- Email: [email protected]
Stabilization of the glucan-binding lectin of Streptococcus sobrinus by specific ligand
PERGAMON
Archives of Oral Biology 43 (1998) 33±38
ARCHIVES OF ORAL BIOLOGY
Stabilization of the glucan-binding lectin of Streptococcus sobrinus by speci®c ligand Andrea M. Denson, R. J. Doyle * Department of Microbiology and Immunology, University of Louisville, Health Science Center, Louisville, KY 40292, USA Accepted 29 August 1997
Abstract Cell suspensions of Streptococcus sobrinus can be aggregated by high molecular-weight a-1,6 glucans. The aggregation depends on the ®delity of a cell wall-bound, glucan-binding lectin (GBL). It is thought that the lectin may play a part in the sucrose-dependent accretion of streptococci in dental plaques. Results showed that the anionic detergent, sodium dodecyl sulphate (SDS) was a potent inhibitor of the lectin. When cells were incubated in SDS and washed to remove the detergent, lectin activity was diminished. Following incubation of the cells with SDS in the presence of glucan T-10, a low molecular-weight a-1,6 glucan, the loss of activity was less pronounced, suggesting that the glucan aorded partial protection against denaturation. Urea and guanidine hydrochloride were good inhibitors of the lectin, but, unlike SDS, were not able to inhibit it irreversibly, except at very high concentrations. Cationic detergents, such as cetylpyridinium bromide (and chloride), also irreversibly denatured the streptococcal lectin, but were not as eective as SDS in abolishing its activity. The results suggest that a-1,6 glucan stabilizes the GBL of S. sobrinus, rendering it more resistant to the eect of chaotropes. This may be one reason why dental plaques tend to resist detergents in dentri®ces. # 1998 Elsevier Science Ltd. All rights reserved.
1. Introduction Many bacteria are now known to possess lectins. A lectin is de®ned as a carbohydrate-binding protein that can agglutinate cells or precipitate polysaccharides (reviewed in Doyle, 1994). As far as is known, the only function of bacterial lectins is to participate in adhesion (Ofek and Doyle, 1994). Ma et al. (1996) found that Streptococcus sobrinus elaborated several proteins (adhesons) that could bind to Sephadex G-75, a crosslinked dextran. The proteins could be eluted by chaotropes such as 4 M guanidine hydrochloride and by low molecular-weight a-1,6 glucan (Ma et al. used a commercial glucan of 10,000 mol. wt, called glucan T10, to elute Sephadex-bound proteins). The eluted proteins ranged from 13,000 to 140,000 mol. wt. When a
* Corresponding author. Abbreviations: PBS, phosphate-buered sodium dodecyl sulphate.
saline,
SDS,
SDS±polyacrylamide gel was soaked in Triton X-100 to remove the SDS and incubated with sucrose, the high molecular-weight band became swollen and transparent, suggesting it was glucosyltransferase. Ma et al. grew Strep. sobrinus under conditions that obtain a cell population incapable of aggregation with high molecular-weight a-1,6 glucan (glucan T-2000). The conditions involved growth of the cells in the presence of subinhibitory concentrations of citrate. Presumably, the citrate was able to complex with the manganous ion, an ion essential for the expression of the property of agglutinability by glucan T-2000 (Bauer et al., 1993; LuÈ-LuÈ et al., 1992). When supernatants of the cultures were poured over Sephadex columns and proteins subsequently recovered from them, it was found that a 58±60-KDa protein was missing. This same protein could be found in cultures from cells that were sensitive to aggregation by glucan T-2000. In addition, Ma et al. isolated two isogenic mutants incapable of aggregation with a-1,6 glucans. These mutants were also unable to produce the 58±60-kDa protein. Based on
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 0 8 9 - 7
34
A. M. Denson et al. / Archives of Oral Biology 43 (1998) 33±38
the foregoing results, Ma et al. (1996) suggested the 58±60 kDa protein was glucan- binding lectin. The glucan-binding lectin is distinct from glucan-binding proteins, which are not involved in glucan-dependent aggregation. In a recent paper, Sato et al. (1996) found that the growth of Strep. mutans under ``stressful'' conditions led to its ability to be aggregated by a-1,6 glucans. It thus appears that both Strep. sobrinus and Strep. mutans, the bacteria most frequently isolated from dental carious lesions, can produce glucan-binding lectins. In the case of Strep. sobrinus, the lectin is constitutive, whereas in Strep. mutans it must be induced. Interestingly, the glucan-binding lectin of Strep. mutans was also 60 kDa, but had an amino acid composition distinct from the lectin of Strep. sobrinus. Only limited studies have been performed on the stability of the glucan-binding lectin with respect to denaturation by chaotropes (Bauer et al., 1993; Liang et al., 1989). A part of the present research was directed to de®ning the condition(s) that denature the lectin. Bauer et al. (1993) noted that the lectin was slowly denatured by 4±6 M urea. The work to be described here considers that ligand (a-1,6 glucan) may stabilize the lectin. This consideration was derived from studies in which ligand stabilized or ``protected'' proteins against denaturation or proteolysis (Doyle et al., 1973, 1975, 1976; Glew and Doyle, 1979; Singh et al., 1993; Wang et al., 1996). For example, the lectin concanavalin A is more resistant to aggregation by heat when in the presence of the ligand methyl-a-D-mannoside (Doyle et al., 1973). Drake et al. n T-10 tended to impart resistance of the Strep. sobrinus glucan-binding lectin to trypsin. Liang et al. (1989) provided some preliminary results showing that the glucan-binding lectin of Strep. sobrinus is stabilized by a-1,6 glucan. In the present research, several ways were used to denature proteins, but with emphasis on SDS as this detergent is a component of many dentri®ces. 2. Materials and methods
of autoagglutination and capable of forming even, turbid suspensions after washing (Drake et al., 1988a). The bacteria were harvested by centrifugation, washed twice in cold distilled±deionized water and suspended to an optical density of 0.8±10 in PBS (40 mM sodium phosphate, 150 mM sodium chloride, pH 7.2). The bacteria were grown for 15±18 hr to late exponential phase. Cells from cultures >24 hr are usually acidi®ed, due to production of lactic acid, resulting in low concentrations of glucan-binding lectins (Bauer et al., 1993; Drake et al., 1988b; LuÈ-LuÈ et al., 1992). 2.2. Aggregation assays Procedures developed by Drake et al. (1988a) were followed throughout. A suspension (3 or 4 ml) of bacteria in PBS was mixed vigorously with glucan T-2000 (5 mg/ml ®nal concentration) for 5 sec. The rate of aggregation was followed as loss of turbidity against time. Usually, readings were made every 15 sec on a Bausch and Lomb Spectronic 20 spectrophotometer at 540 nm. Control tubes contained PBS plus cells, but lacked glucan T-2000. Rate constants were calculated from the slopes of ®rst-order plots of ln A/A0 vs time, where A = sample turbidity at some time and A0=starting turbidity. The rate constants, K, are given in units per min or per sec. In some experiments, cells were initially preincubated with detergent or chaotrope for de®ned periods of time at room temperature. The cells were then washed two to ®ve times in distilled± deionized water to remove the detergents and suspended in PBS. 2.3. Reagents and chemicals Lysozyme, mutanolysin, SDS, urea, guanidine hydrochloride, cetylpyridinium chloride, cetylpyridinium bromide, Triton X-100, Brij and glucans T-10 and T2000 were obtained from Sigma. Tween 80 was a product of Difco.
2.1. Bacteria and growth conditions
3. Results
Strep. sobrinus 6715 was used throughout the study. This bacterium possesses glucan-binding lectin activity and has been used in previous studies from this laboratory. The bacteria were maintained on brain±heart infusion agar (Difco Laboratories, Detroit, MI) at 48C. For routine growth the bacteria were cultured in trypticase soy broth (Difco) at 378C in 5% CO2. Before autoclaving the medium was incubated for 2 hr at 558C with 5 mg yeast invertase (Sigma Chemical Company, St. Louis, MO) per g dry medium. The medium was then incubated at 378C for 4 h with yeast dextranase. The enzyme treatments result in cells free
There are various kinds of detergents. Some are positively charged, such as cetylpyridinium salts, others are negatively charged, such as SDS, and others are neutral, such as Triton X-100 and Tween 80. The detergents have an ionic or hydrophilic end and a hydrophobic end. Because SDS is known to be a good denaturant as well as a component of dentri®ces, the present work was primarily directed to determining its eects on a glucan-binding lectin from a cariogenic streptococcus. In Fig. 1, results are shown describing the assay for glucan-binding lectin and the inhibitory eects of SDS on the lectin. In all cases involving SDS
A. M. Denson et al. / Archives of Oral Biology 43 (1998) 33±38
Fig. 1. Inhibition of the glucan-binding lectin (GBL) of Strep. sobrinus by sodium dodecyl sulphate (SDS). Following incubation of the cells in SDS solutions, the suspensions were centrifuged and washed extensively in water before suspension in PBS and assay for GBL.
the detergent was washed out of the cells before suspension of the bacteria for lectin assays. Concentrations of SDS at 1.25 mg/ml completely abolished the activities of glucan-binding lectin, whereas lower concentrations gave lower degrees of lectin loss. In another experiment (Fig. 2), the rate constants for SDS-treated cells were determined as a function of SDS concentration, but under conditions to include glucan T-10 in the mixture. At all inhibiting concentrations of SDS, the glucan aorded ``protection'' of the lectin from denaturation. Control experiments included glycogen and soluble starch, but these a-glucans were not protective. When glucan T-10 was mixed with cells before SDS, eective protection was demon-
Fig. 2. Inhibition of glucan-binding lectin of Strep. sobrinus by sodium dodecyl sulphate (SDS) in the presence and absence of glucan T-10. The glucan T-10 concentration was 10 mg/ml.
35
strated at glucan concentrations above 4±5 mg/ml (Fig. 3). A comparison of the ability of SDS to denature the glucan-binding lectin with that of other detergents is shown in Table 1. The non-ionic detergents Brij, Triton X-100 and Tween 80 had no eect on the lectin, regardless of the presence of glucan T-10. The basic detergents cetylpyridinium chloride and bromide were inhibitors, but were not as eective as SDS. For example, the bromide at 1.5 mg/ml produced a 50% loss of lectin activity; similar concentrations of SDS completely abolished it. At the same 1.5 mg/ml concentration, cetylpyridinium chloride was not an inhibitor, although at a much higher concentration, it resulted in considerable loss of lectin activity. In both cases where cetylpyridinium bromide and chloride caused denaturation of the lectin, glucan T-10 aorded protection. When the cells were mixed with non-detergent chaotropic agents (Table 2), there was irreversible inhibition of glucan-binding lectin by concentrated urea or guanidine hydrochloride only. Iodine had a small eect on the lectin activity; thiocyanate had even less of an eect. In no case did glucan T-10 oer resistance to denaturation. Earlier work had shown that somewhat lower concentrations of urea tended to denature the glucan-binding lectin, but much longer times were required (Doyle and Taylor, 1994; Liang et al., 1989). If glucan T-10 somehow changes the surface of Strep. sobrinus, then it might be expected that the cells would be more or less susceptible to peptidoglycan hydrolysis. Lin-Lin et al. (1995) and Mata et al. (1997) showed that glucan T-10 rendered Strep. sobrinus more hydrophilic, as determined by ammonium sulphate aggregation and by adhesion of the bacteria to hydrocarbons. Oral streptococci are not readily lysed by
Fig. 3. Inhibition of the glutan-binding lectin of Strep. sobrinus by sodium dodecyl sulphate (SDS) as a function of glucan concentration. The SDS concentration was 2 mg/ml. The glucan T-10 concentration was varied as shown.
36
A. M. Denson et al. / Archives of Oral Biology 43 (1998) 33±38
Table 1 Inhibition of the glucan-binding lectin of Strep. sobrinus 6715 by detergents* Detergent
Concentration (mg/ml)
CTAB CTAC CTAC SDS SDS Brij Triton X-100 Tween 80
1.5 1.5 6.5 1.25 0.16 100 20 20
Inhibition (%) 50 0 45 97 12 0 0 0
Protection by glucan T-10 +(1.5) NA +(1.9) +(4.1) +(1.5) NA NA NA
CTAB(-C), cetylpyridinium bromide (chloride). * %Inhibition = inhibition in presence of detergent. Protection by glucan re¯ects 30-min incubation in the detergent (+10 mg/ ml glucan T-10), washing 4±5 with H2O, followed by assay for lectin activity. Numbers in parentheses refer to K-value ratios (presence or absence of glucan T-10).
4. Discussion
lysozyme, an enzyme in normal human saliva, but in buered suspension, Strep. sobrinus is slowly lysed in the presence of lysozyme. When a suspension (OD = 0.8) of Strep. sobrinus was mixed with a ®nal concentration of 100 mg/ml lysozyme, there was 15% loss in turbidity after 2 hr. Glucan T-10 had no eect on the rate of lysis (Table 3). Similarly, mutanolysin lytic activity was not altered by glucan T-10, nor was the activity of a mixture of lysozyme and mutanolysin aected by glucan T-10. In the case of mutanolysin, the lysis was greater than 75% after 2 hr. When the cells were extracted with SDS (cells became glucanbinding lectin-negative) the rates of lysis were unaffected in either the presence or absence of glucan T-10.
Our results can be summarized as follows: (i) the glucan-binding lectin of Strep. sobrinus can be denatured by anionic or cationic detergents; (ii) the lectin is not inactivated by neutral detergents; (iii) the lectin is protected from detergent inactivation by glucan T10, a ligand for the glucan-binding lectin; (iv) polysaccharides, such as starch or glycogen, did not protect the glucan-binding lectin from inactivation by detergents; (v) denaturants such as 8 M urea and 6 M GuCl caused loss of glucan-binding lectin activity without any protection by speci®c ligand. The results clearly show that speci®c ligand confers stability to the lectin when it is subjected to ionic detergents.
Table 2 Eects of chaotropes on the glucan-binding lectin of Strep. sobrinus Agent
Concentration (M)
Guanidine hydrochloride Guanidine hydrochloride + glucan T-10 Guanidine hydrochloride Guanidine hydrochloride + glucan T-10 Urea Urea + glucan T-10 Urea Urea + glucan T-10 Sodium thiocyanate Sodium thiocyanate + glucan T-10 Sodium iodide Sodium iodide + glucan T-10
3 3 6 6 4 4 8 8 3 3 3 3
% inhibition 0 0 >94 >91 10 10 100 100 5 10 20 24
Cells in PBS were mixed with the above agents to reach the ®nal concentrations shown. After 30 min at room temperature, cells were centrifuged, washed 3 in water and suspended in PBS for assay of lectin activity.
A. M. Denson et al. / Archives of Oral Biology 43 (1998) 33±38
37
Table 3 Rates of cellular lysis of Strep. sobrinus 6715 by lysozyme and/or mutanolysin* Enzyme
Cell population
Rate
Lysozyme
Control cells Cells + glucan T-10 SDS-extracted cells SDS-extracted cells + glucan T-10 Control cells Cells + glucan T-10 SDS-extracted cells SDS-extracted cells + glucan T-10 Control cells Cells + glucan T-10 SDS-extracted cells SDS-extracted cells + glucan T-10
1.0 1.1 0.9 0.9 1.0 0.9 0.9 0.9 1.0 1.0 1.2 1.2
Mutanolysin
Lysozyme + mutanolysin
* Rates are normalized to cells suspended in PBS. Lysozyme and mutanolysin concentrations were 100 mg/ml. Glucan T-10 was 4 mg/ml ®nal concentration. Extraction with 1.0 mg/ml SDS in saline for 1 hr at room temperature, followed by four washes with water. Rates are from ln A/A0 vs time curves as before.
It is known that ligands may protect proteins from denaturants or even proteolysis. For example, the lectin concanavalin A assumes a relatively heat-resistant conformation when in the presence of its methyl-a-Dmannoside ligand (Doyle et al., 1973). The autolytic wall amidase of Bacillus subtilis is resistant to proteolysis when cell wall-bound (Jollie et al., 1980). Antihapten antibodies may become more resistant to chaotropes when haptens are added to their solutions. Binding of cytosine monophosphate by ribonuclease confers a heat-stable conformation to the enzyme (Simons, 1971). In addition, serum albumins are stabilized against proteolysis by fatty acids or dyes (Grossberg et al., 1965). In light of these ®ndings, the results for the glucan-binding lectin do not seem profound. In the case of that lectin, however, some denaturants seemed to inactivate it whereas others do not. A 30-min incubation of the cells in 3 M urea did not modify lectin activity if the urea was removed. A 4-hr incubation caused some loss of glucan-binding lectin. In contrast, 8 M urea completely abolished the lectin activity after a 30-min incubation. The presence of glucan T-10 had no eect on the 8 M urea denaturation of the lectin. The dierence in eectiveness of the various chaotropes can be explained. In the case of cetylpyridinium bromide, the bromide is also chaotropic, but the corresponding chloride in cetylpyridinium chloride is not. On a molar basis therefore, cetylpyridinium bromide should be a better denaturant than the chloride. Urea and guanidine hydrochloride appear to denature by ``ordering'' protein-bound water around them (Israelachvili and Wennerstrom, 1996). SDS, on the other hand, required only a small concentration to eect inactivation of the glucan-binding lectin. SDS not only changes water structure, but also has a rela-
tively long hydrocarbon tail that can intercalate into hydrophobic clefts in proteins (Chakrabarti, 1993; Lienado and Jamieson, 1981). It is possible the protective eects of glucan T-10 are related to the ability of the glucan to exclude SDS (as well as cetylpyridinium chloride/bromide) from sites susceptible to denaturation. Hydrocarbon alone is not enough to cause denaturation of glucan-binding lectin, as shown by the inability of Tween-80 and Triton X-100 to modify the lectin binding to glucan T-2000 or to denature glucosyltransferases (Clardi et al., 1978; Figures and Edwards, 1979; Kawabata et al., 1993; Lim et al., 1982; Marsh, 1991; Wittenberger et al. 1978). Mata et al. (1997) found that low molecular-weight a-1,6 glucans reduced the rate of adhesion of Strep. sobrinus to hydrocarbons. In contrast, the glucan-binding lectinnegative Strep. mutans adhered to hydrocarbon at the same rate in the presence or absence of a-1,6 glucan. The stable form of the glucan-binding lectin resists denaturation by ionic detergents. The sequence of the Strep. sobrinus lectin is unknown, so it is impossible to predict its secondary structures. The eects of chaotropic conditions on the lectin will no doubt be better understood when the structure has been solved. The present research may have future consequences of practical signi®cance, about which some possibly important generalizations concern the most common detergent in toothpastes, SDS (another name for which is sodium lauryl sulphate). SDS confers a pleasing detergent eect in dentri®ces and oers a cleansing action characteristic of soaps. Our results suggest, but do not prove, that sucrose-dependent dental plaques are SDS-resistant. That may be one reason why sucrose has been implicated in dental caries. Even the brushing of teeth with an SDS-containing dentri®ce does not assure removal of plaques.
38
A. M. Denson et al. / Archives of Oral Biology 43 (1998) 33±38
References Bauer, P. D., Trapp, C., Drake, D., Taylor, K. G., Doyle, R. J. (1993) Acquisition of manganous ions by mutans group streptococci. J. Bacteriol. 175, 819±825. Chakrabarti, P. (1993) Anion binding sites in protein structures. J. Mol. Biol. 234, 463±482. Ciardi, J. E., Bowen, W. H., Rolla, G. (1978) The eect of antibacterial compounds on glucosyltransferase activity from Streptococcus mutans. Archs Oral Biol. 23, 301±305. Doyle R. J. (1994) Introduction to lectins and their interactions with microorganisms. In Lectin±Microorganism Interactions. (Eds. RJ Doyle and M Slifkin), pp. 1±64. Marcel Dekker, Inc., NY. Doyle, R. J., Nicholson, S. K., Gray, R. D. (1973) Carbohydrate-induced conformational change in concanavalin A. Carbohydrate Research 29, 265±270. Doyle, J. J., Thomasson, D. L., Gray, R. D. (1975) Spectral changes accompanying the interaction between metal ligands and concanavalin A. FEBS Letters 52, 185±187. Doyle, R. J., Thomasson, D. L., Nicholson, S. K. (1976) Stabilization of concanavalin A by metal ligands. Carbohydrate Research 46, 111±118. Doyle, R. J., Taylor, K. G. (1994) Sucrose, glucan-binding proteins and oral steptococcal adhesion. Cells and Materials 4, 91±100. Drake, D., Taylor, K. G., Bleiweis, A. S., Doyle, R. J. (1988a) Speci®city of the glucan-binding lectin of Streptococcus cricetus. Infect. Immun. 56, 1864±1872. Drake, D., Taylor, K. G., Doyle, R. J. (1988b) Expression of the glucan-binding lectin of Streptococcus cricetus requires manganous ion. Infect. Immun. 56, 2205±2207. Figures, W. R., Edwards, J. R. (1979) Resolution of the glycosyltransferase activities from two strains of Streptococcus mutans by polyacrylamide gel electrophoresis in the presence of Tween 80. Biochim. Biophys. Acta 577, 142±146. Glew, R. H., Doyle, R. J. (1979) Protection of sodium dodecyl sulfate-induced aggregation of concanavalin A by saccharide ligands. Carbohydrate Research 73, 219±226. Grossberg, A. L., Markus, G., Pressman, D. (1965) Change in antibody conformation induced by hapten. Proc. Natl. Acad. Sci. 54, 942±945. Israelachvili, J., Wennerstrom, H. (1996) Role of hydration and water structure in biological and colloidal interactions. Nature 379, 219±225. Jollie, L. K., Doyle, R. J., Streips, U. N. (1980) Extracellular proteases modify cell wall turnover in Bacillus subtilis. J. Bacteriol. 141, 119±1208.
Kawabata, S., Torii, M., Minami, T., Fujiwara, T., Hamada, S. (1993) Eects of selected surfactants on puri®ed glucosyltransferases from mutans streptococci and cellular adherence to smooth surfaces. J. Med. Microbiol. 38, 54±60. Liang, L., Drake, D., Doyle, R. J. (1989) Stability of glucanbinding lectin of oral streptococci. J. Dent. Res. 68, 1677. Lienado R. A. and Jamieson R. A. (1981) Surfactants. Anal. Chem. 53: 174R±182R. Lim, J. K., Smith, S., McGlothlin, J., Gerencser, V. F. (1982) Minimum inhibitory concentration of surfactants for plaque antiadherents. Caries Res. 16, 440±442. Lin, L., Rosenberg, M., Taylor, K. G., Doyle, R. J. (1995) Kinetic analysis of ammonium sulfate dependent aggregation of bacteria. Colloids and Surfaces B: Biointerfaces 5, 127±134. Lu, L., Singh, J. S., Galperin, M. Y., Grake, D., Taylor, K. G., Doyle, R. J. (1992) Chelating agents inhibit activity and prevent expression of streptococcal glucan-binding lectins. Infect. Immun. 60, 3807±3813. Ma, Y., Lassiter, M. O., Banas, J. A., Galperin, M. Y., Taylor, K. G., Doyle, R. J. (1996) Multiple glucan-binding proteins of Streptococcus sobrinus. J. Bacteriol. 178, 1572± 1577. Marsh, P. D. (1991) Dentri®ces containing new agents for the control of plaque and gingivitis: microbiological aspects. J. Clin. Periodontol. 18, 462±467. Mata, L. G., Drake, D., Doyle, R. J. (1997) Modi®cation of surface properties of oral streptococci by a-1,6 glucans. Colloids and Surfaces B: Biointerfaces 8, 295±302. Ofek I. and Doyle R. J. (1994) Bacterial Adhesion to Cells and Tissues. p. 561. Chapman and Hall, New York. Sato, Y., Yamamoto, Y., Kizaki, H. (1997) Cloning and sequence analysis of the gbpC gene encoding a novel glucan-binding protein of Streptococcus mutans. Infect. Immun. 65, 669±675. Simons, R. (1971) Conformational changes in some ribonuclease-inhibitor complexes. Biochim. Biophys. Acta 251, 126±131. Singh, J. S., Taylor, K. G., Doyle, R. J. (1993) Essential amino acids involved in glucan-dependent aggregation of Streptococcus sobrinus. Carbohydrate Research 244, 137± 147. Wang, Q., Singh, S., Taylor, K. G., Doyle, R. J. (1996) Antiadhesions of Streptococcus sobrinus. Adv. Explt. Biol. Med. 408, 249±262. Wittenberger, C. L., Beaman, A. J., Lee, L. N. (1978) Tween 80 eect on glucosyltransferase synthesis by Streptococcus salivarius. J. Bacteriol. 133, 231±239.
Archives of Oral Biology 43 (1998) 33±38
ARCHIVES OF ORAL BIOLOGY
Stabilization of the glucan-binding lectin of Streptococcus sobrinus by speci®c ligand Andrea M. Denson, R. J. Doyle * Department of Microbiology and Immunology, University of Louisville, Health Science Center, Louisville, KY 40292, USA Accepted 29 August 1997
Abstract Cell suspensions of Streptococcus sobrinus can be aggregated by high molecular-weight a-1,6 glucans. The aggregation depends on the ®delity of a cell wall-bound, glucan-binding lectin (GBL). It is thought that the lectin may play a part in the sucrose-dependent accretion of streptococci in dental plaques. Results showed that the anionic detergent, sodium dodecyl sulphate (SDS) was a potent inhibitor of the lectin. When cells were incubated in SDS and washed to remove the detergent, lectin activity was diminished. Following incubation of the cells with SDS in the presence of glucan T-10, a low molecular-weight a-1,6 glucan, the loss of activity was less pronounced, suggesting that the glucan aorded partial protection against denaturation. Urea and guanidine hydrochloride were good inhibitors of the lectin, but, unlike SDS, were not able to inhibit it irreversibly, except at very high concentrations. Cationic detergents, such as cetylpyridinium bromide (and chloride), also irreversibly denatured the streptococcal lectin, but were not as eective as SDS in abolishing its activity. The results suggest that a-1,6 glucan stabilizes the GBL of S. sobrinus, rendering it more resistant to the eect of chaotropes. This may be one reason why dental plaques tend to resist detergents in dentri®ces. # 1998 Elsevier Science Ltd. All rights reserved.
1. Introduction Many bacteria are now known to possess lectins. A lectin is de®ned as a carbohydrate-binding protein that can agglutinate cells or precipitate polysaccharides (reviewed in Doyle, 1994). As far as is known, the only function of bacterial lectins is to participate in adhesion (Ofek and Doyle, 1994). Ma et al. (1996) found that Streptococcus sobrinus elaborated several proteins (adhesons) that could bind to Sephadex G-75, a crosslinked dextran. The proteins could be eluted by chaotropes such as 4 M guanidine hydrochloride and by low molecular-weight a-1,6 glucan (Ma et al. used a commercial glucan of 10,000 mol. wt, called glucan T10, to elute Sephadex-bound proteins). The eluted proteins ranged from 13,000 to 140,000 mol. wt. When a
* Corresponding author. Abbreviations: PBS, phosphate-buered sodium dodecyl sulphate.
saline,
SDS,
SDS±polyacrylamide gel was soaked in Triton X-100 to remove the SDS and incubated with sucrose, the high molecular-weight band became swollen and transparent, suggesting it was glucosyltransferase. Ma et al. grew Strep. sobrinus under conditions that obtain a cell population incapable of aggregation with high molecular-weight a-1,6 glucan (glucan T-2000). The conditions involved growth of the cells in the presence of subinhibitory concentrations of citrate. Presumably, the citrate was able to complex with the manganous ion, an ion essential for the expression of the property of agglutinability by glucan T-2000 (Bauer et al., 1993; LuÈ-LuÈ et al., 1992). When supernatants of the cultures were poured over Sephadex columns and proteins subsequently recovered from them, it was found that a 58±60-KDa protein was missing. This same protein could be found in cultures from cells that were sensitive to aggregation by glucan T-2000. In addition, Ma et al. isolated two isogenic mutants incapable of aggregation with a-1,6 glucans. These mutants were also unable to produce the 58±60-kDa protein. Based on
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 0 8 9 - 7
34
A. M. Denson et al. / Archives of Oral Biology 43 (1998) 33±38
the foregoing results, Ma et al. (1996) suggested the 58±60 kDa protein was glucan- binding lectin. The glucan-binding lectin is distinct from glucan-binding proteins, which are not involved in glucan-dependent aggregation. In a recent paper, Sato et al. (1996) found that the growth of Strep. mutans under ``stressful'' conditions led to its ability to be aggregated by a-1,6 glucans. It thus appears that both Strep. sobrinus and Strep. mutans, the bacteria most frequently isolated from dental carious lesions, can produce glucan-binding lectins. In the case of Strep. sobrinus, the lectin is constitutive, whereas in Strep. mutans it must be induced. Interestingly, the glucan-binding lectin of Strep. mutans was also 60 kDa, but had an amino acid composition distinct from the lectin of Strep. sobrinus. Only limited studies have been performed on the stability of the glucan-binding lectin with respect to denaturation by chaotropes (Bauer et al., 1993; Liang et al., 1989). A part of the present research was directed to de®ning the condition(s) that denature the lectin. Bauer et al. (1993) noted that the lectin was slowly denatured by 4±6 M urea. The work to be described here considers that ligand (a-1,6 glucan) may stabilize the lectin. This consideration was derived from studies in which ligand stabilized or ``protected'' proteins against denaturation or proteolysis (Doyle et al., 1973, 1975, 1976; Glew and Doyle, 1979; Singh et al., 1993; Wang et al., 1996). For example, the lectin concanavalin A is more resistant to aggregation by heat when in the presence of the ligand methyl-a-D-mannoside (Doyle et al., 1973). Drake et al. n T-10 tended to impart resistance of the Strep. sobrinus glucan-binding lectin to trypsin. Liang et al. (1989) provided some preliminary results showing that the glucan-binding lectin of Strep. sobrinus is stabilized by a-1,6 glucan. In the present research, several ways were used to denature proteins, but with emphasis on SDS as this detergent is a component of many dentri®ces. 2. Materials and methods
of autoagglutination and capable of forming even, turbid suspensions after washing (Drake et al., 1988a). The bacteria were harvested by centrifugation, washed twice in cold distilled±deionized water and suspended to an optical density of 0.8±10 in PBS (40 mM sodium phosphate, 150 mM sodium chloride, pH 7.2). The bacteria were grown for 15±18 hr to late exponential phase. Cells from cultures >24 hr are usually acidi®ed, due to production of lactic acid, resulting in low concentrations of glucan-binding lectins (Bauer et al., 1993; Drake et al., 1988b; LuÈ-LuÈ et al., 1992). 2.2. Aggregation assays Procedures developed by Drake et al. (1988a) were followed throughout. A suspension (3 or 4 ml) of bacteria in PBS was mixed vigorously with glucan T-2000 (5 mg/ml ®nal concentration) for 5 sec. The rate of aggregation was followed as loss of turbidity against time. Usually, readings were made every 15 sec on a Bausch and Lomb Spectronic 20 spectrophotometer at 540 nm. Control tubes contained PBS plus cells, but lacked glucan T-2000. Rate constants were calculated from the slopes of ®rst-order plots of ln A/A0 vs time, where A = sample turbidity at some time and A0=starting turbidity. The rate constants, K, are given in units per min or per sec. In some experiments, cells were initially preincubated with detergent or chaotrope for de®ned periods of time at room temperature. The cells were then washed two to ®ve times in distilled± deionized water to remove the detergents and suspended in PBS. 2.3. Reagents and chemicals Lysozyme, mutanolysin, SDS, urea, guanidine hydrochloride, cetylpyridinium chloride, cetylpyridinium bromide, Triton X-100, Brij and glucans T-10 and T2000 were obtained from Sigma. Tween 80 was a product of Difco.
2.1. Bacteria and growth conditions
3. Results
Strep. sobrinus 6715 was used throughout the study. This bacterium possesses glucan-binding lectin activity and has been used in previous studies from this laboratory. The bacteria were maintained on brain±heart infusion agar (Difco Laboratories, Detroit, MI) at 48C. For routine growth the bacteria were cultured in trypticase soy broth (Difco) at 378C in 5% CO2. Before autoclaving the medium was incubated for 2 hr at 558C with 5 mg yeast invertase (Sigma Chemical Company, St. Louis, MO) per g dry medium. The medium was then incubated at 378C for 4 h with yeast dextranase. The enzyme treatments result in cells free
There are various kinds of detergents. Some are positively charged, such as cetylpyridinium salts, others are negatively charged, such as SDS, and others are neutral, such as Triton X-100 and Tween 80. The detergents have an ionic or hydrophilic end and a hydrophobic end. Because SDS is known to be a good denaturant as well as a component of dentri®ces, the present work was primarily directed to determining its eects on a glucan-binding lectin from a cariogenic streptococcus. In Fig. 1, results are shown describing the assay for glucan-binding lectin and the inhibitory eects of SDS on the lectin. In all cases involving SDS
A. M. Denson et al. / Archives of Oral Biology 43 (1998) 33±38
Fig. 1. Inhibition of the glucan-binding lectin (GBL) of Strep. sobrinus by sodium dodecyl sulphate (SDS). Following incubation of the cells in SDS solutions, the suspensions were centrifuged and washed extensively in water before suspension in PBS and assay for GBL.
the detergent was washed out of the cells before suspension of the bacteria for lectin assays. Concentrations of SDS at 1.25 mg/ml completely abolished the activities of glucan-binding lectin, whereas lower concentrations gave lower degrees of lectin loss. In another experiment (Fig. 2), the rate constants for SDS-treated cells were determined as a function of SDS concentration, but under conditions to include glucan T-10 in the mixture. At all inhibiting concentrations of SDS, the glucan aorded ``protection'' of the lectin from denaturation. Control experiments included glycogen and soluble starch, but these a-glucans were not protective. When glucan T-10 was mixed with cells before SDS, eective protection was demon-
Fig. 2. Inhibition of glucan-binding lectin of Strep. sobrinus by sodium dodecyl sulphate (SDS) in the presence and absence of glucan T-10. The glucan T-10 concentration was 10 mg/ml.
35
strated at glucan concentrations above 4±5 mg/ml (Fig. 3). A comparison of the ability of SDS to denature the glucan-binding lectin with that of other detergents is shown in Table 1. The non-ionic detergents Brij, Triton X-100 and Tween 80 had no eect on the lectin, regardless of the presence of glucan T-10. The basic detergents cetylpyridinium chloride and bromide were inhibitors, but were not as eective as SDS. For example, the bromide at 1.5 mg/ml produced a 50% loss of lectin activity; similar concentrations of SDS completely abolished it. At the same 1.5 mg/ml concentration, cetylpyridinium chloride was not an inhibitor, although at a much higher concentration, it resulted in considerable loss of lectin activity. In both cases where cetylpyridinium bromide and chloride caused denaturation of the lectin, glucan T-10 aorded protection. When the cells were mixed with non-detergent chaotropic agents (Table 2), there was irreversible inhibition of glucan-binding lectin by concentrated urea or guanidine hydrochloride only. Iodine had a small eect on the lectin activity; thiocyanate had even less of an eect. In no case did glucan T-10 oer resistance to denaturation. Earlier work had shown that somewhat lower concentrations of urea tended to denature the glucan-binding lectin, but much longer times were required (Doyle and Taylor, 1994; Liang et al., 1989). If glucan T-10 somehow changes the surface of Strep. sobrinus, then it might be expected that the cells would be more or less susceptible to peptidoglycan hydrolysis. Lin-Lin et al. (1995) and Mata et al. (1997) showed that glucan T-10 rendered Strep. sobrinus more hydrophilic, as determined by ammonium sulphate aggregation and by adhesion of the bacteria to hydrocarbons. Oral streptococci are not readily lysed by
Fig. 3. Inhibition of the glutan-binding lectin of Strep. sobrinus by sodium dodecyl sulphate (SDS) as a function of glucan concentration. The SDS concentration was 2 mg/ml. The glucan T-10 concentration was varied as shown.
36
A. M. Denson et al. / Archives of Oral Biology 43 (1998) 33±38
Table 1 Inhibition of the glucan-binding lectin of Strep. sobrinus 6715 by detergents* Detergent
Concentration (mg/ml)
CTAB CTAC CTAC SDS SDS Brij Triton X-100 Tween 80
1.5 1.5 6.5 1.25 0.16 100 20 20
Inhibition (%) 50 0 45 97 12 0 0 0
Protection by glucan T-10 +(1.5) NA +(1.9) +(4.1) +(1.5) NA NA NA
CTAB(-C), cetylpyridinium bromide (chloride). * %Inhibition = inhibition in presence of detergent. Protection by glucan re¯ects 30-min incubation in the detergent (+10 mg/ ml glucan T-10), washing 4±5 with H2O, followed by assay for lectin activity. Numbers in parentheses refer to K-value ratios (presence or absence of glucan T-10).
4. Discussion
lysozyme, an enzyme in normal human saliva, but in buered suspension, Strep. sobrinus is slowly lysed in the presence of lysozyme. When a suspension (OD = 0.8) of Strep. sobrinus was mixed with a ®nal concentration of 100 mg/ml lysozyme, there was 15% loss in turbidity after 2 hr. Glucan T-10 had no eect on the rate of lysis (Table 3). Similarly, mutanolysin lytic activity was not altered by glucan T-10, nor was the activity of a mixture of lysozyme and mutanolysin aected by glucan T-10. In the case of mutanolysin, the lysis was greater than 75% after 2 hr. When the cells were extracted with SDS (cells became glucanbinding lectin-negative) the rates of lysis were unaffected in either the presence or absence of glucan T-10.
Our results can be summarized as follows: (i) the glucan-binding lectin of Strep. sobrinus can be denatured by anionic or cationic detergents; (ii) the lectin is not inactivated by neutral detergents; (iii) the lectin is protected from detergent inactivation by glucan T10, a ligand for the glucan-binding lectin; (iv) polysaccharides, such as starch or glycogen, did not protect the glucan-binding lectin from inactivation by detergents; (v) denaturants such as 8 M urea and 6 M GuCl caused loss of glucan-binding lectin activity without any protection by speci®c ligand. The results clearly show that speci®c ligand confers stability to the lectin when it is subjected to ionic detergents.
Table 2 Eects of chaotropes on the glucan-binding lectin of Strep. sobrinus Agent
Concentration (M)
Guanidine hydrochloride Guanidine hydrochloride + glucan T-10 Guanidine hydrochloride Guanidine hydrochloride + glucan T-10 Urea Urea + glucan T-10 Urea Urea + glucan T-10 Sodium thiocyanate Sodium thiocyanate + glucan T-10 Sodium iodide Sodium iodide + glucan T-10
3 3 6 6 4 4 8 8 3 3 3 3
% inhibition 0 0 >94 >91 10 10 100 100 5 10 20 24
Cells in PBS were mixed with the above agents to reach the ®nal concentrations shown. After 30 min at room temperature, cells were centrifuged, washed 3 in water and suspended in PBS for assay of lectin activity.
A. M. Denson et al. / Archives of Oral Biology 43 (1998) 33±38
37
Table 3 Rates of cellular lysis of Strep. sobrinus 6715 by lysozyme and/or mutanolysin* Enzyme
Cell population
Rate
Lysozyme
Control cells Cells + glucan T-10 SDS-extracted cells SDS-extracted cells + glucan T-10 Control cells Cells + glucan T-10 SDS-extracted cells SDS-extracted cells + glucan T-10 Control cells Cells + glucan T-10 SDS-extracted cells SDS-extracted cells + glucan T-10
1.0 1.1 0.9 0.9 1.0 0.9 0.9 0.9 1.0 1.0 1.2 1.2
Mutanolysin
Lysozyme + mutanolysin
* Rates are normalized to cells suspended in PBS. Lysozyme and mutanolysin concentrations were 100 mg/ml. Glucan T-10 was 4 mg/ml ®nal concentration. Extraction with 1.0 mg/ml SDS in saline for 1 hr at room temperature, followed by four washes with water. Rates are from ln A/A0 vs time curves as before.
It is known that ligands may protect proteins from denaturants or even proteolysis. For example, the lectin concanavalin A assumes a relatively heat-resistant conformation when in the presence of its methyl-a-Dmannoside ligand (Doyle et al., 1973). The autolytic wall amidase of Bacillus subtilis is resistant to proteolysis when cell wall-bound (Jollie et al., 1980). Antihapten antibodies may become more resistant to chaotropes when haptens are added to their solutions. Binding of cytosine monophosphate by ribonuclease confers a heat-stable conformation to the enzyme (Simons, 1971). In addition, serum albumins are stabilized against proteolysis by fatty acids or dyes (Grossberg et al., 1965). In light of these ®ndings, the results for the glucan-binding lectin do not seem profound. In the case of that lectin, however, some denaturants seemed to inactivate it whereas others do not. A 30-min incubation of the cells in 3 M urea did not modify lectin activity if the urea was removed. A 4-hr incubation caused some loss of glucan-binding lectin. In contrast, 8 M urea completely abolished the lectin activity after a 30-min incubation. The presence of glucan T-10 had no eect on the 8 M urea denaturation of the lectin. The dierence in eectiveness of the various chaotropes can be explained. In the case of cetylpyridinium bromide, the bromide is also chaotropic, but the corresponding chloride in cetylpyridinium chloride is not. On a molar basis therefore, cetylpyridinium bromide should be a better denaturant than the chloride. Urea and guanidine hydrochloride appear to denature by ``ordering'' protein-bound water around them (Israelachvili and Wennerstrom, 1996). SDS, on the other hand, required only a small concentration to eect inactivation of the glucan-binding lectin. SDS not only changes water structure, but also has a rela-
tively long hydrocarbon tail that can intercalate into hydrophobic clefts in proteins (Chakrabarti, 1993; Lienado and Jamieson, 1981). It is possible the protective eects of glucan T-10 are related to the ability of the glucan to exclude SDS (as well as cetylpyridinium chloride/bromide) from sites susceptible to denaturation. Hydrocarbon alone is not enough to cause denaturation of glucan-binding lectin, as shown by the inability of Tween-80 and Triton X-100 to modify the lectin binding to glucan T-2000 or to denature glucosyltransferases (Clardi et al., 1978; Figures and Edwards, 1979; Kawabata et al., 1993; Lim et al., 1982; Marsh, 1991; Wittenberger et al. 1978). Mata et al. (1997) found that low molecular-weight a-1,6 glucans reduced the rate of adhesion of Strep. sobrinus to hydrocarbons. In contrast, the glucan-binding lectinnegative Strep. mutans adhered to hydrocarbon at the same rate in the presence or absence of a-1,6 glucan. The stable form of the glucan-binding lectin resists denaturation by ionic detergents. The sequence of the Strep. sobrinus lectin is unknown, so it is impossible to predict its secondary structures. The eects of chaotropic conditions on the lectin will no doubt be better understood when the structure has been solved. The present research may have future consequences of practical signi®cance, about which some possibly important generalizations concern the most common detergent in toothpastes, SDS (another name for which is sodium lauryl sulphate). SDS confers a pleasing detergent eect in dentri®ces and oers a cleansing action characteristic of soaps. Our results suggest, but do not prove, that sucrose-dependent dental plaques are SDS-resistant. That may be one reason why sucrose has been implicated in dental caries. Even the brushing of teeth with an SDS-containing dentri®ce does not assure removal of plaques.
38
A. M. Denson et al. / Archives of Oral Biology 43 (1998) 33±38
References Bauer, P. D., Trapp, C., Drake, D., Taylor, K. G., Doyle, R. J. (1993) Acquisition of manganous ions by mutans group streptococci. J. Bacteriol. 175, 819±825. Chakrabarti, P. (1993) Anion binding sites in protein structures. J. Mol. Biol. 234, 463±482. Ciardi, J. E., Bowen, W. H., Rolla, G. (1978) The eect of antibacterial compounds on glucosyltransferase activity from Streptococcus mutans. Archs Oral Biol. 23, 301±305. Doyle R. J. (1994) Introduction to lectins and their interactions with microorganisms. In Lectin±Microorganism Interactions. (Eds. RJ Doyle and M Slifkin), pp. 1±64. Marcel Dekker, Inc., NY. Doyle, R. J., Nicholson, S. K., Gray, R. D. (1973) Carbohydrate-induced conformational change in concanavalin A. Carbohydrate Research 29, 265±270. Doyle, J. J., Thomasson, D. L., Gray, R. D. (1975) Spectral changes accompanying the interaction between metal ligands and concanavalin A. FEBS Letters 52, 185±187. Doyle, R. J., Thomasson, D. L., Nicholson, S. K. (1976) Stabilization of concanavalin A by metal ligands. Carbohydrate Research 46, 111±118. Doyle, R. J., Taylor, K. G. (1994) Sucrose, glucan-binding proteins and oral steptococcal adhesion. Cells and Materials 4, 91±100. Drake, D., Taylor, K. G., Bleiweis, A. S., Doyle, R. J. (1988a) Speci®city of the glucan-binding lectin of Streptococcus cricetus. Infect. Immun. 56, 1864±1872. Drake, D., Taylor, K. G., Doyle, R. J. (1988b) Expression of the glucan-binding lectin of Streptococcus cricetus requires manganous ion. Infect. Immun. 56, 2205±2207. Figures, W. R., Edwards, J. R. (1979) Resolution of the glycosyltransferase activities from two strains of Streptococcus mutans by polyacrylamide gel electrophoresis in the presence of Tween 80. Biochim. Biophys. Acta 577, 142±146. Glew, R. H., Doyle, R. J. (1979) Protection of sodium dodecyl sulfate-induced aggregation of concanavalin A by saccharide ligands. Carbohydrate Research 73, 219±226. Grossberg, A. L., Markus, G., Pressman, D. (1965) Change in antibody conformation induced by hapten. Proc. Natl. Acad. Sci. 54, 942±945. Israelachvili, J., Wennerstrom, H. (1996) Role of hydration and water structure in biological and colloidal interactions. Nature 379, 219±225. Jollie, L. K., Doyle, R. J., Streips, U. N. (1980) Extracellular proteases modify cell wall turnover in Bacillus subtilis. J. Bacteriol. 141, 119±1208.
Kawabata, S., Torii, M., Minami, T., Fujiwara, T., Hamada, S. (1993) Eects of selected surfactants on puri®ed glucosyltransferases from mutans streptococci and cellular adherence to smooth surfaces. J. Med. Microbiol. 38, 54±60. Liang, L., Drake, D., Doyle, R. J. (1989) Stability of glucanbinding lectin of oral streptococci. J. Dent. Res. 68, 1677. Lienado R. A. and Jamieson R. A. (1981) Surfactants. Anal. Chem. 53: 174R±182R. Lim, J. K., Smith, S., McGlothlin, J., Gerencser, V. F. (1982) Minimum inhibitory concentration of surfactants for plaque antiadherents. Caries Res. 16, 440±442. Lin, L., Rosenberg, M., Taylor, K. G., Doyle, R. J. (1995) Kinetic analysis of ammonium sulfate dependent aggregation of bacteria. Colloids and Surfaces B: Biointerfaces 5, 127±134. Lu, L., Singh, J. S., Galperin, M. Y., Grake, D., Taylor, K. G., Doyle, R. J. (1992) Chelating agents inhibit activity and prevent expression of streptococcal glucan-binding lectins. Infect. Immun. 60, 3807±3813. Ma, Y., Lassiter, M. O., Banas, J. A., Galperin, M. Y., Taylor, K. G., Doyle, R. J. (1996) Multiple glucan-binding proteins of Streptococcus sobrinus. J. Bacteriol. 178, 1572± 1577. Marsh, P. D. (1991) Dentri®ces containing new agents for the control of plaque and gingivitis: microbiological aspects. J. Clin. Periodontol. 18, 462±467. Mata, L. G., Drake, D., Doyle, R. J. (1997) Modi®cation of surface properties of oral streptococci by a-1,6 glucans. Colloids and Surfaces B: Biointerfaces 8, 295±302. Ofek I. and Doyle R. J. (1994) Bacterial Adhesion to Cells and Tissues. p. 561. Chapman and Hall, New York. Sato, Y., Yamamoto, Y., Kizaki, H. (1997) Cloning and sequence analysis of the gbpC gene encoding a novel glucan-binding protein of Streptococcus mutans. Infect. Immun. 65, 669±675. Simons, R. (1971) Conformational changes in some ribonuclease-inhibitor complexes. Biochim. Biophys. Acta 251, 126±131. Singh, J. S., Taylor, K. G., Doyle, R. J. (1993) Essential amino acids involved in glucan-dependent aggregation of Streptococcus sobrinus. Carbohydrate Research 244, 137± 147. Wang, Q., Singh, S., Taylor, K. G., Doyle, R. J. (1996) Antiadhesions of Streptococcus sobrinus. Adv. Explt. Biol. Med. 408, 249±262. Wittenberger, C. L., Beaman, A. J., Lee, L. N. (1978) Tween 80 eect on glucosyltransferase synthesis by Streptococcus salivarius. J. Bacteriol. 133, 231±239.