Binding properties of streptococcal glucosyltransferases for hydroxyapatite, saliva-coated hydroxyapatite, and bacterial surfaces

Archives of Oral Biology 43 (1998) 103±110

ARCHIVES OF ORAL BIOLOGY

Binding properties of streptococcal glucosyltransferases for hydroxyapatite, salivacoated hydroxyapatite, and bacterial surfaces A. M. Vacca-Smith *, W. H. Bowen Department of Dental Research, Rochester Caries Research Center, University of Rochester, 601 Elmwood Avenue, Box 611, Rochester, New York 14642 Accepted 1 October 1997

Abstract The binding speci®cities of Streptococcus glucosyltransferase (Gtf) B, C and D for hydroxyapatite (HA), salivacoated hydroxyapatite (SHA), and bacterial surfaces were examined. For HA beads the following values were obtained: (K = anity; N = number of binding sites) GtfB, K = 46  105 ml/mmol, N = 0.65  10ÿ6 mmol/m2; GtfC, K = 86  105 ml/mmol, N = 4.42  10ÿ6 mmol/m2; GtfD, K = 100  105 ml/mmol, N = 0.83  10ÿ6 mmol/m2. For SHA beads, the following values were obtained: GtfB, K = 14.7  105 ml/mmol, N = 1.03  10ÿ6 mmol/m2; GtfC, K = 21.3  105 ml/mmol, N = 3.66  10ÿ6 mmol/m2; GtfD, K = 1.73  105 ml/mmol, N = 8.88  10ÿ6 mmol/ m2. The binding of GtfB to SHA beads was reduced in the presence of parotid saliva, but the binding of GtfC and D was una€ected. The binding of GtfB to SHA in the presence of parotid saliva supplemented with GtfC and D was reduced when compared with its binding to SHA in the presence of parotid saliva alone. In contrast, the binding of GtfC and D to SHA was una€ected when parotid saliva was supplemented with the other Gtf enzymes. GtfB bound to several bacterial strains (Strep. mutans GS-5, Actinomyces viscosus OMZ105E and Lactobacillus casei 4646) in an active form, while GtfC and D did not bind to bacterial surfaces. It is concluded that of the three Gtf enzymes, GtfC has the highest anity for HA and SHA surfaces and can adsorb on to the SHA surface in the presence of the other two enzymes. GtfD also binds to SHA in the presence of the other enzymes but has a very low anity for the surface. GtfB does not bind to SHA in the presence of the other Gtf enzymes but binds avidly to bacterial surfaces in an active form. Therefore, GtfC most probably binds to apatitic surfaces, while GtfB binds to bacterial surfaces. # 1998 Elsevier Science Ltd. All rights reserved Key words: Glucosyltransferase, Hydroxyapatite, Streptococci, Binding

1. Introduction Enzymatically active glucosyltransferase enzymes (EC 2.4.1.5) have been identi®ed in arti®cial and in vivo-derived salivary pellicles (RoÈlla et al., 1983; Scheie et al., 1987). These enzymes synthesize glucans in situ, which serve as binding sites for streptococci (Gibbons et al.,

* To whom all correspondence should be addressed. Abbreviations: PMSF, phenylmethylsulphonyl ¯uoride. 0003-9969/98/$19.00 # 1998 Elsevier Science Ltd. All rights reserved PII: S 0 0 0 3 - 9 9 6 9 ( 9 7 ) 0 0 1 1 1 - 8

1986; Schilling et al., 1989; Schilling and Bowen, 1992). The glucosyltransferases, which are produced by cariogenic streptococci such as Streptococcus mutans (Hamada and Slade, 1980; Loesche, 1986), are essential for the expression of virulence (Hamada et al., 1984; Yamashita et al., 1993) Strep. mutans produces at least three types of these enzymes (Loesche, 1986; Hanada and Kuramitsu, 1989). Glucosyltransferase B synthesizes an insoluble glucan from sucrose, composed mostly of a1,3-linked glucose moieties. Glucosyltransferase D produces a soluble glucan that

104

A. M. Vacca-Smith, W. H. Bowen / Archives of Oral Biology 43 (1998) 103±110

has predominantly a1,6-linked glucosyl units, and glucosyltransferase C produces a polymer with both a1,3linked and a1,6-linked glucans (Loesche, 1986; Hanada and Kuramitsu, 1989). The salivary pellicle, also termed the acquired enamel pellicle (Dawes et al., 1963), forms on tooth surfaces from salivary molecules (Turner, 1958a,b). The pellicle is formed by the selective adsorption of salivary proteins and glycoproteins on to tooth and apatitic surfaces (Mayhall, 1970; Hay, 1967, 1973; Sonju and RoÈlla, 1973). However, it has also been shown that bacterial-derived enzymes are incorporated into the pellicle (RoÈlla et al., 1983, Scheie et al., 1987). We have shown that glucosyltransferase from a mixture of glucosyltransferase enzymes and other products can adsorb on to saliva-coated hydroxyapatite in an active form, and also that glucosyltransferase C from Strep. mutans is the predominant glucosyltransferase f in in vitro-derived salivary pellicles (Vacca-Smith et al., 1996). Nevertheless, the binding speci®cities of the various glucosyltransferases for the saliva-coated hydroxyapatite beads has not been demonstrated. Our purpose now was to determine the binding properties of the three glucosyltransferase enzymes of Strep. mutans GS-5 for bare and saliva-coated hydroxyapatite beads. The binding properties of the three enzymes for various plaque micro-organisms were also evaluated. 2. Materials and methods 2.1. Bacterial strains and culture conditions The strains used included: Strep. mutans GS-5, Actinomyces viscosus OMZ105E, and Lactobacillus casei 4646. Bacteria were grown for 18 hr in TYG medium (1.4% tryptone/0.8% yeast extract/1.0% glucose) at 378C in a 5%CO2 incubator. After growth, bacteria were harvested by centrifugation at 6,000 g and were washed three times with imidazole bu€er by centrifugation at 6,000 g. The bacteria were suspended in 25 mmol/l imidazole bu€er, pH 6.5, supplemented with sodium azide (0.02%, ®nal concentration) and the protease inhibitor PMSF (1.0 mmol/l, ®nal concentration). The samples were placed in an ice-bath and were sonicated with three 30-sec pulses at 300 W (Braun-Sonic 1510, Braun Biotech, Allentown, PA) to disperse aggregates. After sonication, the optical density of the preparation was adjusted to 1.520.05 at 540 nm. The bacteria were diluted, and the absorbance of each diluted sample was determined at 540 nm. The number of bacterial cells/ml of each dilution was counted on a Petro€Hausser cell-counting chamber (Hausser Scienti®c Partnership, Horsham, PA).

2.2. Collection of parotid saliva Parotid saliva was collected on ice from one donor using a Lashley cup (Lashley, 1916). The saliva was diluted 1:1 with 25.0 mmol/l imidazole bu€er, pH 6.5, supplemented with sodium azide (0.02%, ®nal concentration) and PMSF (1.0 mmol/l, ®nal concentration). 2.3. Hydroxyapatite beads In each sample, 10.0 mg of hydroxyapatite beads were used (Integration Separation Systems, Hyde Park, MA). The surface area of the beads was 0.24 m2, and the particle size was 60.0±100.0 mm. 2.4. Glucosyltransferase enzymes Glucosyltransferase enzymes were prepared from constructs of Strep. milleri (Fukushima et al., 1992) that contained the genes for glucosyltransferases of Strep. mutans GS-5. The strains used were Strep. milleri KSB8, which harbours the gene for glucosyltransferase B, Strep. milleri KSC43, which has the gene for glucosyltransferase C, and Strep. milleri NH5, which contains the gene for glucosyltransferase D. The strains were gifts from Dr H. K. Kuramitsu, SUNY Bu€alo, NY. Bacteria were grown in brain±heart infusion medium (Difco Laboratories, Detroit, MI). The medium was supplemented with 0.2 mmol/l MnSO4, 1.0% Tween-20, 0.1% glucose, 2.0% sorbitol, 10.0 mmol/l PMSF and 10.0 mg/l of erythromycin. Enzymes were puri®ed according to methods previously described, and a single band was visualized on sodium dodecyl sulphate±polyacrylamide electrophoretic gels (Venkitaraman et al., 1995). Enzymes were stored at ÿ86.08C in imidazole bu€er supplemented with 10.0% glycerol. Protein concentration was determined by the 260±280 nm absorbance method outlined in Suelter (1985). Glucosyltransferase activity was determined in the presence of a dextran primer as described elsewhere (Schilling and Bowen, 1988). Puri®cation of glucosyltransferase B resulted in 1.22 mg/ml (spec. act. 0.00355 mmol of glucose incorporated into glucan/min per m g protein), puri®cation of glucosyltransferase C yielded 2.66 mg/ml (spec. act. 0.00149 mmol of glucose incorporated into glucan/min per m g protein) and puri®cation of glucosyltransferase D produced 800 mg/ml (spec. act. 0.000025 mmol of glucose incorporated into glucan/min per m mg protein). One unit of enzyme was de®ned as the amount of enzyme needed to incorporate 1 mmol of glucose into glucan over a 2-hr period. Enzymes were radiolabelled by [14C] formaldehyde methylation (Jentoft and Dearborn, 1979). The enzymes were mixed with bu€er (100 mmol/l Hepes, pH 7.5, 20 mmol/l NaCNBH3 and 250 mCi [14C] formaldehyde (NEN/Dupont, Boston, MA). The samples

A. M. Vacca-Smith, W. H. Bowen / Archives of Oral Biology 43 (1998) 103±110

were incubated overnight, and trichloroacetic acid was added to precipitate protein. The protein was pelleted by centrifugation at 3,000 g, suspended in imidazole bu€er, and concentrated by ultra®ltration in an Amicon Centricon-10 according to the manufacturer's instructions. The sample volume was increased to 1.0 ml with imidazole bu€er and mixed with unlabelled enzyme. A sample of this was counted by liquid-scintillation spectrometry, and speci®c activity was de®ned as mg protein/counts per min. 2.5. Binding of enzymes to hydroxyapatite beads Hydroxyapatite beads were incubated with doubling dilutions of radiolabelled enzymes (glucosyltransferase B, C, and D) for 30 min, at 378C. After incubation the supernatant ¯uids were removed and the amount of radioactive enzyme in them was measured by liquidscintillation spectrometry. The amount of unbound enzyme in the supernatant ¯uids was determined from the speci®c activity of the enzyme. The quantity of enzyme in the supernatant ¯uid was subtracted from the quantity of enzyme added to the beads, which resulted in the quantity of enzyme bound to the beads. This method can be summarized as follows: Amount of enzyme added to the beads ÿ the amount of enzyme in supernatant ¯uids = the amount of enzyme bound to the beads. To determine the nature of the binding, data were analysed in Langmuir binding isotherms (Adamson, 1960), in which the amount adsorbed per unit of surface area of adsorbant (mmol/m2) was plotted against the equilibrium concentration of adsorbate in solution(mmol/ml). The data were also plotted in Scatchard plots (Scatchard, 1948) to determine the number of binding sites (N, mmol/m2) and anity (K, ml/mmol), and Hill plots (Hill, 1913), to assess cooperativity, nH=Hill coecient). 2.6. Binding of enzymes to saliva-coated hydroxyapatite beads Hydroxyapatite beads were coated with 250 ml of parotid saliva (1:1 imidazole bu€er) for 30 min at 378C. After the incubation the beads were washed with imidazole bu€er and exposed to various concentrations of each enzyme for 30 min at 378C. After incubation, the quantity of bound enzyme was determined as described above, and data were analysed as above. 2.7. Binding of enzymes to saliva-coated hydroxyapatite beads in the presence of parotid saliva Hydroxyapatite beads coated with parotid saliva were exposed to various concentrations of glucosyltransferase enzymes in the presence of either parotid

105

saliva (1:1 dilution with imidazole bu€er) or imidazole bu€er. The beads were incubated with the enzymes and the amount of bound enzyme was determined as previously described. 2.8. Binding of enzymes to saliva-coated hydroxyapatite beads in the presence of parotid saliva and other glucosyltransferase enzymes Glucosyltransferase B (0.645  10ÿ6 mmol) was added to the coated beads in the presence of parotid saliva and either 2.06  10ÿ6 mmol of glucosyltransferase C, 2.06  10ÿ6 mmol of glucosyltransferase D, or imidazole bu€er and incubated for 30 min at 378C. The quantity of bound enzyme was determined as described earlier. Glucosyltransferase C (0.543  10ÿ6 mmol) was added to coated beads in the presence of parotid saliva and either glucosyltransferase B or glucosyltransferase D (2.06  10ÿ6 mmol) or imidazole bu€er, and the assay was as described as above. Glucosyltransferase D (1.34  10ÿ6 mmol) was added to coated beads in the presence of parotid saliva and either glucosyltransferase C, glucosyltransferase B (2.06  10ÿ6 mmol each) or imidazole bu€er, and the assay was carried out as described above. 2.9. Binding of enzymes to bacteria Radiolabelled glucosyltransferase B, C or glucosyltransferase D (0.645  10ÿ6 mmol of protein) were incubated with 1  107 bacterial cells (Strep. mutans GS-5, A. viscosus OMZ105E, or L. casei 4646) for 30 min at 378C. After incubation, the bacteria were pelleted by centrifugation at 6,000 g and washed three times by centrifugation with imidazole bu€er. The bacterial cells were transferred to scintillation vials, and the quantity of bound enzyme was determined by direct liquid-scintillation spectrometry. To determine the e€ect of parotid saliva on the binding of the enzymes to bacterial surfaces, the enzymes (0.645  10ÿ6 mmol each of glucosyltransferase B, C and D) were added to bacteria in the presence of either parotid saliva or imidazole bu€er, incubated, and assayed as described above. 2.10. Activity of glucosyltransferase enzymes bound to bacterial surfaces Glucosyltransferase-coated bacterial cells were suspended in bu€er and exposed to 14C-labelled glucosylsucrose (100 mmol/l ®nal concentration in imidazole bu€er) for 2 hr at 378C. After glucan formation, bacteria were suspended in ice-cold ethanol and activity was assayed as described elsewhere (Schilling and Bowen, 1988). Some of the organisms have intrinsic glucosyltransferase activity. Therefore, controls con-

106

A. M. Vacca-Smith, W. H. Bowen / Archives of Oral Biology 43 (1998) 103±110

Table 1 Binding properties of glucosyltransferase (Gtf ) enzymes for hydroxyapatite beadsa Gtf enzyme

Anity, K (ml/mmol)

Number of binding sites Nc (mmol/m2)

GtfB GtfC GtfD

46  105 (0.87) 86  105 (0.99) 100  105 (0.95)

0.65  10ÿ6 (0.11) 4.42  10ÿ6 (0.98) 0.83  10ÿ6 (0.34)

b

Hill coecient ndH 1.11 (0.99) [(+) Coop.] 3.04 (0.95) [(+) Coop.] 6.14 (0.98) [(+) Coop.]

a

The data shown are the mean values of three samples of a single experiment. K, anity, calculated from Scatchard plots, R-values are indicated in parentheses. c N, number of binding sites, calculated from Scatchard plots. SDs are indicated in parentheses. d nH, Hill coecient, calculated for Hill plots. R-values are indicated in parentheses. b

sisted of bacteria without any exogenous glucosyltransferase (non-exposed cells) incubated with radioactive sucrose. The amount of glucan formed in these control samples was then subtracted from that amount formed in experimental samples.

3. Results 3.1. Binding of enzymes to hydroxyapatite beads The binding properties of the glucosyltransferase enzymes for bare hydroxyapatite beads were studied to determine the number of binding sites and the apparent anities of the enzymes for the beads. The data are shown in Table 1. Glucosyltransferase C had the highest anity for the bare beads and the most binding sites (Table 1). The anity of glucosyltransferase D for bare hydroxyapatite beads was high (Table 1). Glucosyltransferase B had a lower number of binding sites and a lower anity for the beads than did glucosyltransferase C (Table 1). Hill plots revealed that all three enzymes displayed positive cooperativity, suggesting that the binding of enzyme to the beads favoured the further binding of enzyme to the beads (Table 1).

3.2. Binding of enzymes to saliva-coated hydroxyapatite beads The binding properties of glucosyltransferase B, C, and D for saliva-coated hydroxyapatite beads were determined. The data are shown in Table 2. Glucosyltransferase C had the highest anity for the saliva-coated beads, followed by glucosyltransferase B and then glucosyltransferase D. Glucosyltransferase D had the highest number of binding sites, followed by glucosyltransferase C and glucosyltransferase B (Table 2). Hill plots indicated that all three enzymes displayed positive cooperativity (Table 2). 3.3. Binding of enzymes to saliva-coated hydroxyapatite beads in the presence of parotid saliva We also determined whether the binding of the glucosyltransferase enzymes to the saliva coated beads was a€ected by the presence of parotid saliva. The results indicated that parotid saliva had no e€ect on the binding of glucosyltransferase C and glucosyltransferase D to coated beads, but reduced the binding of glucosyltransferase B to coated beads by 55%2 6% when compared with the binding of glucosyltransferase enzymes to saliva-coated hydroxyapatite in the presence of bu€er (data not shown).

Table 2 Binding properties of glucosyltransferase (Gtf ) enzymes for saliva-coated hydroxyapatite beadsa Gtf enzyme

Anity, K (ml/mmol)

Number of binding sites Nc (mmol/m2)

GtfB GtfC GtfD

14.7  105 (0.94) 21.3  105 (0.97) 1.73  105 (0.98)

1.03  10ÿ6 (0.62) 3.66  10ÿ6 (0.12) 8.88  10ÿ6 (0.24)

a

b

The data shown are the mean values of three samples of a single experiment. K, anity, calculated from Scatchard plots, R-values are indicated in parentheses. c N, number of binding sites, calculated from Scatchard plots. SDs are indicated in parentheses. d nH, Hill coecient, calculated for Hill plots. R-values are indicated in parentheses. b

Hill Coecient ndH 2.42 (0.94) [(+) Coop.] 2.40 (0.96) [(+) Coop.] 1.58 (0.95) [(+) Coop.]

A. M. Vacca-Smith, W. H. Bowen / Archives of Oral Biology 43 (1998) 103±110

107

Fig. 1. Binding of glucosyltransferase (Gtf)B to saliva-coated hydroxyapatite in the presence of GtfC, GtfD, and parotid saliva. Radiolabelled GtfB was added to SHA beads in the presence of parotid saliva supplemented with GtfC and GtfD. After incubation, the amount of GtfB bound to the beads was determined. The data shown are the mean values for three samples2the SD of a single experiment.

Fig. 2. Binding of glucosyltransferase (Gtf) enzymes to bacterial strains. Radiolabelled enzymes were added to Strep. mutans GS-5, A. viscosus OMZ105E, and L. casei 4646. After incubation, the amount of Gtf bound to the bacterial cells was determined. The data shown are the mean values for three samples2the SD of a single experiment.

3.4. Binding of enzymes to saliva-coated hydroxyapatite beads in the presence of parotid saliva and other glucosyltransferase enzymes

organisms. However, the binding of glucosyltransferase B was signi®cantly more avid than was the binding of the other glucosyltransferase enzymes. Although not shown here, additional experiments revealed that the binding of glucosyltransferase B to the bacterial surfaces was saturable and una€ected by the presence of parotid saliva. The data in Fig. 3 indicate that gluco-

The glucosyltransferase enzymes were added to coated beads in the presence of other glucosyltransferase enzymes and parotid saliva. The binding of glucosyltransferase B, when added to coated beads in the presence of parotid saliva, glucosyltransferase C and glucosyltransferase D, was reduced by 60%2 17% (Fig. 1) when compared with the controls. The binding of glucosyltransferase C to coated beads was unaffected by the presence of parotid saliva, glucosyltransferase B and glucosyltransferase D (data not shown). Also, the binding of glucosyltransferase D to coated beads was una€ected by the presence of parotid saliva, glucosyltransferase B and glucosyltransferase C (data not shown). Collectively, these results suggest that the binding activity of glucosyltransferase B to salivacoated hydroxyapatite beads in the presence of glucosyltransferase C and glucosyltransferase D is greatly reduced, but the binding of glucosyltransferase C and glucosyltransferase D to such beads is una€ected by the presence of the other glucosyltransferase enzymes. 3.5. Binding of enzymes to bacteria and activity of glucosyltransferase enzymes bound to bacterial surfaces We also studied the abilities of the glucosyltransferase enzymes to adsorb on to bacterial surfaces. As seen in Fig. 2, all of the enzymes bound to the micro-

Fig. 3. Activity of glucosyltransferase (Gtf)B adsorbed on to bacterial strains. Bacterial cells coated with GtfB were exposed to sucrose, and glucan formation was assayed after a 2-hr incubation. The data shown are the mean values for three samples2the SD of a single experiment, and were corrected for any endogeneous Gtf activity.

108

A. M. Vacca-Smith, W. H. Bowen / Archives of Oral Biology 43 (1998) 103±110

syltransferase B retained enzymatic activity once adsorbed on to bacterial surfaces. 4. Discussion The binding of several salivary proteins to hydroxyapatite beads has been explored in great detail (Hay, 1967; Bennick and Cannon, 1978; Moreno et al., 1978; Bennick et al., 1979; Tabak et al., 1985). As glucosyltransferase has been identi®ed in in vitro and in vivoderived salivary pellicles (RoÈlla et al., 1983; Scheie et al., 1987; Schilling and Bowen, 1988), we explored the binding properties of three glucosyltransferase enzymes of Strep. mutans GS-5 when exposed to hydroxyapatite surfaces. The binding of salivary proteins to various bare hydroxyapatite surfaces usually followed the Langmuirian model in terms of binding speci®cities (Moreno et al., 1978; Tabak et al., 1985). Langmuir adsorption isotherms are used to describe independent and non-interacting sites (Adamson, 1960). In these models, a rapid rise in the bound protein occurs with increasing protein concentrations, up to a limiting value (Adamson, 1960). However, we found that the binding of glucosyltransferase enzymes to bare and saliva-coated hydroxyapatite beads did not ®t the Langmurian models but rather the Scatchard model (data not shown), indicating binding-site heterogeneity or cooperative interactions (Scatchard, 1948; Levitzki and Koshland, 1976; Dahlquist, 1978; Jose and Larralde, 1982; Thakur et al., 1980). In cases of site heterogeniety, multilayer formation on surfaces can occur, depending on the type of cooperativity. If positive cooperativity exists, the binding of one molecule will produce favourable conditions for the binding of others. If negative cooperativity exists, the binding of one molecule produces conditions not favourable to the deposition of other similar molecules. Results from these experiments and from other studies (Vacca-Smith et al., 1996) support our hypothesis that the predominant glucosyltransferase on the coated beads has glucosyltransferase C-like qualities and may be the glucosyltransferase that binds coated beads from whole saliva. However, the source of glucosyltransferase in saliva is unknown and remains to be demonstrated. Our results also show that glucosyltransferase B may be the predominant glucosyltransferase on bacterial surfaces. These conclusions are drawn form the following observations. Glucosyltransferase C bound to saliva-coated beads favoured the deposition of other glucosyltransferase C enzymes on to the beads, and was able to adsorb the beads in the presence of glucosyltransferase B and D, and in the presence of saliva; glucosyltransferase C has a higher anity for the coated beads than do the other glucosyltransferase enzymes. Glucosyltransferase B showed

the best binding capacity for bacterial surfaces. This is signi®cant because glucosyltransferase enzymes bind to various bacterial surfaces (McCabe and Donkersloot, 1977), glucosyltransferase B interacts with streptococcal cell surfaces (Kato and Kuramitsu, 1991), and glucosyltransferase enzymes have glucan-binding domains and interact with glucans (Mooser and Wong, 1988). This interaction of glucosyltransferase B with streptococcal cell surfaces was based on the structure of the C-terminus of the enzyme and on de novo glucan synthesis. We showed that the interaction of active glucosyltransferase B with cell surfaces occurred in the absence of any glucan synthesis from exogeneous sucrose. However, there is a possibility that glucan synthesis arose from any trace amounts of sucrose present in the brain±heart infusion medium. At any rate, the interaction of glucosyltransferase B with bacterial cell surfaces is signi®cant and could explain why early studies on the histology of dental plaque show bacteria embedded in a mesh of carbohydrate (Critchley et al., 1967, 1968). The C-terminus of glucosyltransferase B di€ers from that of glucosyltransferase C, and this may explain the selective adsorption of glucosyltransferase C on to the saliva-coated hydroxyapatite beads and the selective adsorption of glucosyltransferase B on to bacterial surfaces. All glucosyltransferase enzymes examined thus far have the same basic structure (Russell, 1994): a signal peptide of approx. 38 amino acids at the amino terminus followed by a highly variable domain of about 200 amino acids, distinctive for each enzyme. The C-terminus contains a series of repeat motifs of amino acids, and the pattern of the motif varies for each enzyme. Glucosyltransferase B contains 1475 amino acids and is highly hydrophilic (Shiroza et al., 1987). The structure of glucosyltransferase D is similar to that of glucosyltransferase B, with 1430 amino acids, and is highly hydrophilic (Honda et al., 1990). Glucosyltransferase C, which contains 1375 amino acids, is also highly hydrophilic (Ueda et al., 1988). However, unlike glucosyltransferase B and D, glucosyltransferase C also has small hydrophobic domains in the direct repeat units at the C-terminus (Ueda et al., 1988). Perhaps the presence of these hydrophobic domains confers a conformation on glucosyltransferase C which is di€erent from that of B and D and thus favours the interaction of glucosyltransferase C with the saliva-coated surface of the beads. Interestingly, the deletion of C-terminal hydrophobic residues from glucosyltransferase enzymes of Strep. gordonii resulted in decreased abilities of the enzymes to bind to biotin± dextran, suggesting that C-terminal domains of glucosyltransferase may mediate binding activities of glucosyltransferase enzymes (Vickermann et al., 1996). While the role of glucosyltransferase D in the mouth and in plaque formation remains unclear, results from

A. M. Vacca-Smith, W. H. Bowen / Archives of Oral Biology 43 (1998) 103±110

the studies presented here suggest that glucosyltransferase D binds loosely to apatitic surfaces. Perhaps glucosyltransferase D may serve a transient role in plaque formation by providing a primer for other glucosyltransferase enzymes bound to apatitic surfaces and bacteria. What is certain, however, from these studies is that glucosyltransferase C can associate with apatitic surfaces, and that glucosyltransferase B can be associated with bacterial surfaces. Knowledge of these associations is important, for this information could help in the development of antiplaque agents that reduce the activities of glucosyltransferase C adsorbed on to apatitic surfaces and GtfB adsorbed onto bacterial surfaces. Acknowledgements This study was supported by USPHS Grant DEO7907.

References Adamson A. W. (1960) Physical chemistry of surfaces. New York: Wiley Interscience. Bennick, A., Cannon, M. (1978) Quantitative study of the interaction of salivary acidic proline-rich proteins with hydroxyapatite. Caries Res. 12, 159±169. Bennick, A., Cannon, M., Madapallimattam, G. (1979) The nature of hydroxyapatite-binding sites in salivary acidic proline-rich proteins. Biochem. J. 183, 115±126. Critchley, P., Saxton, C. A., Kolendo, A. B. (1968) The histology and histochemistry of dental plaque. Caries Res. 2, 115±119. Critchley, P., Wood, J. M., Saxton, C. A., Leach, S. A. (1967) The polymerisation of dietary sugars by dental plaque. Caries Res. 1, 112±129. Dahlquist, F. W. (1978) The meaning of Scatchard and Hill plots. Methods Enzymol. 48, 270±299. Dawes, C., Jenkins, G. N., Tinge, C. H. (1963) The nomenclature of the integuments of the enamel surface of the teeth. Br. Dent. J. 115, 65±68. Fukushima, K., Ikeda, T., Kuramitsu, H. K. (1992) Expression of Streptococcus mutans gtf genes in Streptococcus milleri. Infect. Immun. 60, 2815±2822. Gibbons, R. J., Cohen, L., Hay, D. I. (1986) Strains of Streptococcus mutans and Streptococcus sobrinus attach to di€erent pellicle receptors. Infect. Immun. 52, 555±561. Hamada, S., Koga, T., Ooshima, T. (1984) Virulence factors of Streptococcus mutans and dental caries prevention. J. Dent. Res. 63, 407±411. Hamada, S., Slade, H. D. (1980) Biology, immunology, and cariogenicity of Streptococcus mutans. Microbiol Rev. 44, 331±384. Hanada, N., Kuramitsu, H. K. (1989) Isolation and characterization of the Streptococcus mutans gtf D gene, coding for synthesis of primer dependent soluble glucan synthesis. Infect. Immun. 57, 2079±2085.

109

Hay, D. I. (1967) The adsorption of salivary proteins by hydroxyapatite and enamel. Archs oral Biol. 12, 937±946. Hay, D. I. (1973) The interaction of human parotid salivary proteins with hydroxyapatite. Archs oral Biol. 20, 1517± 1529. Hill, A. V. (1913) The combinations of haemoglobin with oxygen and with carbon monoxide. I. Biochem. J. 7, 471± 480. Honda, O., Kato, C., Kuramitsu, H. K. (1990) Nucleotide sequence of the Streptococcus mutans gtf D gene encoding the glucosyltransferase-S enzyme. J. Gen. Micro. 136, 2099± 2105. Jentoft, N., Dearborn, D. G. (1979) Labeling of proteins by reductive methylation using sodium cyanoborohydride. J. Biol. Chem. 254, 4359±4365. Jose, M. V., Larralde, C. (1982) Alternative interpretations of unusual Scatchard plots: contribution of interactions and heterogeneity. Math Biosc. 58, 159±170. Kato, C., Kuramitsu, H. K. (1991) Molecular basis for the association of glucosyltransferase with the cell surface of oral streptococci. FEMS Microbiol Lett. 79, 153±158. Lashley, K. S. (1916) Re¯ex secretion of the parotid gland. J. Exp. Psychol. 1, 46±465. Levitzki, A., Koshland, D. E., Jr (1976) The role of negative cooperativity and half-of-the sites reactivity in enzyme regulation. Curr. Top. Cell. Regul. 10, 1±40. Loesche, W. J. (1986) Role of Streptococcus mutans in human dental decay. Microbial Rev. 50, 353±380. Mayhall, C. W. (1970) Concerning the composition and source of the acquired pellicle of human teeth. Archs oral Biol. 15, 1327±1341. McCabe, R. M., Donkersloot, J. A. (1977) Adherence of Veilonella species mediated by extracellular glucosyltransferase from Streptococcus salivarious. Infect. Immun. 18, 726± 734. Mooser, G., Wong, C. (1988) Isolation of a glucan-binding domain of glucosyltransferase (1,6-a-glucan synthase) from Streptococcus sobrinus. Infect. Immun. 56, 880±884. Moreno, A. C., Kresak, M., Hay, D. I. (1978) Adsorption of two human parotid salivary macromolecules on hydroxy-, ¯uorhydroxy- and ¯uorapatites. Archs oral. Biol. 23, 525± 533. RoÈlla, G., Ciardi, J. E., Bowen, W. H. (1983) Identi®cation of IgA, IgG, lysozyme, albumin, amylase and glucosyltransferase in the protein layer adsorbed to hydroxyapatite from whole saliva. Scand. J. Dent. Res. 91, 186±190. Russell, R. R. B. (1994) The application of molecular genetics to the microbiology of dental caries. Caries Res. 28, 69±82. Scatchard, G. (1948) The attractions of proteins for small molecules and ions. Ann. N.Y. Acad. Sci. 51, 660±672. Scheie, A. A., Eggen, K. H., RoÈlla, G. (1987) Glucosyltransferase activity in human in vivo formed enamel pellicle and in whole saliva. Scand. J. Dent. Res. 95, 212±215. Schilling, K. M., Blitzer, M. H., Bowen, W. H. (1989) Adherence of Streptococcus mutans to glucans formed in situ in salivary pellicle. J. Dent. Res. 68(Spec. Iss.), 678± 1680. Schilling, K. M., Bowen, W. H. (1988) The activity of glucosyltransferases adsorbed onto saliva-coated hydroxyapatite. J. Dent. Res. 67, 2±8.

110

A. M. Vacca-Smith, W. H. Bowen / Archives of Oral Biology 43 (1998) 103±110

Schilling, K. M., Bowen, W. H. (1992) Glucan synthesized in situ in experimental pellicle functions as speci®c binding sites for Streptococcus mutans. Infect. Immun. 60, 284±295. Suelter C. H. (1985) A practical guide to enzymology. New York: John Wiley & Sons. Shiroza, T., Ueda, S., Kuramitsu, H. K. (1987) Sequence analysis of the glucosyltransferase B gene from Streptoccus mutans. J. Bacteriol. 169, 4263±4270. Sonju, T., RoÈlla, G. (1973) Chemical analyses of the acquired pellicle formed in two hours on cleaned teeth in vivo. Caries Res. 7, 30±38. Tabak, L. A., Levine, M. J., Jain, N. K., Bryan, A. R., Cohen, R. E., Monte, L. A., Zawacki, S., Nancollas, G. H., Slomiany, A., Slominay, B. L. (1985) Adsorption of human salivary mucins to hydroxyapatite. Archs oral Biol. 30, 423± 427. Thakur, A. K., Ja€e, M. L., Rodbard, A. (1980) Graphical analysis of ligand binding systems: evaluation by Monte Carlo studies. Anal. Biochem. 107, 279±295. Turner, E. P. (1958a) The integument of the enamel surface of the human tooth. 1. Dent. Practnr. dent. Rec. 8, 341± 348.

Turner, E. P. (1958b) The integument of the enamel surface of the human tooth. 2. Dent. Practnr. dent. Rec. 8, 373± 382. Ueda, S., Shiroza, T., Kuramitsu, H. K. (1988) Sequence analysis of the glucosyltransferase C gene from Streptococcus mutans GS-5. Gene 69, 101±109. Vacca-Smith, A. M., Venkitaraman, A. R., Schilling, K. M., Bowen, W. H. (1996) Characterization of glucosyltransferase of human saliva adsorbed onto hydroxyapatite surfaces. Caries Res. 30, 354±360. Venkitaraman, A. R., Vacca-Smith, A. M., Kopec, L. K., Bowen, W. H. (1995) Characterization of glucosyltransferase B, GtfC, and GtfD in solution and on the surface of saliva-coated hydroxyapatite. J. Dent. Res. 74, 1695±1701. Vickermann, M. M., Sluvak, M. C., Minick, P. E., Clewell, D. B. (1996) Changes in the carboxyl-terminal repeat region a€ect extracellular activityand glucan products of Streptoccus gordonii glucosyltransferase. Infect. Immun. 64, 5117±5128. Yamashita, Y., Bowen, W. H., Burne, R. A., Kuramitsu, H. K. (1993) Role of the Streptococcus mutans glucosyltransferase genes in caries induction on the speci®c-pathogen-free rat model. Infect. Immun. 61, 3811±3817.