Staphylococcus aureus SasA Is Responsible for Binding to the Salivary Agglutinin gp340, Derived from Human Saliva

Abstract

Staphylococcus aureus is a major human pathogen that can colonize the nasal cavity, skin, intestine, and oral cavity as a commensal bacterium. gp340, also known as DMBT1 (deleted in malignant brain tumors 1), is associated with epithelial differentiation and innate immunity. In the oral cavity, gp340 induces salivary aggregation with several oral bacteria and promotes bacterial adhesion to tissues such as the teeth and mucosa. S. aureus is often isolated from the oral cavity, but the mechanism underlying its persistence in the oral cavity remains unclear. In this study, we investigated the interaction between S. aureus and gp340 and found that S. aureus interacts with saliva- and gp340-coated resin. We then identified the S. aureus factor(s) responsible for binding to gp340. The cell surface protein SasA, which is rich in basic amino acids (BR domain) at the N terminus, was responsible for binding to gp340. Inactivation of the sasA gene resulted in a significant decrease in S. aureus binding to gp340-coated resin. Also, recombinant SasA protein (rSasA) showed binding affinity to gp340, which was inhibited by the addition of N-acetylneuraminic acid. Surface plasmon resonance analysis showed that rSasA significantly bound to the NeuAcα(2-3)Galβ(1-4)GlcNAc structure. These results indicate that SasA is responsible for binding to gp340 via the N-acetylneuraminic acid moiety.

INTRODUCTION

Human tissue fluids, such as saliva and tears, contain several glycoproteins, which induce bacterial aggregation and clearance, acting as components of the innate immune system (1, 2, 3). gp340 (also known as DMBT1 [deleted in malignant brain tumors 1]) is a glycoprotein found in saliva, tears, the small intestine, and the lungs (3, 4, 5, 6). gp340 is composed of 13 highly homologous scavenger receptor cysteine-rich (SRCR) domains and is heavily O- and N-glycosylated (7, 8). This glycoprotein is associated with tumors, epithelial differentiation, and innate immunity (3, 8, 9). Epithelial cells express this glycoprotein, which protects them against the adhesion to or invasion into the host cells (3, 10). Also, gp340 is upregulated by inflammatory cytokines (10). Therefore, gp340 is thought to be an important factor for mediating the host immune response against microbial infection. gp340 is known to bind many pathogens, including Gram-positive and -negative bacteria and viruses (4, 11, 12, 13). This binding is mediated by a specific interaction between bacteria and gp340. The specific binding of some bacterial species has been investigated. For example, in Streptococcus gordonii, the sialic acid-binding protein, Hsa or GspB, interacted with the carbohydrate portion of gp340 (11, 14, 15), while Streptococcus mutans AgI/II adhesion involves binding to the consensus motif (VEVLXXXW) in the SRCR domain of gp340 (11, 16). Also, leucine-rich proteins (Lrr), such as BspA in Tannerella forsythia and LrrG in Streptococcus agalactiae, are known to bind another consensus motif (RX[R/Q]XR) in the SRCR domain (11, 17, 18). However, in many bacterial species, including Staphylococcus aureus, the specific interaction has not been characterized in detail.

S. aureus is a well-known pathogen in humans and produces many toxins and exoenzymes that cause various suppurative diseases, food poisoning, and toxic shock syndrome (19, 20). Furthermore, clinical isolates, especially methicillin-resistant S. aureus (MRSA), exhibit multiple antibiotic resistances, resulting in serious problems with antimicrobial therapy against S. aureus infection (21, 22). Besides these virulence factors, S. aureus produces several factors that protect against host-derived innate immune factors, such as complement, antibody, and neutrophils, as well as recently recognized innate immune antimicrobial peptides, such as LL-37 and defensins (23, 24). Although S. aureus is a major pathogen, this organism can localize as a commensal bacterium in the nasal cavity, oral cavity, skin, and intestine (25, 26, 27). Therefore, S. aureus is exposed to several types of tissue fluid in the infectious lesion. Previously, Jumblatt et al. reported S. aureus binding to gp340 in tears (5). However, the specific interaction between S. aureus and gp340 has not been elucidated. S. aureus can be isolated from the oral cavity, suggesting that this bacterium is exposed to gp340 in saliva. To explore S. aureus persistence in the oral cavity, it is important to investigate its interaction with saliva. In this study, we first investigated the interaction of S. aureus with saliva and then tried to identify the factor responsible for binding to gp340.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

The bacterial strains used in this study are listed in Table 1. S. aureus was cultured in Trypticase soy broth (TSB) (Becton-Dickinson Microbiology Systems, Cockeysville, MD) at 37°C. Escherichia coli was grown in Luria-Bertani (LB) broth at 37°C. Tetracycline (Tc; 5 μg/ml) or ampicillin (Amp; 100 μg/ml) was added when appropriate.

Table 1.

Bacterial strains used in this study

Strains	Gene identitya	Gene name	Characteristicsb	Reference
Staphylococcus aureus
RN4220			Laboratory strain, methicillin sensitive	47
MW2			Clinical strain, methicillin resistant (mecA+)	48
srtA mutant	MW2448	srtA	srtA:: pCL52.1 in MW2, Tcr	This study
sasA mutant	MW2575	sasA	sasA:: pCL52.1 in MW2, Tcr	This study
sasA complement	MW2575	sasA	sasA/pCL8 in sasA mutant, Tcr Cpr	This study
sdrE mutant	MW0518	sdrE	sdrE::pCL52.1 in MW2, Tcr	This study
MS23002			Clinical strain, methicillin resistant (mecA+)	This study
MS23290			Clinical strain, methicillin sensitive	This study
TF2947			Clinical strain, methicillin resistant (mecA+)	This study
TF3303			Clinical strain, methicillin resistant (mecA+)	This study
TF3053			Clinical strain, methicillin sensitive	This study
TY234			Clinical strain, methicillin sensitive	This study
TY186			Clinical strain, methicillin sensitive	This study
TY349			Clinical strain, methicillin resistant (mecA+)	This study
TY934			Clinical strain, methicillin sensitive	This study
TY1638			Clinical strain, methicillin sensitive	This study
Escherichia coli
MM1072	MW2575	sasA	sasA/pQE30 in E. coli for rSasA-N XL-II, Ampr	This study
MM1073	MW2575	sasA	sasA/pQE30 in E. coli for rSasA-BR XL-II, Ampr	This study
MM1074	MW2575	sasA	sasA/pQE30 in E. coli for rSasA-L XL-II, Ampr	This study
MM1075	MW2575	sasA	sasA/pQE30 in E. coli for rSasA-N-ΔL XL-II, Ampr	This study

^aGene identity in S. aureus MW2.

^bTc^r, tetracycline resistance; Cp, chloramphenicol resistance; Amp^r, ampicillin resistance.

Purification of gp340 from human saliva.

Human saliva was obtained from a healthy volunteer, which was approved by the ethics committee of Kagoshima University. Unstimulated whole saliva was obtained from a single donor (male, 54 years of age) in an ice-chilled plastic tube and clarified by centrifugation at 12,000 × g for 20 min. Purification of gp340 from saliva was performed as described previously (16). Briefly, clarified saliva (50 ml) was first diluted 2-fold with aggregation buffer. Aggregation buffer components were as follows: 1.5 mM KH₂PO₄, 6.5 mM Na₂HPO₄, 2.7 mM KCl, and 137 mM NaCl (pH 7.2). An overnight culture (60 ml) of S. mutans UA159 was collected and washed with aggregation buffer. The cell density was then adjusted to 5 × 10⁹ cells/ml in 50 ml of aggregation buffer. Equal volumes (50 ml) of bacterial suspension and saliva were mixed and incubated at 37°C for 60 min. Bacterial cells were collected by centrifugation at 5,000 × g for 10 min and washed once with aggregation buffer. Salivary components were eluted with 1.5 ml of aggregation buffer containing 5 mM EDTA. The eluate was subsequently filtered (0.45-μm pore size) and dialyzed against aggregation buffer at 4°C. The dialyzed eluate was then subjected to gel filtration chromatography on a Superdex 200 HR column (GE Healthcare, Uppsala, Sweden) equilibrated with aggregation buffer. The eluate at void volume was collected and used as salivary agglutinin, gp340. Purification of gp340 was verified by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) in a 7.5% polyacrylamide gel followed by silver staining. Also, since gp340 forms a complex with secretory IgA, immunoblotting was performed using anti-human IgA (heavy chain) (Funakoshi, Tokyo, Japan) as the primary antibody. After SDS-PAGE, proteins were transferred to a nitrocellulose membrane. After blocking with 2% skim milk in Tris-buffered saline (TBS; 20 mM Tris, 137 mM NaCl [pH 8.0]) containing 0.05% Tween 20 (TBS-T), antiserum specific for IgA (diluted 1:1,000 in 1% skim milk in TBS-T) was added, followed by incubation at 37°C for 1 h. After the membrane was washed with TBS-T, anti-mouse IgG conjugated with horseradish peroxidase (HRP) (diluted 1:1,000 in TBS-T) (Promega, Madison, WI) was added, and the membrane was incubated at 37°C for 1 h. The membrane was washed five times, and the band that reacted with the antiserum was detected using a chemiluminescence detection system (PerkinElmer, Waltham, MA).

Construction of the srtA, sasA, and sdrE mutants.

Inactivation of srtA, sasA, and sdrE in S. aureus was achieved as described previously (28). Briefly, DNA fragments containing an internal region of each orf (srtA fragment, 273 bp at position 55 to 327 from the 5′ terminus of the srtA gene [full length of the orf gene, 621 bp]; sasA fragment, 793 bp at position 243 to 1035 from the 5′ terminus of the sasA gene [6,828 bp]; sdrE fragment, 572 bp at position 320 to 891 from the 5′ terminus of the sdrE gene [3,426 bp]) were amplified. DNA fragments were digested with BamHI and HindIII and then cloned into a pCL52.1 vector, which is thermosensitive and can replicate at 30°C but not at 42°C. After electroporation of the plasmid into S. aureus RN4220, the bacteria were cultured at 30°C with Tc (10 μg/ml) overnight. The plasmid in RN4220 was then transduced into the MW2 strain using phage 80α. Both strains containing the plasmid were cultured overnight at 30°C. Appropriate dilutions of the culture were poured onto Trypticase soy agar (TSA) plates containing Tc (5 μg/ml), followed by incubation at 42°C overnight. Ten colonies were collected and replated on TSA containing Tc. Disruption of the target gene was confirmed by PCR. The mutant obtained was grown at 42°C for preculture. Then a small portion of the preculture was inoculated to fresh medium containing Tc (5 μg/ml) and grown at 37°C. We checked that the plasmid integrated was not excised when the bacteria were incubated at 37°C overnight.

For the complementation experiment, the DNA fragment of sasA amplified with specific primers was cloned into pCL8, which was an E. coli-S. aureus shuttle vector (29). The plasmid was electroporated to RN4220, and then the plasmid was transduced to an S. aureus MW2 sasA mutant by the method described above. The plasmids used in this study are listed in Table 2.

Table 2.

Plasmids used in this study

Plasmid	Characteristicsa
pCL52.1	E. coli-S. aureus shuttle vector, thermosensitive plasmid, Ampr Tcr (29)
pCL8	E. coli-S. aureus shuttle vector, Ampr Cpr (30)
pQE30	Expression vector for His-tagged protein, Ampr (Qiagen)
pMM1070	PCR fragment (srtA-F+srtA-R)/pCL52.1
pMM1071	PCR fragment (sasA-F+MW0621sasA-R)/pCL52.1
pMM1025	PCR fragment (sdrE-F+sdrE-R)/pCL52.1
pMM1072	PCR fragment (rSasA-N-F+rSasA-N-R)/pQE30
pMM1073	PCR fragment (rSasA-BR-F+rSasA-BR-R)/pQE30
pMM1074	PCR fragment (rSasA-L-F+rSasA-L-R)/pQE30
pMM1075	PCR fragment (rSasA-N-ΔL-F+rSasA-N-ΔL-R)/pQE30

^aAmp^r, ampicillin resistance; Tc^r, tetracycline resistance; Cp^r, chloramphenicol resistance.

Construction of recombinant SasA.

Recombinant SasA proteins were expressed as 6×His-tagged proteins. Since the BR (region that is rich in basic amino acid residues) domain of GspB in S. gordonii, which is a member of the same group of serine-rich glycoproteins as SasA, was demonstrated to be involved in binding to salivary proteins (30), we constructed a truncated protein with the N-terminal one-third, which retains BR domain, and named it rSasA-N (see Fig. 5). Also, we constructed a recombinant protein that comprised only the BR domain (named SasA-BR). Furthermore, since we found a lectin-like domain in the BR domain (see Fig. 5), we constructed the recombinant protein that comprised only the lectin-like domain (named rSasA-L) and rSasA-N with a deletion of the lectin-like domain (named rSasA-N-ΔL) (see Fig. 5). The primers used are listed in Table 3. DNA fragments encoding the sasA gene, which was amplified using specific primers, were cloned into the pQE30 vector (Qiagen, Tokyo, Japan) to generate pMM1072 (rSasA-N), pMM1073 (rSasA-BR), and pMM1074 (rSasA-L). The pQE30 vector was used for the construction of histidine-tagged recombinant protein. For construction of rSasA-N-ΔL, two DNA fragments covering the LRR1 domain or BR domain (deleting the lectin-like domain) and amplified with specific primers were fused and then cloned into the pQE30 vector to generate pMM1075. Recombinant proteins were purified according to the manufacturer's instructions. The purified recombinant proteins were verified by SDS-PAGE in a 7.5% polyacrylamide gel, followed by Coomassie blue staining.

Fig 5.

Schematic representation of S. aureus SasA. Shown is a DNA map of the sasA region; predicted functional domains of SasA and regions of recombinant proteins are indicated. The fragment used for gene inactivation in pCL52.1 is represented by the bold line. AA, amino acids.

Table 3.

Primers used in this study

Purpose and gene identity	Gene name or location	Forward primer	Reverse primer
Quantitative PCR
MW0006	gyrA	5′-AAGGTGTTCGCTTAATTCGC-3′	5′-ATTGCATTTCCTGGTGTTTC-3′
MW2575	sasA	5′-GTGCCGCAGTAGGTATTGGT-3′	5′-ACGTTGTCGCAACACCATAA-3′
Gene inactivation
MW2448	srtA	5′-TTGGATCCGTGGCAGCATATTTGTTT-3′	5′-TGATTCGTTTTCTTCTGC-3′
MW2575	sasA	5′-TATGTTGCATGACCAGC-3′	5′- CGGGATCCCCTGTTTCACCTAATACA-3′
MW0518	sdrE	5′- TAAAGCTTAATCAACTAAATCAAGTAC-3′	5′- TCGGATCCTGCTGCCACATTATCTTTA-3′
Recombinant protein expression
rSasA-N		5′-ATGGATCCATGAGTAAAAGACAGAAAGA-3′	5′- CACCCGGGTCGTCACCGCATTTCTTC-3′
rSasA-BR		5′-ATGGATCCACTGGATTGCCATCTGGACT-3′	5′- CACCCGGGTCGTCACCGCATTTCTTC-3′
rSasA-L		5′- GTGGATCCGCTATGTCAACATTTGCG-3′	5′- CCAAGCTTTACATCAACGTATGTCACT-3′
rSasA-N-ΔL	N terminal	5′- ATGGATCCATGAGTAAAAGACAGAAAGA-3′	5′- GACTCTGTTTTATTAACTGTAATTGT-3′
	C terminal	5′- TTAATAAAACAGAGTCTGCTGTTACA-3′	5′- CACCCGGGTCGTCACCGCATTTCTTC-3′

Antiserum and immunoblotting.

Antiserum against rSasA-N was obtained by immunizing mice as described previously (31). For immunoblotting, exponential-phase S. aureus cells were collected from a 50-ml culture and were washed with phosphate-buffered saline (PBS). A cell wall fraction was prepared as described previously (32). The cells were suspended in 1 ml of digestion buffer (30% raffinose in 50 mM Tris [pH 7.5] with 0.145 M NaCl) containing 1 mg of lysostaphin (Sigma-Aldrich, Tokyo, Japan), 100 μg of DNase (Nippon Gene, Tokyo, Japan), and phenylmethylsulfonyl fluoride (1 mM). The cell suspension was incubated for 30 min at 37°C. The protoplasts were removed by centrifugation at 6,000 × g for 10 min. The supernatant obtained was used as a cell wall fraction. The concentration of protein in each whole-cell lysate was quantified using a bicinchoninic acid (BCA) protein assay kit (Pierce Biotechnology, Rockford, IL). An identical amount of each sample was resolved by 7.5% SDS-PAGE. Immunoblotting analysis was performed with the method described above. Anti-SasA serum (1st antibody) and horseradish peroxidase-conjugated anti-mouse IgG (2nd antibody) (Promega, Madison, WI) were used for 1,000-fold dilution.

Assay for binding of S. aureus to saliva- and gp340-coated resin.

Clarified saliva, clarified saliva after treatment of S. mutans cells, and purified gp340 at various concentrations (5, 10, 20, and 40 μg/ml) were used in a binding assay. Bovine serum albumin (BSA) at 20 μg/ml was also used as a control. An overnight culture of the S. aureus strain was collected and washed with PBS (8.1 mM Na₂HPO₄, 1.5 mM KH₂PO₄, 137 mM NaCl, and 2.7 mM KCl). The bacterial suspension was then adjusted to 1 × 10⁶ cells/ml with PBS. A uniform, disk-shaped resin material (ACRON; GC, Tokyo, Japan) (10-mm diameter, 1.2-mm height) was prepared as per the manufacturer's instructions. The resin was soaked in 250 μl of clarified saliva or purified gp340 in 24-well culture plates at 4°C for 24 h. After the resin was washed with PBS three times, bacterial suspension (5 × 10⁵ cells/500 μl) was added and the plates were incubated at 37°C for 1 h. The bacteria were then removed, and the coated resin was washed three times with PBS. PBS (500 μl) containing 0.05% Triton X-100 and trypsin (100 μg/ml) was added, and the mixture was allowed to stand for 10 min at room temperature. The suspension was then diluted in PBS, and the appropriate dilutions (100 μl each) were plated on TSA plates. After overnight incubation at 37°C, the number of CFU was determined.

To investigate the effect of trypsin treatment on S. aureus binding to gp340, S. aureus MW2 was treated with 0.5, 1, 2, or 4 μg/ml of trypsin for 10 min at 37°C. After five washes with PBS, S. aureus was subjected to the binding assays described above.

To investigate the inhibitory effect of monosaccharides, lectins, or rSasA on S. aureus binding to gp340-coated resin, we used 10 mM fucose, galactose (Gal), mannose, N-acetylglucosamine, N-acetylgalactosamine, and N-acetylneuraminic acid (NeuAc), which are components of this glycoprotein (gp340) (16). The lectins Sambucus sieboldiana agglutinin (SSA), which is specific for NeuAcα(2-6)Gal/GalNAc (20 or 40 μg/ml), and Maackia amurensis agglutinin (MAM), which is specific for NeuAcα(2-3)Gal (20 or 40 μg/ml), were also used. Various concentration of rSasA-N (5 to 100 μg/ml) and various types of rSasA (each at 20 μg/ml) were used. Prior to the addition of bacterial suspension to the resin, 10 mM each sugar was added to the suspension, followed by incubation at 37°C for 10 min. The sugar-containing suspension was then added to the resin. For the lectin or rSasA inhibition assay, lectin or rSasA was reacted with gp340 at 37°C for 10 min, followed by the addition of bacterial cells. Binding efficiency was determined using the method described above.

Assay for binding of recombinant SasA to gp340.

Purified rSasA-N (0.25, 0.5, 1.0, 2.0, or 2.5 μg/ml), rSasA-BR (5 μg/ml), rSasA-L (5 μg/ml), and rSasA-N-ΔL (5 μg/ml) were used. gp340 (100 μl; 20 μg/ml) was added to a 96-well plate and incubated at 37°C for 1.5 h. gp340 was then removed, and each well was washed with PBS-T three times. Then 1% BSA in PBS containing 0.05% Tween 20 (PBS-T) was added to each well, followed by incubation for 1.5 h at 37°C. After a washing, various concentrations of rSasA in PBS-T were added to wells. After incubation for 1.5 h at 37°C, the solution was removed and each well was washed with PBS-T five times. Next, an anti-6×His monoclonal antibody (Wako, Osaka, Japan) (1,000-fold dilution in PBS-T) was added, followed by incubation for 1.5 h at 37°C. After a washing with PBS-T, an anti-mouse IgG-HRP conjugate (Promega) (2,000-fold dilution) was added and the plate was incubated for 1.5 h at 37°C. After three washings with PBS-T, OPD solution (133.3 mM citric acid, 66.7 mM Na₂HPO₄, 23.1 mM o-phenylenediamine, and 5 × 10⁻⁴% H₂O₂) was added to each well, and the absorbance at 492 nm was determined. The amount of bound rSasA-N was calculated from a standard curve generated by applying rSasA-N (0.125 to 4 μg/ml) to wells. The dissociation constant (K_d) was calculated by Scatchard plot analysis.

To investigate the inhibitory effects of mannose and N-acetylneuraminic acid on rSasA-N binding to gp340, each sugar (10 mM) was added to the rSasA-N, followed by incubation at 37°C for 10 min. The sugar-containing suspension was then added to each well.

SPR analysis.

The sugar chain structures immobilized on the array-type sugar chips are listed below (see Fig. 8; note that the saccharide unit of all sugar chains at the reducing end, which loses pyranose structure during the reductive amination reaction, serves as a hydrophilic spacer). To immobilize the sugar chains, the ligand conjugates were first prepared by reacting sugar chains with linker compounds and were immobilized on gold-coated chips, as previously reported (33). Briefly, gold-coated chips (Toyobo, Shiga, Japan) were washed with UV ozone cleaner (Structure Probe Inc., West Chester, PA). Next, 1-μl aliquots of ligand conjugates (500 μM) were spotted onto the gold-coated chips using an automatic spotter (Toyobo) according to the manufacturer's instructions. Every ligand conjugate was spotted twice on a fixed place on the chip. The sugar chain-immobilized gold chips were incubated for 1 h at room temperature and then washed with distilled water and PBS-T. Surface plasmon resonance (SPR) analysis was performed using MultiSprinter (Toyobo), according to the manufacturer's instructions, at room temperature. The array-type sugar chip prepared as described above was mounted on a prism with refraction oil (refractive index at 589 nm [n_D] = 1.700; Cargill Laboratories Inc., Cedar Grove, NJ) and equilibrated with running buffer (PBS-T) at a flow rate of 150 μl/min. After one analysis period, the sugar chip was regenerated by washing with 10 mM NaOH containing 0.5% sodium lauryl sulfate. rSasA-N in PBS-T at a concentration of 11, 33, or 100 μM was loaded onto the sugar chip at 150 μl/min, respectively, and the signal data were collected using an analytic software program (Toyobo).

Fig 8.

SPR analysis was performed using MultiSprinter (Toyobo), according to the manufacturer's instructions, at room temperature. The array-type sugar chip, prepared as described in Materials and Methods, was mounted on a prism with refraction oil and equilibrated with running buffer (PBS-T) at a flow rate of 150 μl/min. After one analysis period, the sugar chip was regenerated by washing with 10 mM NaOH containing 0.5% sodium lauryl sulfate. rSasA-N dissolved in PBS-T at a concentration of 0.11, 0.33, or 1 μM was loaded onto the sugar chip at 150 μl/min, and the signal data were collected using an analytic software program. GlcNAc, N-acetylglucosamine; GalNA, N-acetylgalactosamine; Glc, glucose; Man, mannose; Fuc, fucose; Xyl, xylose.

Analysis of sasA expression in clinical isolates.

Eleven clinical S. aureus isolates, including S. aureus MW2, were used in this study. A small portion (10⁸ cells) of S. aureus cultured overnight was inoculated into 10 ml of fresh TSB and then incubated at 37°C with shaking. When the optical density at 660 nm reached 0.5, cells were harvested. Total RNA was extracted from the bacterial cells with a FastRNA Pro Blue kit (MP Biomedicals, Solon, OH) in accordance with the manufacturer's protocol. An aliquot of total RNA (1 μg) was reverse transcribed to cDNA using a first-strand cDNA synthesis kit (Roche, Tokyo, Japan). Using cDNA as the template, quantitative PCR was performed using a LightCycler system (Roche, Tokyo, Japan). Primers for sasA were constructed and used to determine the optimal conditions for expression analysis; gyrA was used as an internal control. Three independent experiments were performed, and the average result was calculated. Statistical analysis was performed by Pearson's correlation coefficient test. Primers are listed in Table 3. Also, the binding of S. aureus clinical isolates to gp340-coated resin was investigated using the method described above.

RESULTS

Purification of gp340.

We purified gp340 from saliva in two steps: rough purification from S. mutans cells and further purification by gel filtration chromatography on a Superdex 200 HR column. In Fig. 1, purified gp340 subjected to 7.5% acrylamide gel analysis is shown. Previous reports showed that the protein purified by this method is a complex of gp340 and secretory IgA (sIgA), known as salivary agglutinin. gp340 complexed with sIgA had a molecular mass of more than 300 kDa and reacted strongly with an anti-IgA antibody. Treatment of purified gp340 with mercaptoethanol resulted in degradation of sIgA, indicating that this agglutinin is a gp340 protein complexed with sIgA (Fig. 1). In this study, we used this purified protein as gp340 (complex of gp340 and sIgA).

Fig 1.

Purification of gp340 from saliva. gp340 purification was verified by SDS-PAGE in a 7.5% polyacrylamide gel, followed by silver staining (A). Immunoblotting was then performed using anti-human IgA (heavy chain) as the primary antibody (B).

Saliva and gp340 binding of S. aureus strains.

We first investigated S. aureus binding to saliva-coated resin, gp340 (20 μg/ml)-coated resin, BSA-coated resin, and noncoated resin. S. aureus MW2 significantly bound to saliva- and gp340-coated resin compared to BSA-coated and noncoated resin (Fig. 2A). However, S. aureus strain MW2 did not bind to the resin coated with saliva after treatment of S. mutans cells. Upon exposure of 5 × 10⁵ cells to the resin, 3.33 × 10³ and 5.51 × 10³ cells were bound to saliva- and purified gp340-coated resin, respectively. We then investigated the optimal gp340 concentration for coating of the resin. Resin was exposed to gp340 at 5, 10, 20, and 40 μg/ml, and a binding assay was performed. As shown in Fig. 2B, application of >5 μg/ml of gp340 resulted in no increase in binding efficiency, so we used 20 μg/ml of gp340 for resin coating.

Fig 2.

(A) S. aureus MW2 binding to saliva- and gp340-coated resin was evaluated. The resin was soaked in 250 μl of clarified saliva, clarified saliva after treatment of S. mutans cells, or purified gp340 in 24-well culture plates at 4°C overnight. After three washes with PBS, S. aureus was added, followed by incubation at 37°C for 1 h. Finally, the number of S. aureus cells bound to the resin was evaluated as described in Materials and Methods. *, P < 0.01 compared to the value with PBS, as determined by the Dunnett method. (B) Purified gp340 (5, 10, 20, or 40 μg/ml) was added to the resin, and then an S. aureus binding assay was performed. *, P < 0.01 compared to the value with nonpurified gp340, as determined by the Dunnett method. (C) S. aureus MW2 was exposed to various trypsin concentrations (0.5, 1, 2, or 4 μg/ml) for 10 min at 37°C. After three washes with PBS, S. aureus was subjected to binding assay as described in Materials and Methods. *, P < 0.01 compared to the value without trypsin, as determined by the Dunnett method.

Identification of the S. aureus factor responsible for binding to gp340.

To identify the factor involved in gp340 binding, we first investigated the effect of trypsin digestion of S. aureus cells on binding to gp340-coated resin. Trypsin-digested S. aureus cells exhibited dose-dependently reduced binding efficiency (Fig. 2C). This indicates that the candidate(s) for the factor responsible for S. aureus binding is a cell surface protein.

We then constructed an srtA knockout mutant and investigated its binding efficiency. The srtA gene encodes sortase, which mediates the covalent binding of surface proteins to peptidoglycan (34). Therefore, when srtA is inactivated, several surface proteins (including adhesins) cannot bind to peptidoglycan, reducing their localization to the cell surface. The srtA mutant of the MW2 strain exhibited significantly decreased efficiencies of binding to saliva- and gp340-coated resin (Fig. 3).

Fig 3.

Binding of sortase-deficient S. aureus to gp340. A sortase-deficient S. aureus strain was subjected to a binding assay, as described in Materials and Methods. *, P < 0.01; **, P < 0.001; both P values are for comparisons to values obtained with wild-type (WT) MW2, as determined by t test.

SrtA plays a role in the localization of 20 proteins to the cell surface (34); thus, one of these could be involved in gp340 binding. Since gp340 is a glycoprotein, we investigated the effects of sugars on S. aureus binding to gp340. N-Acetylneuraminic acid significantly inhibited the binding of S. aureus to gp340, while other sugars had only a weak inhibitory effect (Table 4). Among 20 sortase A-dependent proteins, we found that one cell surface protein, SasA (also named SraP), was a member of the family of serine-rich glycoproteins including GspB and Hsa of Streptococcus gordonii (35, 36, 37), showing binding to the N-acetylneuraminic acid moiety. We then constructed an sasA mutant of the MW2 strain and performed a binding assay. Immunoblotting analysis showed that SasA protein showing approximately 240 kDa was absent in the sasA mutant (Fig. 4A). As a control, we also constructed the sdrE mutant. SdrE is one of the sortase-dependent serine-aspartate repeat proteins and has reported to be associated with human platelet aggregation and also binding to bone sialoprotein (38, 39). The sasA mutant showed significant decreases in efficiency of binding to saliva- and gp340-coated resins, while the sdrE mutant showed similar efficiencies of binding to the two resins (Fig. 4B). Compared to that of the wild type, the efficiency of binding of the mutant to saliva-coated and gp340-coated resin showed 13.8- and 5.7-fold reductions, respectively. The complementation strain showed increased binding efficiency compared to that of the sasA mutant.

Table 4.

Effects of sugars on S. aureus binding to gp340^a

Sugar	Binding (×10 CFU/resin)	% Inhibition
Control	471.3 ± 42.3
Fucose	396.3 ± 28.0	15.9
Galactose	400.8 ± 15.8	15.0
Mannose	225.0 ± 18.5	52.3
Glucose	338.1 ± 26.3	28.2
N-Acetylgalactosamine	287.0 ± 20.3	39.1
N-Acetylglucosamine	273.8 ± 44.8	41.9
N-Acetylneuraminic acid	130.0 ± 18.8	72.4

^aThe control and sugars were all used at 10 mM. Binding values are means ± standard deviations.

Fig 4.

Binding of SasA-deficient S. aureus to gp340. (A) Immunoblotting analysis of SasA in the wild type and the sasA mutant. Cell wall fractions (the wild type and the mutant) and the recombinant SasA were resolved by 7.5% SDS-PAGE, and then immunoblotting analysis was performed described in Materials and Methods. (B) A SasA-deficient S. aureus strain was used for the binding assay, as described in Materials and Methods. *, P < 0.01; **, P < 0.005; both P values are for comparison to values obtained with WT MW2, as determined by t test (saliva) and by Dunnett's two-way analysis of variance (gp340).

Effect of recombinant SasA on S. aureus binding to gp340.

To demonstrate the binding of SasA to gp340, we generated recombinant SasA (rSasA). We first attempted to construct full-length rSasA (228.4 kDa) but failed due to the large molecular mass. Next, we constructed a truncated rSasA-N with the N-terminal one-third, which retains the BR domain (Fig. 5). Finally, we purified the truncated rSasA-N. We then determined the effects of various rSasA-N concentrations on S. aureus resin binding. Pretreatment with rSasA-N inhibited the efficiency of binding to gp340-coated resin in a dose-dependent manner (Fig. 6A). This effect was also observed upon use of saliva-coated resin (Fig. 6B). Furthermore, we constructed an rSasA protein that comprised only BR domain, named rSasA-BR (Fig. 6B). Use of rSasA-BR in an experiment similar to that described above yielded a result (59.6% inhibition) similar to that for rSasA-N (data not shown). However, rSasA-L (lectin-like domain in BR domain) had a slight inhibitory effect on S. aureus binding to gp340 (17.5% inhibition), while rSasA-N-ΔL still had a 50.8% inhibitory effect on binding.

Fig 6.

Inhibitory effect of recombinant SasA protein on S. aureus binding to gp340. Various concentrations of rSasA-N (A) or various types of rSasA, including rSasA-N, rSasA-BR, rSasA-L, and rSasA-N-ΔL (B), were reacted with gp340-coated resin for 10 min at 37°C. S. aureus cells were then added to each well, and binding assays were performed as described in Materials and Methods. *, P < 0.01 compared to the value obtained without rSasA-N, as determined by the Dunnett method.

Binding affinity of rSasA for gp340.

We investigated the binding affinity of rSasA-N for gp340 and found that rSasA-N bound to gp340 in a dose-dependent manner (Fig. 7A). Scatchard analysis of rSasA-N saturation binding revealed a straight line and a dissociation constant (K_d) calculated as 4.65 × 10⁻¹³ M (Fig. 7B). Use of >2 μg/ml of rSasA-N yielded similar binding affinities, and so we used 20 μg/ml rSasA-N in further experiments.

Fig 7.

Binding of recombinant SasA to gp340. Various concentrations of rSasA-N (A) or various types of rSasA, including rSasA-N, rSasA-BR, rSasA-L, and rSasA-N-ΔL (C), were applied to each gp340-coated well. After incubation for 1.5 h at 37°C, the solution was removed and wells were washed with PBS-T five times. Anti-His tag antibody (1,000-fold dilution in PBS-T) was then added, followed by incubation for 1.5 h at 37°C. After a wash with PBS-T, an anti-mouse IgG-HRP conjugate (2,000-fold dilution) was added, and plates were incubated for 1.5 h at 37°C. After three washes with PBS-T, OPD was added to each well and the absorbance at 492 nm was determined. Based on the data of rSasA-N binding, Scatchard plot analysis was performed (B). *, P < 0.05 compared to the value obtained with rSasA-N, as determined by the Dunnett method.

Next, we investigated the gp340 binding using several kinds of rSasA. rSasA-BR showed the binding affinity similar to that of rSasA-N, while rSasA-N-ΔL showed a 50% decrease of binding affinity compared to rSasA-N, and rSasA-L almost eliminated the binding affinity to gp340 (Fig. 7C). We then investigated the effects of various sugars on rSas-N binding affinity to gp340. Since mannose and N-acetylneuraminic acid significantly inhibited S. aureus binding to gp340-coated resin, we used mannose, N-acetylneuraminic acid, and glucose (negative control) in this assay. We found that mannose (19.0% inhibition) and N-acetylneuraminic acid (67.9% inhibition) suppressed rSasA binding to gp340, while glucose did not (data not shown).

SPR analysis.

rSasA-N was dissolved in PBS-T and loaded onto the array-type sugar chip, and binding to immobilized sugar chains was monitored as signal intensity using the SPR imaging system (Fig. 8). rSasA-N bound to NeuAcα(2-3)Galβ(1-4)GlcNAc. However, rSasA-N did not exhibit significant binding to NeuAcα(2-3)Galβ(1-3)GlcNAc, NeuAcα(2-6)Galβ(1-3)GlcNAc, NeuAcα2-6Galβ1-4GlcNAc, or other oligosaccharides that have glucose, galactose, fucose, mannose, or GlcNAc as the terminal sugar. Based on the binding curves, the dissociation constant (K_D) between rSasA-N and NeuAcα(2-3)Galβ(1-4)GlcNAc was determined to be 3.1 × 10⁻⁷ M. The K_D values of rSasA-N and NeuAcα(2-3)Galβ(1-4)GlcNAcβ(1-6)Glc were both 4.9 × 10⁻⁷ M.

Furthermore, we investigated the effect of the lectins SSA [specific for NeuAcα(2-6)Gal/GalNAc] and MAM [specific for NeuAcα(2-3)Gal] on S. aureus binding to gp340-coated resin (Table 5). Addition of 20 and 40 μg/ml of MAM resulted in 57.7 and 72.1% inhibition of S. aureus binding to gp340, respectively, while 20 and 40 μg/ml of SSA resulted in 16.6 and 39.1% inhibition, respectively. Regarding the sasA mutant, addition of these lectins had a weak inhibitory effect on gp340 binding.

Table 5.

Effects of lectins on S. aureus binding to gp340

Lectin	Dose (μg/ml)	Binding (×10 CFU/resin)a	% Inhibition
Wild type
Control		471.9 ± 28.6
MAM	20	199.4 ± 24.5	57.7
	40	131.9 ± 11.1	72.1
SSA	20	393.8 ± 40.5	16.6
	40	287.5 ± 33.1	39.1
ΔsasA
Control		106.9 ± 13.6
MAM	20	100.6 ± 5.012.5	5.8
	40	81.3 ± 7.8	24.0
SSA	20	135.0 ± 9.6	−26.3
	40	107.5 ± 12.4	−0.6

^aBinding values are means ± standard deviations.

Expression of sasA in S. aureus clinical isolates.

Quantitative PCR analysis showed that sasA expression differed among strains. Also, S. aureus binding to gp340-coated resin varied among strains. We identified a positive correlation between sasA expression and binding ability (data not shown). A statistically significant (P < 0.005) correlation (r = 0.7946) between sasA expression and gp340 binding was found. Consequently, strains that exhibited high sasA expression levels also revealed a high efficiency of binding to gp340-coated resin and vice versa.

DISCUSSION

In this study, we demonstrated that S. aureus could bind to gp340 derived from saliva. Previously, gp340 formed the complex with secretory IgA (16), and we found that the purified gp340 in this study formed the complex. Also, we demonstrated that S. aureus bound to the resin coated with clarified saliva but did not bind to that with the saliva after the treatment of S. mutans cells. Since we found that secretory IgA was still contained in the saliva after the treatment (data not shown), this result indicates that S. aureus could not bind to secretory IgA. Since gp340 is a salivary agglutinin, which aggregates bacterial cells in saliva (4), we first investigated the effect of gp340 on salivary aggregation of S. aureus. S. mutans is known to induce aggregation in the presence of gp340. Compared to S. mutans, S. aureus exhibited weak salivary aggregation activity (data not shown), while S. aureus could adhere to gp340-coated resin. Gibbons et al. reported that proline-rich protein (PRP) on the cell surface promoted bacterial binding, while the same protein in solution had no binding affinity (40), indicating that the characteristics of PRP differed between the liquid and cell surface states. Also, Loimaranta et al. reported that gp340 in the fluid and surface states exhibits different bacterial recognition properties (15). Furthermore, fibronectin and laminin in fluid and on the cell surface showed different bacterial recognition properties (41, 42). These results suggest that S. aureus binds to the gp340 receptor on the surface more strongly than in solution. S. aureus can colonize the oral cavity in healthy individuals as well as patients with systemic or oral diseases. The presence of prosthetic devices, such as acrylic dentures, within the oral cavity might be conducive for growth of staphylococcus species (43). In this study, we used denture acrylic resin and found that gp340 bound to the resin and mediated S. aureus binding. This suggests that gp340 on acrylic resin is associated with the persistence of S. aureus in the oral cavities of patients with dentures.

SasA was first identified as SraP, which bound to human platelets (44). Siboo et al. demonstrated that the N-terminal region of SraP was associated with binding of platelets, and that inactivation of the encoding gene reduced virulence in an endovascular infection model (44). SasA is a member of the family of serine-rich O-linked glycoproteins such as GspB and Hsa of Streptococcus gordonii and Fap1 of S. sanguinis (35, 36, 37). Previous reports also demonstrated the glycosylation of SasA (44). O-linked glycosylation, which forms oligosaccharide linkages to the hydroxyl group of serine or threonine residues of peptide chains, is found in streptococcal and staphylococcal species. In this study, we demonstrated that the BR domain (a region that is nonglycosylated and rich in basic amino acid residues) could bind to gp340. Also, Takamatsu et al. reported that the BR domain of GspB bound to human salivary proteins (30, 45). These results suggest that the non-serine-rich region of this protein family has binding affinity to glycoproteins such as gp340 and a platelet membrane glycoprotein (GPIbα). In this study, we also clearly showed that the BR domain of SasA contributed to bind gp340. Since we found that SasA contains a typical lectin-like region in the BR domain (Fig. 5), which is found in legume-type lectins, including arcelin and concanavalin A, we first considered that this lectin-like domain in BR of SasA is important for gp340 binding. Unexpectedly, rSasA-L, which covered only the lectin-like domain, could not inhibit S. aureus binding to gp340, and it also failed to bind to gp340 (Fig. 6B and 7C). Also, rSasA-N-ΔL still had the binding affinity to gp340, although the binding affinity of rSasA-N-ΔL was almost 50% decreased compared to that of rSasA-N. These results indicate that the whole BR domain, not only the lectin-like region, is required for the binding to gp340. Since the sasA mutation caused an 82.4% loss of the binding affinity to gp340, SasA is considered to play a central role in S. aureus binding to gp340. However, the srtA mutant showed a 98% loss of the binding, implying than another sortase-dependent protein(s) is slightly involved in the binding to gp340.

In this study, we found that N-acetylneuraminic acid (NeuAc) inhibited the binding of SasA to gp340 (Table 4). Previously, gp340 was shown to contain several saccharides, including N-acetylneuraminic acid, N-acetylglucosamine, fucose, and galactose (16). Also, Issa et al. reported that salivary agglutinin contained partially sialylated O-linked oligosaccharides (46). SRP analysis clearly demonstrated that rSasA-N could bind to NeuAc but not glucose, galactose, N-acetylglucosamine, mannose, or fucose (Fig. 8). Furthermore, rSasA bound to NeuAcα(2-3)Galβ(1-4)GlcNAc significantly (K_D, 3.1 × 10⁻⁷M) but weakly bound to NeuAcα(2-3)Galβ(1-3)GlcNAc and NeuAcα(2-6)Galβ(1-4)GlcNAc. These results are in agreement with the observation that MAM [NeuAcα(2-3)-type lectin] significantly inhibited S. aureus binding to gp340, but SSA [NeuAcα(2-6)-type lectin] showed a weak inhibitory effect (Table 5). Hsa of S. gordonii has been shown to bind to the NeuAcα(2-3)Galβ(1-4)GlcNAc structure of N-linked oligosaccharides and NeuAcα(2-3)Galβ(1-4)GlcNAc/NeuAcα(2-3)Galβ(1-3)GalNAc structures of O-linked oligosaccharides, whereas GspB bound to only the NeuAcα(2-3)Galβ(1-4)GlcNAc/NeuAcα(2-3)Galβ(1-3)GalNAc structure of O-linked oligosaccharides of GPIbα in platelets (46). These results suggest that the non-serine-rich region in the N terminus of this protein family functions as a lectin-like protein, which binds to the NeuAc portion of glycoproteins such as gp340 and GPIbα. However, the binding specificity for the NeuAc moiety may differ among the serine-rich glycoproteins of various bacteria.

In conclusion, we found that SasA of S. aureus bound to gp340 and that the lectin-like domain in SasA is important for this interaction. Also, we found that NeuAc is an important structure for the SasA ligand. Since the NeuAc moiety is found in many glycoproteins expressed in human tissues, SasA of S. aureus may be associated with adherence and colonization to various regions in humans. These results indicate that SasA plays an important role in S. aureus infection.