|Home | About | Journals | Submit | Contact Us | Français|
Streptococcus suis, one of the most important and prevalent pathogens in swine, presents a major challenge to global public health. HP0197 is an S. suis surface antigen that was previously identified by immunoproteomics and can bind to the host cell surface. Here, we investigated the interaction between HP0197 and the host cell surface glycosaminoglycans (GAGs) using indirect immunofluorescence and cell adhesion inhibition assays. In addition, we determined that a novel 18-kDa domain in the N-terminal region of HP0197 functions as the GAG-binding domain. We then solved the three-dimensional structures of the N-terminal 18-kDa and C-terminal G5 domains using x-ray crystallography. Based on this structural information, the GAG-binding sites in HP0197 were predicted and subsequently verified using site-directed mutagenesis and indirect immunofluorescence. The results indicate that the positively charged residues on the exposed surface of the 18-kDa domain, which are primarily lysines, likely play a critical role in the HP0197-heparin interaction that mediates bacterium-host cell adhesion. Understanding this molecular mechanism may provide a basis for the development of effective drugs and therapeutic strategies for treating streptococcal infections.
Streptococcus suis is a porcine and human pathogen that is associated with a spectrum of diseases, including meningitis, septicemia, pneumonia, endocarditis, and arthritis (1). Thirty-three serotypes (types 1–31, 33, and 1/2) have been described based on their capsular polycarbohydrates. S. suis serotype 2 (SS2)4 is considered the most pathogenic and most prevalent capsular type in diseased pigs. Since its induction of major outbreaks in China and Vietnam (2, 3), S. suis has elicited considerable public health concern worldwide because it has resulted in substantial economic losses and emerged as a zoonotic pathogen with novel variants that cause streptococcal toxic shock-like syndrome in humans. It has been shown to be the primary cause of adult meningitis in Vietnam, the secondary cause in Thailand, and the third most common cause of community-acquired bacterial meningitis in Hong Kong (3–6). Adhesion to human and porcine epithelial and endothelial cells is a critical step during S. suis colonization and invasion (7). Uncovering the molecular mechanisms that mediate S. suis adherence to host cells may contribute to the development of effective vaccines and therapeutic strategies.
The hypothetical protein HP0197, which was identified as a surface antigen via an immunoproteomic method in our previous study (8), was demonstrated to be a candidate target for a novel vaccine (9). The HP197 gene is present in most pathogenic S. suis strains. This protein contains a C-terminal cell wall sorting signal sequence, which features a conserved LPXTG motif (see Fig. 2). SrtA (sortase A), an enzyme involved in the covalent linkage of surface proteins to peptidoglycans, specifically cleaves the LPXTG sequence between the Thr and Gly residues and links the protein to the cell wall through the nucleophilic attack of the amino group of the lipid II peptidoglycan precursor (10, 11). This property makes HP0197 a potentially vital virulence factor for S. suis. Furthermore, HP0197 displays no significant sequence homology to any known protein, and its function remains unknown (supplemental Fig. S1). Therefore, we initiated a structural and functional study of HP0197, which contributes to the understanding of the pathologic mechanism of S. suis.
The WT SS2 strain 05ZY was isolated from the brain of a diseased piglet; the expression of muramidase-released protein, extracellular protein factor, and suilysin was confirmed (12). This S. suis strain was grown either in Todd-Hewitt broth or on Todd-Hewitt agar at 37 °C under aerobic conditions. Escherichia coli DH5α was cultured in either LB broth or on agar at 37 °C for 8 h. E. coli BL21 B834 was cultured in LeMaster medium (see supplemental Table S1 for composition) at 37 °C. The gene fragments containing HP0197 (residues 41–531), the HP0197 18-kDa domain (residues 55–200), and the HP0197 loop domain (residues 201–417) were amplified by PCR from the SS2 strain 05ZY genome and cloned into an in house-modified version of the pET32a vector (Novagen). In this vector, the thioredoxin tag was deleted, and the S-Tag and thrombin recognition site were replaced with PreScission protease-cleavable segments. The HP0197 G5 domain (residues 418–472) was PCR-amplified from the SS2 strain 05ZY genome and mutated (V470M) using standard PCR-based mutagenesis. The mutation was confirmed by DNA sequencing, and the mutant was cloned into an in house-modified version of the pET28a vector (Novagen). In this vector, the yeast small ubiquitin-like modifier (SUMO) homolog Smt3 was inserted into the plasmid using the NheI and BamHI restriction sites, and the thrombin recognition site was replaced with SUMO protease (Ulp1)-cleavable segments. The pET32a/18-kDa domain plasmid was used as a template to design and produce constructs containing site-directed mutations in which the positively charged residue clusters (Group A, Lys55, Lys114, and Lys118; Group B, Lys61, Lys64, and Arg177; Group C, Lys131 and Lys192; Group D, Lys108 and Lys110; and Group E, Arg143, Lys144, and Lys145) were replaced with alanine residues using the Multipoints mutagenesis kit (TaKaRa) according to the manufacturer's instructions.
The native, mutant, and selenomethionyl (SeMet) recombinant proteins of the HP0197 18-kDa domain were expressed in E. coli BL21(DE3) and E. coli BL21 B834 cells. Protein expression was induced with 0.2 mm isopropyl β-d-thiogalactopyranoside at 25 °C for 16–18 h. The His6-tagged 18-kDa domain fragment that was expressed in the bacterial cells was purified by Ni2+-NTA-agarose (Qiagen) affinity chromatography and then digested with the PreScission enzyme to cleave the N-terminal His6 tag. The cleaved tag was removed by passing the digested mixture over a Superdex 200 size exclusion column (GE Healthcare). The SeMet-HP0197 G5 domain fusion protein was expressed in E. coli BL21 B834 cells that were induced with 0.2 mm isopropyl β-d-thiogalactopyranoside at 25 °C for 16–18 h. The His6-SUMO-tagged 18-kDa fragment that was expressed in the bacterial cells was purified by Ni2+-NTA-agarose affinity chromatography. After removing the imidazole from the eluted protein solution using a HiPrep 26/10 desalting column (GE Healthcare), the fusion protein was digested with the Ulp1 enzyme to cleave the N-terminal His6-SUMO tag. The cleaved tag was removed by passing the digested mixture over a Ni2+-NTA-agarose column. The target proteins were further purified on a Superdex 200 size exclusion column. The recombinant HP0197 loop protein was expressed in E. coli BL21(DE3) cells. Protein expression was induced with 0.2 mm isopropyl β-d-thiogalactopyranoside at 37 °C for 4–6 h. The His6-tagged 18-kDa domain fragment that was expressed in the bacterial cells was purified by Ni2+-NTA-agarose affinity chromatography and then digested with the PreScission enzyme to cleave the N-terminal His6 tag. The cleaved tag was removed by passing the digested mixture over a Superdex 200 size exclusion column.
The wild-type and SeMet-substituted HP0197 18-kDa domain crystals were obtained by mixing the protein (40 mg/ml in 20 mm Tris-HCl (pH 7.5) containing 100 mm NaCl and 1 mm DTT) with equal volumes of the reservoir solution (containing 30% PEG 3350 and 0.4 m NaNO3) via the sitting drop vapor diffusion method for 1 week at 20 °C. The crystals were then frozen in cryoprotectant, which consisted of the reservoir solution supplemented with 25% (v/v) ethylene glycol. The diffraction data were collected at beamline BL17U1 of the Shanghai Synchrotron Radiation Facility (SSRF) and processed using the HKL2000 software program (13). The wild-type and SeMet-substituted HP0197 18-kDa domain crystals belonged to space group P22212, with unit cell dimensions of a = 49.235, b = 87.601, and c = 38.540 Å. These crystals diffracted to 2.8 and 2.441 Å, respectively.
The SeMet-substituted HP0197 G5 domain mutant V53M crystals were obtained by mixing the protein (30 mg/ml in 20 mm Tris-HCl (pH 7.5) containing 100 mm NaCl and 1 mm DTT) with equal volumes of reservoir solution containing 38% 2-methyl-2,4-pentanediol, 0.1 m acetate (pH 4.5), and 0.1 m NBSD-256 using the sitting drop vapor diffusion method for 1 week at 20 °C. The crystals were then frozen in cryoprotectant, which was identical to the reservoir solution. The diffraction data were collected at beamline BL17U1 and processed using the HKL2000 software program. The SeMet-substituted HP0197 G5 domain mutant V53M crystals belonged to space group P4122, with unit cell dimensions of a = b = 114.592 and c = 128.545 Å. These crystals diffracted to 2.487 Å.
The HKL2MAP program (14) yielded four selenium sites and six selenium sites in one asymmetric unit of The HP0197 18-kDa and G5 domains, respectively, and the initial single anomalous dispersion phases were calculated using the PHENIX software program (15). The residues of the two structures were first built automatically using the PHENIX software package; they were then built manually using the Coot software program (16) based on 2Fobs − Fcalc and Fobs − Fcalc difference Fourier maps. The structural models were refined using both the CNS program (17) and the PHENIX program. The final structures of the HP0197 18-kDa and G5 domains had Rcrystal values of 22.4 and 23.8%, respectively, and Rfree values of 27.9 and 27.8%, respectively. The detailed data collection and refinement statistics are summarized in Table 1. The structural figures were generated with the PyMOL software program (version 1.5, Schrödinger LLC).
The human laryngeal epithelial cell line HEp-2 (CCTCC GDC004) was grown to confluence in 6-well plates in 2 ml of RPMI 1640 medium supplemented with 10% fetal calf serum and 1% penicillin/streptomycin. Briefly, on the day prior to the assay, the medium was replaced with 1 ml of fresh medium, and the cells were incubated overnight at 37 °C in a 5% CO2 atmosphere. On the following day, 0.1 ml of HP0197, 18-kDa domain, 18-kDa domain mutant constructs, loop region, G5 domain, or BSA was added to the wells at a final concentration of 0.1 μm. The 6-well plates were incubated at 37 °C in a 5% CO2 atmosphere for 1.5 h. The medium was removed from the wells, and the monolayers were washed three times with 1 ml of PBS to remove any non-adherent proteins. The adherent cells were dislodged and harvested by centrifugation at 3000 × g for 10 min. The cells were washed with 1 ml of PBS and resuspended in 0.1 ml of 2% paraformaldehyde in PBS at 4 °C overnight. On the following morning, the samples were washed three times with PBS to remove the fixative; they were then resuspended in 1 ml of PBS and incubated with either an anti-His tag monoclonal antibody (Sigma-Aldrich) or HP0197 antiserum, which was raised against the purified recombinant HP0197 protein, at 37 °C for 1.5 h. After three washes with PBS, the samples were incubated with an FITC-labeled mouse IgG antibody (Kirkegaard & Perry Laboratories, Inc.) and washed five times with PBS. A minimum of 105 cells/sample was analyzed by flow cytometry (BD Biosciences). The data were plotted as the cell count versus fluorescence intensity. The cell population of interest was assessed using the BSA sample to define the nonspecific fluorescence and/or the autofluorescence. The positive staining region for the remaining samples was defined to include the highest 1–5% of the stained cells in the BSA-stained sample.
For the inhibition studies, hyaluronate (HA), N-acetylglucosamine, glucuronic acid, sialic acid, glucose, galactose, mannose, heparin, heparan sulfate (HS), chondroitin sulfate (CS), and dermatan sulfate (DS) (Sigma-Aldrich) were preincubated with variant proteins at 4 °C for 1 h. The mixtures were then added to the wells to produce an inhibitor concentration of 0–500 μg/ml, and the mixture was incubated for 1.5 h at 37 °C. The cells were washed and analyzed by flow cytometry as described above.
For the heparinase treatment assay, HEp-2 cells were pretreated with 3, 6, or 12 units/ml heparinase I (Sigma-Aldrich) for 0.5 h at 37 °C in an atmosphere of 5% CO2 prior to incubation with HP0197 or its 18-kDa domain. The cells were then washed and analyzed by flow cytometry as described above. All of the assays were performed three times.
HEp-2 cells were cultured in 24-well plates in DMEM supplemented with 10% fetal calf serum at 37 °C in a humidified incubator. S. suis was grown in Todd-Hewitt broth for 6 h at 37 °C, harvested by centrifugation, washed twice with PBS (pH 7.4), and resuspended in fresh DMEM. The HEp-2 cells (105 cells/well) were infected with 1-ml aliquots of the bacterial suspensions (106 cfu/ml; multiplicity of infection (host/microbe) of 1:10). The plates were centrifuged at 800 × g for 10 min and incubated for 1 h at 37 °C in a 5% CO2 atmosphere. After incubation, the cells were thoroughly washed five times with PBS and then lysed with sterile water before dilution, agar plating, and bacterial colony counting. The assays were performed at least three times.
HP0197, the 18-kDa domain, and mutants A–E (in which the residues in each group were replaced with neutral alanine residues; 0.02 μm) were diluted in DMEM and added to the epithelial cells 30 min prior to addition of the bacterial inoculum. The adhesion assay was then performed as described above.
Before infection, the bacterial suspension was incubated with 200 μg of one of the following glycosaminoglycans (GAGs) for 30 min in an ice bath: heparin, HS, CS, DS, or HA. The adhesion assay was then performed as described above.
We analyzed the predicted secondary structure of HP0197 and discovered an N-terminal α-helix-rich region referred to as the 18-kDa domain based on its molecular mass and a C-terminal G5 domain connected by a proline-rich loop (see Fig. 2A). The G5 domain is found in a variety of enzymes, including streptococcal IgA peptidases (18) and various bacterial glycosyl hydrolases (19). An indirect immunofluorescence assay of epithelium-like HEp-2 cells confirmed the binding of HP0197 to the host cells (Fig. 1A). Therefore, we hypothesized that HP0197 may associate with certain carbohydrates on the host cell surface.
To evaluate the carbohydrate-binding activity of HP0197, we conducted an HEp-2 cell adhesion inhibition assay. HP0197 was preincubated with HA, N-acetylglucosamine, glucuronic acid, sialic acid, glucose, galactose, mannose, or heparin, and the binding of HP0197 to the HEp-2 cells was tested by indirect immunofluorescence. Flow cytometry analysis (Fig. 1A and supplemental Fig. S2) indicated that only heparin inhibited the binding of HP0197 to the HEp-2 cells, suggesting that HP0197 can bind to heparin. To provide further evidence of this interaction, HEp-2 cells were pretreated with a serial dilution series of heparinases and then tested for HP0197 binding. The treatment of the HEp-2 cells with heparinases prior to incubation with HP0197 or the 18-kDa domain diminished the binding of both in a dose-dependent manner (Fig. 1B).
Because heparin is a member of the GAG family, we investigated whether other GAGs can affect the interaction between HP0197 and HEp-2 cells. Additional inhibition studies were performed by pretreating HP0197 with varying concentrations of the following GAGs: heparin, HS, CS, and DS. HA was included as a negative control. Heparin, HS, CS, and DS all attenuated the binding of HP0197 to the HEp-2 cells in a dose-dependent manner (Fig. 1C), indicating that HP0197 associates with GAGs other than heparin.
To identify the region(s) of HP0197 that are responsible for binding cellular GAGs, we evaluated the heparin-binding activity of individual regions using a flow cytometry-based HEp-2 cell adhesion inhibition assay. The purified 18-kDa domain, G5 domain, loop region, and full-length HP0197 were individually preincubated in the presence or absence of heparin and added to an equal volume of HEp-2 cells. After incubation, the cell surface-binding proteins were detected by flow cytometry using an anti-His tag monoclonal antibody. Unexpectedly, the data indicated that the 18-kDa domain adhered to the HEp-2 cell surface similar to HP0197, whereas the G5 domain and loop region did not bind to the HEp-2 cells, suggesting that the 18-kDa domain mediates HP0197 heparin-binding activity (supplemental Fig. S3). Notably, heparinase treatment effectively destroyed the binding site for the 18-kDa fragment; however, full-length HP0197 was much less affected, which indicates that, in addition to the 18-kDa domain, other regions of HP0197 also contribute to its host cell attachment.
To further understand the role of HP0197 in bacterial pathogenesis, we determined the three-dimensional structures of the N-terminal 18-kDa and C-terminal G5 domains using x-ray crystallography. The crystal structure of the HP0197 18-kDa domain (residues 55–200) was determined at 2.4 Å resolution by single anomalous dispersion (Fig. 2A and Table 1). The HP0197 18-kDa domain is composed of six antiparallel α-helices (α1–α6), and its surface is dominated by acidic residues, with the exception of several small, scattered, positively charged clusters (Fig. 2B).
The crystal structure of the HP0197 G5 domain (residues 418–472) was determined at 2.5 Å resolution by single anomalous dispersion (Fig. 2B and Table 1). The HP0197 G5 domain reveals a topology that is characteristic of the G5 domain with two triple-stranded β-sheets. A central intertwined motif connects the two sheets (Fig. 2B), which leads to a strand switch. As a result, sheet β1 contains an antiparallel β-strand arrangement, and sheet β2 displays a mixed parallel/antiparallel β-strand arrangement. The HP0197 G5 domain has a well organized structure. There is a high percentage of residues with strong β-sheet propensity (Val and Thr compose 20.2 and 8.0% of the residues in this domain, respectively), which contributes to this structure's stability. Because β-sheets exist in a conformation that facilitates their interaction with other β-strands, they are considered to be adhesive (20, 21). The structure of the HP0197 G5 domain is very similar to that of the G5 domain in the resuscitation-promoting factor ΔDUFRpfB from Mycobacterium tuberculosis (Protein Data Bank code 3EO5) (22), with root mean square deviations of 0.886 Å for the N-terminal loop and 0.756 Å for the C-terminal loop (supplemental Fig. S4).
Several structures of heparin-binding proteins have been reported, and the correlation of these structures with their functions, such as heparin-thrombin binding (23), heparin-fibroblast growth factor binding (24, 25), and heparin-viral protein binding (26, 27), has been analyzed by site-directed mutagenesis. These studies have demonstrated that clusters of positively charged basic amino acids on the protein surfaces form ion pairs with spatially defined, negatively charged sulfo or carboxyl groups on the heparin chain (28). In support of this conclusion, our data showed that sulfated GAGs, including heparin, HS, CS, and DS, but not HA, could bind to HP0197 (Fig. 1C). Therefore, we postulated that the charge interactions between the sulfated GAGs and the residues in the 18-kDa domain mediate the binding of the GAGs to HP0197.
On the basis of the high resolution structure of the 18-kDa domain, we predicted the GAG-binding sites of this domain to be primarily basic residues; we then divided these sites into five groups (A–E) (Fig. 3) according to their structural relevance. To illustrate the role of each of the five groups of residues in the 18-kDa domain-GAG interaction, we performed several assays. Site-directed mutagenesis was utilized to construct variant 18-kDa proteins (mutants A–E) in which the residues in each group were replaced with neutral alanine residues. The purified variant proteins were incubated with HEp-2 cells and detected with an anti-His tag monoclonal antibody. Flow cytometry revealed significantly lower levels of cell binding for mutants A, B, and D compared with the native 18-kDa construct. Specifically, HP0197 bound to 93.1% of the cells, whereas mutants A, B, and D bound to 28.2, 44.3, and 33.2% of the cells, respectively (Fig. 4). This result suggests that Lys55, Lys61, Lys64, Lys114, Lys118, Arg143, Lys144, Lys145, and Arg177 may form the heparin-binding sites of the 18-kDa domain. Furthermore, the residues in Groups A, B, and D are adjacent to one another at the same end of the molecule in space; therefore, a GAG wrapping around the molecule could potentially bind all three residue groups.
To assess the contribution of HP0197 to S. suis host cell attachment, soluble HP0197, the 18-kDa domain, and 18-kDa domain mutants were preincubated with HEp-2 cells prior to the SS2 adhesion assay. Compared with the control, incubation of the epithelial cells with HP0197 and the 18-kDa domain reduced the SS2 adhesion by 61.7 and 50.8%, respectively. With mutants A, B, and D, the adhesion was inhibited by 5, 10, and 8.3%, whereas mutants C and E decreased the adhesion by 49.2 and 46.7%, respectively (Fig. 5).
On the other hand, the involvement of GAGs in the adherence of bacteria to the host cells was investigated using heparin, HS, CS, DS, and HA to preincubate HEp-2 cell monolayers. These polysulfated polysaccharides reduced (by up to 56%) the adherence of S. suis to the epithelial cell cultures (Fig. 6). The highest inhibition was obtained by using the GAGs heparin (56%) and HS (52%). CS and DS also decreased the level of adherence of the bacteria by 20.2 and 23.8%, respectively. Binding of bacteria to HEp-2 cells was not significantly affected by HA (5.6%). To summarize, the positively charged residues on the 18-kDa domain surface, which are primarily lysines, are critical for the HP0197-heparin interaction and contribute to the host cell adhesion of S. suis.
Several S. suis surface components, including FBPS (29), enolase (30), GAPDH (31), and SadP (32), contribute to the host cell adhesion of S. suis. FBPS and enolase bind to fibronectin; GAPDH was determined to be an albumin-binding protein; SadP recognizes CD77; and in this study, HP0197 was shown to bind to the host cell surface GAGs, particularly HS.
The data implicating the host cell surface GAG structures, in particular, heparin and HS, as receptors for HP0197 include the following. 1) HP0197 binds to HEp-2 cells. 2) This interaction is inhibited in a dose-dependent manner when HP0197 is pretreated with heparin, HS, or other sulfated GAGs, but not with other carbohydrates. 3) This interaction is inhibited in a dose-dependent manner when HEp-2 cells are pretreated with heparinases.
Our data demonstrated that the N-terminal 18-kDa domain of HP0197 is the GAG-binding domain, which suggested that positively charged basic amino acids on the surface of the 18-kDa domain represent the possible GAG-binding sites. The contribution of these charged basic amino acids to the host cell attachment of the 18-kDa domain was evaluated by site-directed mutagenesis and an indirect immunofluorescence assay. The incubation of HP0197 and the 18-kDa domain with HEp-2 cells attenuated the adhesion of S. suis, whereas the potential GAG-binding site mutants barely influenced this adhesion. Correspondingly, inhibition of bacterial binding was observed in the presence of soluble heparin and HS. Therefore, our data demonstrate that the positively charged basic residues on the 18-kDa domain are critical for the HP0197-GAG interaction and S. suis attachment to the host cell.
The adhesion of pathogenic bacteria to mammalian cells is regarded as a key mechanism of bacterial infection. GAG is an important component of the extracellular matrix and is a typical target for microbial pathogens that invade host cells (33, 34). Glycoproteins, another component of the extracellular matrix, have previously been reported to interact with SS2 (35). The pathogenic enterococci and streptococci Enterococcus faecalis (36), Streptococcus agalactiae (37), Streptococcus pneumonia (38), and Streptococcus pyogenes (39) have been shown to interact with various human GAGs. In the leading meningitis pathogen group B streptococcus, the host sulfated GAGs are significant to its blood-brain barrier penetration ability and central nervous system infection (40).
To our knowledge, this study is the first to demonstrate that an LPXTG-anchored cell wall surface protein in S. suis recognizes and binds to GAGs and that S. suis adheres to the host cell surface through this GAG-binding interaction. Additionally, we identified a definite novel GAG-binding domain, the 18-kDa domain of HP0197, and determined its three-dimensional structure. The design of inhibitors targeting specific pathogen virulence mechanisms represents an attractive strategy in an era of increasing resistance to conventional antibiotics. The results from this study could contribute to the development of novel strategies or drugs that prevent and treat streptococcal infections.
We thank the staff at beamline BL17U1 of the Shanghai Synchrotron Radiation Facility for excellent technical assistance during the data collection process and Dr. Guiqing Peng for careful revision of this manuscript.
*This work was supported by National Natural Science Foundation of China Grants 31121004 and 30901076, 973 Program Grant 2012CB518805, 863 Program Grant 2011AA10A210, and Chinese Major Special Science and Technology Project Grant 2012ZX10004214.
This article contains supplemental Figs. S1–S4 and Table S1.
4The abbreviations used are: