Correspondence: Marcelo Gottschalk, Centre de Recherche en Infectiologie Porcine (CRIP), Faculté de Médecine Vétérinaire, Université de Montréal, 3200 rue Sicotte, Saint-Hyacinthe, QC, Canada J2S 2M2. Tel.: +450 773 8521, ext. 1 8374; fax: +450 778 8108; e-mail: email@example.com
In our ongoing efforts to develop a vaccine against Streptococcus suis infection, we tested the potential of S. suis enolase (SsEno), a recently described S. suis adhesin with fibronectin-binding activity, as a vaccine candidate in a mouse model of S. suis-induced septicemia and meningitis. Here, we show that SsEno is highly recognized by sera from convalescent pigs and is highly immunogenic in mice. Subcutaneous immunization of mice with SsEno elicited strong immunoglobulin G (IgG) antibody responses. All four IgG subclasses were induced, with IgG1, IgG2a and IgG2b representing the highest titers followed by IgG3. However, SsEno-vaccinated and nonvaccinated control groups showed similar mortality rates after challenge infection with the highly virulent S. suis strain 166′. Similar results were obtained upon passive immunization of mice with hyperimmunized rabbit IgG anti-SsEno. We also showed that anti-SsEno antibodies did not increase the ability of mouse phagocytes to kill S. suis in vitro. In conclusion, these data demonstrate that although recombinant SsEno formulated with Quil A triggers a strong antibody response, it does not confer effective protection against infection with S. suis serotype 2 in mice.
Despite increasing research in recent years, Streptococcus suis continues to cause a variety of diseases in pigs worldwide, including septicemia, meningitis, arthritis and endocarditis (Higgins & Gottschalk, 2005). Among the 35 serotypes described, serotype 2 is considered the most virulent and is most frequently isolated from diseased pigs (Higgins & Gottschalk, 2005). In addition, S. suis has also been described as an important zoonotic agent, especially in Europe and Asia (Lun et al., 2007). Human infections with S. suis are most frequently manifested as purulent meningitis, but septic shock with multiple organ failure, endocarditis, pneumonia, arthritis, and peritonitis have also been reported (Lun et al., 2007).
The pathogenesis of S. suis infection is not fully understood. In swine, the potential portals of entry for S. suis are the palatine and pharyngeal tonsils and thereafter bacteria can spread via the hematogenous or lymphogenous route (Madsen et al., 2002). Once in the bloodstream, S. suis has to resist phagocytosis and killing by phagocytic cells to cause acute septicemia that may lead to septic shock. Bacteria can reach different organs, including the central nervous system (CNS), via different mechanisms that are only partially elucidated to date (Gottschalk & Segura, 2000).
Several potential virulence factors have been implicated in the infection process. The most promising virulence factors to date are the capsule polysaccharide and serum opacity factor (OFS) (Baums et al., 2006), as isogenic mutants lacking either of these factors are rapidly cleared and eliminated from circulation (Smith et al., 1999; Charland et al., 2000; Baums et al., 2006). However, some nonvirulent strains are also encapsulated or have allelic variations of OFS (Takamatsu et al., 2008), indicating that virulence of S. suis likely involves multiple factors (Gottschalk & Segura, 2000). Other virulence candidates have been proposed, but most of them are not present in all virulent strains or are present in nonvirulent strains (Gottschalk et al., 2007). So far, whole-cell vaccines or bacterins (commercial and autogenous) have been used in the field to prevent S. suis disease, however, with disappointing results (Halbur et al., 2000; Lapointe et al., 2002). Vaccination with these bacterins does not induce high levels of antibodies and causes, at most, serotype-specific responses (Higgins & Gottschalk, 2005). We recently identified S. suis enolase (SsEno), a new S. suis surface fibronectin-binding protein (Esgleas et al., 2008) that might participate in the pathogenesis of S. suis infection by mediating bacterial attachment to and internalization into brain microvascular endothelial cells (BMEC) (Esgleas et al., 2008). SsEno might be an attractive vaccine candidate against S. suis infections as it possesses highly conserved epitopes (Ehinger et al., 2004) and is expressed at the surface of all S. suis serotypes described to date (Esgleas et al., 2008). In addition, a recent study identified S. suis enolase as an important antigenic protein that contributes to the virulence of S. suis (Jing et al., 2008). The objective of this study was to determine whether the immune response induced by immunization with purified SsEno can confer protection against challenge with the homologous strain of S. suis serotype 2 in a mouse model of infection.
Materials and methods
Streptococcus suis serotype 2 highly virulent strain 166′ (Berthelot-Herault et al., 2005) was kindly provided by Dr M. Kobisch, AFSSA, Ploufragan, France. Working log-phase cultures were prepared in Todd Hewitt broth (THB) (BD, Sparks, MD) as described previously (Dominguez-Punaro et al., 2007). Growth was allowed until the suspension reached c. 5 × 108 CFU mL−1. Final inoculum corresponded to 107 CFU mL−1 (for pigs) and 108 CFU mL−1 (for mice). Escherichia coli strain BL21DE3 (Novagen, Madison, WI) was used for expression experiments as described elsewhere (Esgleas et al., 2008).
Cloning and expression of the α-enolase gene
Cloning and purification of SsEno were performed as described previously (Esgleas et al., 2008). Briefly, the coding sequence of SsEno was amplified by PCR using chromosomal DNA from S. suis SS166 as template and the complete gene was cloned into pET-32a vector (Novagen). The plasmid pET-32a-SsEno was introduced into E. coli Bl21DE3 for an isopropyl-β-d-thiogalactoside-inducible expression of recombinant S. suis enolase (rSsEno). The His-tagged fusion protein was purified by chromatography under native conditions on HisTrap according to the manufacturer's instructions (Amersham Biosciences AB, Uppsala, Sweden). Protein concentrations were determined by the Lowry method (Lowry et al., 1951). Before vaccination assay, the purity (>95%) of recombinant SsEno preparations was determined by scanning densitometry of the protein on an sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel stained with Coomassie blue and with a silver-stained SDS-PAGE gel (data not shown) as reported previously (Esgleas et al., 2008).
Presence of anti-SsEno antibodies in convalescent animals
All animal experiments were conducted according to the guidelines and protocols set forth by the Canadian Council on Animal Care and approved by the Université de Montréal committee on animal care. Seven 4-week-old piglets from a herd free of S. suis serotype 2 disease were infected intravenously with 107 CFU from a log-phase culture of S. suis strain 166′. Animals were monitored for clinical signs and treated with antibiotics if needed to avoid death. Serum samples were taken before and 3 weeks postinfection. Antibody titers against rSsEno from these convalescent pig sera were measured by direct enzyme-linked immunosorbent assay (ELISA). Maxisorp® flat-bottom microtiter 96-well plates (Nunc, VWR, Mississauga, ON, Canada) were coated overnight at 4 °C with 5 μg mL−1 of purified recombinant SsEno. The plates were further incubated with 1/1000 dilution of pig sera and bound antibodies were detected by incubation with peroxidase-conjugated goat anti-swine immunoglobulin G (IgG) (Jackson Immunoresearch Laboratories, West Grove, PA) for 1 h at room temperature. The plates were developed with TMB substrate (Zymed, S. San Francisco, CA) and absorbance was measured at 450 nm.
Active protection assay of mice with rSsEno
Six-week-old female CD-1 mice (Charles River Laboratories, Wilmington, MA) were immunized subcutaneously twice, 2 weeks apart, with either 20 μg of purified SsEno mixed with 20 μg of Quil-A adjuvant (Brenntag Biosector, Frederikssund, Denmark) or 20 μg of Quil-A only as a control in 100 μL of phosphate-buffered saline (PBS) per mouse. Ten days after the second vaccination, animals were challenged intraperitoneally with 108 CFU per mouse of log-phase S. suis, strain 166′ in 1 mL of THB. Sera collected from each mouse before immunization, before the second dose and before challenge infection were assayed for anti-SsEno antibody titers by ELISA, as described below. Mice were monitored daily for clinical signs such as abnormal behavior (hyperexcitation, episthotonus, opisthotonus, bending of the head toward one side, walking in circles or strong locomotive problems), rough hair coat, ataxia and mortality until day 10 postinfection (Dominguez-Punaro et al., 2007). This mouse model of infection was recently used to reproduce septic shock and meningitis that might be considered to be similar to those induced by S. suis in pigs (Higgins & Gottschalk, 2005; Dominguez-Punaro et al., 2007).
Passive protection of mice with rabbit antibodies against rSsEno
Groups of 12 female CD-1 mice (Charles River, 6 weeks old) were injected intraperitoneally with 0.5 mL of rabbit anti-rSsEno serum (Esgleas et al., 2008) or 0.5 mL of normal rabbit serum as a control. Hyperimmune sera against rSsEno was produced as described previously (Li et al., 2006) and the titer (<1/100 000) evaluated by ELISA (Shah & Swiatlo, 2006). Three hours later, 10 mice per group were injected intraperitoneally with 108 CFU per mouse of log-phase S. suis, strain 166′ in 1 mL of THB. Sera were collected 24 h after the serum administration to measure anti-SsEno antibodies by ELISA assay as described below. Mice were monitored daily for weight loss, clinical signs and mortality.
Determination of active and passive antibody titers in mice by ELISA
Titers of SsEno-specific total IgG in mice sera were determined by ELISA as described previously (Li et al., 2006). Briefly, Polysorb plates (Nunc-Immunoplates, Rochester, NY) were coated overnight at 4 °C with purified recombinant SsEno. Because of an extremely high antibody response, and comparing to what was used to measure antibodies in swine, a significant reduced concentration of SsEno (0.3 μg mL−1) was used to coat the plates. After incubation with serial dilutions of test, bound antibodies were detected by incubation with peroxidase-conjugated goat anti-mouse IgG, IgG1, IgG2a, IgG2b or IgG3 antisera (Serotec, Kidlington, Oxford, UK) for 1 h at room temperature. For determination of antibodies titers in the passive protection assay, the same protocol was used but with peroxidase-conjugated anti-rabbit IgG (Jackson ImmunoResearch Laboratories) as a secondary antibody. Plates were then developed as described above. The antibody titer was considered as the serum dilution that resulted in an OD450 nm reading of 0.1 after subtracting background.
Killing of S. suis serotype 2 strain 166′ by mouse phagocytes in the presence of anti-SsEno antibodies
In vitro killing of S. suis serotype 2 strain 166′ by mouse phagocytes in the presence of polyclonal anti-SsEno antibodies was measured as described previously (Lancefield, 1957; Pancholi & Fischetti, 1998) with some modifications. Briefly, 0.3 mL (from five different animals) of freshly heparinized blood was mixed with 0.5 mL of appropriately THB diluted bacteria (200–500 CFU mL−1) in the presence of 25 μg mL−1 of protein G purified rabbit anti-SsEno antibodies (Esgleas et al., 2008). Similar concentrations of purified normal rabbit antibodies were used as negative control and of purified hyperimmune rabbit anti-S. suis (whole cell) antibodies as a positive control, as described previously (Chabot-Roy et al., 2006). Mixtures were incubated at 37 °C for 3 h with constant slow rotation (Lancefield, 1957; Pancholi & Fischetti, 1998). At the end of the incubation, an appropriately diluted aliquot was plated onto THB agar and incubated overnight at 37 °C to count surviving bacteria. Results are expressed as mean percentage±SD of bacterial survival, with survival of bacteria opsonized with normal rabbit antibodies considered as 100%. Data are representative of four independent experiments. An internal control for the growth of S. suis serotype 2 strain 166′ in the presence of the different conditions was also included. Briefly, the same concentration of diluted bacteria (200–500 CFU mL−1) were grown in THB in the presence of 25 μg mL−1 of the different purified antibodies or no antibodies for 3 h in stationary conditions and the number of CFU in each condition were counted.
Antibody titers and percentage of killed bacteria of experimental groups were compared using Student's t-test (***P<0.0005, **P<0.005, *P<0.05). Survival curves were evaluated using the Kaplan–Meier method and the significance of the difference was tested using the Log-rank test.
Immunogenicity of SsEno in pigs
None of the seven animals used in this study had SsEno-specific antibodies before the experimental infection (Fig. 1), as only background values were detected in these animals. In contrast, pigs infected with a log-phase culture of S. suis strain 166′ showed significant anti-SsEno-IgG responses at 3 weeks postinfection (Fig. 1).
SsEno-specific IgG and IgG subclasses in active protection assay
There were no SsEno-specific antibodies in sera from any of the animals before the first vaccination (Fig 2a). Immunization of mice with SsEno elicited a strong antigen-specific response. At 14 days after the first vaccination, SsEno elicited a significant IgG response that was further increased after the second immunization (Fig. 2a). Analysis of sera demonstrated that although SsEno-immunized animals produced all IgG subclasses, IgG1, IgG2a and IgG2b responses were predominantly followed by IgG3 (Fig. 2b). In contrast, animals vaccinated only with adjuvant did not show any antibody responses (Fig. 2a and b).
Clinical signs and mortality in the active protection assay
A few hours after challenge infection with S. suis serotype 2 strain 166′, all mice (control and SsEno immunized group) exhibited clinical signs, such as ruffled hair coat suggesting fever and slow response to stimuli. From day 0 to 3 postinfection, 25% of immunized animals died from septicemia (Fig. 3a). After day 3 postinfection, 58% of the immunized animals that survived septicemia developed severe CNS signs such as running in circles and opisthotonos and died or met criteria for euthanasia due to the severity of their condition (Fig. 3a). Similar results were observed in the control group; although all the animals resisted the septicemic phase, 33% of animals died after day 3 postinfection due to meningitis (Fig. 3a). In all dead animals, S. suis serotype 2 was isolated from different organs (data not shown).
During the septicemic phase, the immunized group lost c. 15% of their body weight (Fig. 3b). Similar results were seen in the control group (Fig. 3b). In both groups, animals that survived the septicemic phase were able to recover their initial body weight at the end of the experiment (Fig. 3b).
Clinical signs and mortality in the passive protection assay
A few hours after infection, passively immunized animals presented similar clinical signs as those vaccinated in the active protection assay. During the first 12 h, 30% of animals immunized with anti-SsEno antibody died from septicemia compared with 70% of the control group. However, all passively immunized animals as well as control mice died within the first 6 days postinfection from either septicemia or meningitis (Fig. 4). Analysis of sera clearly indicated that there was a high titer of rabbit anti-SsEno only in the passive immunized group (data not shown).
Killing of S. suis serotype 2 strain 166′ by mouse phagocytes
To further evaluate the bactericidal/opsonic capacity of anti-rSsEno antibodies, a killing assay was performed using mouse whole blood. As shown in Fig. 5, S. suis treated with normal rabbit IgGs grew well in mouse blood (negative control). In contrast, S. suis treated with rabbit anti-whole S. suis IgGs was rapidly killed as expected (positive control). Incubation of S. suis with affinity purified anti-SsEno IgG antibodies did not enhance bacterial killing, confirming results obtained in the passive protection assay.
Several approaches have been used to develop vaccines for S. suis. However, little success has been achieved thus far because the protection elicited was either serotype or strain dependent, and results in most instances have been equivocal (Haesebrouck et al., 2004). For example, killed whole cells or live avirulent vaccines can provide partial protection but only with repeated immunization (Busque et al., 1997; Wisselink et al., 2001). Protein-based subunit vaccines have also been tested using virulence markers such the hemolysin (Jacobs et al., 1996), the muramidase-released protein and extracellular protein factor (Wisselink et al., 2001) that have been shown to protect pigs from homologous and heterologous serotype 2 strains. However, their use is hindered by the fact that a substantial number of virulent strains in some geographical regions do not express these proteins (Gottschalk & Segura, 2000). More recently, a surface expressed protein (SAO) has been observed to confer protection against experimental infection in mice and pigs (Li et al., 2007). However, as this protection is not complete, other proteins could be combined with SAO to optimize protection.
Fibronectin-binding proteins have been suggested as potential vaccine targets for preventing bacterial infections because antibodies directed against such adhesins may prevent bacterial attachment and also enhance opsonization and killing by leukocytes (Wizemann et al., 1999; Rennermalm et al., 2001). In this study, we tested the potential of the recently described fibronectin-binding protein, SsEno, as a vaccine candidate in a mouse model of S. suis-induced septicemia and meningitis (Dominguez-Punaro et al., 2007). SsEno is a good vaccine candidate based on the following characteristics: (1) it is present at the surface of all 35 different S. suis serotypes (Esgleas et al., 2008); (2) it contributes to S. suis adhesion to and invasion of host cells (Esgleas et al., 2008); and (3) it is a highly conserved protein (Ehinger et al., 2004). In addition, we further showed that sera from convalescent animals strongly recognize this protein, suggesting expression of SsEno in vivo. Very recently, another study also suggested an important role of enolase in the virulence of S. suis (Jing et al., 2008). However, the potential of this protein to confer protection against other pathogens is controversial. Mice immunized with antibodies against Plasmodium falciparum enolase are protected from challenge infection with a lethal mouse malaria strain (Pal-Bhowmick et al., 2007). In contrast, recombinant enolase from Candida albicans induces only modest protection against disseminated candidiasis (Montagnoli et al., 2004). To our knowledge, a protection test against streptococci using enolase as an antigen to elicit an immune response has not been reported yet.
To evaluate SsEno as a vaccine candidate against S. suis infection, we used a highly virulent strain for swine in a well-standardized mouse model, which presents two different phases: (1) a septicemic phase, (24–48 h postinfection) and (2) a meningitis/encephalitis phase, where animals that survive septicemia will subsequently die from a serious infection of the CNS (Dominguez-Punaro et al., 2007). Our results indicated that the antibodies actively elicited by SsEno at the concentration used in combination with Quil-A adjuvant was not protective from either phase of challenge infection with S. suis. The Th1-type responses and IgG subclasses induced by the SAO protein are two main components of host immunity against S. suis infection (Li et al., 2007). In this study, all SsEno-specific IgG subclasses (IgG1, IgG2a, IgG2b and IgG3) were induced, with IgG1 and the Th1-type antibodies, IgG2a and IgG2b, as the predominant subclasses. Therefore, other components of the host immune response besides nonopsonic antibodies may be necessary for effective protection against S. suis infection. We attributed the fact that the vaccinated group presented a higher level of mortality than the control group to a simple variation in an animal assay. Although it would have been tempting to interpret the results as a possible role of antibodies against enolase playing a certain role in autoimmunity as described (Terrier et al., 2007), animals from both groups were perfectly healthy at the moment of challenge.
To confirm these results, we tested whether rabbit anti-SsEno serum provides passive protection against challenge infection with the same strain of S. suis in mice. As expected, and although there was a slight delay in the appearance of severe clinical signs, passive immunization with anti-SsEno serum did not confer significant protection against S. suis infection. As it has been shown that antibodies derived from hyperimmune sera against whole bacteria (including antibodies against the capsule) can protect mice against infection (Charland et al., 1997) and induce bacteria killing (Chabot-Roy et al., 2006), it is possible that anti-SsEno antibodies are not opsonic, allowing bacteria to reproduce in high numbers to cause disease. On the other hand, antibodies against other fibronectin-binding proteins have already been described to enhance phagocyte killing in other bacterial species (Rennermalm et al., 2001). To test this, we evaluated the capacity of the purified anti-SsEno IgGs to opsonize bacteria and promote their killing by murine phagocytes. Although most of the produced antibodies were of the Th1 type, results obtained in this study showed that the anti-SsEno IgG subclasses were not able to induce S. suis killing by mouse phagocytes.
Because anti-SsEno has been shown to reduce bacterial adhesion and invasion to BMEC in vitro (Esgleas et al., 2008), we had expected a certain degree of protection against the CNS phase of infection. However, SsEno is not the only receptor involved in such process (Vanier et al., 2007). It is possible that the high concentration of S. suis in blood, as a consequence of a lack of bacterial killing by leukocytes, may overcome the partial inhibition of bacterial-BMEC interactions mediated by anti-SsEno antibodies.
In summary, although we demonstrated that SsEno elicits an important antibody response in convalescent pigs and immunized mice, this response as evaluated in the present study is inadequate for effective protection against S. suis infection.
This work was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC) grant 0680154280.