Streptococcus mutans Serotypes a and d: Characterization of

Transcription

1 INFCTION AND IMMUNITY, OCt. 1975, p Copyright X) 1975 American Society for Microbiology Vol. 12, No. 4 Printed in U.SA. Serological Purification of Polysaccharide Antigens from Streptococcus mutans Serotypes a and d: Characterization of Multiple Antigenic Determinants ROSEMARY LINZER, HIDEHIKO MUKASA,' AND HUTTON D. SLADE* Department of Microbiology, Northwestern University Medical and Dental Schools, Chicago, Illinois Received for publication 13 May 1975 The polysaccharide antigen preparations from serotype a and serotype d strains of Streptococcus mutans contained both a serotype-specific antigenic determinant and a common a-d antigenic determinant, as demonstrated by agar gel diffusion studies and a quantitative cross-precipitin assay. The chromatographically purified antigens were isolated by a method which depended on their serological specificity to determine if these two antigenic determinants were located on the same molecule. The a and d polysaccharides were recovered from specific antigen-antibody complexes and characterized with respect to their immunological specificity and chemical composition. Agar gel diffusion tests demonstrated that, in both the a and the d preparations, the serotype-specific antigenic determinant and the common a-d antigenic determinant were present in one molecule. An increased knowledge of the ability of ing the synthesis of cell-associated glucans necessary for cell adherence (26). This result lent Streptococcus mutans to form plaque on the smooth surfaces of teeth and its role in the support to an earlier proposal that the purified development of dental caries (4, 6, 14, 32, 34) serotype antigens of S. mutans might be used has stimulated interest in the development of as a possible vaccine against these cariogenic vaccines against this organism. By immunizing organisms (35). with whole formalin-killed cells ofs. mutans in A vaccine based on a serotype antigen is by the vicinity of the parotid glands, a correlation definition limited in scope. Seven serotypes of between increased salivary immunoglobulin A S. mutans (serotypes a through g) have been (IgA) titers and reduced caries incidence has demonstrated to date (1, 30). These serotypes been demonstrated in rats (35). Recently, an are distinct from the Lancefield serogroups of increase in serum IgG titer has been related to streptococci, except for minor cross-reactions reduction of caries in rhesus monkeys after subcutaneous and submucous immunization with antigens have been characterized as polysaccha- with group E. The Lancefield group and type heat-killed cells of S. mutans (17). In the later rides, teichoic acids, or proteins (7, 15, 21, 36). study, serum IgG was thought to enter the oral Serotype a (24), b (23), c (R. Linzer, K. Gill, and cavity via the crevicular fluid. In these studies, H. Slade, J. Dent. Res., in press; J. Wetherell, the role ofthe antibodies is probably not bactericidal but rather involves interference with the crobiol., p. 53, 1975), and d (13, 18) antigens of A. Bleiweis, Abstr. Annu. Meet. Am. Soc. Mi- adherence of S. mutans to the tooth surface. S. mutans have been characterized as polysaccharides located in or associated with the bac- Secretory IgA from saliva has been shown to specifically inhibit the adherence of some terial cell wall. Several partial cross-reactions strains of S. salivarius and S. mitis to human have been described among the serotypes. For buccal epithelial cells (37). In vitro studies have example, the a and d strains demonstrate partial cross-reactions (1, 2, 18, 24) as do the c and e shown inhibition of the adherence of S. mutans to smooth glass surfaces by antibodies against (2; Hamada and Slade, J. Dent. Res. [suppl. C], whole cells of S. mutans (8, 28). Some of these in press) and c, e, and f strains (Hamada and antibodies appeared serotype specific (8). Indeed, antibodies to the serotype a polysaccha- antigens may be of importance in the develop- Slade, unpublished data). These cross-reacting ride antigen have been shown to inhibit adherence by blocking the binding ofglucosyltransfer- Previous work with the a and d antigens, ment of a purified vaccine. ase to the cell wall and, consequently, inhibit- using agar gel diffusion and immunoelectrophoresis, suggested that the serotype-specific deter- I Present address:bdei Medical School, Saitama-Ken, Japanminant and the cross-reacting a-d determinant 791

2 792 LINZER, MUKASA, AND SLADE were on one polysaccharide molecule (18, 24). However, several streptococci have been shown to contain distinct group and type polysaccharides in their cell wall or capsular material, e.g., groups B and F (16, 36). Therefore, the possibility of multiple polysaccharides existing in the preparations from S. mutans was considered. These polymers would have to be of similar size and electrical charge to have remained undetected through column fractionations and immunoelectrophoresis. Purification of polysaccharides by their serological specificity has proved helpful in the study of streptococcal group and type antigens (21, 36). Therefore, serotype a and serotype d antigens of S. mutans were isolated by a method which depended on their serological specificity to verify previous data concerning their composition and multiple antigenic determinants. This report describes the characterization of the a and d polysaccharides recovered from specific antigen-antibody complexes. Anti-a serumi MATERIALS AND METHODS S. mutans antigens. S. mutans strains HS6 (serotype a) and B13 (serotype d) were grown in Todd- Hewitt broth (Difco Laboratories, Detroit, Mich.) supplemented with 1.8% glucose and salts (12). The serotype antigens were extracted and purified by repeated column chromatography as previously described (18, 24). Serological procedures. Anti-HS6 and anti-b13 sera were prepared in New Zealand white rabbits as reported (18). Specific anti-a and anti-d sera were prepared by adsorbing whole anti-hs6 serum with B13 cells (serotype d) and, conversely, by adsorbing whole anti-b13 serum with HS6 cells (serotype a) (18, 24). This procedure removed cross-reacting antia-d globulins. The anti-a-d globulins were recovered by incubating the cells at ph 2 for 1 h at 5 C. A quantitative cross-precipitin reaction was performed with the anti-a-d preparation. Anti-a-d globa a, a-d determiiinants) Anti-B13S serum (d. a-d globulins) ulins, prepared from whole anti-hs6 serum, were reacted with increasing concentrations of a antigen at 37 C for 1 h and 5 C for 2 h. The precipitates were removed by centrifugation and washed initially with 50,ul of saline. The precipitates were then washed and assayed according to the standard quantitative precipitin assay (33). The supernatant and the initial wash (100 Il, total volume) were incubated with 0.3 ug of d antigen as above and assayed as previously reported (33). Agar gel diffusion techniques have been described (29). Serological purification of antigens. The scheme for the serological purification of the a and d antigen preparations is shown in Fig. 1. Samples of chromatographically purified HS6-antigen (15,tg) were combined with 0.15 ml of anti-a serum or 0.3 ml of whole anti-b13 serum plus saline to a final volume of 0.45 ml. Similarly, samples of chromatographically purified B13-antigen (10,ug) were combined with 0.15 ml of whole anti-hs6 serum or 0.3 ml of anti-d serum plus saline to a final volume of 0.45 ml. The reaction mixtures were incubated at 37 C for 1 h and at 5 C for 2 h. The precipitated antigen-antibody complexes were collected by centrifugation at 6,000 x g for 20 min and were washed three times with 0.5 ml of saline. The antigen-antibody complexes were solubilized using a method based on procedures developed by McCarty and Lancefield for the study of group A streptococcal antigens (21). The washed complexes were dissolved in 0.01 N HCl (0.1 ml) with intermittent stirring at 25 C for 10 min. An equal volume of 5% trichloroacetic acid was added, and the incubation was continued at 25 C for 10 min and 5 C for an additional 10 min. The precipitated globulins were removed using a membrane filter (MF type, 0.22,um pore size) fitted in a Swinnex-13 filter holder (Millipore Corp., Bedford, Mass.). Samples for serological analysis were neutralized with 1 N NaOH; samples for gas-liquid chromatographic analysis were extracted three times with equal volumes of ether to remove trichloroacetic acid and were lyophilized. Analytical methods. Total sugars were measured by the phenol-sulfuric acid method (5). Gas-liquid Aniti-d serum d (d, a-d determiiinants) INFECT. IMMUN. Anti-HS6 serum (a, a-( globulins) -AB 0.01 N HCI AlB 5', trichlouracetic aci(d A Bi PP-t -A B N HCI AB 5' trichloroacetic acid ABi PPv -AB t 0.01 N HCI AB 5e, tr-ichlorouacetic acidi AB(PPi) -AB N HCI AB 5 trichloroacetit acid FIG. 1. Serological purification scheme. Antigens were recovered from specific antigen-antibody complexes (Ag-Ab) as described in Materials and Methods. AB(PP

3 VOL. 12, 1975 S. MUTANS POLYSACCHARIDE ANTIGENS 793 chromatography was used for the identification and quantitative analysis of specific monosaccharides (18, 22). Proteins were determined by the method of Lowry et al. (19). RESULTS Characterization of chromatographically purified antigens. The a and d antigen preparations that had been purified by repeated column chromatography were examined by agar gel diffusion, a quantitative cross-precipitin reaction, and gas-liquid chromatography. In Fig. 2A, the preparations were tested in agar with a mixture of specific anti-a and anti-d sera. The resulting precipitin bands had double spurs which characterized the polysaccharides as serologically distinct molecules. Serotype a antigen did not react with adsorbed anti-d serum, nor did serotype d antigen react with adsorbed anti-a serum (see Fig. 4 and 5). However, the a and d antigen preparations demonstrated a reaction of serological identity when reacted with anti-a-d globulins (Fig. 2B). The reactions of the a and d antigens with anti-a-d globulins were examined further by the quantitative precipitin assay. The precipitin curves resulting from the reaction of anti-a-d globulin (5,ul) with increasing concentrations of d and a antigens are shown in Fig. 3A and B, respectively. At their equivalence points, the two antigens precipitated 15.9 and 15.4,g of antibody protein. After incubation of the anti-a-d globulins with a antigen and removal of the precipitate by centrifugation, the supernatant and initial wash (100,ul total) were incubated with 0.3 ug of d antigen (Fig. 3B). If the anti-a-d preparation was a mixture of anti-a and anti-d globulins, the reaction with the d antigen would have remained constant with increasing concentrations ofa antigen. However, this was not the case. The cross-precipitin assay yielded reciprocal curves and thus confirmed that the a and b antigens were reacting with the same anti-a-d globulin. Therefore, the antigen preparations contained both a serotype-specific determinant and a cross-reacting antigenic determinant. Analysis on gas-liquid chromatography revealed the ratios of galactose to glucose in the a and d preparations to be 5.5:1 and 1.9:1, respectively (Table 1). The recovery of sugars by gasliquid chromatographic procedures averaged 95% with respect to total sugar determinations by the phenol-sulfuric acid method. Characterization of serologically purified a antigen. Samples of serotype a antigen were recovered from specific antigen-antibody complexes to determine whether the a and a-d antigenic determinants were on the same or dis- FIG. 2. Agar diffusion study with chromatographically purified a and d antigens. Samples of a antigen (5 pg) and d antigen (3 pg) were assayed (A) with a mixture ofspecific anti-a (30 pi) and anti-d (40 p) sera and (B) with anti-a-d globulins (20 4d) prepared from whole anti-hs6 serum.

4 794 LINZER, MUKASA, AND SLADE li ride molecule. What appear to be slight spurs 15 A ^ B can be seen between the precipitin bands of '53 wells 2 and 3 and wells 4 and 5. These are z -\ / - lo probably due to the higher globulin and albumin content of the whole serum preparations. O0 / / Anti-a-d globulins consistently produced L 5_ sharper precipitin bands, presumably because they are relatively free of serum albumin. Nei- / X ' os L L. ther antigen preparation reacted with specific a4 0.5 a2 QA anti-d serum. ANTIGEN CU9) The composition of the antigen which had.... ~~been released and purified from anti-a globu- FIG. 3. Quantitative cross-precipitin assay with anti-a-d globulin. Anti-a-d globulin (from anti-hs6 serum, 5 p1) was reacted with increasing concentra- A tions of d antigen (A) and a antigen (B, *). After removing the antibody-a antigen complexes by centrifugation, the supernatants were incubated in the presence of03 pg of d antigen (B, 0) (see Materials and Methods). TABLE 1. Composition of a and d antigens Reducing Galactose Glucose Reov- Antigen sugars (mg/100 (mg ery (mg/100 mg) mg) mg) Serotype a Ia (5.5)b 8 (1) 85 IIC 38 (5.4) 7 (1) 74 Serotype d Ia (1.9) 34 (1) 104 IIC 33 (1.7) 19 (1) 54 a Chromatographically purified. b Parentheses indicate ratio to glucose concentration. c Purified from antigen-antibody complex with serotype-specific serum. tinct polysaccharide molecules. Chromatographically purified a antigen from S. mutans strain HS6 was reacted with specific anti-a serum and with anti-b13 serum which contained cross-reacting anti-a-d globulins. After dissolving the antigen-antibody complexes in 0.01 N HCI, the antibody globulins were removed by trichloroacetic acid precipitation. The recovered antigens were analyzed for their serological activity and composition. In Fig. 4, the serological activity of the polysaccharides recovered from specific complexes with anti-a serum and anti-b13 serum were tested by agar gel diffusion. The antigens were placed in the center wells and six antisera were distributed in the outer wells. Both samples reacted with anti-a serum and with anti-a-d sera prepared from whole anti-hs6 and anti- B13 sera. Therefore, the a and a-d antigenic determinants appeared to be on one polysaccha- B2 *i '5.. INFECT. IMMUN. FIG. 4. Agar diffusion study with recovered a antigen. Samples of serotype a antigen, recovered from complexes with (A) anti-a serum and (B) anti-b13 each). The outer wells contained various sera: (1) anti-a, 20 p1; (2) whole anti-hs6, 20 p1; (3) anti-a-d (from anti- HS6), 20 jd; (4) anti-a-d (from anti-b13), 80 /X1; (5) whole anti-b13, 80 p1; and (6) anti-d, 40 p1. Photograph was taken after 42 h. serum, were placed in the center wells (100 p1 Ji:

5 VOL. 12, 1975 lins was analyzed by gas-liquid chromatography. The analysis showed an identical profile with the starting material (Table 1). The galactose-to-glucose ratios of the preparations were 5.4:1. Galactosamine and glucosamine may also be present in trace amounts. Recovery of sugars after the serological procedures and gasliquid chromatography was 74%. Characterization of serologically purified d antigen. Samples ofchromatographically purified serotype d antigen from S. mutans strain B13 were precipitated with anti-d serum and whole anti-hs6 serum which contained anti-ad globulins. After dissolving the complexes and removing the antibody globulins by trichloroacetic acid precipitation, the recovered antigens were tested to determine whether the d and a-d antigenic determinants were present on the same molecule. The serological activity of the antigens recovered from specific complexes with anti-d serum and anti-hs6 serum were tested by agar gel diffusion (Fig. 5). Both samples reacted with anti-d serum and with anti-a-d sera prepared from whole anti-b13 and anti-hs6 sera. Therefore, the type d polysaccharide appeared to carry both the d and a-d antigenic determinants on one molecule. The samples did not react with anti-a serum. A quantitative precipitin assay was used to measure the recovered polysaccharides against standard curves of the purified B13 antigen (Fig. 6). Each sample was assayed with both anti-d and anti-hs6 sera. The two preparations averaged 0.12,.g of antigen per 4,ul tested. This calculated to 7.2,g (final volume 240 1l) or a total recovery of 72%. In a parallel experiment, antigen which had been released and purified from anti-d globulins was analyzed by gas-liquid chromatography (Table 1). Analysis revealed the presence of only two sugars, galactose and glucose. These sugars were present in proportions similar to those in the starting preparation, i.e., galactose to glucose, 1.7:1. Recovery of sugars after the serological procedures and gas-liquid chromatography was 54%. DISCUSSION The polysaccharide antigen preparations from serotype a and d strains of S. mutans contained both the serotype-specific antigenic determinant and a common a-d antigenic determinant, as demonstrated by agar gel diffusion studies (Fig. 2) and a quantitative cross-precipitin assay (Fig. 3). The object of this study was to determine if these two antigenic determinants were located on the same molecule, as sug- S. MUTANS POLYSACCHARIDE ANTIGENS 795 FIG. 5. Agar diffusion study with recovered d antigen. Samples of serotype d antigen, recovered from complexes with (A) anti-d globulin and (B) anti-hs6 serum, were placed in the center wells (100 Ad each). The outer wells contained various sera as described in Fig. 4, except (5) contained 20 pi. gested by previous studies using agar diffusion and immunoelectrophoresis (18, 24). A serological purification technique was used for this purpose. The polysaccharide antigen purified from S. mutans strain HS6 (serotype a) could be recovered from complexes with anti-a serum and with anti-b13 serum which contained anti-a-d globulins (Fig. 4). The recovered antigens carried both the specific serotype a and the common a-d determinants. Antigen recovered from

6 796 LINZER, MUKASA, AND SLADE,". s U) =Z z 0 cr OA 0.6 d ANTIGEN (iq) FIG. 6. Quantitative precipitin assay with serotype d antigen. Standard precipitin curves were established for anti-d serum (20 /.d) and anti-hs6 serum (10 p1) with increasing concentrations of serotype d antigen. Antigens (4 pi) recovered from complexes with anti-d globulin (0) and anti-hs6 serum (Ol) were tested with both sera. Final volumes were adjusted to 50 pi with saline. a complex with anti-a globulin was analyzed using gas-liquid chromatography (Table 1). The sugar composition of the recovered antigen appeared identical to the starting material; i.e., it contained galactose and glucose in a ratio of 5.4:1 with possible trace amounts of galactosamine and glucosamine. The recovery of sugars was 74%. In an experiment with antigen from S. mutans strain B13 (serotype d), samples recovered from complexes with anti-d and anti-hs6 sera contained both the serotype-specific d determinant and the cross-reacting a-d determinant (Fig. 5). The composition of the polysaccharide precipitated with anti-d serum was similar to that of the starting material; i.e., galactose and glucose were the only sugars detected and these were present in a 1.7:1 ratio (Table 1). Thus, both the serotype a and d antigens are polysaccharides containing at least two distinct antigenic determinants on one molecule, i.e., either the a and a-d or the d and a-d. Recently, two additional serotypes of S. mutans were proposed by Perch et al. (30). These new serotypes were designatedfandg in accordance with the Bratthall serotypes a through e (1, 2). Anti-g serum prepared against 0MZ65 was reported to cross-react with serotype a and serotype d extracts. The chromatographically purified a and d polysaccharide antigens characterized here were strongly precipitated by anti- 0MZ65 serum provided by B. Perch. However, INFECT. IMMUN. sufficient serum was not available to determine if these cross-reactions represented a common a-d-g determinant or the a-g, d-g determinants described by Perch et al. (30). The structure of the a and d serotype antigens may be similar to the structure of the polysaccharide 0 antigens of Salmonella. In this genus, variations of a basic trisaccharide repeating sequence result in a large number of distinguishable serotypes (20, 31). Variations may be related to (1) changes in the position or configuration of a linkage within the basic chain, (2) attachment of a monosaccharide at specific points on the basic sequence, or (3) deletion or substitution of one of the saccharides in the basic repeating unit (31). Since specificities to both terminal and nonterminal residues exist, multiple antigenic determinants are present on one polysaccharide molecule. Although galactose and glucose are the only major saccharides of the a and d serotype antigens of S. mutans, variations in the configuration and position of linkages between these monosaccharides could easily account for the several distinct antigenic determinants which have been serologically identified..a more extensive elucidation of these determinants will have to await complete structural analysis of the antigens. The existence of multiple antigenic determinants on the polysaccharide antigens of S. mutans may be helpful in the search for a vaccine to aid in the reduction of dental caries. Adherence of S. mutans to smooth surfaces, such as those on teeth, has been correlated with the production of insoluble glucans by a glucosyltransferase (8, 9, 11, 25, 27, 28). Antiserum which was specific for the serotype a cell wall polysaccharide antigen has been shown to block the binding of glucosyltransferase to the surface of heat-killed cells, thus inhibiting adherence of the cell to a glass surface (26). A similar study with serotype d antiserum has also shown inhibition of cell adherence (Linzer and Slade, unpublished observations). The serotype a and d antigens deserve particular consideration in a vaccine program for several reasons. First, serotypes c and d were found to be the most frequently occurring serotypes of S. mutans in a survey of dental plaque samples from 14 regions of 10 countries (3). In another study using fluorescein-conjugated antisera, serotype d was the most frequently occurring group in 51 plaque samples (10). Second, no serotype a, d, or g strains were isolated in a study of blood samples from 54 patients with subacute endocarditis. However, 50 strains of S. mutans serotype c were isolated from these 54 patients (30). Third, the multiple antigenic

7 VOL. 12, 1975 S. MUTANS POLYSACCHARIDE ANTIGENS 797 determinants on either the a or d antigen increase its spectrum of immunogenic activity to include antibodies against S. mutans strains of serotypes a, d, and g. The present report has firmly established the presence of both the serotype-specific and the common a-d antigenic determinants on the a and d antigen molecules, i.e., a and a-d or d and a-d. ACKNOWLEDGMENTS We gratefully acknowledge the technical assistance of Kamal Gill in preparing anti-hs6 and anti-b13 sera for this study, and we thank B. Perch for a sample of anti-omz65 serum. This investigation was supported by Public Health Service research grants HE from the National Heart and Lung Institute and DE from the National Institute of Dental Research; and by grants from the Grainger Fund, the Pioneer Fund, and the Hemac Fund. H.D.S. is the recipient of Public Health Service research career award K6-GM from the National Institute of General Medical Sciences. LITERATURE CITED 1. Bratthall, D Demonstration of five serological groups of streptococcal strains resembling Streptococcus mutans. Odontol. Revy 21: Bratthall, D Immunofluorescent identification of Streptococcus mutans. Odontol. Revy 23: Bratthall, D Demonstration of Streptococcus mutans strains in some selected areas of the world. Odontol. Revy 23: Carlsson, J Plaque formation and streptococcal colonization on teeth. Odontol. Revy 19(Suppl. 14): Dubois, M., K. A. Gilles, J. K. Hamilton, P. A. Rebers, and F. Smith Colorimetric method for determination of sugars and related substances. Anal. Chem. 28: Fitzgerald, R. J., H. V. Jordan, and H. R. Stanley Experimental caries and gingival pathologic changes in the gnotobiotic rat. J. Dent. Res. 39: Fox, E. N M proteins of group A streptococci. Bacteriol. Rev. 38: Genco, R. J., R. T. Evans, and M. A. Taubman Specificity of antibodies to Streptococcus mutans; significance in inhibition of adherence, p In J. Mestecky and A. R. Lawton (ed.), The immunoglobulin A system. Plenum Press, New York. 9. Gibbons, R. J., and M. Nygaard Synthesis of insoluble dextran and its significance in the formation of gelatinous deposits by plaque-forming streptococci. Arch. Oral Biol. 13: Grenier, E. M., W. C. Eveland, and W. J. Loesche Identification ofstreptococcus mutans serotypes in dental plaque by fluorescent antibody techniques. Arch. Oral Biol. 18: Guggenheim, B., and H. E. Schroeder Biochemical and morphological aspects of extracellular polysaccharides produced by cariogenic streptococci. Helv. Odontol. Acta 11: Hess, E. L., and H. D. Slade An electrophoretic examination of cell-free extracts from various serological types of group A hemolytic streptococci. Biochim. Biophys. Acta 16: Iacono, V. J., M. A. Taubman, D. J. Smith, and M. J. Levine Isolation and immunochemical characterization of the group-specific antigen of Streptotoccus mutans Infect. Immun. 11: Jordan, H. V., and P. H. Keyes In vitro methods for the study of plaque formation and various lesions. Arch. Oral Biol. 11: Krause, R. M The antigens of group D streptococci, p In L. W. Wannamaker and J. M. Matsen (ed.), Streptococci and streptococcal diseases. Academic Press Inc., New York. 16. Lancefield, R. C., and E. H. Freimer Type-specific polysaccharide antigens of group B streptococci. J. Hyg. 64: Lehner, T., S. J. Challacombe, and J. Caldwell Immunological and bacteriological basis for vaccination against dental caries in rhesus monkeys. Nature (London) 254: Linzer, R., and H. D. Slade Purification and characterization ofstreptococcus mutans group d cell wall polysaccharide antigen. Infect. Immun. 10: Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193: Luderitz, O., K. Jann, and R. Wheat Somatic and capsular antigens of Gram-negative bacteria, p In M. Florkin and E. H. Stotz (ed.), Comprehensive biochemistry, vol. 26A. Elsevier Publishing Co., Amsterdam. 21. McCarty, M., and R. C. Lancefield Variation in the group-specific carbohydrate of group A streptococci. I. Immunochemical studies on the carbohydrates of various strains. J. Exp. Med. 102: Mukasa, H., and H. D. Slade Chemical composition and immunological specificity of the streptococcal group 0 cell wall polysaccharide antigen. Infect. Immun. 5: Mukasa, H., and H. D. Slade Structure and immunological specificity of the Streptococcus mutans group b cell wall antigen. Infect. Immun. 7: Mukasa, H., and H. D. Slade Extraction, purification, and chemical and immunological properties of the Streptococcus mutans group a polysaccharide cell wall antigen. Infect. Immun. 8: Mukasa, H., and H. D. Slade Mechanism of adherence of Streptococcus mutans to smooth surfaces. I. Roles of insoluble dextran-levan synthetase enzymes and cell wall polysaccharide antigen in plaque formation. Infect. Immun. 8: Mukasa, H., and H. D. Slade Mechanism of adherence of Streptococcus mutans to smooth surfaces. II. Nature of the binding site and the adsorption of dextran-levan synthetase enzyme on the cell wall surface of the streptococcus. Infect. Immun. 9: Mukasa, H., and H. D. Slade Mechanism of adherence of Streptococcus mutans to smooth surfaces. III. Purification and properties of the enzyme complex responsible for adherence. Infect. Immun. 10: Olson, G. A., A. S. Bleiweis, and P. A. Small, Jr Adherence inhibition ofstreptococcus mutans: an assay reflecting a possible role of antibody in dental caries prophylaxis. Infect. Immun. 5: Ouchterlony, Diffusion in gel methods for immunological analysis. Progr. Allergy 5: Perch, B., E. Kjems, and T. Ravn Biochemical and serological properties of Streptococcus mutans from various human and animal sources. Acta Pathol. Microbiol. Scand. 82: Robbins, P. W., and T. Uchida Determinants of specificity in Salmonella: changes in antigenic structure mediated by bacteriophage. Fed. Proc. 21: Scherp, H. W Dental caries; prospects for prevention. Science 173:

8 798 LINZER, MUKASA, AND SLADE 33. Slade, H. D Extraction of cell-wall polysaccharide antigen from streptococci. J. Bacteriol. 90: Stoppelaar, J. D. De, K. G. Konig, A. J. M. Plasschaert, and J. S. Van Der Hoeven Decreased cariogenicity of a mutant of Streptococcus mutans. Arch. Oral Biol. 16: Taubman, M. A Role of immunization in dental disease, p In S. E. Mergenhagen and H. W. Scherp (ed.), Comparative immunology of the oral INFECT. IMMUN. cavity. U.S. Government Printing Office, Washington, D.C. 36. Willers, J. M. N., M. F. Michel, M. J. Sysma, and K. C. Winkler Chemical analysis and inhibition reactions of the group and type antigens of group F streptococci. J. Gen. Microbiol. 36: Williams, R. C., and R. J. Gibbons Inhibition of bacterial adherence by secretory immunoglobulin A: a mechanism of antigen disposal. Science 177: