Tonsillar Application of Formalin-Killed Cells of Streptococcus sobrinus Reduces Experimental Dental Caries in Rabbits

Abstract

Living Streptococcus sobrinus cells were orally inoculated into nonimmune rabbits and rabbits immunized with formalin-killed cells of S. sobrinus through tonsillar application to examine the anticaries potential of this method of immunization. The living S. sobrinus cell numbers and the caries areas in the rabbits immunized by tonsillar application decreased to a level one-fifth of that in nonimmune rabbits.

In recent years, we have demonstrated a new immunization route, tonsillar application, for inducing mucosal immunity (2–8). This immunization induced antibodies more effectively than nasal application and oral administration. The tonsillar application of formalin-killed whole cells of Streptococcus sobrinus and S. mutans in rabbits induced salivary immunoglobulin A (IgA) and blood plasma IgG against these cells (7). It was also found that these antibodies in the blood plasma induced by tonsillar application did not contain antibodies cross-reacting with human cardiac muscles; such antibodies have been known to be induced in the blood plasma by intramuscular injection. Further, the antibodies induced by tonsillar application of S. sobrinus AHT-k (serotype g) selectively reacted with S. sobrinus AHT-k and serologically related mutans group streptococci (serotypes a, d and h), whereas those induced by intramuscular injection reacted with a variety of oral streptococci, including unrelated mutans group streptococci (serotypes b, c, e, and f) in addition to the aforementioned streptococci (unpublished observations). In our unpublished observation, the antibodies induced by tonsillar application of S. sobrinus inhibited S. sobrinus colonization onto enamel slices in test tubes.

In the present study, we inoculated living S. sobrinus cells into rabbits immunized with formalin-killed cells of S. sobrinus by tonsillar application and the anticaries potential was examined.

Immunization and antibody induction.

S. sobrinus AHT-k was cultured in brain heart infusion broth for 18 h at 37°C and killed with 10% formalin for 24 h. Twenty-four 6-week-old rabbits were used. Three hundred microliters of the killed-S. sobrinus suspension (10¹⁰ cells/ml) was dropped by a micropipette onto the surface of the palatine tonsil of nine rabbits, instilled by a conductor into the stomach of three other rabbits, and intramuscularly injected without any adjuvants into the femoral region of three more rabbits every week for 6 weeks. In the three control rabbits, phosphate-buffered saline (PBS) alone was applied to the tonsil. There was a further group of six nonimmune rabbits. The saliva and blood plasma of these rabbits were collected once a week after immunization. Immunoglobulin concentrations in the saliva and blood plasma were examined once a week for 21 weeks by enzyme-linked immunosorbent assay using plates coated with the ultrasonic fragments of S. sobrinus as described before (7). In the case of IgA, IgA concentration in saliva was represented as that of a monomer calculated by the value based on the standard curve obtained by enzyme-linked immunosorbent assay using purified rabbit IgA from serum (Inter-cell Technologies Inc., Hopewell, N.J.).

Inoculation of living S. sobrinus cells.

Three days after the final immunization, the living S. sobrinus cells (10⁵ cells) in 1 ml of saline were orally inoculated into the rabbits every 3 days for 15 weeks, 35 times in total, with 1 ml of 3 M sucrose solution. In three of the nine rabbits immunized by tonsillar application, 1 ml of 1 mM galactose solution was orally administered every 3 days in addition to the above-described inoculation. We confirmed that indigenous mutans group streptococci were negligible. All rabbits were maintained on a solid diet (Cler Japan, Inc., Tokyo, Japan).

Detection of living S. sobrinus and caries regions.

The cells were separately collected three times from one of the five lower left mandibular tooth surfaces in each rabbit by swabbing with a cotton bud and were directly cultured for 18 h at 37°C on plates of mitis salivarius agar medium, and 100 random colonies were blotted onto nitrocellulose membranes. The blotted strips were reacted with rat antiserum against S. sobrinus (1:100 diluted) and horseradish peroxidase-labeled goat IgG antibody against rat IgG (1:200 diluted; ICN Pharmaceuticals Inc.). The anti-S. sobrinus antiserum used was prepared from rats immunized with S. sobrinus whole cells by intramuscular injection every week for 6 weeks. The serum was absorbed by whole cells of S. sanguis ATCC 10556, S. salivarius IFO13956, S. pyogenes ATCC 12344, and serologically related S. downei ATCC 33748 (MFe28). The absorbed serum was confirmed to react selectively against S. sobrinus and not to react with other oral streptococci. The positive spots on the blotted strips were counted as S. sobrinus cells. The proportion of S. sobrinus colonies in 100 random colonies was determined as the mean of nine specimens separately obtained from three rabbits in each group. We confirmed that S. sobrinus proportions obtained by this method were similar to the proportion of mutans group streptococci obtained with Dentcult-SM (Orion Diagnostica, Espoo, Finland) or mitis salivarius-bacitracin (0.2 U/ml) agar medium supplemented with 15% sucrose, both of which allow selective growth of mutans group streptococci from the oral cavity.

The area of the caries regions was calculated with a computer programed to trace the caries region on a photograph of the lower right mandibular teeth, taken horizontally from the buccal side, of each rabbit, because caries on the occlusal surface are rare and recovery from decay is possible by continuous tooth growth in rabbits. The total areas of caries regions in each rabbit were shown as the mean of the three rabbits in each group.

Statistical analysis.

The twenty-four rabbits used were divided into eight groups of three rabbits each. Three of the groups were immunized by tonsillar application, one group was immunized by intramuscular injection, and another was immunized by intragastric instillation. Also, there were one control group (PBS applied) and two nonimmune groups. All experimental results were expressed as the means ± standard deviations (SD). Statistical analysis was carried out by using the Friedman test. When significances were observed, the differences were determined by using the post hoc test (Scheffe’s F method).

In rabbits immunized by tonsillar application, the concentration of anti-S. sobrinus salivary IgA was significantly higher than that in rabbits immunized by intramuscular injection and intragastric instillation (P < 0.01) and was maintained at high levels for at least 15 weeks after the final tonsillar application (Fig. 1). The concentrations of anti-S. sobrinus salivary IgA in rabbits immunized by intragastric instillation and intramuscular injection were 1/4 and 1/10 that in rabbits immunized by tonsillar application, respectively. The concentration of anti-S. sobrinus IgG in blood plasma in these immunized rabbits was significantly higher than that in the control rabbits (PBS applied) (P < 0.005 or P < 0.01).

FIG. 1.

Concentrations of anti-S. sobrinus IgA antibody in saliva (a) and anti-S. sobrinus IgG antibody in blood plasma (b) after tonsillar application (circle), intragastric instillation (triangle), and intramuscular injection (solid square) of formalin-killed S. sobrinus cells. The results of control rabbits (PBS applied) are also shown (open square). The results of antibody concentrations show the average for the three rabbits of each group. Error bars indicate SD. Symbols for statistical significance: ∗, P < 0.005; ∗∗, P < 0.01 (5 to 21 weeks versus control with PBS applied); †, P < 0.01 (5 to 21 weeks versus intramuscular injection); ††, P < 0.01 (5 to 21 weeks versus intragastric instillation).

In nonimmune rabbits, experimental dental caries appeared at the 6th week after the initial inoculation. The area of the dental caries increased to nearly half the area of the tooth surface by the 15th week after the initial inoculation (Fig. 2). Intramuscular injection had no effect on the reduction of dental caries. In rabbits immunized by tonsillar application, the area of the caries region was 25% of that in nonimmune rabbits (P < 0.01) (Fig. 3). In rabbits immunized by intragastric instillation, the caries area was 72% of that in nonimmune rabbits (P < 0.05), and this is similar to the results found for prevention by intragastric instillation in previous studies (11–13). Our results show that tonsillar application was highly effective for preventing dental caries, and this prevention, in the rabbits immunized by tonsillar application, appeared to be related to the level of anti-S. sobrinus IgA in saliva.

FIG. 2.

Teeth after inoculation with S. sobrinus for 15 weeks in nonimmune rabbits (A), rabbits immunized by intramuscular injection (B), rabbits immunized by intragastric instillation (C), rabbits immunized by tonsillar application (D), rabbits inoculated with S. sobrinus with galactose (E), and noninoculated rabbits (F). The dark parts of the tooth surface are the caries regions. In noninoculated rabbits, no caries regions were found.

FIG. 3.

Proportion of living S. sobrinus colonies in 100 random colonies (open column) and area of caries regions (solid column) on the tooth surface after immunization and subsequent S. sobrinus inoculation for 15 weeks. The proportion of living S. sobrinus colonies and the area of caries regions are shown as an average of three individual rabbits in each group. Error bars indicate SD. Abbreviations: non-immun., nonimmune rabbits; i.m. inject., rabbits immunized by intramuscular injection; i.g. instil., rabbits immunized by intragastric instillation; ton. appl., rabbits immunized by tonsillar application; gal., galactose. Symbols for statistical significance: ∗, P < 0.05; ∗∗, P < 0.01 (versus S. sobrinus-inoculated nonimmune rabbits); †, P < 0.01 (versus S. sobrinus-inoculated rabbits receiving tonsillar application).

The changes in the proportion of living S. sobrinus colonies in 100 random colonies cultured from the tooth surface was nearly parallel to the changes in the areas of the caries regions (Fig. 3). The decrease in caries area seems to be caused by the specific elimination of living cells by anti-S. sobrinus IgA in saliva. In the rabbits immunized by tonsillar application and in which the galactose was administered after inoculation with S. sobrinus, the S. sobrinus cell proportion and the caries areas were significantly greater than those in the rabbits inoculated with S. sobrinus alone (P < 0.01). These increases caused by galactose were not observed in nonimmune rabbits (data not shown). Since galactose appeared to decrease the potential of the caries prevention by anti-S. sobrinus salivary IgA, it appears that the antibody recognizes a galactose-containing antigen of S. sobrinus, such as the serotype antigen (18).

Nasal application is known to be another method for inducing salivary IgA (15). Both the antibody concentration and the level of agglutination in saliva of rabbits immunized by nasal application were half of those obtained by tonsillar application in the case in which sheep erythrocytes (3–6) and S. sobrinus (8) were used as antigens. Since the specific IgA concentration seems to be important for caries prevention, as suggested above, nasal application appears to have a lower potential for caries prevention than tonsillar application. Intragastric instillation with and nasal application of bacterial cell components are other successful methods for preventing caries without causing a cross-reaction in human cardiac muscle (9, 10, 16). The major problem with these methods of immunization is that they often require adjuvants to enhance the induction of the salivary antibodies (15, 19, 21). In human vaccines, cholera toxin B subunit is often included as an adjuvant (1, 14). Cholera toxin B subunit may induce oral tolerance and delayed-type hypersensitivity (17, 20). Tonsillar application of the bacterial cells induces enough antibodies to prevent dental caries without using such adjuvants.