Vol. 20, No. 3
INFECTION AND IMMUNITY, June 1978, p. 652-659 0019-9567/78/0020-0652$02.00/0 Copyright i 1978 American Society for Microbiology
Printed in U.S.A.
Interaction of Glucosyltransferases with the Cell Surface of Streptococcus mutans HOWARD K. KURAMITSU AND L. INGERSOLL Department of Microbiology-Immunology, Northwestern University Medical-Dental Schools, Chicago, Illinois 60611
Received for publication 20 January 1978
The partially purified glucosyltransferase (GTF) fraction synthesizing primarily water-insoluble glucans, GTF-A, and the homogeneous fraction synthesizing water-soluble glucans, GTF-B, were utilized to assess the binding of GTF activity to the cell surface of Streptococcus mutans GS-5. Growth of the cells in either Todd-Hewitt broth or a chemically defined medium did not appear to affect the ability of the cells to bind either enzyme fraction. Heat inactivation of the cells did not significantly reduce the interaction of the enzymes with the cells. Cell surface glucan molecules appear to be involved in GTF binding to the cells because: (i) dextranase or a-1,3-glucanase treatment of the cells markedly reduced enzyme binding; (ii) the inclusion of soluble dextrans in the binding assays reduced both GTF-A and GTF-B binding to the cells; and (iii) pretreatment of the cells or the GTF-B fraction with soluble dextrans before binding significantly reduced enzyme binding to the cells. In addition, enzyme binding appears to require a cell surface protein component because Pronase, but not trypsin, treatment of the cells reduced enzyme binding. Furthermore, the removal of a portion of the cell surface GTF-glucan complex with 3 N NaCl appears to provide additional binding sites for the enzymes. These results are interpreted in terms of the mechanism of the conversion of extracellular GTF to the cell-associated form. A variety of experimental approaches have suggested an important role for Streptococcus mutans in dental caries formation (11). The ability of this oral streptococcus to colonize tooth surfaces appears to be in part dependent on its ability to synthesize adherent glucan molecules from sucrose (9). The glucosyltransferases (GTFs) (EC 2.4.1.5) catalyzing this reaction exist in both a cell-associated and an extracellular form (10). The results from both in vitro (24) and in vivo (29) experiments suggest that the cell-associated form of the activity is primarily responsible for the colonization of hard surfaces. Recent results also suggest that the extracellular GTF activity serves as a precursor to the cell-
associated form (14). In addition, both enzymatic (2, 4) and immunological (19) evidence imply that more than one species of GTF may be elaborated by a single strain of S. mutans. Furthermore, characterization of the GTF activity of several different strains ofS. mutans indicated that enzyme species synthesizing primarily water-soluble or -insoluble glucans could be isolated from culture fluids (4, 6, 18, 26). The conversion of the extracellular GTF to the cell-associated form requires binding of the enzymes to the cell surface of S. mutans. Pre652
vious investigations concerning the interaction of the enzymes with the cell surface suggested that glucans (25) and protein components (15, 25) were required for enzyme binding. In addition, the interaction of purified GTF preparations with glucans further suggested a role for these polysaccharides in enzyme binding to the cell surface (7). Because previous cell binding experiments involved the utilization of relatively crude GTF preparations (15, 25), it was of interest to investigate the interaction of S. mutans with more purified enzyme preparations. Furthermore, the earlier use of cells grown in complex media (25) yields relatively large amounts of cell-associated glucan molecules (17, 18) and possibly other components of the media which could play a major role in GTF binding to these cells. Therefore, the S. mutans GS-5 partially purified fraction synthesizing primarily insoluble adherent glucans, GTF-A, and the highly purified, soluble-glucan-synthesizing fraction, GTFB (18), were examined with respect to their interaction with the cell surface of strain GS-5 grown in both complex and chemically defined (CD) media. The results are discussed relative to the mechanism of conversion of the extracellular GTF activity to the cell-associated form.
VOL. 20, 1978
GLUCOSYLTRANSFERASE BINDING
653
MATERLALS AND METHODS glucanases or proteolytic enzymes with 0.02 M potasOrganisms. S. mutans GS-5 was maintained essen- sium phosphate buffer, pH 6.0, for 18 h at 37°C. The tially as previously described (16). The cells utilized in treated cells were washed with 5 ml of 0.9% NaCl the binding experiments were routinely grown for 18 (three times) to remove residual exogenous enzymes. h at 370C in either Todd-Hewitt broth (THB) or CD The cells were then resuspended in SCA and utilized media (15), harvested, washed three times with 0.9% for the GTF binding experiments. Materials. [U-'4C]glucose-sucrose (212 mCi/ NaCl, and suspended in one-tenth the original volume mmol) was obtained from New England Nuclear Corp. in 0.9% NaCI-0.02% sodium azide (SCA). Enzymes. The GTF-A and GTF-B fractions were Dextranase and Pronase were purchased from Sigma isolated by a modification (Kuramitsu and Ingersoll, Chemical Corp., whereas the unfractionated CladosArch. Oral Biol., in press) of the previously described porium resinae a-1,3-glucanase preparation was a genprocedure (18) involving the growth of the organism erous gift of E. T. Reese (U.S. Army Natick Labs, in 0.5% Casamino Acids-vitamins-salts (8) medium Natick, Mass.). The water-soluble carboxymethyl-acontaining 1% glucose. The GTF-A fraction represents 1,3-glucan was kindly provided by H. J. Phaff (Unithe high-molecular-weight complex isolated after gel versity of California, Davis), and the nigerose was filtration chromatography (18), whereas the essen- provided by J. H. Nordin (University of Massachutially homogeneous GTF-B fraction was isolated after setts). gel filtration and hydroxylapatite chromatography RESULTS (18). Enzyme activity was measured utilizing the standard ['4C]glucose-sucrose assay in the presence of Binding of GTF activities to viable and primer dextran T10, and enzyme units are defined as heat-inactivated cells. Several previous invespreviously described (18). tigations (15, 24) have indicated that heat-inacThe GTF-B-dextran T150 complex was isolated after gel filtration chromatography on Bio-Gel A-0.5 tivated cells of S. mutans bind crude GTF prepcolumns (2.5 by 90 cm). Enzyme (23.6 U, 6.48 U/mg) arations as well as viable cells. To determine and dextran T150 (1.0 mg) were incubated for 30 min whether the purified enzyme fractions responsiat 370C and layered onto the top of the resin previ- ble for water-soluble and -insoluble glucan synously equilibrated with 0.01 M potassium phosphate thesis behave in a similar manner, GTF-A and buffer (pH 6.0)-0.02% sodium azide. The enzyme com- GTF-B binding to viable and heat-inactivated plex was eluted from the column with the same buffer cells was measured. Heat treatment appeared to at 220C, concentrated with an Amicon PM-10 ultra- increase the number of GTF-A binding sites on filter, and represents GTF activity eluting identically the cell surface while decreasing somewhat the with dextran T150. of the sites for the enzyme complex (Fig. Dextranase (EC 3.2.1.11) activity was measured by affinity incubating enzyme preparations, 10 mM sodium cit- 1A). However, this conclusion is dependent on rate-potassium phosphate buffer (pH 5.0), dextran T80 (100 ug), and SCA (total volume, 1.0 ml) at 37°C for 1 to 18 h depending on the enzyme activity. Appropriate portions were assayed for reducing sugar formation by the Somogyi-Nelson (1) procedure. a-1,3-Glucanase (EC 3.2.1.59) activity was measured in the same manner, except that nigerose tetrasaccharide (100 gLg) or soluble carboxymethyl-a-1,3-glucan (100,ug) was utiL. lized as the substrate. 0 Proteolytic activity was determined utilizing azoal4 2 bumin as the substrate (28). GTF-A(mU) E Enzyme binding. Cells, either viable or heat in1.0 activated, were incubated with the appropriate enB zyme fractions, 0.05 M potassium phosphate buffer a (pH 6.0), and SCA (total volume, 2.0 ml) for 30 min at 37°C followed by further incubation at 4°C for 18 h. 0.5 After centrifugation at 12,000 x g for 10 min, the cells were washed with 5.0 ml of 0.09% NaCl and resuspended in 2.0 ml of SCA. Appropriate portions were then utilized to determine enzyme binding. Control s |S samples without exogenous enzyme additions were 6 4 2 treated identically, and cell-associated binding activiGTFB(mU) ties were corrected accordingly. The activity bound is FIG. 1. Binding of GTF-A and GTF-B to viable expressed as the change of GTF activity per 108 colony-forming units (CFU) compared with the control and heat-inactivated THB-grown cells. Cells (1.9 x samples. I09 CFU) were incubated with (A) GTF-A (0.57 U/mg) and (B) GTF-B (6.48 U/mg) and treated as described Cell treatments. Heat-inactivated cells were prepared by heating viable cell suspensions at 100°C for in the text. Symbols: 0, viable cells; 0, heat-inacti30 min. Cells were treated with the indicated levels of vated cells. LL
654
KURAMITSU AND INGERSOLL
the assumption that the interaction of added exogenous GTF activity with the cell surface does not inhibit endogenous or the added GTF activities. This appears to be the case because there is no decrease in total GTF activity (supernatant plus cell associated) when exogenous GTF-A is added to viable cells. This latter observation also suggests that endogenous cell-associated dextranase (12, 27) or proteolytic activity does not interfere with GTF activity. Therefore, the heat-inactivated cells possess more binding sites for the GTF-A fraction than do viable cells. The primer-dependent GTF-B fraction (18) appeared to bind to viable cells to a greater extent than to the heat-inactivated cells (Fig. 1B) in regard to both the total number of GTFB binding sites and the affinity of each site for the enzyme. However, at higher enzyme concentrations, both cell populations bound nearly equivalent levels of the enzyme. These results suggest that the number of cell surface sites for binding GTF-A and GTF-B fractions are not significantly decreased after heat inactivation. Furthermore, both GTF fractions appear to bind to the same sites on the cell surface because pretreatment of cells with GTF-A followed by heat inactivation of the cell-enzyme complex markedly reduced the subsequent binding of GTF-B to the cells (data not shown). The utilization of cells grown in CD media resulted in GTF-B binding similar to that observed with THB-grown cells (Table 1). Furthermore, heatinactivated CD-grown cells bound slightly higher levels of GTF-B activity when compared with viable cells. Similar results were also observed for GTF-A binding to CD-grown cells (data not shown). The utilization of heat-inactivated cells in the binding experiments has the advantages of minimizing any contributions ofthe endogenous cellassociated activities of viable cells to the meaTABLE 1. Interaction of GTF-B with CD- and THBgrown cells A GTF-B (mU/10 CFU)
GTF-B added
(mU) 16.2 32.3 48.5
Viable THB"
Viable
CDh
Heat-iac-
tivated CDb
0.55 1.13 1.23 3.10 3.63 4.85 3.61 4.27 5.60 Cells (2.3 x 10' CFU, 0.05 mU/108 CFU) grown in THB medium were incubated with GTF-B (5.9 U/mg) and treated as described in the text. bViable or heat-inactivated cells (9.4 X 109 CFU, 0.30 mU/108 CFU) grown in CD-glucose medium were incubated with GTF-B (5.9 U/mg) and treated as described in the text. a
INFECT. IMMUN.
surement of bound enzyme activity and also eliminates the effects of any cell-associated hydrolytic activities on either the binding process or resultant enzyme activity. Effects of proteolytic treatment on the binding of GTF activities. That the binding of unfractionated GTF preparations to S. mutans may involve a cell surface protein component was initially suggested by the reduced enzyme binding capacity of cells of S. mutans HS6 (serotype a) pretreated with proteolytic enzymes (25). In contrast, when either viable or heat-inactivated cells of strain GS-5 were pretreated with trypsin, no significant reduction in GTF binding capacity was observed. Trypsinization of the cells does lead to some inactivation of cell-associated GTF activity because viable cells treated in this manner lose approximately 17% of the cell-associated activity. This observation also suggests that the association of GTF with the cell surface protects most of the enzyme from trypsinization because the extracellular GTF loses approximately 50% activity with a similar treatment. However, when heat-inactivated cells grown in THB were preincubated with a nonspecific proteolytic enzyme, Pronase, a marked reduction in cell binding of both GTFA and GTF-B was observed (Fig. 2). The reduced binding capacity of Pronase-treated cells is not due to residual proteolytic activity associated wth the cells because reheating (30 min at 100°C) Pronase-treated cells yields identical GTF binding capacity and mixing viable cells with Pronase-treated cells yields additive cellassociated activities. In addition, the Pronase preparation does not contain detectable a-1,3or a-1,6-glucanase activities when assayed for 18 h. Furthermore, the substitution of CD-grown cells for THB-grown cells yielded similar results. As demonstrated previously for strain HS-6 cells (25), pretreatment of GS-5 cells with pepsin also reduced the subsequent binding of the GTF fractions (data not shown). These results suggest that a protein component, stable to the action of trypsin, plays an important role in GTF binding to the cell surface of strain GS-5. Effects of glucanase treatment on the binding of GTF activities. Previous results (25) have also demonstrated that pretreatment of heat-inactivated cells of strain HS-6 with commercial dextranase markedly reduced the subsequent binding of relatively crude GTF preparations. To assess the effects of altering cell surface glucans on the binding of both GTF-A and GTF-B, heat-inactivated cells of strain GS5 were pretreated with either dextranase or a1,3-glucanase. The ability of both GTF fractions to bind to the treated cells was then compared
VOL. 20, 1978
655
GLUCOSYLTRANSFERASE BINDING
with control cell suspensions (Table 2). Treatment of heat-inactivated cells grown in THB with either glucanase preparation markedly reduced the ability of the cells to bind either GTFA or GTF-B. Treatment of the cells with a combination of both glucanases was not any more effective in reducing enzyme binding than treatment with each glucanase alone. The reduced GTF binding of cells pretreated with either glucanase is not the result of hydrolysis of THB medium components absorbed to the cell surface, because similar results were obtained with cells grown in the CD medium (data not shown). These latter observations also indicate that cells grown in CD medium possess some cell surface glucans associated with GTF binding. Although neither glucanase preparation utilized is homogeneous, the commercial dextranase is devoid of detectable a-1,3-glucanase activ-
ity, whereas the C. resinae a-1,3-glucanase does not hydrolyze dextrans. Only trace amounts of proteolytic activity could be detected in either glucanase preparation. However, the reduced binding produced by the glucanases could not result primarily from contaminant proteolytic activity because the glucanase treatments (Ta-
ble 2) reduced GTF binding more than proteolytic treatment (Fig. 2). These results, therefore, suggest that branched glucans containing both a-1,3- and a-1,6-glucose linkages may be involved in the binding of both purified GTF preparations to the cell surface of S. mutans GS-5 under these conditions. Effects of dextrans on GTF binding. Because the pretreatment of viable or heat-inactivated cells of strain GS-5 with glucanases markedly reduced the subsequent binding of both GTF preparations, it appeared that the GTF fractions were interacting with cell surface glucan molecules. If this were the case, the addition of glucan molecules to the binding mixtures 0.3 might be expected to inhibit enzyme interaction with the cell surface. The addition of low-molecular-weight dextran T10 to the binding mixtures U. U 0.2markedly decreased cell binding of both the I0 GTF-A and GTF-B fractions (Fig. 3). The binding of GTF-A to heat-inactivated THB-grown cells was completely abolished by the addition U.~ 0.1 of as little as 50 ,ug of dextran T10, whereas GTF-B binding was inhibited approximately LOW. 60% by high concentrations of the low-molecular-weight dextran. Similar effects were also observed when high-molecular-weight dextran 6 2 4 0 T2000 was included in the binding assays. Unlike GTF(mU) dextrans, the addition of soluble carboxymethylFIG. 2. Effects of Pronase on GTF-A and GTF-B a-1,3-glucan to the binding mixtures did not binding to heat-inactivated cells. Cells (4 x 10"' CFU) inhibit enzyme binding. Soluble dextrans also grown in THB medium were incubated with and markedly inhibited the binding of the GTF fracwithout Pronase (1.3 U) as described in the text. GTFA (0.57 U/mg) and GTF-B (5.9 U/mg) binding was tions to CD-grown cells (data not shown). The decreased enzyme binding observed in measured as described in the text. Symbols: *, GTFA binding to control cells; 0, GTF-A binding to the presence of dextrans could be explained in E
Pronase-treated cells; A, GTF-B binding to control cells; A, GTF-B binding to Pronase-treated cells.
terms of either dextran interaction with the cells and/or with the GTF preparations. To assess
TABLE 2. Effects of glucanases on the binding of GTF-A and GTF-B to heat-inactivated cells' GTF-B
GTF-A Treatment
Added
(mU/10" CFU)
None Dextranase (7.6 U)
0.17 0.17
Bound
mU/1b CU
0.22 0.015
Bound
Added
(mU/10 %CFU)
129
1.7
mU/1O" CFU
0.48
28
3.5 8.8 1.7 0.06 2.9 0.17 0.017 10.0 1.7 0.05 a-1,3-Glucanase (0.68 mU) a Heat-inactivated THB-grown cells (1.9 X 10" CFU) were incubated with the indicated levels of glucanases and SCA (total volume, 2.0 ml) for 18 h at 37"C. After washing with 5 ml of 0.9% NaCl, the cells were resuspended in 2.0 ml of SCA and portions (1.9 x 109 CFU) were mixed with the indicated levels of GTF-A (0.57 U/mg) or GTF-B (5.9 U/mg). Incubations for binding and the enzyme assays were carried out as described earlier.
INFECT. IMMUN.
KURAMITSU AND INGERSOLL
656
the relative contribution of both types of interactions on enzyme binding, heat-inactivated cells of strain GS-5 were initially treated with dextran T10, the cells were washed thoroughly, and GTF-B binding was measured (Fig. 4). Prior incubation of the cells with dextran T10 markedly decreased the subsequent binding of GTFB to the cell surface. Inhibition of enzyme binding was also observed with dextran T2000. A similar reduction in GTF-A binding was observed when heat-inactivated cells were pre100 75 a
c 3 0
50
m LA.
HD
25
0
0.2
0.4
Dextron TIO (mg) FIG. 3. Effects of dextran T10 on GTF-A and GTFB binding to heat-inactivated cells. Cells (2.3 x 109 CFU) grown in THB medium were incubated in the presence of GTF-A (3.5 m U, 0.57 U/mg) or GTF-B (27 mU, 6.48 U/mg) with the indicated amounts of dextran T10 as described in the text. Binding of enzymes is expressed relative to the values obtained in the absence of dextran TlO for GTF-A (0.12 mU/11 CFU) and GTF-B (0.55 mU/J10 CFU). Symbols: 0, GTF-B binding; 0, GTF-A binding.
TABLE 3. Elution of cell-associated GTF activity and glucans
0.4
U.
Elution Elution
0.3
0
E
treated with dextrans (data not shown). Therefore, the interaction of dextran molecules with the cell surface of S. mutans markedly inhibited the subsequent binding of exogenous GTF preparations. Pretreatment of cells with dextrans could affect subsequent binding of GTFs by eluting off cell-associated GTF-glucan complexes as demonstrated previously (21). The removal of cellassociated glucans by the addition of dextrans could be demonstrated by measuring the loss of endogenously synthesized ['4C]glucan from the cell surface in the presence of dextran T10 (Table 3). In addition, a significant decrease in the cell-associated GTF activity was also observed for viable cells after dextran elution. These results suggest that soluble dextrans are capable of removing GTF-glucan complexes from the cell surface of strain GS-5. The interaction of soluble dextrans with GTFB to form dextran-GTF-B complexes also inhibited the subsequent binding of the enzyme to heat-inactivated cells (Fig. 5). The enzyme-dextran complex, isolated after gel filtration chromatography, bound very poorly to the cells relative to the uncomplexed homogeneous GTF-B preparation. Similar comparisons were not carried out with the GTF-A fraction because this high-molecular-weight fraction contains glucanlike molecules and the enzyme activity of this fraction is not stimulated by the addition of exogenous glucans (18). These results suggest that dextran interaction with both the cells and GTF-B fraction can play a role in reducing enzyme binding to the cells.
0.2
La: I-
0.1I 0
GTF-B (mU)
FIG. 4. Effects of dextran TIOpretreatment of cells on GTF-B binding. Heat-inactivated cells (2.7 x 1010 CFU) grown in THB medium were incubated with and without dextran T1O (5 mg) for 15 min at 37C, washed three times with 0.9% NaCl, and resuspended in SCA. GTF-B (6.48 U/mg) was added to the cell suspensions, and binding was measured as described earlier. Symbols: *, GTF-B binding to control cells; 0, GTF-B binding to dextran-treated cells.
Cell-associated GTF
Cell-associated glucan
CFU)a 0.15 (100) 0.085 (56) 0.088 (59)
sion)
(mU/108
(cpm/suspen-
475 (100)b Saline 250 (53)c 3 N NaCl 250 (53)b Dextran T10 (1.0 mg) a Cells (2 x 1010 CFU) grown in THB medium were incubated with saline, dextran T10, or 3 N NaCl for 30 min at 37°C and washed three times with 5.0 ml of saline. Appropriate portions were then assayed for total GTF activity. b Cells (2.3 x 109 CFU) were incubated with [14C]sucrose in the standard assay mixture for 60 min. The cells containing cell-associated ['4C]glucan were then incubated with saline or dextran T10 for 30 min at 37°C and filtered and counted as described previously (18). e Cells incubated with ['4C]sucrose were filtered and washed directly with 3 N NaCl (three times) before counting.
GLUCOSYLTRANSFERASE BINDING
VOL. 20, 1978
Effects of NaCl extraction on GTF binding. Previous results have indicated that cellassociated GTF activity can be extracted from the cell surface of strain GS-5 utilizing hypertonic salt solutions (17). This procedure also removes glucans from the cell surface (Table 3). If cell surface glucan-GTF complexes are involved in enzyme binding, the removal of the complexes might be expected to affect the subsequent binding of GTF fractions. Heat-inactivated cells extracted with 3 N NaCl contained a greater number of binding sites with lowered affinity for GTF-B than did unextracted cells (Fig. 6). These results suggest that the removal 0.4L.
0.3
01 E 0.2
-0
E/ UA. a0o1~~~ 0/
-
2 3 4 GTF - B (mU) FIG. 5. Effects of dextran T150 pretreatment of GTF-B on enzyme binding to heat-inactivated cells. The dextran T150-GTF-B complex was prepared and isolated as described in the text. The complex as well as GTF-B alone was added to heat-inactivated cells (6.8 x I& CFU), and binding was determined as described in the text. Symbols: 0, GTF-B binding; 0, dextran T150-GTF-B binding.
0
1
0.3
0
0.2
0
LL
0.1
I GTF- B (mU) FIG. 6. Effects of NaCl extraction on GTF-B binding. Heat-inactivated cells (6.2 x 1iO CFU) grown in THB medium were washed with saline or 3 N NaCl 0
(three times) as describedpreviously (17). GTF-B (6.48 U/mg) binding was determined as previously described. Symbols *, GTF-B binding by saline-washed cells; 0, GTF-B binding by NaCl-extracted cells.
657
of a portion, but not all (17; Table 3), of the GTF-glucan cell surface complexes produces additional "sites" for enzyme binding. However, it is also possible that other molecules affecting GTF binding were also removed or altered by NaCl extraction. Similar results were also observed when the interaction of GTF-A with NaCl-extracted cells was measured (data not shown). Removal of a portion of the glucan-GTF complex from CD-grown cells with 3 N NaCl also yielded comparable results. DISCUSSION Both enzymatic (17, 21, 23) and immunological (13, 19) characterizations as well as recent kinetic data (14) suggest that the cell-associated GTF activity of S. mutans may be derived from the extracellular enzymes. However, the cell surface structures involved in the binding of the enzymes to the cells have not been identified. Mukasa and Slade (25), utilizing relatively crude GTF preparations and THB-grown cells of strain HS-6, have proposed that GTF binding occurred in the vicinity of the serotype a antigen and glucan molecules associated with a heatstable protein on the cell surface. In contrast, Germaine and Schachtele (7) have suggested that GTF molecules become associated with the cell surface by interacting directly with other GTF molecules bound to cell surface glucans. The glucan molecules are in turn associated with a glucan-binding protein on the cell surface. Recently, several potentially different glucan binding proteins have been described in extracts of S. mutans (20; M. M. McCabe and R. M. Hamelik, Proceedings of the International Symposium on The Secretory Immune System and Caries Immunity, in press). Furthermore, the results of agglutination and flocculation studies could be interpreted as suggesting that binding sites for different types of glucans (linear or branched) may exist on the surface of S. mutans (22; C. Wu-Yuan, S. Tai, and H. D. Slade, in press). However, it is not yet clear whether any of the glucan-binding proteins described are involved in GTF binding to the cells. The results of the present investigation indicate that heat treatment of GS-5 cells does not significantly decrease the number of binding sites for the GTF-A or GTF-B fraction. However, the affinity of the sites for each GTF fraction is somewhat reduced (Fig. 1; Table 1). Therefore, the binding site(s) for the GTF enzymes are composed of a heat-stable component in agreement with previous suggestions utilizing crude GTF preparations (15, 24). Furthermore, both GTF fractions appear to bind to the same cell surface components of S. mutans GS-5.
INFECT. IMMUN. 658 KURAMITSU AND INGERSOLL However, it is not clear why heat-inactivated nents which either retard GTF binding or intercells appear to bind more GTF-A molecules than fere with enzyme activity. do viable cells even though both cell suspensions In contrast to cells grown in complex media contain a similar number of GTF-B binding (25) recent results (15) have suggested that cellsites. Perhaps heat treatment might expose ad- associated glucans do not appear to be required ditional low-affinity binding sites for GTF-A but for the binding of GTF activity to cells grown in not for GTF-B. a CD medium devoid of possible sucrose or Although heat stable, the putative GTF bind- polysaccharide contaminants. However, the ing sites appear to be proteinaceous or protein- present results utilizing CD-grown cells and associated because Pronase treatment of the more purified enzyme preparations do suggest a cells decreased GTF-A and GTF-B binding to role for cell-associated glucans in GTF binding. cells grown in either complex (Fig. 2) or chemi- The difference between the present and earlier cally defined media (data not shown). The heat (15) investigations may be due to the utilization stability and resistance to trypsin of the binding of more purified enzyme preparations in the sites may result from the association of the present study containing less glucan contamiprotein component with polysaccharide or other nants. Thus, the glucans present in crude enmacromolecules of the cell surface. Previous re- zyme preparations (15) may have obviated the sults (25) demonstrating that trypsin or pepsin requirement for cell surface glucans. Recently, treatment of S. mutans HS-6 decreased GTF Montville et al. (23) have also demonstrated that binding indicate that the cell surface architec- strain GS-5 grown in a different chemically deture may vary between different strains of S. fined medium also possesses relatively high levmutans. els of cell-associated GTF activity. It is clear from the present investigation (TaBecause previous results have suggested that ble 2) and previous observations (25) that cell- receptors for branched glucans may be relatively associated glucans can be involved in the inter- heat stable (15, 22), it is tempting to speculate action of GTF activity with the surface of S. that these molecules may play a role in convertmutans. The results of adding water-soluble dex- ing extracellular GTF activity to the cell-assotrans to the binding mixtures (25; Fig. 3 through ciated form. In addition, glucans produced by 5) are also compatible with this view. In addi- cell-associated GTF activity would be heat station, recent studies (15) indicate that water-sol- ble. Separate heat-labile receptors for linear gluuble dextran molecules can bind directly to the cans may also exist on the cell surface of S. cell surface of strain GS-5. This binding might mutans (24) and could also be involved in GTF also interfere with subsequent GTF interaction binding (7). Thus, it is possible that extracellular with the cell surface mediated by glucan mole- GTF activity (containing glucans) can become cules present in the exogenous enzyme com- cell associated after interaction with multiple plexes (3, 5, 18, 26). The cell surface glucan glucan-binding components of the cell surface as involved in GTF binding appears to be the well as to cell-associated glucan molecules. Rebranched glucan product synthesized by the cell- cent results from this laboratory (15) have demassociated GTF activity because treatment of onstrated that CD-grown cells of strain GS-5 the cells with either a-1,6- or a-1,3-glucanase which display reduced ability to interact with significantly reduced GTF binding (Table 2). glucans bind exogenous unfractionated GTF Furthermore, naturally occurring a-1,3-glucans preparations as well as the parental cells. Therehave not been positively identified as a normal fore, this nonaggregating variant with normal cell surface component in streptococci. However, GTF binding may be altered in the glucan rebecause neither glucanase preparation utilized ceptors required for agglutination but not those in the present or earlier (25) investigations is capable of interacting with glucans present in homogeneous, these conclusions are not un- the crude GT preparations. Once enzyme-glucan equivocal. complexes become attached to the cell surface It has recently been suggested (15) that the via glucan receptors or cell-associated glucans, number of S. mutans cell surface sites capable subsequent enzyme molecules can become cell of binding GTF activity appears to be limited. bound via protein-protein or protein-glucan inSaturation of these sites prevents further bind- teractions (7). Thus, the binding of GTF activity ing of enzyme activity. The results of the present to the cell surface of S. mutans may involve investigation suggest that extraction of a portion multiple binding sites with different types of of the cell surface glucan-GTF complex with protein-polysaccharide interactions. Final proof hypertonic salt solutions may provide additional of this proposal awaits the direct isolation and sites for enzyme binding. In addition, such ex- purification of the cell surface receptor moletraction might also remove cell surface compo- cules involved in GTF binding.
659
VOL. 20, 1978
GLUCOSYLTRANSFERASE BINDING
ACKNOWLEDGMENT This investigation was supported by Public Health Service grant DE-03258 from the National Institute of Dental Research.
of a variant of Streptococcus mutans altered in its ability to interact wich glucans. Infect. Immun. 16:575-586. Kuramitsu, H. K. 1973. Characterization of invertase activity from cariogenic Streptococcus mutans. J. Bacteriol. 115:1003-1010. Kuramitsu, H. K. 1974. Characterization of cell-associated dextransucrase from glucose-grown cells of Streptococcus mutans. Infect. Immun. 10:227-235. Kuramitsu, H. K. 1975. Characterization of extracellular glucosyltransferase activity of Streptococcus mutans. Infect. Immun. 12:738-749. Kuramitsu, H. K., and L. Ingersoll. 1976. Differential inhibition of Streptococcus mutans in vitro adherence by anti-glucosyltransferase antibodies. Infect. Immun. 13:1775-1777. McCabe, M. M., R. M. Hamelik, and E. E. Smith. 1977. Purification of dextran-binding protein from cariogenic Streptococcus mutans. Biochem. Biophys. Res. Commun. 78:273-278. McCabe, M. M., and E. E. Smith. 1973. Origin of the cell-associated dextransucrase of Streptococcus mutans. Infect. Immun. 7:829-838. McCabe, M. M., and E. E. Smith. 1975. Relationship between cell-bound dextransucrase and the agglutination of Streptococcus mutans. Infect. Immun.
LITERATURE CITED 1. Ashweli, G. 1957. Colorimetric analysis of sugars. Methods Enzymol. 3:85-86. 2. Ceska, M., K. Granath, B. Norman, and B. Guggenheim. 1972. Structural and enzymatic studies on glucans synthesized with glucosyltransferases of some strains of oral streptococci. Acta Chem. Scand. 26:2223-2230. 3. Ciardi, J. E., A. J. Beaman, and C. L. Wittenberger. 1977. Purification, resolution, and interaction of glucosyltransferases of Streptococcus mutans 6715. Infect. Immun. 18:237-246. 4. Fukui, K., Y. Fukui, and T. Moriyama. 1974. Purification and properties of dextransucarase and invertase from Streptococcus mutans. J. Bacteriol. 118:796-804. 5. Germaine, G. R., A. M. Chludzinski, and C. F. Schachtele. 1974. Streptococcus mutans dextransucrase: requirement for primer dextran. J. Bacteriol. 120:287-294. 6. Germaine, G. R., S. K. Harlander, W.-L. S. Leung, and C. F. Schachtele. 1977. Streptococcus mutans dextransucransucrase: functioning of primer dextran and endogenous dextranase in water-soluble and waterinsoluble glucan synthesis. Infect. Immun. 16:637-648. 7. Germaine, G. R., and C. F. Schachtele. 1976. Streptococcus mutans dextransucrase: mode of interaction with high-molecular-weight dextran and role in cellular aggregation. Infect. Immun. 13:365-372. 8. Gibbons, R. J. 1972. Presence of an invertase-like enzyme and a sucrose permeation system in strains of Streptococcus mutans. Caries Res. 6:122-131. 9. Gibbons, R. J., and S. Banghart. 1967. Synthesis of extracellular dextran by cariogenic bacteria and its presence in human dental plaque. Arch. Oral Biol. 12:11-24. 10. Gibbons, R. J., and M. Nygaard. 1968. Synthesis of insoluble dextran and its significance in the formation of gelatinous deposits by plaque-forming streptococci. Arch. Oral Biol. 13:1249-1262. 11. Gibbons, R. J., and J. van Houte. 1975. Dental caries. Annu. Rev. Med. 26:121-136. 12. Guggenheim, B., and J. J. Burckhardt. 1974. Isolation and properties of a dextranase from Streptococcus mutans OMZ 176. Helv. Odontol. Acta 18:101-113. 13. Guggenheim, B., and E. Newbrun. 1969. Extracellular glucosyltransferase activity of an HS strain of Streptococcus mutans. Helv. Odontol. Acta 13:84-97. 14. Janda, W. M., and H. K. Kuramitsu. 1976. Regulation of extracellular glucosyltransferase production and the relationship between extracellular and cell-associated activities in Streptococcus mutans. Infect. Immun. 14:191-202. 15. Janda, W. M., and H. K. Kuramitsu. 1977. Properties
16. 17. 18. 19.
20.
21. 22.
12:512-520. 23. Montville, T. J., C. L Cooney, and A. J. Sinskey. 1977. Distribution of dextransucrase in Streptococcus mutans and observations on the effect of soluble dextran on dextransucrase activities. Infect. Immun. 18:629-635. 24. Mukasa, H., and H. D. Slade. 1973. 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. 8:555-562. 25. Mukasa, H., and H. D. Slade. 1973. Mechanism of adherence of Streptococcus mutans to smooth surfaces. II. Nature of the binding site and the adsorption of dextran-levan synthetase enzymes on the cell-wall surface of the streptococcus. Infect. Immun. 9:419-429. 26. Mukasa, H., and H. D. Slade. 1974. Mechanism of adherence of Streptococcus mutans to smooth surfaces. III. Purification and properties of the enzyme complex responsible for adherence. Infect. Immun. 10:1135-1145. 27. Staat, R. H., and C. F. Schachtele. 1974. Evaluation of dextranase production by the cariogenic bacterium Streptococcus mutans. Infect. Immun. 9:467-469. 28. Tomarelli, R. M., J. Charney, and M. L. Harding. 1949. The use of azoalbumin as a substrate in the colorimetric determination of peptic and tryptic activity. J. Lab. Clin. Med. 34:428433. 29. van Houte, J., and V. N. Upeslacis. 1976. Studies on the mechanism of sucrose-associated colonization of Streptococcus mutans on teeth of conventional rats. J. Dent. Res. 55:216-222.
INFECTION AND IMMUNITY, June 1978, p. 652-659 0019-9567/78/0020-0652$02.00/0 Copyright i 1978 American Society for Microbiology
Printed in U.S.A.
Interaction of Glucosyltransferases with the Cell Surface of Streptococcus mutans HOWARD K. KURAMITSU AND L. INGERSOLL Department of Microbiology-Immunology, Northwestern University Medical-Dental Schools, Chicago, Illinois 60611
Received for publication 20 January 1978
The partially purified glucosyltransferase (GTF) fraction synthesizing primarily water-insoluble glucans, GTF-A, and the homogeneous fraction synthesizing water-soluble glucans, GTF-B, were utilized to assess the binding of GTF activity to the cell surface of Streptococcus mutans GS-5. Growth of the cells in either Todd-Hewitt broth or a chemically defined medium did not appear to affect the ability of the cells to bind either enzyme fraction. Heat inactivation of the cells did not significantly reduce the interaction of the enzymes with the cells. Cell surface glucan molecules appear to be involved in GTF binding to the cells because: (i) dextranase or a-1,3-glucanase treatment of the cells markedly reduced enzyme binding; (ii) the inclusion of soluble dextrans in the binding assays reduced both GTF-A and GTF-B binding to the cells; and (iii) pretreatment of the cells or the GTF-B fraction with soluble dextrans before binding significantly reduced enzyme binding to the cells. In addition, enzyme binding appears to require a cell surface protein component because Pronase, but not trypsin, treatment of the cells reduced enzyme binding. Furthermore, the removal of a portion of the cell surface GTF-glucan complex with 3 N NaCl appears to provide additional binding sites for the enzymes. These results are interpreted in terms of the mechanism of the conversion of extracellular GTF to the cell-associated form. A variety of experimental approaches have suggested an important role for Streptococcus mutans in dental caries formation (11). The ability of this oral streptococcus to colonize tooth surfaces appears to be in part dependent on its ability to synthesize adherent glucan molecules from sucrose (9). The glucosyltransferases (GTFs) (EC 2.4.1.5) catalyzing this reaction exist in both a cell-associated and an extracellular form (10). The results from both in vitro (24) and in vivo (29) experiments suggest that the cell-associated form of the activity is primarily responsible for the colonization of hard surfaces. Recent results also suggest that the extracellular GTF activity serves as a precursor to the cell-
associated form (14). In addition, both enzymatic (2, 4) and immunological (19) evidence imply that more than one species of GTF may be elaborated by a single strain of S. mutans. Furthermore, characterization of the GTF activity of several different strains ofS. mutans indicated that enzyme species synthesizing primarily water-soluble or -insoluble glucans could be isolated from culture fluids (4, 6, 18, 26). The conversion of the extracellular GTF to the cell-associated form requires binding of the enzymes to the cell surface of S. mutans. Pre652
vious investigations concerning the interaction of the enzymes with the cell surface suggested that glucans (25) and protein components (15, 25) were required for enzyme binding. In addition, the interaction of purified GTF preparations with glucans further suggested a role for these polysaccharides in enzyme binding to the cell surface (7). Because previous cell binding experiments involved the utilization of relatively crude GTF preparations (15, 25), it was of interest to investigate the interaction of S. mutans with more purified enzyme preparations. Furthermore, the earlier use of cells grown in complex media (25) yields relatively large amounts of cell-associated glucan molecules (17, 18) and possibly other components of the media which could play a major role in GTF binding to these cells. Therefore, the S. mutans GS-5 partially purified fraction synthesizing primarily insoluble adherent glucans, GTF-A, and the highly purified, soluble-glucan-synthesizing fraction, GTFB (18), were examined with respect to their interaction with the cell surface of strain GS-5 grown in both complex and chemically defined (CD) media. The results are discussed relative to the mechanism of conversion of the extracellular GTF activity to the cell-associated form.
VOL. 20, 1978
GLUCOSYLTRANSFERASE BINDING
653
MATERLALS AND METHODS glucanases or proteolytic enzymes with 0.02 M potasOrganisms. S. mutans GS-5 was maintained essen- sium phosphate buffer, pH 6.0, for 18 h at 37°C. The tially as previously described (16). The cells utilized in treated cells were washed with 5 ml of 0.9% NaCl the binding experiments were routinely grown for 18 (three times) to remove residual exogenous enzymes. h at 370C in either Todd-Hewitt broth (THB) or CD The cells were then resuspended in SCA and utilized media (15), harvested, washed three times with 0.9% for the GTF binding experiments. Materials. [U-'4C]glucose-sucrose (212 mCi/ NaCl, and suspended in one-tenth the original volume mmol) was obtained from New England Nuclear Corp. in 0.9% NaCI-0.02% sodium azide (SCA). Enzymes. The GTF-A and GTF-B fractions were Dextranase and Pronase were purchased from Sigma isolated by a modification (Kuramitsu and Ingersoll, Chemical Corp., whereas the unfractionated CladosArch. Oral Biol., in press) of the previously described porium resinae a-1,3-glucanase preparation was a genprocedure (18) involving the growth of the organism erous gift of E. T. Reese (U.S. Army Natick Labs, in 0.5% Casamino Acids-vitamins-salts (8) medium Natick, Mass.). The water-soluble carboxymethyl-acontaining 1% glucose. The GTF-A fraction represents 1,3-glucan was kindly provided by H. J. Phaff (Unithe high-molecular-weight complex isolated after gel versity of California, Davis), and the nigerose was filtration chromatography (18), whereas the essen- provided by J. H. Nordin (University of Massachutially homogeneous GTF-B fraction was isolated after setts). gel filtration and hydroxylapatite chromatography RESULTS (18). Enzyme activity was measured utilizing the standard ['4C]glucose-sucrose assay in the presence of Binding of GTF activities to viable and primer dextran T10, and enzyme units are defined as heat-inactivated cells. Several previous invespreviously described (18). tigations (15, 24) have indicated that heat-inacThe GTF-B-dextran T150 complex was isolated after gel filtration chromatography on Bio-Gel A-0.5 tivated cells of S. mutans bind crude GTF prepcolumns (2.5 by 90 cm). Enzyme (23.6 U, 6.48 U/mg) arations as well as viable cells. To determine and dextran T150 (1.0 mg) were incubated for 30 min whether the purified enzyme fractions responsiat 370C and layered onto the top of the resin previ- ble for water-soluble and -insoluble glucan synously equilibrated with 0.01 M potassium phosphate thesis behave in a similar manner, GTF-A and buffer (pH 6.0)-0.02% sodium azide. The enzyme com- GTF-B binding to viable and heat-inactivated plex was eluted from the column with the same buffer cells was measured. Heat treatment appeared to at 220C, concentrated with an Amicon PM-10 ultra- increase the number of GTF-A binding sites on filter, and represents GTF activity eluting identically the cell surface while decreasing somewhat the with dextran T150. of the sites for the enzyme complex (Fig. Dextranase (EC 3.2.1.11) activity was measured by affinity incubating enzyme preparations, 10 mM sodium cit- 1A). However, this conclusion is dependent on rate-potassium phosphate buffer (pH 5.0), dextran T80 (100 ug), and SCA (total volume, 1.0 ml) at 37°C for 1 to 18 h depending on the enzyme activity. Appropriate portions were assayed for reducing sugar formation by the Somogyi-Nelson (1) procedure. a-1,3-Glucanase (EC 3.2.1.59) activity was measured in the same manner, except that nigerose tetrasaccharide (100 gLg) or soluble carboxymethyl-a-1,3-glucan (100,ug) was utiL. lized as the substrate. 0 Proteolytic activity was determined utilizing azoal4 2 bumin as the substrate (28). GTF-A(mU) E Enzyme binding. Cells, either viable or heat in1.0 activated, were incubated with the appropriate enB zyme fractions, 0.05 M potassium phosphate buffer a (pH 6.0), and SCA (total volume, 2.0 ml) for 30 min at 37°C followed by further incubation at 4°C for 18 h. 0.5 After centrifugation at 12,000 x g for 10 min, the cells were washed with 5.0 ml of 0.09% NaCl and resuspended in 2.0 ml of SCA. Appropriate portions were then utilized to determine enzyme binding. Control s |S samples without exogenous enzyme additions were 6 4 2 treated identically, and cell-associated binding activiGTFB(mU) ties were corrected accordingly. The activity bound is FIG. 1. Binding of GTF-A and GTF-B to viable expressed as the change of GTF activity per 108 colony-forming units (CFU) compared with the control and heat-inactivated THB-grown cells. Cells (1.9 x samples. I09 CFU) were incubated with (A) GTF-A (0.57 U/mg) and (B) GTF-B (6.48 U/mg) and treated as described Cell treatments. Heat-inactivated cells were prepared by heating viable cell suspensions at 100°C for in the text. Symbols: 0, viable cells; 0, heat-inacti30 min. Cells were treated with the indicated levels of vated cells. LL
654
KURAMITSU AND INGERSOLL
the assumption that the interaction of added exogenous GTF activity with the cell surface does not inhibit endogenous or the added GTF activities. This appears to be the case because there is no decrease in total GTF activity (supernatant plus cell associated) when exogenous GTF-A is added to viable cells. This latter observation also suggests that endogenous cell-associated dextranase (12, 27) or proteolytic activity does not interfere with GTF activity. Therefore, the heat-inactivated cells possess more binding sites for the GTF-A fraction than do viable cells. The primer-dependent GTF-B fraction (18) appeared to bind to viable cells to a greater extent than to the heat-inactivated cells (Fig. 1B) in regard to both the total number of GTFB binding sites and the affinity of each site for the enzyme. However, at higher enzyme concentrations, both cell populations bound nearly equivalent levels of the enzyme. These results suggest that the number of cell surface sites for binding GTF-A and GTF-B fractions are not significantly decreased after heat inactivation. Furthermore, both GTF fractions appear to bind to the same sites on the cell surface because pretreatment of cells with GTF-A followed by heat inactivation of the cell-enzyme complex markedly reduced the subsequent binding of GTF-B to the cells (data not shown). The utilization of cells grown in CD media resulted in GTF-B binding similar to that observed with THB-grown cells (Table 1). Furthermore, heatinactivated CD-grown cells bound slightly higher levels of GTF-B activity when compared with viable cells. Similar results were also observed for GTF-A binding to CD-grown cells (data not shown). The utilization of heat-inactivated cells in the binding experiments has the advantages of minimizing any contributions ofthe endogenous cellassociated activities of viable cells to the meaTABLE 1. Interaction of GTF-B with CD- and THBgrown cells A GTF-B (mU/10 CFU)
GTF-B added
(mU) 16.2 32.3 48.5
Viable THB"
Viable
CDh
Heat-iac-
tivated CDb
0.55 1.13 1.23 3.10 3.63 4.85 3.61 4.27 5.60 Cells (2.3 x 10' CFU, 0.05 mU/108 CFU) grown in THB medium were incubated with GTF-B (5.9 U/mg) and treated as described in the text. bViable or heat-inactivated cells (9.4 X 109 CFU, 0.30 mU/108 CFU) grown in CD-glucose medium were incubated with GTF-B (5.9 U/mg) and treated as described in the text. a
INFECT. IMMUN.
surement of bound enzyme activity and also eliminates the effects of any cell-associated hydrolytic activities on either the binding process or resultant enzyme activity. Effects of proteolytic treatment on the binding of GTF activities. That the binding of unfractionated GTF preparations to S. mutans may involve a cell surface protein component was initially suggested by the reduced enzyme binding capacity of cells of S. mutans HS6 (serotype a) pretreated with proteolytic enzymes (25). In contrast, when either viable or heat-inactivated cells of strain GS-5 were pretreated with trypsin, no significant reduction in GTF binding capacity was observed. Trypsinization of the cells does lead to some inactivation of cell-associated GTF activity because viable cells treated in this manner lose approximately 17% of the cell-associated activity. This observation also suggests that the association of GTF with the cell surface protects most of the enzyme from trypsinization because the extracellular GTF loses approximately 50% activity with a similar treatment. However, when heat-inactivated cells grown in THB were preincubated with a nonspecific proteolytic enzyme, Pronase, a marked reduction in cell binding of both GTFA and GTF-B was observed (Fig. 2). The reduced binding capacity of Pronase-treated cells is not due to residual proteolytic activity associated wth the cells because reheating (30 min at 100°C) Pronase-treated cells yields identical GTF binding capacity and mixing viable cells with Pronase-treated cells yields additive cellassociated activities. In addition, the Pronase preparation does not contain detectable a-1,3or a-1,6-glucanase activities when assayed for 18 h. Furthermore, the substitution of CD-grown cells for THB-grown cells yielded similar results. As demonstrated previously for strain HS-6 cells (25), pretreatment of GS-5 cells with pepsin also reduced the subsequent binding of the GTF fractions (data not shown). These results suggest that a protein component, stable to the action of trypsin, plays an important role in GTF binding to the cell surface of strain GS-5. Effects of glucanase treatment on the binding of GTF activities. Previous results (25) have also demonstrated that pretreatment of heat-inactivated cells of strain HS-6 with commercial dextranase markedly reduced the subsequent binding of relatively crude GTF preparations. To assess the effects of altering cell surface glucans on the binding of both GTF-A and GTF-B, heat-inactivated cells of strain GS5 were pretreated with either dextranase or a1,3-glucanase. The ability of both GTF fractions to bind to the treated cells was then compared
VOL. 20, 1978
655
GLUCOSYLTRANSFERASE BINDING
with control cell suspensions (Table 2). Treatment of heat-inactivated cells grown in THB with either glucanase preparation markedly reduced the ability of the cells to bind either GTFA or GTF-B. Treatment of the cells with a combination of both glucanases was not any more effective in reducing enzyme binding than treatment with each glucanase alone. The reduced GTF binding of cells pretreated with either glucanase is not the result of hydrolysis of THB medium components absorbed to the cell surface, because similar results were obtained with cells grown in the CD medium (data not shown). These latter observations also indicate that cells grown in CD medium possess some cell surface glucans associated with GTF binding. Although neither glucanase preparation utilized is homogeneous, the commercial dextranase is devoid of detectable a-1,3-glucanase activ-
ity, whereas the C. resinae a-1,3-glucanase does not hydrolyze dextrans. Only trace amounts of proteolytic activity could be detected in either glucanase preparation. However, the reduced binding produced by the glucanases could not result primarily from contaminant proteolytic activity because the glucanase treatments (Ta-
ble 2) reduced GTF binding more than proteolytic treatment (Fig. 2). These results, therefore, suggest that branched glucans containing both a-1,3- and a-1,6-glucose linkages may be involved in the binding of both purified GTF preparations to the cell surface of S. mutans GS-5 under these conditions. Effects of dextrans on GTF binding. Because the pretreatment of viable or heat-inactivated cells of strain GS-5 with glucanases markedly reduced the subsequent binding of both GTF preparations, it appeared that the GTF fractions were interacting with cell surface glucan molecules. If this were the case, the addition of glucan molecules to the binding mixtures 0.3 might be expected to inhibit enzyme interaction with the cell surface. The addition of low-molecular-weight dextran T10 to the binding mixtures U. U 0.2markedly decreased cell binding of both the I0 GTF-A and GTF-B fractions (Fig. 3). The binding of GTF-A to heat-inactivated THB-grown cells was completely abolished by the addition U.~ 0.1 of as little as 50 ,ug of dextran T10, whereas GTF-B binding was inhibited approximately LOW. 60% by high concentrations of the low-molecular-weight dextran. Similar effects were also observed when high-molecular-weight dextran 6 2 4 0 T2000 was included in the binding assays. Unlike GTF(mU) dextrans, the addition of soluble carboxymethylFIG. 2. Effects of Pronase on GTF-A and GTF-B a-1,3-glucan to the binding mixtures did not binding to heat-inactivated cells. Cells (4 x 10"' CFU) inhibit enzyme binding. Soluble dextrans also grown in THB medium were incubated with and markedly inhibited the binding of the GTF fracwithout Pronase (1.3 U) as described in the text. GTFA (0.57 U/mg) and GTF-B (5.9 U/mg) binding was tions to CD-grown cells (data not shown). The decreased enzyme binding observed in measured as described in the text. Symbols: *, GTFA binding to control cells; 0, GTF-A binding to the presence of dextrans could be explained in E
Pronase-treated cells; A, GTF-B binding to control cells; A, GTF-B binding to Pronase-treated cells.
terms of either dextran interaction with the cells and/or with the GTF preparations. To assess
TABLE 2. Effects of glucanases on the binding of GTF-A and GTF-B to heat-inactivated cells' GTF-B
GTF-A Treatment
Added
(mU/10" CFU)
None Dextranase (7.6 U)
0.17 0.17
Bound
mU/1b CU
0.22 0.015
Bound
Added
(mU/10 %CFU)
129
1.7
mU/1O" CFU
0.48
28
3.5 8.8 1.7 0.06 2.9 0.17 0.017 10.0 1.7 0.05 a-1,3-Glucanase (0.68 mU) a Heat-inactivated THB-grown cells (1.9 X 10" CFU) were incubated with the indicated levels of glucanases and SCA (total volume, 2.0 ml) for 18 h at 37"C. After washing with 5 ml of 0.9% NaCl, the cells were resuspended in 2.0 ml of SCA and portions (1.9 x 109 CFU) were mixed with the indicated levels of GTF-A (0.57 U/mg) or GTF-B (5.9 U/mg). Incubations for binding and the enzyme assays were carried out as described earlier.
INFECT. IMMUN.
KURAMITSU AND INGERSOLL
656
the relative contribution of both types of interactions on enzyme binding, heat-inactivated cells of strain GS-5 were initially treated with dextran T10, the cells were washed thoroughly, and GTF-B binding was measured (Fig. 4). Prior incubation of the cells with dextran T10 markedly decreased the subsequent binding of GTFB to the cell surface. Inhibition of enzyme binding was also observed with dextran T2000. A similar reduction in GTF-A binding was observed when heat-inactivated cells were pre100 75 a
c 3 0
50
m LA.
HD
25
0
0.2
0.4
Dextron TIO (mg) FIG. 3. Effects of dextran T10 on GTF-A and GTFB binding to heat-inactivated cells. Cells (2.3 x 109 CFU) grown in THB medium were incubated in the presence of GTF-A (3.5 m U, 0.57 U/mg) or GTF-B (27 mU, 6.48 U/mg) with the indicated amounts of dextran T10 as described in the text. Binding of enzymes is expressed relative to the values obtained in the absence of dextran TlO for GTF-A (0.12 mU/11 CFU) and GTF-B (0.55 mU/J10 CFU). Symbols: 0, GTF-B binding; 0, GTF-A binding.
TABLE 3. Elution of cell-associated GTF activity and glucans
0.4
U.
Elution Elution
0.3
0
E
treated with dextrans (data not shown). Therefore, the interaction of dextran molecules with the cell surface of S. mutans markedly inhibited the subsequent binding of exogenous GTF preparations. Pretreatment of cells with dextrans could affect subsequent binding of GTFs by eluting off cell-associated GTF-glucan complexes as demonstrated previously (21). The removal of cellassociated glucans by the addition of dextrans could be demonstrated by measuring the loss of endogenously synthesized ['4C]glucan from the cell surface in the presence of dextran T10 (Table 3). In addition, a significant decrease in the cell-associated GTF activity was also observed for viable cells after dextran elution. These results suggest that soluble dextrans are capable of removing GTF-glucan complexes from the cell surface of strain GS-5. The interaction of soluble dextrans with GTFB to form dextran-GTF-B complexes also inhibited the subsequent binding of the enzyme to heat-inactivated cells (Fig. 5). The enzyme-dextran complex, isolated after gel filtration chromatography, bound very poorly to the cells relative to the uncomplexed homogeneous GTF-B preparation. Similar comparisons were not carried out with the GTF-A fraction because this high-molecular-weight fraction contains glucanlike molecules and the enzyme activity of this fraction is not stimulated by the addition of exogenous glucans (18). These results suggest that dextran interaction with both the cells and GTF-B fraction can play a role in reducing enzyme binding to the cells.
0.2
La: I-
0.1I 0
GTF-B (mU)
FIG. 4. Effects of dextran TIOpretreatment of cells on GTF-B binding. Heat-inactivated cells (2.7 x 1010 CFU) grown in THB medium were incubated with and without dextran T1O (5 mg) for 15 min at 37C, washed three times with 0.9% NaCl, and resuspended in SCA. GTF-B (6.48 U/mg) was added to the cell suspensions, and binding was measured as described earlier. Symbols: *, GTF-B binding to control cells; 0, GTF-B binding to dextran-treated cells.
Cell-associated GTF
Cell-associated glucan
CFU)a 0.15 (100) 0.085 (56) 0.088 (59)
sion)
(mU/108
(cpm/suspen-
475 (100)b Saline 250 (53)c 3 N NaCl 250 (53)b Dextran T10 (1.0 mg) a Cells (2 x 1010 CFU) grown in THB medium were incubated with saline, dextran T10, or 3 N NaCl for 30 min at 37°C and washed three times with 5.0 ml of saline. Appropriate portions were then assayed for total GTF activity. b Cells (2.3 x 109 CFU) were incubated with [14C]sucrose in the standard assay mixture for 60 min. The cells containing cell-associated ['4C]glucan were then incubated with saline or dextran T10 for 30 min at 37°C and filtered and counted as described previously (18). e Cells incubated with ['4C]sucrose were filtered and washed directly with 3 N NaCl (three times) before counting.
GLUCOSYLTRANSFERASE BINDING
VOL. 20, 1978
Effects of NaCl extraction on GTF binding. Previous results have indicated that cellassociated GTF activity can be extracted from the cell surface of strain GS-5 utilizing hypertonic salt solutions (17). This procedure also removes glucans from the cell surface (Table 3). If cell surface glucan-GTF complexes are involved in enzyme binding, the removal of the complexes might be expected to affect the subsequent binding of GTF fractions. Heat-inactivated cells extracted with 3 N NaCl contained a greater number of binding sites with lowered affinity for GTF-B than did unextracted cells (Fig. 6). These results suggest that the removal 0.4L.
0.3
01 E 0.2
-0
E/ UA. a0o1~~~ 0/
-
2 3 4 GTF - B (mU) FIG. 5. Effects of dextran T150 pretreatment of GTF-B on enzyme binding to heat-inactivated cells. The dextran T150-GTF-B complex was prepared and isolated as described in the text. The complex as well as GTF-B alone was added to heat-inactivated cells (6.8 x I& CFU), and binding was determined as described in the text. Symbols: 0, GTF-B binding; 0, dextran T150-GTF-B binding.
0
1
0.3
0
0.2
0
LL
0.1
I GTF- B (mU) FIG. 6. Effects of NaCl extraction on GTF-B binding. Heat-inactivated cells (6.2 x 1iO CFU) grown in THB medium were washed with saline or 3 N NaCl 0
(three times) as describedpreviously (17). GTF-B (6.48 U/mg) binding was determined as previously described. Symbols *, GTF-B binding by saline-washed cells; 0, GTF-B binding by NaCl-extracted cells.
657
of a portion, but not all (17; Table 3), of the GTF-glucan cell surface complexes produces additional "sites" for enzyme binding. However, it is also possible that other molecules affecting GTF binding were also removed or altered by NaCl extraction. Similar results were also observed when the interaction of GTF-A with NaCl-extracted cells was measured (data not shown). Removal of a portion of the glucan-GTF complex from CD-grown cells with 3 N NaCl also yielded comparable results. DISCUSSION Both enzymatic (17, 21, 23) and immunological (13, 19) characterizations as well as recent kinetic data (14) suggest that the cell-associated GTF activity of S. mutans may be derived from the extracellular enzymes. However, the cell surface structures involved in the binding of the enzymes to the cells have not been identified. Mukasa and Slade (25), utilizing relatively crude GTF preparations and THB-grown cells of strain HS-6, have proposed that GTF binding occurred in the vicinity of the serotype a antigen and glucan molecules associated with a heatstable protein on the cell surface. In contrast, Germaine and Schachtele (7) have suggested that GTF molecules become associated with the cell surface by interacting directly with other GTF molecules bound to cell surface glucans. The glucan molecules are in turn associated with a glucan-binding protein on the cell surface. Recently, several potentially different glucan binding proteins have been described in extracts of S. mutans (20; M. M. McCabe and R. M. Hamelik, Proceedings of the International Symposium on The Secretory Immune System and Caries Immunity, in press). Furthermore, the results of agglutination and flocculation studies could be interpreted as suggesting that binding sites for different types of glucans (linear or branched) may exist on the surface of S. mutans (22; C. Wu-Yuan, S. Tai, and H. D. Slade, in press). However, it is not yet clear whether any of the glucan-binding proteins described are involved in GTF binding to the cells. The results of the present investigation indicate that heat treatment of GS-5 cells does not significantly decrease the number of binding sites for the GTF-A or GTF-B fraction. However, the affinity of the sites for each GTF fraction is somewhat reduced (Fig. 1; Table 1). Therefore, the binding site(s) for the GTF enzymes are composed of a heat-stable component in agreement with previous suggestions utilizing crude GTF preparations (15, 24). Furthermore, both GTF fractions appear to bind to the same cell surface components of S. mutans GS-5.
INFECT. IMMUN. 658 KURAMITSU AND INGERSOLL However, it is not clear why heat-inactivated nents which either retard GTF binding or intercells appear to bind more GTF-A molecules than fere with enzyme activity. do viable cells even though both cell suspensions In contrast to cells grown in complex media contain a similar number of GTF-B binding (25) recent results (15) have suggested that cellsites. Perhaps heat treatment might expose ad- associated glucans do not appear to be required ditional low-affinity binding sites for GTF-A but for the binding of GTF activity to cells grown in not for GTF-B. a CD medium devoid of possible sucrose or Although heat stable, the putative GTF bind- polysaccharide contaminants. However, the ing sites appear to be proteinaceous or protein- present results utilizing CD-grown cells and associated because Pronase treatment of the more purified enzyme preparations do suggest a cells decreased GTF-A and GTF-B binding to role for cell-associated glucans in GTF binding. cells grown in either complex (Fig. 2) or chemi- The difference between the present and earlier cally defined media (data not shown). The heat (15) investigations may be due to the utilization stability and resistance to trypsin of the binding of more purified enzyme preparations in the sites may result from the association of the present study containing less glucan contamiprotein component with polysaccharide or other nants. Thus, the glucans present in crude enmacromolecules of the cell surface. Previous re- zyme preparations (15) may have obviated the sults (25) demonstrating that trypsin or pepsin requirement for cell surface glucans. Recently, treatment of S. mutans HS-6 decreased GTF Montville et al. (23) have also demonstrated that binding indicate that the cell surface architec- strain GS-5 grown in a different chemically deture may vary between different strains of S. fined medium also possesses relatively high levmutans. els of cell-associated GTF activity. It is clear from the present investigation (TaBecause previous results have suggested that ble 2) and previous observations (25) that cell- receptors for branched glucans may be relatively associated glucans can be involved in the inter- heat stable (15, 22), it is tempting to speculate action of GTF activity with the surface of S. that these molecules may play a role in convertmutans. The results of adding water-soluble dex- ing extracellular GTF activity to the cell-assotrans to the binding mixtures (25; Fig. 3 through ciated form. In addition, glucans produced by 5) are also compatible with this view. In addi- cell-associated GTF activity would be heat station, recent studies (15) indicate that water-sol- ble. Separate heat-labile receptors for linear gluuble dextran molecules can bind directly to the cans may also exist on the cell surface of S. cell surface of strain GS-5. This binding might mutans (24) and could also be involved in GTF also interfere with subsequent GTF interaction binding (7). Thus, it is possible that extracellular with the cell surface mediated by glucan mole- GTF activity (containing glucans) can become cules present in the exogenous enzyme com- cell associated after interaction with multiple plexes (3, 5, 18, 26). The cell surface glucan glucan-binding components of the cell surface as involved in GTF binding appears to be the well as to cell-associated glucan molecules. Rebranched glucan product synthesized by the cell- cent results from this laboratory (15) have demassociated GTF activity because treatment of onstrated that CD-grown cells of strain GS-5 the cells with either a-1,6- or a-1,3-glucanase which display reduced ability to interact with significantly reduced GTF binding (Table 2). glucans bind exogenous unfractionated GTF Furthermore, naturally occurring a-1,3-glucans preparations as well as the parental cells. Therehave not been positively identified as a normal fore, this nonaggregating variant with normal cell surface component in streptococci. However, GTF binding may be altered in the glucan rebecause neither glucanase preparation utilized ceptors required for agglutination but not those in the present or earlier (25) investigations is capable of interacting with glucans present in homogeneous, these conclusions are not un- the crude GT preparations. Once enzyme-glucan equivocal. complexes become attached to the cell surface It has recently been suggested (15) that the via glucan receptors or cell-associated glucans, number of S. mutans cell surface sites capable subsequent enzyme molecules can become cell of binding GTF activity appears to be limited. bound via protein-protein or protein-glucan inSaturation of these sites prevents further bind- teractions (7). Thus, the binding of GTF activity ing of enzyme activity. The results of the present to the cell surface of S. mutans may involve investigation suggest that extraction of a portion multiple binding sites with different types of of the cell surface glucan-GTF complex with protein-polysaccharide interactions. Final proof hypertonic salt solutions may provide additional of this proposal awaits the direct isolation and sites for enzyme binding. In addition, such ex- purification of the cell surface receptor moletraction might also remove cell surface compo- cules involved in GTF binding.
659
VOL. 20, 1978
GLUCOSYLTRANSFERASE BINDING
ACKNOWLEDGMENT This investigation was supported by Public Health Service grant DE-03258 from the National Institute of Dental Research.
of a variant of Streptococcus mutans altered in its ability to interact wich glucans. Infect. Immun. 16:575-586. Kuramitsu, H. K. 1973. Characterization of invertase activity from cariogenic Streptococcus mutans. J. Bacteriol. 115:1003-1010. Kuramitsu, H. K. 1974. Characterization of cell-associated dextransucrase from glucose-grown cells of Streptococcus mutans. Infect. Immun. 10:227-235. Kuramitsu, H. K. 1975. Characterization of extracellular glucosyltransferase activity of Streptococcus mutans. Infect. Immun. 12:738-749. Kuramitsu, H. K., and L. Ingersoll. 1976. Differential inhibition of Streptococcus mutans in vitro adherence by anti-glucosyltransferase antibodies. Infect. Immun. 13:1775-1777. McCabe, M. M., R. M. Hamelik, and E. E. Smith. 1977. Purification of dextran-binding protein from cariogenic Streptococcus mutans. Biochem. Biophys. Res. Commun. 78:273-278. McCabe, M. M., and E. E. Smith. 1973. Origin of the cell-associated dextransucrase of Streptococcus mutans. Infect. Immun. 7:829-838. McCabe, M. M., and E. E. Smith. 1975. Relationship between cell-bound dextransucrase and the agglutination of Streptococcus mutans. Infect. Immun.
LITERATURE CITED 1. Ashweli, G. 1957. Colorimetric analysis of sugars. Methods Enzymol. 3:85-86. 2. Ceska, M., K. Granath, B. Norman, and B. Guggenheim. 1972. Structural and enzymatic studies on glucans synthesized with glucosyltransferases of some strains of oral streptococci. Acta Chem. Scand. 26:2223-2230. 3. Ciardi, J. E., A. J. Beaman, and C. L. Wittenberger. 1977. Purification, resolution, and interaction of glucosyltransferases of Streptococcus mutans 6715. Infect. Immun. 18:237-246. 4. Fukui, K., Y. Fukui, and T. Moriyama. 1974. Purification and properties of dextransucarase and invertase from Streptococcus mutans. J. Bacteriol. 118:796-804. 5. Germaine, G. R., A. M. Chludzinski, and C. F. Schachtele. 1974. Streptococcus mutans dextransucrase: requirement for primer dextran. J. Bacteriol. 120:287-294. 6. Germaine, G. R., S. K. Harlander, W.-L. S. Leung, and C. F. Schachtele. 1977. Streptococcus mutans dextransucransucrase: functioning of primer dextran and endogenous dextranase in water-soluble and waterinsoluble glucan synthesis. Infect. Immun. 16:637-648. 7. Germaine, G. R., and C. F. Schachtele. 1976. Streptococcus mutans dextransucrase: mode of interaction with high-molecular-weight dextran and role in cellular aggregation. Infect. Immun. 13:365-372. 8. Gibbons, R. J. 1972. Presence of an invertase-like enzyme and a sucrose permeation system in strains of Streptococcus mutans. Caries Res. 6:122-131. 9. Gibbons, R. J., and S. Banghart. 1967. Synthesis of extracellular dextran by cariogenic bacteria and its presence in human dental plaque. Arch. Oral Biol. 12:11-24. 10. Gibbons, R. J., and M. Nygaard. 1968. Synthesis of insoluble dextran and its significance in the formation of gelatinous deposits by plaque-forming streptococci. Arch. Oral Biol. 13:1249-1262. 11. Gibbons, R. J., and J. van Houte. 1975. Dental caries. Annu. Rev. Med. 26:121-136. 12. Guggenheim, B., and J. J. Burckhardt. 1974. Isolation and properties of a dextranase from Streptococcus mutans OMZ 176. Helv. Odontol. Acta 18:101-113. 13. Guggenheim, B., and E. Newbrun. 1969. Extracellular glucosyltransferase activity of an HS strain of Streptococcus mutans. Helv. Odontol. Acta 13:84-97. 14. Janda, W. M., and H. K. Kuramitsu. 1976. Regulation of extracellular glucosyltransferase production and the relationship between extracellular and cell-associated activities in Streptococcus mutans. Infect. Immun. 14:191-202. 15. Janda, W. M., and H. K. Kuramitsu. 1977. Properties
16. 17. 18. 19.
20.
21. 22.
12:512-520. 23. Montville, T. J., C. L Cooney, and A. J. Sinskey. 1977. Distribution of dextransucrase in Streptococcus mutans and observations on the effect of soluble dextran on dextransucrase activities. Infect. Immun. 18:629-635. 24. Mukasa, H., and H. D. Slade. 1973. 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. 8:555-562. 25. Mukasa, H., and H. D. Slade. 1973. Mechanism of adherence of Streptococcus mutans to smooth surfaces. II. Nature of the binding site and the adsorption of dextran-levan synthetase enzymes on the cell-wall surface of the streptococcus. Infect. Immun. 9:419-429. 26. Mukasa, H., and H. D. Slade. 1974. Mechanism of adherence of Streptococcus mutans to smooth surfaces. III. Purification and properties of the enzyme complex responsible for adherence. Infect. Immun. 10:1135-1145. 27. Staat, R. H., and C. F. Schachtele. 1974. Evaluation of dextranase production by the cariogenic bacterium Streptococcus mutans. Infect. Immun. 9:467-469. 28. Tomarelli, R. M., J. Charney, and M. L. Harding. 1949. The use of azoalbumin as a substrate in the colorimetric determination of peptic and tryptic activity. J. Lab. Clin. Med. 34:428433. 29. van Houte, J., and V. N. Upeslacis. 1976. Studies on the mechanism of sucrose-associated colonization of Streptococcus mutans on teeth of conventional rats. J. Dent. Res. 55:216-222.
