Characterization of the invasion of porcine endothelial cells by Streptococcus suis serotype 2

Abstract

Streptococcus suis serotype 2 is an important swine pathogen associated mainly with meningitis. In a previous study, we demonstrated the ability of S. suis serotype 2 to adhere to and invade immortalized porcine brain microvascular endothelial cells (PBMECs) forming the blood–brain barrier. The aim of the current work was to further characterize the mechanism(s) by which S. suis invades porcine endothelial cells. The ability of several S. suis strains to interact with PBMECs was not found to correlate with their geographic origin, virulence, host of origin, or suilysin production. Characterization studies demonstrated that proteinaceous adhesins/invasins, cell wall components, lipoteichoic acid, and serum components (including fibronectin) were involved in interactions between S. suis and PBMECs. In addition to PBMECs, S. suis was able to adhere to and invade 2 porcine aortic endothelial cell lines and primary PBMECs.

Introduction

Streptococcus suis serotype 2 is an important swine bacterial pathogen associated mainly with meningitis but also with other infections, such as endocarditis, arthritis, septicemia, and pneumonia (1). Until recently, 35 serotypes were commonly accepted, but a study by Hill and colleagues (2) demonstrated that serotypes 32 and 34 should be clustered with Streptococcus orisratti. Serotype 2 is the most frequent serotype recovered from diseased animals (1). As a zoonotic agent, S. suis has been isolated from human cases of meningitis, endocarditis, and recurrent toxic shock (1). Although human cases are rare, in July 2005 an outbreak of S. suis that affected more than 200 people, mostly Chinese farmers who had slaughtered diseased pigs, was reported; the mortality rate was more than 20% (3).

Several virulence factors, such as hemolysin (suilysin) (4), a fibronectin- and fibrinogen-binding protein (FBPS) (5), adhesins (6), proteases (7), and other proteins (8,9), have been proposed. The only factor identified thus far as playing a critical role in pathogenesis is capsular polysaccharide (CPS), a finding based on evidence that nonencapsulated isogenic mutants are nonpathogenic and are rapidly cleared from the bloodstream in both pig and mouse models of infection (10,11). However, natural nonvirulent strains are encapsulated, which indicates that other virulence factors are also involved. Interestingly, European and North American strains of S. suis serotype 2 differ in their expression of virulence-associated markers and the production of suilysin (1). In addition, it has been suggested that European strains are more virulent than North American strains (12).

The pathogenesis of S. suis infection is still unclear. Once in the bloodstream, bacteria are able to reach the central nervous system (CNS). Indeed, S. suis is frequently isolated from the brain of diseased pigs with clinical signs of meningitis (1). One way by which bacteria could gain access to the CNS is by crossing the blood–brain barrier, which separates the CNS from the bloodstream and is formed in part by the endothelium lining the brain capillaries. This endothelium is composed of brain microvascular endothelial cells (BMECs), is characterized by tight intercellular junctions and the presence of pericytes within the capillary basement membrane, and is surrounded by a sheath formed by astrocyte foot processes (12,13).

In a recent study, we demonstrated the ability of S. suis serotype 2 to adhere to and invade immortalized porcine brain microvascular endothelial cells (PBMECs) (14). Interestingly, in most cases of bacterial meningitis, the BMECs seem to be the primary site of breakdown of the blood–brain barrier (13). Thus, the invasion of porcine endothelial cells forming the blood–brain barrier by S. suis is likely an important step in the pathogenesis of meningitis caused by S. suis. To better understand the interactions between S. suis and PBMECs, we aimed to further characterize the mechanisms used by S. suis to invade porcine endothelial cells.

Materials and methods

Bacterial strains and growth conditions

In this study, S. suis serotype 2 suilysin-positive virulent strain 31533 (14) served as the reference strain. In selected experiments, S. suis serotype 2 suilysin-negative virulent strain 89-1591 (15), used in a previous study with PBMECs, was also included (14). In addition, several S. suis isolates were used in comparative studies. The present study also involved an isogenic mutant derived from strain 31533: the nonencapsulated mutant B218, which was produced in our laboratory by allelic exchange and corresponded to a previously reported transposon-derived mutant 2A (10). Mutant B218 was shown to possess the same characteristics as mutant 2A, with a complete absence of capsular material at the bacterial surface. We also evaluated an isogenic knockout FBPS mutant and its wild-type S. suis strain, both kindly provided by Dr. Astrid de Greeff, Institute for Animal Science and Health, Lelystad, the Netherlands (5).

The bacteria were grown as previously described (14), were washed twice in phosphate-buffered saline (PBS), pH 7.3, and were appropriately diluted in cell culture medium before inoculation. An accurate determination of the number of colony-forming units (CFUs) per milliliter in the final suspension was made by plating onto Todd Hewitt Broth (THB) agar (Difco Laboratories, Detroit, Michigan, USA).

Cell culture

The porcine brain microvascular endothelial cell line PBMEC/C1-2 (16) was cultured as previously described (14). Briefly, cells were grown in complete IF medium, which is a mixture of 1:1 Iscove’s modified Dulbecco’s medium and Ham’s F-12 (Invitrogen, Burlington, Ontario) supplemented with 7.5% (v/v) heat-inactivated fetal bovine serum (FBS), penicillin–streptomycin (Invitrogen), sodium bicarbonate, l-glutamine, human transferrin (MP Biomedicals, Solon, Ohio, USA), N-acetylcysteine, hypoxanthine, porcine heparin, human recombinant fibroblast growth factor- basic (Sigma, Oakville, Ontario), and β-mercaptoethanol (BioRad Laboratories, Hercules, California, USA). The porcine aortic endothelial cell line PAEC₁₁, kindly provided by Dr. Bernard Weill, Université Paris V, Paris, France (17), and the PAEC line AOC, kindly provided by Dr. José Yélamos, Hospital Universitario Virgen de la Arrixaca, Murcia, Spain (18), were grown in RPMI 1640 medium (Invitrogen) supplemented with 10% (v/v) FBS, penicillin–streptomycin, and l-glutamine (MP Biomedicals). Flasks (Falcon; Becton Dickinson, Mississauga, Ontario) and 24-well tissue culture plates (Primaria; Falcon) were precoated with 1% (w/v) type A gelatin from porcine skin (Sigma) to support the cells. Cells were incubated at 37°C with in a humid atmosphere and used before the 20th passage 5% CO₂ in all experiments. For assays, we trypsinized PBMEC, PAEC₁₁, and AOC cells by adding a trypsin–ethylene diamine tetraacetic acid (EDTA) solution (Invitrogen) and then diluted the mixture in the culture medium at 8 × 10⁴, 1.5 × 10⁵, and 1.2 × 10⁵ cells/mL, respectively. The cell suspension was distributed in tissue culture plates and incubated to confluence. Primary PBMECs (pPBMECs) were purchased from Cell Systems Corporation (Kirkland, Washington, USA) and grown according to the supplier’s recommendations. For pPBMEC assays, the cell suspension was distributed in tissue culture plates at a concentration of 1 × 10⁵ cells/mL and incubated to confluence. Immediately before the experiments, medium was removed from the plates and replaced by the respective medium without antibiotics.

Cell adhesion and invasion assays

The adhesion assay to quantify total cell-associated (intracellular plus surface-adhered) bacteria was performed as previously described (14), with some modifications. Briefly, bacteria were pelleted, washed twice with PBS, and resuspended at 10⁶ CFU/mL in fresh cell culture medium without antibiotics. Confluent cell monolayers were inoculated with 1-mL aliquots of bacterial suspension. The plates were centrifuged at 800 × g for 10 min to bring bacteria to the surface of the monolayer and incubated for 2 h at 37°C with 5% CO₂. The monolayers were then vigorously washed 5 times to eliminate nonspecific bacterial attachment. The presence of S. suis was verified in the last wash: no significant numbers of S. suis were found (data not shown). The plates were then incubated for 10 min at 37°C in the presence of 200 μL of 0.05% trypsin–0.03% EDTA. Next, 800 μL of ice-cold deionized water was added, and cells were disrupted by scraping the bottom of the well and by repeated pipetting to liberate cell-associated bacteria. Serial dilutions of this cell lysate were plated onto THB agar and incubated overnight at 37°C, after which the bacteria were counted. To ensure that the use of gelatin in the well did not provoke an artificially high estimate of the number of adherent S. suis, an adhesion assay was performed directly into the gelatin-coated well (without endothelial cells): S. suis was shown not to adhere to this protein (data not shown). Levels of adhesion were expressed as the total number of CFUs recovered per well. Cytotoxicity controls were performed with the CytoTox 96 nonradioactive cytotoxicity assay (Promega, Madison, Wisconsin, USA): no cytotoxicity was detected under the assay conditions (data not shown).

The invasion assay to quantify intracellular bacteria was performed in a similar manner. However, to kill extracellular and surface-adhered bacteria after the initial infection period, cells were washed twice with PBS, and 1 mL of cell culture medium containing 100 μg/mL of gentamicin and 5 μg/mL of penicillin G (Sigma) was added to each well. After incubation for 1 h at 37°C with 5% CO₂, monolayers were washed 3 times with PBS before incubation with 0.05% trypsin–0.03% EDTA. Levels of invasion were expressed as the total number of CFUs recovered per well. The efficiency of antibiotic treatment in killing extracellular bacteria was confirmed by plating a 100-μL sample of the last PBS wash suspension onto THB agar (data not shown).

Studies to characterize PBMEC invasion by S. suis

To study the role of proteins in bacterial invasion, we treated S. suis strain 31533 (10⁶ CFU/mL) with 0.1 to 2 mg/mL of proteinase K (Roche Diagnostics, Laval, Quebec), 50 to 500 μg/mL of pronase (Roche), or 0.5 to 4 mg/mL of trypsin (Invitrogen) for 1 h at 37°C, then washed the bacteria twice in PBS and resuspended them in culture medium before PBMEC inoculation. To evaluate the role of bacterial cell surface components, we performed competitive studies by pretreating PBMECs at 37°C with 10 to 200 μg/mL of purified S. faecalis lipoteichoic acid (LTA; Sigma) for 90 min, 100 μg/mL of sialic acid (Sigma) for 60 min, or 1 to 100 μg/mL of purified cell wall of S. suis serotype 2 for 30 min. Since LTA from S. suis is currently not available, but S. suis LTA and S. faecalis LTA both react with group D antiserum and have some structural similarities (19), we used LTA from S. faecalis as an appropriate alternative. Purified cell wall of S. suis was produced as previously described (20) and contained less than 7% (w/w) of protein, as detected with the Micro BCA Protein Assay Kit (Pierce, Rockford, Illinois, USA). To verify the role of soluble fibronectin in bacterial invasion, S. suis strain 31533 (10⁶ CFU/mL) was also preincubated with 250 μg/mL of human plasma fibronectin (Roche Diagnostics) for 1 h at 37°C, washed in PBS, and resuspended in culture medium before PBMEC inoculation. A standard PBMEC invasion assay was also performed in the absence of serum, in the presence of 50% (v/v) FBS, or in the presence of 7.5% to 50% (v/v) heat-inactivated swine serum (free of anti-S. suis antibodies).

Statistical analysis

All data were expressed as mean (and standard deviation). Data were analyzed by a 2-tailed, unpaired t test. To compare the adhesion and invasion capacities of several S. suis strains, we used a linear mixed model. A P-value of less than 0.05 was considered significant. All assays were repeated at least 3 times.

Results and discussion

Comparison of adhesion and invasion capacities of several S. suis strains

Several S. suis serotype 2 strains (Table I) were compared for their ability to adhere to and invade PBMECs according to their host and geographic origins, production of suilysin, and virulence. Strains of human or swine origin showed similar levels of adhesion and invasion, in agreement with previous studies showing no differences among human and swine strains in the level of adhesion to human BMECs or human and swine epithelial cells (21,22). Although it has been reported that serotype 2 strains of different geographic origins are phenotypically and genotypically different (1), there were no significant differences in mean adhesion and invasion levels between the European and North American strains that we evaluated (Figure 1). In addition, most European strains produce suilysin (1). As was reported from a previous study using a suilysin-negative mutant (14), we observed no correlation between suilysin production and level of adhesion or invasion in a comparative analysis of the different strains. Collectively, these data suggest that factors additional to the capacity for adhesion and invasion of PBMECs may account for the observation that serotype 2 strains from Europe are more virulent than North American strains (12,15).

Table I.

Capacities of strains of Streptococcus suis to adhere to and invade porcine brain microvascular endothelial cells

					Colony-forming units recovered per well; mean (and standard deviation)
Strain	Geographic origin	Virulence	Suilysin production	Host of origin	Adhesion assay	Invasion assay
LEF95	France	+	+	Human	9.73 (1.42) × 105	9.30 (4.27) × 102
H11/1	United Kingdom	+	+	Human	2.76 (1.38) × 105	4.83 (1.59) × 102
AR770353	The Netherlands	+	−	Human	8.87 (1.46) × 105	1.05 (0.30) × 103
166’	France	+	+	Swine	7.38 (4.74) × 104	7.18 (4.11) × 102
90-1330	Canada	−	−	Swine	7.83 (1.32) × 105	6.53 (2.32) × 102
94-623	France	−	−	Swine	7.10 (1.71) × 104	1.04 (0.15) × 103
SX332	United States	+	+	Swine	3.37 (2.73) × 105	2.31 (1.54) × 103
95-8242	Canada	+	+	Swine	2.38 (1.40) × 105	2.10 (1.39) × 103
98-B575	Canada	+	−	Swine	1.63 (0.76) × 105	1.02 (0.42) × 103
98-B099	Canada	+	−	Swine	1.49 (0.64) × 106	2.52 (1.10) × 103
98-8993	Canada	+	−	Swine	3.30 (2.46) × 105	6.03 (0.60) × 103
S735	The Netherlands	+	+	Swine	4.10 (0.92) × 104	3.70 (1.66) × 102
D282	The Netherlands	+	+	Swine	6.82 (2.19) × 104	2.28 (0.82) × 103
24	France	+	+	Swine	3.58 (2.44) × 105	1.37 (0.31) × 104
89-999	Canada	+	−	Swine	6.03 (2.70) × 105	4.47 (3.97) × 102
AAH4	United States	+	−	Swine	3.07 (0.20) × 105	7.00 (0.36) × 102
T15	The Netherlands	−	−	Swine	1.41 (0.39) × 104	3.97 (1.56) × 102
3889	The Netherlands	−	−	Swine	8.20 (3.22) × 104	8.53 (4.77) × 102
4005	The Netherlands	+	+	Swine	1.10 (0.75) × 105	1.54 (0.64) × 103
89-1591	Canada	+	−	Swine	1.11 (0.52) × 104	2.80 (1.30) × 102
31533a	France	+	+	Swine	4.31 (2.30) × 104	1.18 (0.33) × 103

^aReference strain used in this study

Figure 1.

Interactions between porcine brain microvascular endothelial cells (PBMECs) and Streptococcus suis according to geographic origin and phenotype (virulent or not) of the S. suis strain. In assays, adhesion to or invasion of PBMECs by S. suis strains was expressed as the mean number (bar) of colony-forming units (CFUs) recovered per well for each group of strains 2 h after inoculation of 10⁶ CFU/mL.

Although the number of nonvirulent strains that we evaluated was low, these strains showed mean levels of adhesion and invasion similar to those observed with virulent strains (Figure 1). However, as described for Escherichia coli K1 (23), a high-grade bacteremia was required for S. suis to reach the CNS, thus suggesting a critical role for bacterial survival and dissemination in blood in the process of S. suis invasion of the blood–brain barrier at high levels (24).

Characterization of PBMEC invasion by S. suis

To verify if bacterial proteins play a role in the invasion of PBMECs, bacteria were pretreated with several proteases with different proteolytic activities. As shown in Figure 2, proteinase K, pronase, and trypsin inhibited the invasion of PBMECs by S. suis strain 31533 in a dose-dependent manner. Proteases were also used in the adhesion assays. Bacteria treated with 2 mg/mL of proteinase K, 500 μg/mL of pronase, and 4 mg/mL of trypsin showed 53%, 55%, and 29% adhesion to PBMECs, respectively, compared with non-treated bacteria (considered as showing 100% adhesion). None of the proteases affected the bacterial viability at the concentrations used (data not shown). These results provide evidence that proteinaceous adhesins/invasins may be important for the invasion of PBMECs by S. suis. However, the specific adhesin(s) involved in S. suis interactions with host cells may differ among distinct cell types. Thus, although proteases similar to those used in this study were reported to inhibit the interaction of the Galα1–4 Gal-binding P adhesin with erythrocytes (25), these proteases fail to block the adhesion of S. suis to J774 macrophages and porcine epithelial cells (22,26).

Figure 2.

Inhibition of invasion of PBMECs by S. suis strain 31533 by pretreatment of the bacteria with the indicated concentrations of proteases at 37°C for 1 h before inoculation of 10⁶ CFU/mL. An asterisk indicates a significant difference (P < 0.05) compared with the level of invasion without protease treatment (considered as 100%).

Cell wall components have been demonstrated to be responsible for the adhesion of pathogens to endothelial cells (27). Therefore, we investigated whether these components were also involved in the adhesion of S. suis to PBMECs. Adhesion was reduced to 44% by pretreatment of PBMECs with 100 μg/mL of purified cell wall, and invasion was inhibited in a dose-dependent manner (Figure 3). Pretreatment of PBMECs with 200 μg/mL of LTA, a cell wall component, reduced adhesion to 43% and inhibited invasion in a dose-dependent manner (Figure 3). Thus, our results suggest that the bacterial cell wall may contain adhesins in addition to proteinaceous adhesins/invasins that mediate interactions between S. suis and PBMECs. A previous study also demonstrated a high level of inhibition of bacterial adhesion to porcine epithelial cells with similar doses of S. suis purified cell wall (22). Indeed, cell wall components of other streptococcal pathogens, such as S. pneumoniae and group A and B streptococci, have been previously described to be responsible for bacterial adhesion to host cells (27–29). Our results indicate that LTA may play a role in the interactions between S. suis and PBMECs. Thus, similar to S. pyogenes, S. suis could possess multiple adhesins (LTA and proteins) that mediate its interactions with host cells (28).

Figure 3.

Inhibition of invasion of PBMECs by S. suis strain 31533 by pretreatment of the PBMEC monolayers with the indicated concentrations of purified S. faecalis lipoteichoic acid (LTA) or purified cell wall at 37°C for 90 and 30 min, respectively, before inoculation of 10⁶ CFU/mL. An asterisk indicates a significant difference (P < 0.05) compared with the level of invasion without purified bacterial component (considered as 100%).

Since S. suis capsule contains sialic acid (12), which has been shown to inhibit adhesion to macrophages (26), PBMECs were also pretreated with sialic acid to determine the role of this capsular component in S. suis interactions with PBMECs. Pretreatment with sialic acid did not significantly inhibit invasion at a concentration of 100 μg/mL (data not shown), the concentration that significantly inhibited adhesion of S. suis to macrophages (26). Thus, different adhesins may be involved in the adhesion of S. suis to distinct types of cells.

Pretreatment of bacteria with human plasma fibronectin induced strong increases in adhesion (553% [190%]) and invasion (703% [214%]) compared with the absence of fibronectin (considered as 100%). As was recently reported, S. suis is able to bind to plasma and cellular fibronectins (30) and possesses an FBPS (5). Our results suggest that S. suis uses plasma fibronectin as a bridge between bacteria and the cell surface. Interactions mediated by bridging fibronectin were reported for Staphylococcus aureus with human umbilical vein endothelial cells (HUVECs) and for S. pyogenes with lung epithelial cells (31,32). However, other pathogens, such as Enterococcus faecalis, bind to fibronectin attached to the surface of the target host cell (33).

Interestingly, the described FBPS (5) may not be the only S. suis fibronectin-binding protein, since we observed no significant differences in levels of adhesion and invasion between the FBPS mutant and its wild-type strain (data not shown). It is possible that S. suis possesses multiple fibronectin receptors, as do other streptococci (28).

To test if serum components play a role in the invasion of PBMECs by S. suis, we performed invasion assays in the absence of serum or the presence of different concentrations of FBS or swine serum. An absence of serum or different concentrations of FBS did not significantly affect levels of invasion; however, swine serum significantly increased the level of invasion, in a dose-dependent manner (Figure 4). Serum components (including fibronectin) could act as molecular bridges between the bacteria and cell surface receptor(s). Given its abundance in blood, plasma fibronectin would be readily available for use by circulating S. suis (34). However, these factors may not be absolutely necessary for invasion, since S. suis was able to invade PBMECs in the absence of serum. It is possible that factors present in FBS (including plasma fibronectin) were not available at sufficient levels to mediate interactions with S. suis or were not recognized by S. suis owing to their bovine origin. Thus, interactions of S. suis with PBMECs likely involve multiple adhesins as well as serum components that act as bridges. Consistent with our results, group A streptococci possess several adhesins of different types that mediate their complex interactions with host cells (28).

Figure 4.

Effect of serum type and concentration on the level of PBMEC invasion by S. suis strain 31533 (10⁶ CFU/mL). A standard invasion assay was performed in the presence of different concentrations of fetal bovine serum (FBS) or swine serum as well as in the absence of serum. An asterisk indicates a significant difference (P < 0.05) compared with the level of invasion with 7.5% FBS (considered as 100% invasion).

Interactions between S. suis and pPBMECs

Culturing primary cells is time-consuming, and results obtained from these cells are subject to variation among batches. To overcome these problems, immortalized cell lines are frequently used as in vitro models. However, it was reported that cell immortalization strongly enhanced Listeria monocytogenes invasion of cells from the porcine ileum (35). Our results show that this is not the case with S. suis. As shown in Figure 5, S. suis was able to adhere to and invade pPBMECs. Indeed, strain 31533 showed similar levels of binding to and invasion of both immortalized and pPBMECs, whereas strain 89–1591 showed slightly higher levels of binding and invasion of pPBMECs compared with immortalized PBMECs. In addition, as was previously reported with the PBMEC cell line (14), adhesion to and invasion of pPBMECs depended on bacterial concentration (data not shown). Confirming results obtained with PBMECs (14), a suilysin-negative mutant showed levels of adhesion to and invasion of pPBMECs similar to those observed with the wild-type strain (unpublished observations).

Figure 5.

Interactions of S. suis strains 31533 and 89–1591 with immortalized PBMECs or primary PBMECs (pPBMECs) 2 h after inoculation of 10⁶ CFU/mL.

Interactions between S. suis and PAEC cell lines

Cells from large vessels differ from those of microvascular origin in the expression of receptors and adhesion molecules (36). Moreover, microvascular endothelial cells form tight junctions, whereas the endothelium of large vessels is fenestrated (13). These differences may help explain the observation that group B streptococci invade microvascular endothelial cells at higher levels than pulmonary artery cells (37). It is not known whether the level of invasion by S. suis also varies according to the type of blood vessel. Since S. suis has been reported to cause diseases in addition to meningitis, such as endocarditis and pneumonia (1), the ability of S. suis to adhere to and invade porcine endothelial cells from different tissues may be relevant to bacterial pathogenesis. To investigate this, we inoculated 2 different PAEC cell lines, AOC and PAEC₁₁, with different S. suis strains. As shown in Figure 6, S. suis adhered to and invaded the 2 cell lines at similar levels, and these interactions depended on the bacterial concentration (data not shown). These results suggest that receptor(s) for interactions between S. suis and porcine cells are present in both microvascular endothelial cells from the blood–brain barrier and aortic endothelial cells. Interestingly, these results differ from those obtained with human cells. Our laboratory has previously reported that S. suis neither adheres to nor activates HUVECs but is able to adhere to and induce the release of proinflammatory cytokines from human BMECs (21,38). Therefore, unlike porcine endothelial cells, human BMECs but not HUVECs may have receptor(s) for interactions with S. suis. In the present work, similar results were obtained with a suilysin-negative mutant and its wild-type strain (unpublished observations). These results are in agreement with those previously reported for PBMECs (14). Taken together, our results demonstrate that suilysin is not required for S. suis interactions with porcine endothelial cells.

Figure 6.

Interactions of S. suis strains 31533 and 89–1591 with porcine aortic endothelial cell lines AOC and PAEC₁₁ 2 h after inoculation of 10⁶ CFU/mL.

Survival of bacteria in blood is a critical step to enable the bacteria to access and invade the CNS. There is much evidence that CPS is an important factor that promotes bacterial resistance to phagocytic clearance (10,11). Compared with its wild-type strain 31533, the nonencapsulated mutant B218 showed higher levels of adhesion to and invasion of AOC (P < 0.05; data not shown). Similarly, the levels of adhesion to and invasion of PBMECs were significantly increased in the absence of CPS (14). The higher levels of interactions of nonencapsulated S. suis with aortic endothelial cells as well as PBMECs may be attributed to the exposition of cell wall components required for bacterial interaction with receptors present in both cell types. These data, together with those obtained in the inhibition studies (Figures 2 and 3), suggest that CPS is not involved in the adhesion or invasion process and that cell wall components (proteins and LTA, for example) may play an important role in these interactions. Although the ability of the capsule to interfere with bacteria– endothelial cell interactions has been reported for Haemophilus influenzae (39) and S. pneumoniae (40), encapsulated S. suis is still able to interact with host cells, which suggests that adhesins/invasins are exposed at the cell surface.

Conclusion

Many questions remain unsolved concerning the pathogenesis of meningitis caused by S. suis serotype 2. One critical gap in our knowledge is the mechanism(s) by which the bacteria enter the CNS (12). Recently, we demonstrated the capacity of S. suis serotype 2 to bind to and penetrate immortalized PBMECs (14), and Tenenbaum and associates (41) demonstrated that S. suis induces a loss of the blood–cerebrospinal fluid barrier function in an in vitro model. Thus, identifying the mechanisms used by S. suis to gain access into the CNS may have a significant impact on our understanding of the pathogenesis of meningitis caused by this bacterium.

Our results suggest that multifactorial mechanisms are used by S. suis to interact with PBMECs. These include cell surface protein(s) and cell wall components (mainly LTA) that mediate bacterial adhesion and invasion. Our results also show that S. suis is able to interact with endothelial cells derived from various host tissues, suggesting a potential mechanism for bacterial dissemination. Studies are ongoing to further identify the molecules responsible for attachment and subsequent invasion of porcine endothelial cells by S. suis.