ABSTRACT
The interplay between mucosal immune responses to natural exposure to mutans streptococci and the incorporation and accumulation of these cariogenic microorganisms in oral biofilms is unclear. An initial approach to explore this question would be to assess the native secretory immunity emerging as a consequence of Streptococcus mutans infection. To this end, we analyzed salivary immunoglobulin A (IgA) antibody to mutans streptococcal glucosyltransferase (Gtf) and glucan binding protein B (GbpB) and to domains associated with enzyme function and major histocompatibility complex (MHC) class II binding in two experiments. Salivas were collected from approximately 45-day-old Sprague-Dawley rats, which were then infected with S. mutans SJ32. Infection was verified and allowed to continue for 2 to 2.5 months. Salivas were again collected following the infection period. Pre- and postinfection salivas were then analyzed for IgA antibody activity using peptide- or protein-coated microsphere Luminex technology. S. mutans infection induced significant levels of salivary IgA antibody to Gtf (P < 0.002) and GbpB (P < 0.001) in both experiments, although the levels were usually far lower than the levels achieved when mucosal immunization is used. Significantly (P < 0.035 to P < 0.001) elevated levels of postinfection salivary IgA antibody to 6/10 Gtf peptides associated with either enzyme function or MHC binding were detected. The postinfection levels of antibody to two GbpB peptides in the N-terminal region of the six GbpB peptides assayed were also elevated (P < 0.031 and P < 0.001). Interestingly, the patterns of the rodent response to GbpB peptides were similar to the patterns seen in salivas from young children during their initial exposure to S. mutans. Thus, the presence of a detectable postinfection salivary IgA response to mutans streptococcal virulence-associated components, coupled with the correspondence between rat and human mucosal immune responsiveness to naturally presented Gtf and GbpB epitopes, suggests that the rat may be a useful model for defining mucosal responses that could be expected in humans. Under controlled infection conditions, such a model could prove to be helpful for unraveling relationships between the host response and oral biofilm development.
Initial colonization of the human oral cavity by commensal bacteria, such as Streptococcus mitis (28, 33), is quickly followed by secretory immune responses to these microorganisms, which can be detected in saliva (6, 35). Significant heterogeneity is observed in these early responses (35). As additional oral habitats develop, colonization by a broader spectrum of commensal microorganisms takes place. Each species is likely to induce unique secretory immune responses, as well as enhance the response to cross-reacting epitopes already present. Mutans streptococci, which are strongly associated with dental caries (14), usually colonize children during the second year of life, when oral habitats are more complex (3, 19). Exposure to these cariogenic streptococci also results in a detectable, although varied, set of secretory immune responses to several of the mutans streptococcal components associated with successful acquisition (34).
The significance of these so-called natural secretory immune responses to members of the oral biofilm is unclear. Cole and coworkers (6) have suggested that a portion of the response is polyclonal and of lower affinity and has an innate immunity-like rapid response role. It is also likely that adaptive, higher-affinity secretory immune responses are also generated during colonization, especially at the outset of colonization. The variety of specificities seen in secretory (salivary) immunoglobulin A (IgA) responses to earlier-colonizing oral streptococci (e.g., S. mitis and Streptococcus salivarius) and later-colonizing oral streptococci (e.g., Streptococcus mutans) supports this concept. Some of the natural higher-affinity responses might also modulate the success or duration of colonization. For example, recent evidence suggests that early formation of salivary IgA antibody to S. mutans glucan binding protein B (GbpB) (2, 18) may delay colonization of young children who are being heavily challenged with this cariogenic streptococcus (25, 26). The effect of salivary IgA antibody on S. mutans acquisition is compatible with observations indicating that experimental provision of antibody to mutans streptococcal virulence components by either active or passive immunization can provide protective immunity to experimental infection with S. mutans or Streptococcus sobrinus (for reviews, see references 12 and 31).
Rats have frequently been used to explore the effect of increasing the secretory (salivary) immune response by immunization with a variety of putative virulence epitopes on subsequent oral disease, especially dental caries (31). However, the use of this model to study the effect which salivary IgA antibody, naturally induced by infection, could have on the success or extent of colonization with S. mutans is relatively unexplored. This approach should help us to understand the role of natural immunity in acquisition of the cariogenic streptococci in the human oral biofilm. In order to apply the model in this way, the characteristics of natural salivary IgA immune responses to components associated with adhesion or accumulation of S. mutans following infection alone would be required. In this regard one could follow development of salivary IgA antibody to glucosyltransferases (Gtf) and glucan binding protein B (GbpB), as well as intrinsic peptides with putative functional or major histocompatibility complex (MHC) binding roles, which have also been shown to elicit significant immune responses and, in some cases, protective immunity in experimental models. Such epitopes include several sequentially distinct residues in the N-terminal half of glucosyltransferase which are associated with its catalytic activity (8, 9, 16, 22-24, 42). Also included would be natural immune responses to regions associated with the ability of Gtf to bind glucan through a series of repeating sequences in the C-terminal third of the molecule (1, 17, 43). Although functional domains in S. mutans GbpB have not yet been identified (20), several peptide sequences in the GbpB structure (21) have potential for MHC binding.
Thus, the purpose of the present investigation was to study the attributes of the natural secretory immune response to components of S. mutans important to its colonization by analysis of the presence and patterns of salivary IgA immune responses to Gtf and GbpB and to 16 associated peptides which are related either to function or to MHC binding after experimental challenge of rats with S. mutans.
MATERIALS AND METHODS
Animals for infection experiments.Pathogen-free Sprague-Dawley CD strain rats (Kingston facility, Charles River) were used in the infection studies. Prior to infection, all rats were initially housed in single-animal cages and were fed pelleted laboratory chow (Harlan Teklad Corp.). All rats in the infection experiments had previously served as sham-immunized controls for immunization experiments. Before infection, each rat in all experiments received two subcutaneous injections in the vicinity of the salivary gland of phosphate-buffered saline when it was approximately 24 days old (in complete Freund adjuvant [Difco Laboratories, Detroit, MI]) and when it was 45 days old (in incomplete Freund adjuvant [Difco]). Infection with S. mutans was then initiated approximately 3 to 5 days after the second sham immunization at an age when all rat molars were erupted.
Rat infection.All rats in the infection studies were challenged with S. mutans strain SJ32 when they were 48 to 50 days old. Rats were placed in tubs (six rats/tub), given cariogenic diet 2000 (56% confectioner's sugar), and orally infected with 108 streptomycin-resistant S. mutans SJ32 cells for three consecutive days. Rats were again housed in single-animal cages after the infection protocol was completed and given diet 2000 for the duration of the experiment. Infection was verified for all rats after systematic swabbing of rat molars and plating on mitis salivarius agar and mitis salivarius agar with streptomycin as previously described (41). In experiment 1, 74 days after infectious challenge, S. mutans accounted for 22 to 93% (median, 39%) of the total streptococcal flora. In experiment 2, 62 days after infectious challenge, S. mutans accounted for 30 to 88% (median, 57%) of the total streptococcal flora. Postinfection salivas were collected 65 (experiment 2) or 79 (experiment 1) days after the initial challenge with S. mutans.
Group comparisons.Pre- and postinfection salivas from experiment 1 (n = 11) and experiment 2 (n = 11) were used for measurement of salivary IgA antibody to Gtf, Gtf peptides, and GbpB. Salivas were also analyzed for salivary IgA antibody to GbpB peptides. Since many of the salivas from experiment 1 and all salivas from experiment 2 were depleted, 13 pre- and postinfection salivas from two additional experiments (rats infected on day 50 and salivas collected 52 to 59 days later) in which similar sham immunization and S. mutans infection protocols were used were screened together with five pre- and postinfection salivas from experiment 1 for antibody to GbpB peptides (n = 18).
Animals for the immunization experiment.To compare the secretory (salivary) immune response to S. mutans infection with mucosal induction of salivary IgA antibody by immunization, salivas from the following immunization experiment were assayed for antibody to Gtf. S. sobrinus Gtf antigen was incorporated into a poly(d,l-lactide-co-glycolide) (PLGA)-based (ratio of lactide to glycolide, 75:25) biodegradable carrier copolymer (Boehringer Ingelheim Chemicals, Inc.) using 1% gelatin as a bioadhesive agent, as previously described (36). The dose used for this immunization experiment released approximately 10 μg of Gtf. Alternatively, 50 μg of S. sobrinus Gtf was mixed with aluminum phosphate. Forty-day-old Sprague-Dawley rats were inoculated intranasally three times at 1-week intervals with Gtf-PLGA microparticles (n = 8) or with Gtf mixed with aluminum phosphate (n = 8) or were sham immunized with empty PLGA microparticles (n = 8). Salivas from all 24 rats were collected 15 days after the last intranasal inoculation, clarified, and stored frozen until antibody analysis.
Saliva collection.Saliva samples were collected from all rats in the infection experiments when the rats were approximately 45 days old (preinfection saliva) and after approximately 2 to 2.5 months of infection with S. mutans SJ32. To collect saliva, rats were first momentarily anesthetized with a gas mixture containing 50% carbon dioxide and 50% oxygen and then anesthetized by intraperitoneal injection of a mixture (0.65 ml/kg) containing 3 parts ketamine (Ketaset; 100 mg/ml; Fort Dodge Lab, Ft. Dodge, IA) and seven parts xylazine (Rompun; 20 mg/ml; Bayer Corp., Shawnee Mission, KS). Saliva secretion was stimulated by subcutaneous injection of 0.6 ml carbachol (0.1 mg/ml in saline; Sigma Chemical Co., St. Louis, MO) per kg of rat. After saliva collection, rats were awakened by subcutaneous injection of 0.1 ml/kg of atropine sulfate (0.4 mg/ml; American Pharmaceutical Partners, Inc., Los Angeles, CA) and then yohimbine (yobine; 2.0 mg/ml; Lloyd Laboratories, Shenandoah, IA) using a volume that was 1.4 times the volume used for anesthesia. Clarified saliva was stored at −70°C.
Protein preparation. S. mutans strain SJ23 GbpB and S. sobrinus strain 6715 Gtf were prepared as previously described (32, 40). The S. mutans GbpB was homogeneous in Western blots. The S. sobrinus Gtf preparation contained a mixture of Gtf isotypes, approximately 30% of which was Gtf-I, which shares >80% homology with S. mutans GtfB and GtfC. S. sobrinus Gtf-S isotypes comprised the balance of the Gtf preparation.
Peptides.Peptide constructs were synthesized by AnaSpec, San Jose, CA, as multiple antigenic peptides using the stepwise solid-phase method of Merrifield with a core matrix of lysines, which yielded macromolecules with four identical peptides, as previously described (39). Six monoepitopic peptide construct sequences (QGQ, VAR, SYI, QAA, ANY, and SIG) were based on predicted human MHC class II epitopes (11) that were in the GbpB sequence, using the ProPred algorithm (30). Three monoepitopic peptide construct sequences (Pep 7, Pep 11, and Pep 16) were based on predicted MHC class II epitopes that were in the Gtf sequence, using the ProPred algorithm. Five monoepitopic peptide construct sequences (GGY, CAT, HDS, LVK, and VMAD) were based on Gtf sequences associated with putative catalytic interactions (8, 9, 22-24, 42). Two monoepitopic peptide construct sequences (GLB and GLU) were based on Gtf sequences associated with putative glucan binding domains in the C-terminal third of the Gtf protein (1, 17, 43). Of further interest, several of these peptides (SYI, GGY, CAT, GLU, and Pep 11) have been shown to induce protective immune responses to the cariogenic effects of S. mutans and/or S. sobrinus infection (7, 31, 38, 41). The Gtf/GbpB locations and sequences of the peptides are shown in Fig. 1 and Table 1.
Location of peptides in the sequences of Gtf and GbpB. Putative functional domains of Gtf are indicated above the Gtf diagram.
Peptides used for detection of salivary IgA antibody
IgA antibody measurement.Salivas were diluted 1:16 in PBS-BN (phosphate-buffered saline (Sigma Chemical), 1% bovine serum albumin [BSA] (Sigma Chemical), 0.05% sodium azide [pH 7.4]) and then filtered with 0.22-μm centrifuge tube filters (Spin-X; CoStar). Salivary IgA antibody reactivities with Gtf or GbpB or with peptides derived from these proteins were tested using a particle-based multiplex fluorescent immunoassay (Luminex). Fluorescently tagged microspheres with different fluorescent signatures were coated with Gtf, GbpB, or 1 of 16 protein-derived peptides. Protein and peptide antigens were covalently attached to these beads via their amines using the “protein coupling protocol” (www.luminexcorp.com/support/protocols/protein.html ). Optimal coating of 3.75 × 106 beads required 7.5 μg of protein and 3.75 μg (SIG peptide), 7.5 μg, or 37.5 μg (HDS and LVK) of peptide. However, at this bead concentration 190 μg of the VMAD peptide and 260 μg of the ANY peptide were used, and 600 μg of the QAA peptide coated beads at less than saturation. Each saliva assay well contained 2,500 beads with each antigen. The mixtures used for assays contained (i) a combination of all GbpB peptide-coated microspheres, (ii) a combination of all Gtf peptide-coated microspheres, or (iii) native Gtf protein- and GbpB protein-coated microspheres. To avoid cross-reactions between antibody and similar peptide epitopes, GLU and GLB peptide-coated beads were always tested in separate wells, as were LVK and VMAD peptide-coated beads. Uncoated microspheres were added to each mixture as controls. Bead mixtures (50 μl) were added to wells in 1.2-μm-pore-size filter plates (Millipore Corp., Bedford, MA). Microspheres were then washed twice with PBS-BN under a vacuum and resuspended in 50 μl of the same buffer. Fifty microliters of each diluted saliva sample (diluted 1:16 in PBS-BN) was then added and incubated at 37°C with shaking for 120 min. After the buffer was drained under a vacuum, microspheres were incubated with a 1:250 solution of mouse monoclonal anti-rat IgA (MARA-2 clone; Accurate Chem Scientific Corp.) in PBS-BN for 30 min at 37°C. Microspheres were then incubated with a 1:250 solution of R-phycoerythrin-conjugated goat anti-mouse IgG (Invitrogen) for 30 min at 37°C. All incubations were performed in the dark. After washing, the samples were resuspended in PBS-BN and read with a Luminex 100 analyzer to obtain a median fluorescence intensity (FI). The appropriate value was subtracted to account for the background in the wells without saliva for each antigen. The median coefficient of variation for positive replicates was 14.8%.
The total salivary IgA relative to the IgA in a standard rat saliva was measured in a separate plate by incubating a 1/2,000 dilution of saliva with beads coated with affinity-purified goat antibody to rat IgA (United States Biological, Swampscott, MA). The reaction was detected with monoclonal mouse anti-rat IgA and then with phycoerythrin-conjugated goat anti-mouse IgG, as described above.
Individual salivary postinfection IgA antibody FIs were considered positive if they were greater than the mean preinfection FI plus 2 standard deviations (SD) for the group, when assays were performed using the corresponding peptide- or protein-coated beads. The results are expressed as the median FI divided by the total IgA concentration.
Statistical analysis.The statistical significance of differences between the increase in postinfection levels for groups of salivas assayed with spheres coated with Gtf/GbpB epitopes and the corresponding increases for salivas tested with the spheres coated with BSA was analyzed by one-way analysis of variance (ANOVA), followed by the Tukey pairwise multiple-comparison test, in which data were normally distributed. Alternatively, data were analyzed by Kruskal-Wallis one-way ANOVA on ranks, followed by Dunn's multiple-comparison procedure when nonparametric distributions were encountered.
RESULTS
Salivary IgA antibody reactivity with Gtf following S. mutans infection.Salivas were collected from 22 44- to 46-day-old rats (experiments 1 and 2) before infection with S. mutans, and then salivas were collected again following an approximately 2-month infection period. Salivary IgA antibody to Gtf was measured in pre- and postinfection salivas, after which the antibody activity (expressed as FI) was normalized for individual salivary IgA concentrations. As shown in Table 2, the mean increase in IgA antibody activity with Gtf after 2 months of exposure to S. mutans was more than 10-fold (P < 0.002) in both experiments. Figure 2 shows the increases in salivary IgA antibody (FI/IgA) to Gtf for both experiments. Appreciable variation in the ability of rats to mount a mucosal immune response to infection was observed. However, 7/11 salivas in experiment 1 and 10/11 salivas in experiment 2 contained levels of IgA antibody to Gtf that were greater than the mean plus 2 SD of the reactivity of all salivas tested against BSA. Thus, infection was usually sufficient to induce a detectable mucosal immune response to Gtf under the infection and analysis conditions used.
Salivary IgA antibody reactivity to Gtf in rats whose salivas were taken before infection and after 79 days (experiment 1) or 66 days (experiment 2) of infection with S. mutans SJ32. The results are expressed as increases in postinfection salivary IgA antibody activity compared with preinfection salivary IgA antibody activity, measured as the Luminex FI for reactions of salivas with Gtf- or BSA (control)-coated microspheres divided by the IgA concentration of the saliva. The cross-hatched bars indicate the mean plus 2 SD for the BSA control run in duplicate for each saliva.
IgA antibody in salivas obtained before and after S. mutans infection
Salivary IgA antibody reactivity with GbpB following S. mutans infection.Salivary IgA antibody to GbpB/IgA was also measured in the pre- and postinfection salivas in experiments 1 and 2. Table 2 shows that despite appreciable variation in FI/IgA antibody levels among rats, significant (P < 0.001) increases in IgA antibody activity for GbpB were observed after 2 months of exposure to S. mutans. Figure 3 shows the increases in the levels of salivary IgA antibody (FI/IgA) to GbpB for each saliva in both experiments. More than one-half of the postinfection salivas (6/11 salivas in experiment 1 and 8/11 salivas in experiment 2) contained levels of IgA antibody to GbpB that were greater than the mean plus 2 SD for the reactivity of all salivas tested against BSA. Thus, infection was also sufficient to induce a detectable mucosal immune response to GbpB in many animals. Interestingly, 59% of the salivas had significant postinfection responses to both Gtf and GbpB.
Salivary IgA antibody reactivity with GbpB in rats whose salivas were taken before infection and after 79 days (experiment 1) or 66 days (experiment 2) of infection with S. mutans SJ32. The results are expressed as the increases in postinfection salivary IgA antibody activity compared with preinfection salivary IgA antibody activity, measured as the Luminex FI for reaction of salivas with GbpB- or BSA (control)-coated microspheres, divided by the IgA concentration of the saliva. The cross-hatched bar indicates the mean plus 2 SD for the BSA control run in duplicate for each saliva.
Salivary IgA antibody reactivity with Gtf- or GbpB-derived peptides following infection. (i) Gtf peptides.Since infection with S. mutans resulted in the appearance of salivary IgA antibody to Gtf and to GbpB in many animals, it was of interest to identify the molecular epitopes to which natural mucosal IgA antibodies were induced. To this end, the pre- and postinfection salivas of all rats in experiments 1 and 2 were assayed for IgA antibody activity against a panel of 10 Gtf-derived peptides (Table 1 and Fig. 1) that had been associated either with putative regions containing functional activity or with peptides that had MHC class II binding activity. Several of these peptides (GGY, CAT, Pep 11, and GLU) previously had been shown to induce a protective immune response to infection with cariogenic mutans streptococci (7, 31, 41). Figure 4 shows the median postinfection increases in the FI/IgA antibody value for each of the peptides assayed with all 22 salivas from both experiments. This group of salivas showed significant postinfection increases in FI/IgA antibody levels when the reactivities with the CAT and HDS peptides from the N-terminal third of the Gtf molecule, the VMAD peptide from the central region, and three peptides from the glucan-binding C-terminal third of the molecule (GLB, GLU, and Pep 16) were compared with the change in reactivity when salivas were tested against BSA. The HDS peptide (residues 549 to 567), which contains at least one residue implicated in catalysis (42), and the two similar peptides derived from the putative glucan binding domain (GLB and GLU) exhibited the highest mean and median increases. Significant postinfection increases in the antibody activities of the saliva groups were not observed with several peptides (GGY, Pep 7, Pep 11, and LVK), although occasional salivas did exhibit apparent increases in responses to these peptides after infection. Thus, following infection “natural” IgA antibody responses to Gtf epitopes that are associated with the catalytic and glucan binding activities of the enzyme can be observed.
Salivary IgA antibody reactivity with Gtf-derived peptides in combined rat salivas from experiments 1 and 2 (22 rats). The results are expressed as the median increases in postinfection salivary IgA antibody activity compared with preinfection salivary IgA antibody activity for each of the peptides indicated on the abscissa, measured as the Luminex FI for reaction of salivas with peptide-, Gtf-, or BSA-coated microspheres, divided by the IgA concentration of the saliva. The increases for each set of 22 salivas tested with each peptide were compared by one-way ANOVA with the increases for the same set of salivas tested with BSA. Peptide reactivities that were meaningfully elevated above the BSA reactivity are indicated by the black bars, within which the levels of statistical significance are shown.
(ii) GbpB peptides.Eighteen pre- and postinfection saliva pairs were assayed for IgA antibody activity against a panel of six GbpB-derived peptides (Table 1 and Fig. 1) that had been associated with peptides that had MHC class II binding activity. Figure 5 shows the median postinfection increases in FI/IgA salivary antibody values when the salivas were assayed against each of the GbpB-derived peptides. Significant postinfection increases in the FI/IgA antibody values were observed when the reactivities of these salivas with the QGQ and VAR peptides were compared with the changes in reactivity when salivas were tested against BSA. The postinfection increase in the group salivary antibody activity did not reach significance with several peptides (SYI, QAA, ANY, and SIG), although after infection occasional salivas did exhibit apparent increases with the SYI and SIG peptides. Thus, certain GbpB-derived peptides apparently induced a mucosal immune response after infection with GbpB-bearing and -secreting S. mutans.
Salivary IgA antibody reactivity with GbpB-derived peptides in combined rat salivas from experiment 1 and two additional infection experiments (18 rats). The results are expressed as the median increases in postinfection salivary IgA antibody activity compared with preinfection salivary IgA antibody activity for each of the peptides indicated on the abscissa, measured as the Luminex FI for reaction of salivas with peptide-, GbpB-, or BSA-coated microspheres, divided by the IgA concentration of the saliva. The increases for each set of 18 salivas tested with each peptide were compared by one-way ANOVA with the increases for the same set of salivas tested with BSA. Peptide reactivities that were meaningfully elevated above the BSA reactivity are indicated by the black bars, within which the levels of statistical significance are shown.
Comparison of salivary IgA antibody levels following S. mutans infection or mucosal (intranasal) immunization with Gtf.We investigated the relative expression of salivary IgA antibody to Gtf following infection or following intranasal immunization with Gtf delivered in PLGA microparticles. For the latter condition, the levels of IgA antibody to Gtf were measured in the salivas of rats 2 weeks after completion of an immunization regimen consisting of three weekly intranasal inoculations of S. sobrinus Gtf. Previous studies indicated that this collection time corresponded to the peak mucosal response (36). Figure 6 shows a comparison of the salivary IgA antibody levels of rats immunized intranasally with the corresponding IgA antibody levels of the infected rats described above. Mucosal immunization induced a >10-fold-higher level of salivary IgA antibody to Gtf than infection induced. Interestingly, however, salivas from 4 of the 22 infected rats had IgA antibody levels within the range observed for the immunized group.
Comparison of salivary IgA antibody induction by S. mutans infection or by mucosal immunization with Gtf. Salivary IgA antibody to Gtf in the collective infection experiments (22 rats) was compared with antibody reactivity in eight rats which were sham immunized intranasally (sham-IN) and 16 rats which were intranasally immunized with PLGA-Gtf or Gtf and aluminum phosphate (IN). The results are expressed as the group mean ± standard error (SE) of the Luminex FI of saliva IgA antibody reactivity with Gtf-coated microspheres. FIs of individual rat salivas are indicated by open circles. preSM, preinfection salivas; postSM, postinfection salivas.
DISCUSSION
The results of this study clearly demonstrate that infection of pathogen-free rats with S. mutans can induce a detectable mucosal IgA antibody response to at least two streptococcal proteins that are associated with the ability of this cariogenic streptococcus to accumulate in the oral biofilm. Similar to the pattern seen in young children of salivary IgA antibody specificities following primary exposure to immunogenic levels of S. mutans infection (25, 26), responses to both glucosyltransferases and glucan binding protein B were observed in a majority of rats under the infection conditions used in these experiments. Although the IgA antibody levels covered a wide range and were, on average, 10-fold less than those that could be obtained by intranasal immunization with Gtf (Fig. 6) or GbpB (data not shown) during the same time frame, the responses following infection were significant and occasionally fell within the range of the responses arising from immunization.
Induction of secretory immunity by oral infection is not surprising and has often been demonstrated in humans for many mutans streptococcal antigens, including Gtf (5, 10), GbpB (25, 34), adhesins (4, 13), and other components (for a review, see reference 31). Salivary antibody to several of these components can be detected soon after mutans streptococci become established in the oral biofilm (34). Much of the initially detectable salivary IgA antibody to mutans streptococci in young children is antibody to a relatively small number of components. Interestingly, the “response” can be quite variable in terms of the degree and the antigen recognized, even between siblings in the same family who would presumably be exposed to similar infectious doses from the primary caregiver (34). This variability is similar to that seen in this animal study in that there was no significant correlation between antibody to Gtf and antibody to GbpB among the 22 rats following exposure to a single strain of S. mutans. Thus, the rat experimental model reflects the individuality of the adaptive immune response and may mimic the human experience in this regard.
Identification of the epitopes inducing natural immune responses to infection would be helpful not only for understanding the dynamics of mucosal immunity to virulence antigens but also for designing strategies to intervene in pathogen infection. Peptides obtained from several locations within the glucosyltransferase enzyme were shown to react with salivary IgA antibody in S. mutans-infected rats. More than 50% of the infected animals showed a significant increase in responses reactive with the HDS peptide. This 19-mer peptide is associated with the beta7 strand element of the putative (beta, alpha)8 catalytic barrel domain of Gtf (8) and contains a catalytically essential histidine residue (His461 in S. mutans GtfB) (42). Site-directed mutagenesis of a nearby aspartate (Asp467 in S. mutans GtfB) within the HDS-associated Gtf sequence altered the water solubility of the glucan synthesized from sucrose (29). Furthermore, this peptide can induce significant levels of serum IgG and salivary IgA antibodies reactive with Gtf after mucosal (intranasal) immunization of Sprague-Dawley rats (37). The “natural” immunogenicity of the epitope(s) in the HDS sequence may trigger immune responses which can modify the incorporation of mutans streptococci, an hypothesis which can now be explored using the rodent model.
Significant salivary IgA antibody reactivity was also associated with the GLU and GLB peptides in the glucan binding domain. This reactivity may in part be the result of the repeating motifs within these sequences (1, 17, 43), potentially increasing both the frequency of their potential immunogenicity and the subsequent glucan binding within the Gtf molecule. Experimental immunization with the GLU peptide can give rise to significant inhibition of Gtf enzymatic activity (39). Injection of this peptide into a local salivary gland (41) and intranasal immunization with a chimeric vaccine containing a larger portion of the glucan binding domain (15) each resulted in protective immunity to experimental infection with mutans streptococci. This result was not observed when Pep 16 was used to immunize rats (7). It is notable that previous studies showed that there was a significant correlation with adult human parotid IgA antibody reactive with the GLU peptide and Gtf from either S. mutans or S. sobrinus (39). Thus, it is not surprising that infection of rats also gives rise to salivary IgA antibody reactive with epitopes reflected in this peptide's sequence and that the results further support the use of this model to explore correlations of natural immune response characteristics and mutans streptococcal infection.
Salivary IgA antibody from infected rats also showed significant reactivity with peptides derived from the N terminus of the GbpB protein. At least three GbpB epitopes capable of inducing a mucosal immune response following infection were located in an 80-residue region between residues 52 and 113. Experimental immunization with at least one peptide in this region has been shown to give rise to protective immunity (38) and can induce salivary IgA antibody after mucosal immunization (27). The epitope(s) associated with the VAR peptide showed significant reactivity in salivas of more than 70% of the infected rats. Few animals exhibited a change in the levels of salivary IgA antibody to the QAA and ANY peptides following infection. The pattern of salivary IgA antipeptide reactivity in this rat study is quite similar to the pattern recently observed for salivas of young children during their initial exposure to mutans streptococci (26). In the previous study, the levels of salivary IgA antibody to the VAR and QGQ peptides correlated with the levels of salivary IgA antibody reactive with S. mutans GbpB, providing evidence that the peptide-reactive antibody was indeed derived from the mucosal response to GbpB. The levels of salivary IgA antibody to the QAA and ANY peptides were unremarkable in the children, although several children did have elevated levels of antibody to the SIG peptide. Again, this overall correspondence between human and rat salivary IgA responses to GbpB epitopes emphasizes their similarity in mucosal immune recognition. Interestingly, all responsive GbpB peptides have been shown to induce significant levels of antibody following immunization (B. Shen, W. F. King, M. A. Taubman, and D. J. Smith, presented at the 82nd General Session of the IADR/AADR/CADR, 10 to 14 March 2004; C. E. Smith, W. F. King, M. A. Taubman, and D. J. Smith, presented at the 83rd General Session of the IADR/AADR/CADR, 10 to 13 March 2005).
Taken together, the results show that infection alone was sufficient to induce a variety of antibody specificities to epitopes on two proteins which are central to the accumulation of S. mutans in the oral biofilm. The correspondence between rats and humans of mucosal immune responsiveness to naturally presented Gtf and GbpB epitopes suggests that the rat may be a useful model to explore and define mucosal responses that could be expected in humans and, under controlled infection conditions, be helpful in unraveling relationships between the host response and oral biofilm development.
ACKNOWLEDGMENTS
This investigation was supported by NIH grants R37 DE-01653, RO3-TW-06324, and RO1 DE-04733 (United States), by Harvard Medical School, and by FAPESP grants 02/07156-1 and 04/07425-8 (Brazil).
FOOTNOTES
- Received 15 February 2008.
- Returned for modification 2 May 2008.
- Accepted 6 May 2008.
- Copyright © 2008 American Society for Microbiology
