Enhanced Immunogenicity of a Genetic Chimeric Protein Consisting of Two Virulence Antigens of Streptococcus mutans and Protection against Infection

Abstract

The saliva-binding region (SBR) of the cell surface antigen I/II (AgI/II) and the glucan-binding region (GLU) of the glucosyltransferase enzyme of Streptococcus mutans have been implicated in the initial adherence of S. mutans to saliva-coated tooth surfaces and the subsequent sucrose-dependent accumulation of S. mutans, respectively. Here, we describe the construction and characterization of a genetic chimeric protein consisting of the two virulence determinants SBR and GLU (SBR-GLU). The effectiveness of this construct in inducing mucosal and systemic immune responses to each virulence determinant following intranasal immunization was compared to that of each antigen alone or an equal mixture of SBR and GLU (SBR+GLU) in a mouse model. Furthermore, the ability of antibodies induced to SBR-GLU to protect against S. mutans infection was also investigated. Immunization of mice with the chimeric protein SBR-GLU resulted in significantly enhanced (P < 0.001) levels of serum immunoglobulin G (IgG) anti-SBR antibody activity compared to those in the SBR and SBR+GLU groups. The SBR-GLU-immunized mice also demonstrated a significant (P < 0.05) increase in salivary and vaginal IgA antibody responses to SBR and GLU. Analysis of the serum IgG subclass responses to SBR in mice immunized with SBR alone indicated a mixed IgG1 and IgG2a response. A preferential IgG1 response compared to an IgG2a anti-GLU response was induced in mice immunized with GLU alone. Similarly, a preferential IgG1 response was also induced to SBR when GLU was present in either a mixed or conjugated form. Finally, a significant reduction (P < 0.05) in S. mutans colonization was observed only in mice immunized with the SBR-GLU chimeric protein. Taken together, our results indicate that the chimeric protein SBR-GLU significantly enhanced mucosal immune responses to SBR and GLU and systemic immune responses to SBR. The ability of SBR-GLU to induce responses effective in protection against colonization of S. mutans suggests its potential as a vaccine antigen for dental caries.

Streptococcus mutans is an etiologic agent of dental caries, an infectious disease resulting in the demineralization of tooth surfaces. Colonization of tooth surfaces by these microorganisms is considered to be the first important process for the induction of dental caries (23, 34). Two major virulence factors of S. mutans have been implicated in the molecular pathogenesis of dental caries. The cell surface fibrillar protein, originally termed antigen I/II (AgI/II) (33), has been implicated in the initial adherence of S. mutans to the salivary pellicle-coated tooth surface (21, 32). Salivary immunoglobulin A (IgA) antibodies to the whole AgI/II molecule have been shown to inhibit S. mutans adherence in an in vitro system (7) and in S. mutans colonization and dental caries development in vivo (19). A functional domain of AgI/II important for initial adherence is the saliva-binding region (SBR), which is located within the N-terminal one-third of the molecule (2, 5, 26). Studies by Hajishengallis et al. (8) have shown that mucosal immunization of rats with SBR conjugated with the B subunit of cholera toxin (CT) results in the induction of protective immunity against infection by S. mutans and caries formation. Furthermore, immunization of mice with a Salmonella vector expression SBR resulted in mucosal and systemic immune responses to SBR, which corresponded with protection against S. mutans colonization of tooth surfaces (11).

The glucosyltransferase (GTF) enzymes play a major role in the sucrose-dependent accumulation of S. mutans to tooth surfaces through the synthesis of glucans from sucrose (20, 23). GTF has two functional domains: i.e., an N-terminal catalytic sucrose-binding domain involved in hydrolyzing sucrose to glucose and fructose and a C-terminal glucan-binding domain involved in the binding of the synthesized glucan polymer and presumably chain extension of the growing glucan polymers (17, 27, 28, 46). Studies by Smith et al. (39, 40) have shown that antibodies to peptides corresponding to sequences within the catalytic (CAT) or glucan-binding (GLU) regions can interfere with GTF function. Other studies have shown that immunization of rats with these synthetic peptides results in a reduction in the level of smooth surface and sulcal caries after infection with Streptococcus sobrinus and in sulcal caries after infection with S. mutans (43). We have previously subcloned the putative CAT region and GLU of the GTF-I of S. mutans and shown that antibodies to recombinant CAT and especially to GLU inhibit glucan synthesis by GTF (14). In a subsequent study in an experimental mouse model, it was shown that the induction of specific salivary antibodies against GLU could prevent S. mutans colonization of tooth surfaces and caries formation (15).

Since SBR and GLU are important in different stages of caries pathogenesis, it is possible that a vaccine composed of SBR and GLU may have a synergistic protective effect against S. mutans colonization. In this regard, previous studies have shown that rabbit IgG antibodies (47) and bovine milk antibodies (30) against a cell surface protein antigen PAc (AgI/II)-GTF fusion protein (PAcA-GB) inhibited both the initial and the subsequent glucan-mediated adherence of S. mutans in an in vitro tooth surface model. In the present study, we describe the construction and characterization of a genetic chimeric protein consisting of the two previously described (2, 5, 14, 26) virulence determinants SBR and GLU (SBR-GLU). The immunogenicity of this construct was compared to that of each antigen alone or an equal mixture of SBR and GLU. The protective effect of SBR-GLU against S. mutans colonization in a mouse model following intranasal (i.n.) immunization was also investigated. Evidence is provided that the chimeric protein was more effective than coadministered SBR and GLU in inducing immune responses to both components and in protection against S. mutans infection.

MATERIALS AND METHODS

Genetic construction.

The plasmids pET20b(+)-SBR (9) and pET20b(+)-GLU (14), encoding the SBR of AgI/II and the GLU of GTF-I from S. mutans, respectively, were used to construct pET20b(+)-SBR-GLU (Fig. 1). The DNA segment encoding GLU was amplified by PCR with plasmid pET20b(+)-GLU. PCR primers were chosen with the help of the Oligo 4.03 primer analysis program (National Bioscience, Inc., Plymouth, Minn.), and the appropriate restriction sites were introduced for subcloning (XhoI site at the 5′ end of the upper and lower primers). The 0.9-kb gene segment encoding GLU was ligated in frame with the 3′ end of the 1.2-kb gene segment encoding SBR in the pET20b(+)-SBR vector, and the resulting plasmid was named pET20b(+)-SBR-GLU. The pET20b(+)-SBR-GLU was introduced into Escherichia coli BL21(DE3) by electroporation. The transformed colonies were selected on Luria-Bertani (LB) agar plates (1% tryptone, 0.5% yeast extract, 1% NaCl, 0.1% dextrose, 1.8% agar) containing 50 μg of carbenicillin per ml. The transformant was examined for the presence of a 5.8-kb plasmid [pET20b(+)-SBR-GLU] by using the Wizard Miniprep DNA Purification System (Promega, Madison, Wis.). The presence of the insert was confirmed by XhoI digestion followed by gel electrophoresis.

FIG. 1.

Schematic representation of the construction of the pET20b(+)-SBR-GLU encoding the chimeric protein.

Recombinant protein expression and purification.

Expression of SBR-GLU by E. coli BL21(DE3) containing pET20b(+)-SBR-GLU was confirmed by induction of a mid-log-phase culture with 0.36 mM isopropyl-β-d-thiogalactopyranoside (IPTG). The soluble recombinant chimeric SBR-GLU protein was purified on a His-bind resin column according to the manufacturer's instructions (Novagen, Madison, Wis.). Briefly, E. coli BL21(DE3) containing pET20b(+)-SBR-GLU was grown in LB broth containing carbenicillin (50 μg/ml) to the mid-log phase at 30°C and then induced with IPTG for 3 h. Following centrifugation, the cells were suspended in binding buffer (0.5 M NaCl, 20 mM Tris-HCl [pH 7.9], 5 mM imidazole) and stored at −70°C. The cells were then thawed and sonicated, and the supernatant was passed through a 0.45-μm-pore-size filter and loaded onto a His-bind resin column. The column was washed with binding buffer followed by wash buffer (60 mM imidazole, 0.5 M NaCl, 20 mM Tris-HCl [pH 7.9]). The bound SBR-GLU was eluted from the column with 1 M imidazole and dialyzed against phosphate-buffered saline (PBS). SBR was purified from the soluble fraction of E. coli BL21(DE3) containing pET20b(+)-SBR by the same method. GLU was purified under denaturing conditions from the inclusion bodies in the cytoplasmic fraction of E. coli BL21(DE3) containing pET20b(+)-GLU cells, as previously described (15).

The purity and specificity of the SBR-GLU, SBR, and GLU preparations were verified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), followed by Western blot analysis with horseradish peroxidase (HRP)-conjugated rabbit IgG anti-SBR antibodies (10) or rabbit IgG anti-GLU antibodies (14). The protein content of each preparation was estimated by the bicinchoninic acid protein determination assay (Pierce, Rockford, Ill.), with bovine serum albumin (BSA) as the standard.

Mouse immunization and sample collection.

Groups of six BALB/c mice, 10 weeks of age, were immunized i.n. with SBR (28 μg per mouse), GLU (22 μg per mouse), SBR and GLU together (SBR+GLU; 28 μg of SBR and 22 μg of GLU), or SBR-GLU (50 μg per mouse) on days 0, 14, and 28. A sham-immunized group received PBS only. Each dose was applied slowly to both nares, and neither dose exceeded 20 μl at each administration. Saliva, vaginal wash, and blood samples were collected prior to the immunizations and at weeks 6, 8, and 10 postimmunization. Saliva samples were collected after stimulating saliva flow by intraperitoneal injection of 5 μg of carbachol (Sigma Chemical Co., St. Louis, Mo.). Serum was obtained after centrifugation of blood samples collected with a heparinized capillary pipette from the retro-orbital plexus. Vaginal wash samples were collected by flushing the vagina twice with 50 μl of PBS. All samples were stored at −70°C until analyzed for antibody activity by an enzyme-linked immunosorbent assay (ELISA). The saliva samples were centrifuged prior to analysis of antibody activity. The animal work was performed according to the National Institutes of Health guidelines, and protocols were approved by the University of Alabama at Birmingham Institutional Animal Care and Use Committee.

Antibody responses.

The level of specific antibodies against SBR or GLU in samples was determined by ELISA on Maxisorp microtiter plates (Nunc, Roskilde, Denmark) coated with SBR or GLU, as previously described (10, 15). The levels of total IgA in secretions were detected on plates coated with an optimal concentration of goat anti-mouse IgA antibodies. Plates were blocked for 2 h at room temperature with 1% BSA in 0.01 M phosphate buffer (pH 7.2) containing 0.5 M NaCl, 0.15% Tween 20, and 0.02% azide. Peroxidase-conjugated goat anti-mouse IgG, IgG subclass, and IgA (Southern Biotechnology Assoc., Inc., Birmingham, Ala.) were used for detection followed by o-phenylenediamine substrate with H₂O₂. The concentrations of antibodies and total Ig in test samples were calculated by interpolation on standard curves generated by using a mouse Ig reference serum (ICN Biomedicals, Aurora, Ohio) and constructed by a computer program based on four-parameter logistic algorithms (Softmax/Molecular Devices Corp., Menlo, Calif.). The levels of IgA antibody activity in secretions are expressed as the ratio of IgA anti-SBR or anti-GLU per total IgA levels, in order to normalize for variation in total Ig content in the samples.

Mouse infection model.

In another series of experiments, the effectiveness of the chimeric protein SBR-GLU in inducing protective immunity against S. mutans infection in a mouse model was assessed. The experimental groups and the immunization protocol were the same as those described above, except that two additional sham-immunized groups were included. The experimental infection protocol was similar to that reported previously (11, 15). Briefly, 2 weeks after the third immunization, mice were treated with 4,000 U of penicillin G per ml in their drinking water for 4 consecutive days. Mice were also fed powdered diet 301 (containing 1% sucrose) supplemented with 4 mg of tetracycline and 4 mg of erythromycin per g of diet for 5 consecutive days. The suppression of the normal oral flora in mice was checked 1 day after the termination of antibiotic treatment by plating oral swab samples on blood agar plates.

Following antibiotic treatment, the groups of immunized mice and one group of sham-immunized mice were challenged for 5 consecutive days with 2 × 10⁹ CFU of tetracycline- and erythromycin-resistant S. mutans PC3379 (provided by P. J. Crowley and A. S. Bleiweis, Gainesville, Fla.) by swabbing tooth surfaces with Calgiswabs presoaked with the bacterial suspension.

S. mutans infection levels were assessed by swabbing mice at weekly intervals for 5 weeks, beginning 1 week after the last infection. The tip of each Calgiswab was solubilized in saline and appropriate dilutions were plated on mitis-salivarius (MS) agar or MS agar containing tetracycline and erythromycin. The levels of S. mutans and total oral streptococci were determined by counting the colonies on MS agar with antibiotics and MS agar plates, respectively. At the end of the experiment (week 14), mice were sacrificed, and the mandibles were removed and sonicated in saline to assess S. mutans colonization levels as described above.

Statistical analysis.

Each of the data on antibody responses and S. mutans to total streptococci levels was logarithmically transformed, and statistical analysis (one-way analysis of variance in conjunction with the Tukey multiple-comparisons test) was performed by using the InStat program (Graphpad Software, San Diego, Calif.).

RESULTS

Recombinant protein expression.

The GLU gene segment in pET20b(+)-GLU was PCR amplified and ligated into pET20b(+)-SBR. Restriction enzyme digestion of pET20b(+)-SBR-GLU with XhoI revealed a 4.9-kb fragment and a 0.9-kb fragment as predicted. Analysis of the purified recombinant SBR-GLU by SDS-PAGE, followed by Western blotting revealed a predominant band at 75 kDa, which reacted with both anti-SBR (Fig. 2A) and anti-GLU (Fig. 2B) antibodies. A minor band (35 kDa) that reacted with both antibodies was also seen. This may represent the GLU and a portion of the SBR, since the SBR appears to be more susceptible to proteolysis than GLU (Fig. 2) (9, 14). Analysis of recombinant SBR and GLU revealed a predominant band that reacted only with the anti-SBR (Fig. 2A) or anti-GLU (Fig. 2B) antibody, respectively, as previously described (9, 14).

FIG. 2.

Western blot of SBR-GLU (lane 1), GLU (lane 2), and SBR (lane 3). (A) The blot was probed with HRP-conjugated rabbit IgG anti-SBR antibodies. (B) The blot was probed with rabbit IgG anti-GLU antibodies and developed with HRP-conjugated goat anti-rabbit IgG. Molecular mass markers (in kilodaltons) are indicated on the left.

Serum IgG antibody responses.

In order to determine the immunogenic potential of the genetic chimeric protein, mice were immunized with SBR-GLU, SBR+GLU, SBR, or GLU by the i.n. route. Mice immunized with the chimeric protein SBR-GLU had significantly higher (P < 0.001) serum IgG anti-SBR responses than were seen in the SBR-, GLU-, and sham-immunized mice 2 weeks following the second immunization (Fig. 3A). The level of anti-SBR antibody activity persisted in the SBR-GLU-immunized group throughout the experiment and was significantly higher than those obtained in the SBR+GLU group at weeks 4 (P < 0.001), 6 (P < 0.05), and 10 (P < 0.001). An increase in the serum IgG anti-SBR antibody response was seen in the SBR+GLU group after the third immunization, which was significantly different from that obtained in the control group at weeks 6 and 8 (P < 0.001). The SBR-immunized group showed a slight increase in specific serum IgG anti-SBR antibody responses after each immunization; however, it was not significantly different from that obtained in the control group. There was no significant difference in the serum IgG anti-SBR antibody response between the group coimmunized with SBR and GLU and the group immunized with SBR alone. As expected, no serum IgG response against SBR was induced in mice immunized with GLU or PBS.

FIG. 3.

Serum IgG anti-SBR (A) or anti-GLU (B) responses in mice following i.n. immunization with SBR, GLU, SBR+GLU, and SBR-GLU. Values are expressed as the geometric means ± standard deviations for six mice. ∗, significantly different from the control group at P < 0.001; #, significantly different from the SBR-immunized group at P < 0.001; ^, significantly different from the SBR+GLU-immunized group at P < 0.001, except at week 6 (P < 0.05).

The level of serum IgG anti-GLU antibody activity in the GLU-, SBR+GLU-, or SBR-GLU-immunized mice was significantly higher (P < 0.001) than those seen in the SBR- or sham-immunized mice following the second immunization (Fig. 3B). Contrary to the anti-SBR responses obtained in the groups of immunized mice, no significant difference was seen in the serum IgG anti-GLU antibody responses between mice immunized with the chimeric protein SBR-GLU, SBR+GLU, or GLU alone (Fig. 3B). No serum IgG anti-GLU response was detected in the SBR- or sham-immunized groups.

Serum IgG subclass antibody responses.

To better understand the nature of the responses to the vaccine antigens, serum samples were also assessed for the IgG subclass distribution of the responses to SBR and GLU. The SBR, SBR+GLU, and SBR-GLU groups showed an increase in their serum IgG1 anti-SBR antibody responses after the three immunizations (Fig. 4A). The levels in the SBR-GLU-immunized mice were significantly different from the levels seen in the SBR-immunized mice at weeks 6 and 8 (P < 0.01) and week 10 (P < 0.05). A significant difference (P < 0.01) in serum IgG1 anti-SBR levels was also observed between the SBR-GLU-immunized mice and the SBR+GLU-immunized mice at week 4. The three groups also showed an increase in their serum IgG2a anti-SBR antibody levels (Fig. 4B). The SBR-GLU group showed significantly higher serum IgG2a responses to SBR than the SBR group at weeks 4 (P < 0.01) and 8 (P < 0.05). A significantly higher (P < 0.001) serum IgG2a anti-SBR level in the SBR-GLU group than that in the SBR+GLU group was shown at week 4. No significant difference in the serum IgG1 or IgG2a anti-SBR antibody responses between the SBR and SBR+GLU groups was found. The ratio of SBR-specific serum IgG2a/IgG1 responses in the SBR-immunized mice showed a mixed IgG1 and IgG2a response (Fig. 4C). In the SBR+GLU-immunized mice, the IgG2a/IgG1 ratio decreased after the second immunization. The ratio of the SBR-specific serum IgG2a/IgG1 antibodies in the SBR-GLU group decreased after the first immunization and was significantly lower (P < 0.01) than that of the SBR+GLU group at week 2. The ratio increased after the second immunization and was similar to the ratios for the SBR and SBR+GLU groups by week 4.

FIG. 4.

Levels of anti-SBR IgG1 (A), IgG2a (B), and IgG2a/IgG1 ratios (C) in serum in the SBR-, SBR+GLU-, or SBR-GLU-immunized group. Values are expressed as the geometric means ± standard deviations for six mice. Specific antibody levels or ratios were significantly different from those of the SBR+GLU-immunized group at P < 0.01 (^^) or P < 0.001 (^). Specific antibody levels or ratios were significantly different from those of the SBR-immunized group at P < 0.05 (∗) or P < 0.01 (∗∗).

The kinetics of the serum IgG1 and IgG2a anti-GLU antibody responses in the GLU-, SBR+GLU-, and SBR-GLU-immunized groups reflected the serum IgG anti-GLU antibody response patterns shown in Fig. 3B. The serum IgG2a/IgG1 anti-GLU ratios for the three groups decreased after the first immunization and then increased and remained constant at ∼1 throughout the duration of the experiment (data not shown). No significant difference was seen in the anti-GLU IgG2a/IgG1 ratios among the three groups.

Saliva and vaginal IgA antibody responses.

A salivary IgA anti-SBR response was induced in mice immunized with the chimeric protein SBR-GLU, which was significantly higher (P < 0.001) than that seen in the sham-immunized mice after the second and third immunizations (Fig. 5A) and which persisted through week 10. The level of salivary IgA anti-SBR activity was also significantly higher than that seen in the SBR- and SBR+GLU-immunized mice at weeks 4, 6, 8, and 10 (P < 0.001; except at week 8 for the SBR+GLU group, P < 0.01). No difference was seen in the salivary IgA anti-SBR antibody responses in the coimmunized and SBR-only-immunized groups.

FIG. 5.

Saliva IgA anti-SBR (A) or anti-GLU (B) responses in mice following i.n. immunization with SBR, GLU, SBR+GLU, and SBR-GLU. Values are expressed as the geometric means of the percent anti-SBR or anti-GLU IgA antibody per total IgA ± standard deviations for six mice. ∗, significantly different from the control group at P < 0.001.

A salivary IgA anti-GLU antibody response was also induced in the SBR-GLU-immunized group (Fig. 5B). The levels of anti-GLU antibody activity were significantly different from those in the control group at weeks 4, 6, 8, and 10 (P < 0.001). The levels of anti-GLU activity were also significantly different from that seen in the SBR+GLU group at weeks 4 (P < 0.01) and 8 (P < 0.05) and from the GLU group at weeks 4, 8, and 10 (P < 0.001), as well as week 6 (P < 0.01). Although the IgA anti-GLU antibody responses in the GLU and SBR+GLU groups were significantly elevated (P < 0.01) after the third immunization, they were not significantly different from those in the control group. No significant difference was shown between the coimmunized group and the GLU-immunized group.

Specific IgA antibody responses to SBR or GLU were also detected in vaginal wash samples from mice in each group. The vaginal response patterns in the various groups of immunized mice were similar to the salivary responses shown in Fig. 5 (data not shown). However, significant differences were found at week 6 (P < 0.05 for anti-SBR antibody responses and P < 0.001 for anti-GLU antibody responses) in the SBR-GLU-immunized mice compared to the control group.

Protection against S. mutans colonization.

In order to determine the effect of the immune responses to SBR-GLU in protection against S. mutans infection, additional groups of immunized and control mice were challenged orally with this bacterium. Salivary IgA antibody levels to SBR and GLU were determined in each experimental and control group before infection and at the end of the study. Significantly elevated levels of salivary IgA antibodies against SBR and GLU were found in SBR-GLU-immunized mice prior to infection with S. mutans and at the end of the experiment (P < 0.001) (Table 1). Following challenge with S. mutans PC3379, sham-immunized mice and mice immunized with SBR alone maintained high levels of S. mutans throughout the experiment (Fig. 6). Mice immunized with GLU alone showed only a slight decrease in S. mutans levels during the 6-week experimental period. In contrast, the SBR-GLU-immunized mice exhibited a 76% reduction in S. mutans colonization level at week 3, which was significantly lower (P < 0.05) than those in the SBR-immunized and control groups. A 50% decrease in S. mutans colonization level was observed in the SBR+GLU-immunized group at week 5, but this reduction was not significantly different from that seen in the other infected groups. No S. mutans was detected in the control group (data not shown).

TABLE 1.

Salivary IgA responses to SBR and GLU in the infection experiment^a

Treatment group	Antibody activityb:
	Prior to infection		At end of expt
	Anti-SBR	Anti-GLU	Anti-SBR	Anti-GLU
SBR	1.20 ± 0.15		1.49 ± 0.72
GLU		3.85 ± 0.88		1.82 ± 0.21
SBR+GLU	2.25 ± 0.88	4.40 ± 1.56	1.97 ± 0.30	2.97 ± 0.52
SBR−GLU	13.83 ± 2.43*	12.24 ± 2.11*	5.96 ± 1.41*	7.08 ± 1.20*
Controlc	1.48 ± 0.78	1.24 ± 0.44	0.78 ± 0.12	1.10 ± 0.21

^aIndividual saliva samples were collected prior to infection (day 42), which was 14 days following the third immunization, and at the termination of the experiment (day 98) and assessed for antibody activity by ELISA.

^bValues are expressed as the means of the specific anti-SBR or anti-GLU salivary IgA (nanograms per milliliter) per total salivary IgA (micrograms per milliliter) ± standard errors for six mice. *, Significantly different from control groups at P < 0.001.

^cValues are the mean levels of antibody activity obtained from the sham-immunized, sham-infected control group and the sham-immunized infected control group.

FIG. 6.

Percentage of S. mutans PC3379 per total oral streptococci in the oral cavity of immunized or control mice challenged with 2 × 10⁹ CFU of strain PC3379 on 5 consecutive days. The values are expressed as the geometric means ± standard errors for six mice. ∗, significantly different from the control at P < 0.05.

DISCUSSION

AgI/II and GTF are considered to be major virulence factors of S. mutans (3, 6, 13, 20). Antibodies to functional epitopes of AgI/II and GTF have been shown to inhibit S. mutans colonization and caries formation in experimental animal models (8, 15, 38, 43). In the present study, we constructed a genetic chimeric protein comprised of the saliva-binding region of AgI/II (SBR) and the glucan-binding region of GTF (GLU). Furthermore, we determined the effectiveness of the fusion protein SBR-GLU in inducing mucosal and systemic antibody responses and protective immunity against S. mutans infection following i.n. immunization of mice.

Mice immunized with the chimeric SBR-GLU protein had higher levels of serum IgG and salivary and vaginal IgA anti-SBR antibody responses than mice immunized with SBR+GLU or SBR alone. Previous studies have also shown that SBR alone is a poor immunogen, whereas mucosal administration of SBR as a chimeric fusion with the A2/B subunits of CT (CTA2/B) (4) or when coadministered with monophosphoryl lipid A (MPL) (24) resulted in enhanced mucosal and serum antibody responses. The present study provides evidence that the chimeric protein SBR-GLU was also able to induce significant anti-SBR antibody responses in the absence of any adjuvant.

The finding that mice immunized with GLU alone had similar levels of serum IgG antibody activity to GLU as those immunized with SBR-GLU or SBR+GLU supports our previous findings that GLU is a good immunogen and is capable of inducing immune responses that were protective against S. mutans infection (15). In the previous study, it was shown that fusion of thioredoxin to GLU did not result in the induction of significantly higher responses to GLU than were seen with GLU alone. This was an unexpected finding based on previous studies by others (1, 12, 22, 35). Furthermore, when GLU was coadministered with CT or MPL by the i.n. route, no augmentation in responses was seen (unpublished observations). In the present study, no substantial salivary or vaginal IgA anti-GLU antibody response was induced in the GLU- or SBR+GLU-immunized group. One explanation may be that the dose of GLU used in this study was too low to induce a good mucosal response. In this regard, in our previous studies (14, 15), the amount of GLU used was twice as much as in the present study.

One strategy that has been used to enhance host responses to synthetic peptides that are poor immunogens involves coimmunization of antigens (31, 36). Studies by Taubman et al. (44) have demonstrated that coimmunization of rats with the peptides from the functional CAT and GLU domains of GTF-I resulted in significantly higher serum IgG and salivary IgA antibody responses to CAT or GTF than were seen in rats immunized with either peptide alone. These investigators suggested that GLU-responsive T cells can provide bystander help to CAT-responsive T cells, which possibly give rise to memory T cells for CAT. The T-cell bystander help for GLU-specific activated T cells could in turn stimulate CAT-specific B cells, resulting in enhanced antibody production in both the systemic and the mucosal (salivary) immune compartments. In our study, we found that coimmunization with a mixture of SBR and GLU (SBR+GLU) induced serum IgG antibody responses and salivary and vaginal IgA antibody responses to SBR and GLU that were not significantly different from those observed in the mice immunized with an equivalent amount of SBR or GLU alone. The disparity between our results and those of Taubman et al. (44) may be due to the different nature and size of the peptides or polypeptides used. In the study by Taubman et al. (44), the antigen mixture consisted of 21-mer CAT and 22-mer GLU peptides synthesized on a core matrix of three lysine molecules to yield four identical peptides per molecule. Our antigens were the recombinant 42-kDa SBR and 33.5-kDa GLU polypeptides (9, 14). Another explanation for the difference in the results may relate to the immunization regimen used. In the studies by Taubman et al. (44), rats were immunized subcutaneously in the vicinity of the salivary glands with 50 μg of CAT or GLU or 50 μg of CAT plus 50 μg of GLU, each incorporated in complete Freund's adjuvant on day 0 or incomplete Freund's adjuvant on day 7. In our study, a lower dose of antigen without adjuvant (28 μg of SBR plus 22 μg of GLU) was given three times to mice by the i.n. route.

Bystander help is thought to occur through the release of T-cell-derived factors that act nonspecifically on activated B cells and no direct link between the antigenic determinants recognized by the T cell and the B cell is required, whereas cognate help involves a direct interaction between the Th and B cells and results in the transduction of a signal to the B cell in the form of locally released factors and/or cross-linking of small molecules. In this regard, previous studies by Shaw et al. (36) provided evidence that chimeric peptides (covalent linkage of the epitopes) induce enhanced levels of antibodies with significantly higher affinity than those seen following coimmunization of the peptides and suggested that cognate help is more efficient than bystander help for promoting antibody production. Recent studies by Taubman et al. (42) also demonstrated that immunization with a diepitopic construct (CAT-GLU) containing two copies of both CAT and GLU peptides on a lysine backbone was more effective than a mixture of the CAT and GLU peptides in inducing immune responses. Our results provide evidence for the effectiveness of a chimeric protein in promoting immune responses by demonstrating the induction of higher anti-SBR and anti-GLU antibody responses following immunization with SBR-GLU than with a mixture of equivalent amounts of SBR and GLU. A possible explanation may be that the SBR-GLU molecule is more stable than each component alone. It has been shown that SBR is susceptible to proteolysis (Fig. 2) (9, 14). Therefore, it is possible that the C-terminal portion of SBR is protected from proteolysis as a chimeric protein with GLU. This could allow for more effective presentation of epitopes for the induction of responses to SBR and GLU.

In the present study, we report a mixed IgG1 and IgG2a antibody response in mice immunized with SBR alone. In mice, Th1 and Th2 cells mediate the predominant production of IgG2a and IgG1 antibody responses, respectively (18, 29, 41). Therefore, our results suggest the induction of a mixed type 1-type 2 response, which is in agreement with findings by Toida et al. (45). However, we also observed a slightly higher IgG1 than IgG2a antibody response in mice immunized with GLU alone. These results support our recent finding (16) suggesting a preferential type 2 response to GLU based on the IgG2a/IgG1 anti-GLU response ratio. Furthermore, the present results suggest that a preferential type 2 response to SBR was induced when GLU was present in either a mixed or conjugated form, indicating the influence GLU has on the overall nature of the immune response.

The inhibition of S. mutans colonization seen in the present study was in agreement with the salivary antibody responses. Only mice immunized with the chimeric protein SBR-GLU showed both a significant reduction in the S. mutans colonization levels and a significant increase in salivary antibody responses. These results support the importance of inducing salivary responses that could interfere with the two different stages involved in the pathogenesis of S. mutans infection (i.e., attachment to and accumulation on tooth surfaces) in developing a vaccine against dental caries. Recent studies by others using a similar fusion protein, termed “PAcA-GB,” have shown that bovine milk containing antibodies against the fusion protein was effective in controlling the recolonization of S. mutans in the oral cavity of humans (37). Furthermore, passive immunization of rats with bovine milk containing antibodies against the PAcA-GB fusion protein was shown to protect against S. mutans induced dental caries in a rat model (25). However, these investigators did not compare the effectiveness of antibodies to the chimeric protein to that of the antibodies to each component. Our data provide further evidence for the effectiveness of a chimeric SBR-GLU protein in inducing a protective response against S. mutans infection following i.n. immunization and shows that the chimeric protein was more effective than each protein alone or an admixture of both proteins.

In conclusion, we have shown that the chimeric protein SBR-GLU significantly enhanced mucosal immune responses to SBR and GLU, systemic responses to SBR, and protection against S. mutans infection in an experimental mouse model. The use of chimeric proteins could be a valuable approach for developing a vaccine against dental caries. Furthermore, our results suggest a promising way to develop vaccines for nasal administration that induce enhanced salivary IgA and serum IgG antibody responses in the absence of adjuvant. Ongoing studies are investigating the mechanism or mechanisms that regulate induction of immune responses by SBR-GLU.