the P value was Ͻ0.01.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only.