Journal of Oral Rehabilitation 2004 31; 278285

Adherence of Streptococcus mutans to various restorative materials in a continuous ow system

S . E I C K * , E . G L O C K M A N N , B . B R A N D L & W . P F I S T E R *

Departments of *Medical Microbiology and

Conservative Dentistry, University Hospital of Jena, Jena, Germany

SUMMARY A continuous ow system was developed to evaluate the adhesion of Streptococcus mutans ATCC 25175 to lling materials (Ariston, Tetric, Dyract, Compoglass, Vitremer, Aqua Ionol, Ketac Fil, amalgam, Galloy and ceramics as controls). Streptococcus mutans was added to saliva-coated test specimens, and a nutrient broth permanently supplied over a time period of 48 h and then the weight of plaque, the number and viability of the bacteria adhering to the materials were determined. The weights of articial plaque on all lling materials tested were higher than those on ceramics, the highest values were measured on the glassiono-

mers. The amount of plaque correlates with the surface roughness, whereas there was no correlation of the surface roughness with the number of colonyforming units (CFU) of S. mutans. The CFU of adhering S. mutans also depends on the viability of the bacteria. The plaque on Ketac Fil contained a high number of viable bacteria. The uorides of glassionomers do not efciently prevent the attachment and the viability of S. mutans. KEYWORDS: adherence, Streptococcus mutans, lling materials, ow system Accepted for publication 2 June 2003

Introduction
In dentistry many different restorative materials are available. Amalgam, glassionomer cements, resinbased composites, resin-modied glassionomer cements and polyacid-modied resins are used as primary lling materials. They are incorporated in the mouth. So determination of their acute systemic toxicity, allergic properties, cytotoxicity, mutagenicity and cancerogenic effects should be included in biological tests (1). Nevertheless, the effects on oral microorganisms have to be considered. Dental materials release different substances (2), which might be able to inuence directly the growth of bacteria (3, 4). However, the adhesion of bacteria and the formation of plaque are important in pathogenesis of caries and periodontal diseases. Some authors have assessed plaque formation on restorative materials in vivo (5, 6), but different factors may inuence these results. However, in in vitro tests normally static bacterial broth cultures are used, which are suitable for a maximum of 24 h. We ourselves observed a decrease of the number of viable bacteria
2004 Blackwell Publishing Ltd

after 8 h in determining the growth curve of S. mutans ATCC 25175 (3). Then the microorganisms will not have enough nutrients and their toxic products will damage themselves. Therefore, the rst aim of our study was to develop a continuous ow system for determining the adhesion and colonization of Streptococcus mutans on saliva-coated restoratives for 48 h. The second aim was to compare different restorative materials with ceramics as a control, particularly the effect of surface roughness and the materials type were investigated.

Materials and methods

Flow system Infusion systems normally used in intensive care units* were selected and connected with a chamber containing a polytetrauoroethylene disc with four cylindrical cavities used for the specimens (Fig. 1). The bacterial
*Braun, Melsungen, Germany.

278

ADHERENCE OF S. MUTANS TO MATERIALS

279

Fig. 1. Chamber of the ow system. The chamber represents the core of the continuous ow system. On a disc of tetrauorethylene four cavities are located for the test specimens prepared of various dental lling materials. After closing the chamber will be connected with infusion system and bacterial suspension and nutrient broth are applied for forming an articial plaque. Table 1. Dental lling materials used for the in vitro test assay Group Metallic materials Conventional glassionomers Resin-modied glassionomers Polyacid-modied resin Composites Ceramics Material Amalcap Plus Galloy Ketac Fil Aplicap Aqua Ionol Vitremer Compoglass Dyract Tetric Ariston pHc Empress Lot number 918025 608133 737628/04 3345 19930416 701 298 950892 810234 105 619 CE 0123 A 15067 Manufacturer Vivadent, Ellwangen, Germany SDI Espe, Seefeld, Germany Voco, Cuxhaven, Germany 3M, St Paul, MN, USA Vivadent, Ellwangen, Germany De Trey/Dentsply, Konstanz, Germany Vivadent, Ellwangen, Germany Vivadent, Ellwangen, Germany Vivadent, Ellwangen, Germany

lters of the infusion systems prevented contamination of the nutrient broth, which was permanently transferred into the chamber to the specimens. Pilot tests showed that an optimum dropping speed of 10 mL h)1 was selected.

Test specimens The descriptions of the dental restorative materials tested are presented in Table 1. First, the cavities (diameter 8 mm, thickness 1 mm) of the chamber disc were insulated with a very thin layer of the Insulating Gel to prevent a permanent connection between the test specimens and the chamber. The restorative materials were mixed following manufacturers instructions. The conventional glassionomer specimens were

covered with a protective varnish (Vocopal varnish). After hardening for 1 h the samples were removed from the chamber and stored in 95% humidity at 37 C for 24 h. All specimens were polished to produce a surface roughness similar to that under clinical conditions. The metallo plastics amalgam and Galloy were nished and polished, the resin composites were nished with diamonds and Super-Snaps. Ceramics were covered with a glaze in a furnace. The roughness of the material surface before bacterial contamination was recorded by means of an optoelectronic measurement. Three test specimens of each material were selected. Three lines of a specimen were measured in a distance of about 2 mm. So the

Heraeus/Kulzer, Wehrheim, Germany.

Voco, Cuxhaven, Germany. Shofu, Ratingen, Germany. UBM Messtechnik GmbH, Ettlingen, Germany.

2004 Blackwell Publishing Ltd, Journal of Oral Rehabilitation 31; 278285

280

S . E I C K et al.
medians of the nine lines were evaluated, whereas a median represented a result of a very high number of measurement points (distance 1 lm). and then the CFU were enumerated. The viability of S. mutans was assessed by a vital uorescent technique. A mixture of uorescein diacetate and ethidium bromide according to Netuschil et al. (7) was used. Viable bacteria degrade the non-uorescent uorescein diacetate to green uorescent uorescein, this staining is binding by the bacterial membrane. Ethidium bromide penetrates only into non-vital bacterial cells and stains them red. Scanning electron microscopic (SEM) photographs were taken of the test specimens after adherence of S. mutans. The results for weight, CFU and surface roughness were expressed as medians and interquartile ranges. Sixteen specimens of each material were included in the determination of the medians of the weights and the CFU. The KruskalWallis test as a multiple pairwise comparison of medium ranks was used for determination of signicance between all materials. The MannWhitney test as a robust test compared each material with ceramics as control. Correlations between the surface roughness and other parameters were conrmed by Spearman rank correlation.

Microbial suspension Streptococcus mutans ATCC 25175, which is considered to be a type strain, was selected for the test assays. The solid cultivation medium consisted of Columbia-agar** with 8% sheep blood. Cultures were transferred for 16 h prior to being used. The incubation occurred at 37 C with infusion of 10% CO2. The inoculum for the microorganisms was prepared by resuspending with phosphate-buffered saline (PBS) to a microbial concentration of 109 bacteria mL)1.

Test assay After incorporation into the chamber the test specimens were covered with articial saliva (ISO 10993) for 1 h to produce a pellicle. This articial saliva was supplemented by 4 g porcine mucine (type II). Then the chamber was closed and connected with the ow system. First 5 mL of the bacterial suspension was transferred. The bacteria were allowed to attach to the pellicle. After 15 min the nutrient broth (brainheartinfusion-broth with 5% sucrose) was constantly supplied during the whole study period. The temperature of incubation was 37 C. After 48 h the test specimens were taken from the cavities for further investigations. The test specimens were weighted twice, rst immediately after removing from the chamber and secondly after discharging the articial plaque by ultrasonication. For this purpose the specimens were transferred into 2 mL PBS and exposed to an ultrasonic capacity of 130 W for 16 s. The difference between the two measurements was determined as the humid plaque weight. Previous tests had shown that the used capacity had no inuence on the viability of S. mutans. After ultrasonication the colony-forming units (CFU) of S. mutans and the percentage of viability of this species were assessed. To determine the viable bacteria a serial dilution was made and aliquots of 01 mL were plated onto Columbia-agar plates. The media were incubated at 37 C with infusion of 10% CO2 for 48 h
**Oxoid, Basingstoke, UK. Sigma-Aldrich Chemie GmbH, Deisenhofen, Germany.

Results
Figure 2 shows the surface roughness of the various materials tested. Ceramics had the smoothest surface (median: 019 lm, interquartile range 016025 lm), followed by the polyacid-modied composites Dyract (025 lm, interquartile range 016029 lm), Compoglass (026 lm, interquartile range 025028 lm), Ariston (032 lm, interquartile range 029037 lm) and Tetric (042 lm, interquartile range 026044 lm). The highest degree of roughness was found for the glass ionomers (Ketac Fil: 064 lm, interquartile range 059 090 lm; Aqua Ionol: 161 lm, interquartile range 090188 lm). Differences between all materials were signicant (KruskalWallis test; P < 0001). The plaque weight on all dental materials was higher than on ceramics (Fig. 3, each P < 0001). The largest amount was found on the glassionomers Aqua Ionol and Ketac Fil (ceramics: 009 mg, interquartile range 007012 mg, Aqua Ionol: 159 mg, interquartile range 121229 mg, Ketac Fil: 172 mg, interquartile range 107202 mg). A positive correlation between the medians of surface roughness and the medians of weight was stated (Spearman test; P < 005).

2004 Blackwell Publishing Ltd, Journal of Oral Rehabilitation 31; 278285

ADHERENCE OF S. MUTANS TO MATERIALS

281

Fig. 2. Surface roughness of various dental lling materials. The medians and the 25 and 75 percentiles are presented. The surface roughness was determined by optoelectronic measurement.

Fig. 3. Weight of the articial plaque on various dental lling materials. The medians and the 25 and 75 percentiles are presented. The weight was determined as the difference of the weight of test specimen after addition of a bacterial suspension of Streptococcus mutans and 48 h permanent supplying of a nutrient broth and after removing the articial plaque by ultrasonication.

In contrast, a correlation between the surface roughness and the number of viable bacteria of S. mutans (CFU) was not found (Fig. 4). The number of S. mutans colonies were as low on ceramics (18 105, interquartile range 1524 105) as on the composite reins (Tetric: 14 105, interquartile

range 0821 105), Ariston: 16 105, interquartile range 1325 105). Signicantly more viable S. mutans were counted on specimens of the polyacid-modied resin Compoglass (17 106, interquartile range 0623 106) and of the glassionomer cement Ketac Fil (13 106, interquartile range 0918 106) than

2004 Blackwell Publishing Ltd, Journal of Oral Rehabilitation 31; 278285

282

S . E I C K et al.

Fig. 4. Colony-forming units (CFU) of Streptococcus mutans on various dental lling materials. The medians and the 25 and 75 percentiles are presented. The CFU in the articial plaque were determined after addition of a suspension of this bacterial strain and 48 h permanent supplying of a nutrient broth. Table 2. Viability of Streptococcus mutans in the articial plaque on various dental materials Material Amalcap Plus Galloy Ketac Fil Aplicap Aqua Ionol Vitremer Compoglass Dyract Tetric Ariston pHc Empress 25% stages viable 25 25 75 50 25 50 25 25 25 75

on the resin-modied glassionomer Vitremer, and the glassionomers Aqua Ionol and Ketac Fil. The bacterial cells seemed to be covered by extracellular substances, probably polysaccharides. Only a few short chains were seen on composite resins and Dyract as a polyacid-modied resin. Some examples are shown in Fig. 5.

Discussion
Often bacterial adhesion studies are carried out for only a few hours. So only the rst stages of bacterial adhesion are studied. Bacteria reach the surface in a passive or active way. They adhere in a reversible and later in an irreversible way. Specic attachment and detachment processes are not so important in this early stage of microbial retention, so retention of bacteria depends on the initial critical surface energy, zeta potential and the ow rate (8, 9). Later microorganisms attach to the surface by specic interactions and nally they colonize (10). The study performed also considered these later stages of bacterial adhesion. It was possible to establish a dynamic ow system with a high degree of viability of the microorganisms in a relatively easy way. Furthermore, this construction might be also used for different other purposes. Our test chamber was similar to those of Rundegren et al. (11) who studied the effect of plaque inhibiting agents.

on those of ceramics (MannWhitney test, each P < 0001). The viability of S. mutans determined by uorescent staining is shown in Table 2. The highest percentages of viable bacteria were observed on ceramics and Ketac Fil. In contrast, low percentages of viable bacteria were found on amalgam, Galloy, composite materials, Vitremer and Dyract. The SEM features conrmed the results of the surface roughness found by optoelectronic measures. Streptococcus mutans formed large chains on the metallo plastics, ceramics and the polyacid-modied resin Compoglass. Chain forming streptococci were observed

2004 Blackwell Publishing Ltd, Journal of Oral Rehabilitation 31; 278285

ADHERENCE OF S. MUTANS TO MATERIALS

283

Fig. 5. SEM photographs of dental lling materials. The photographs of ceramics, amalgam, Vitremer, Ariston were taken after addition of a bacterial suspension of Streptococcus mutans and 48 h permanent supplying of a nutrient broth. To avoid effects on specimens the preparations were not dehydrated. So an effect on pellicle cannot be excluded.

The purpose of using saliva was to cover the specimens with a pellicle, antibacterial effects of the saliva were excluded. Articial saliva was chosen to avoid individual differences. Streptococcus mutans has a relatively high surface energy and is negatively charged; electrostatic forces play the important role for the unspecic adhesion of S. mutans (12). So the electrolytes had to be similar to the native saliva. However, mucine was a necessary component, because S. mutans is known to attach specically to the acidic-rich proteins of the mucine (13). Differences in the degree of plaque and also of the number of viable S. mutans on the different materials were stated. This nding is contrary to the results of Shahal et al. (14) who assessed no differences among the materials if saliva coating was performed. But Satou et al. (12) found material-specic differences in bacterial adhesion independently of covering the specimens with saliva. Our observations of the degree of S. mutans plaque formation clearly demonstrated the inuence of surface roughness over a time period of 48 h. So

ndings of Quirynen and Bollen (10) and Rimondini et al. (15) about dominance of surface roughness in the process are conrmed. Surface irregularities protect bacteria against shear forces during their initial reversible binding to allow them a stronger attachment. Studies by Kawai and Takaoka (16) showed that these phenomena were more marked after 8 and 24 h incubation than after 3 h incubation. In contrast, there was no correlation between the surface roughness and the number of viable S. mutans. In part these deviations might be due to the viability of the bacteria determined by uorescent staining. Otherwise a different degree of extracellular substances was observed by SEM photographs, but an exact determination both of the extracellular substances and the thickness of the pellicle on the specimens were not performed. It seems the bacteria were able to form extracellular substances from components of the nutrient broth, e.g. proteins, sucrose. A more protein-containing pellicle on Ketac Fil as a glassionomer cement (GIC) than on a composite was

2004 Blackwell Publishing Ltd, Journal of Oral Rehabilitation 31; 278285

284

S . E I C K et al.
reported (17). In addition, differences in glucan adherence in dependence of the restorative material were described (16). The viability of S. mutans showed remarkable differences among the materials tested. Our results of vital uorescence were similar to those of Netuschil et al. (18) who found 2050% viable mutans streptococci on amalgam and 70% on glass as a material without an antibacterial effect. Amalgam inhibits the growth of S. mutans (19, 20). In the present study the viability of S. mutans on the composites, Vitremer and Dyract was as low as on amalgam. Resin-based materials have cytotoxic effects, especially freshly prepared llings release substances, and in articial saliva aged materials can be cytotoxic (21). Surprisingly the CFU and also the viability of S. mutans were higher on glassionomers than on composites. In general, an inhibiting effect of the glassionomers on the growth of S. mutans was observed in association with the release of uorides (4, 20, 22). But in a study by Palenik et al. (23) Ketac Fil did not inhibit the bacterial growth of S. mutans in a diffusion test and also in our in vitro growth studies S. mutans was not inhibited by eluates of Ketac Fil, Aqua Ionol, Vitremer and Dyract either (3). Different observations were also reported by accumulation studies. Contrary to our result of very few S. mutans on composite resins Kawai and Takaoka (16) found a high number of S. sobrinus on Tetric compared with Dyract and Vitremer. The in vitro accumulation of S. mutans on extracted teeth with a Ketac Fil lling was lower than on non-treated teeth (23). Forss et al. (5) found more microbes on glassionomers than on composites in vivo and also Carlen et al. (17) conrmed a high adhesion of S. mutans to Ketac Fil. It is known that conventional and resin-modied glassionomers released the largest amount of uorides followed by the polyacid-modied resins, and the uoride uptake in the surrounding of the composite Tetric is very low (24). This phenomenon may be important for the accumulation of uorides in the surrounding enamel. But in our study the concentration of uorides did not damage S. mutans and prevent attachment. So an in vitro study by Shu et al. (25) showed that continuously supplied uorides elevated the pH in plaque and reduced enamel surface softening but not the accumulation of S. mutans. The plaque formed in vitro on the various lling materials differs in dependence of the material used. The amount of plaque correlates with the surface roughness. The uorides of glassionomers do not efciently prevent the attachment and the viability of S. mutans. To prevent plaque accumulation and especially the colonization of S. mutans a lling material with both a smooth surface and inhibiting effect on the attachment of S. mutans should be used.

Acknowledgments
We are grateful to I. Hermann and R. Kaiser (Department of Ultrastructure Research of the Jena University) for providing us the SEM photographs. N. Buchheim is acknowledged for the excellent determination of the surface roughness. And we thank Professor J.W. Einax (Institute of Inorganic Chemistry and Analytical Chemistry of the Friedrich-Schiller University of Jena) for advice in statistical analysis of the data.

References
1. Schmalz G. Concepts in biocompatibility testing of dental restorative materials. Clin Oral Investig. 1997;1:154. 2. Geurtsen W. Substances released from dental resin composites and glass ionomer cements. Eur J Oral Sci. 1998;106: 687. 3. Eick S, Welker D, Bo hland A, Pro ssel A, Pster W, Straube E. Effekt von Glasionomerwerkstoff-Eluaten auf ausgewa hlte Spezies der Mundora. Swiss Dent. 1996;17:19. 4. Friedl KH, Schmalz G, Hiller KA, Shams M. Resin-modied glass ionomer cements: uoride release and inuence on Streptococcus mutans growth. Eur J Oral Sci. 1997;105:81. 5. Forss H, Seppa L, Alakuijala P. Plaque accumulation on glass ionomer lling materials. Proceed Finnish Dent Soc. 1991;87:343. 6. Hannig M. Transmission electron microscopy of early plaque formation on dental materials in vivo. Eur J Oral Sci. 1999;107:55. 7. Netuschil L, Reich E, Brecx M. Direct measurement of the bactericidal effect of chlorhexidine on human dental plaque. J Clin Periodontol. 1989;16:484. 8. Weerkamp AH, Uyen HM, Busscher HJ. Effect of zeta potential and surface energy on bacterial adhesion to uncoated and saliva-coated human enamel and dentin. J Dent Res. 1988;67:1483. 9. Christersson CE, Dunford RG, Glantz PO, Baier RE. Effect of critical surface tension on retention of oral microorganisms. Scand J Dent Res. 1989;97:247. 10. Quirynen M, Bollen CM. The inuence of surface roughness and surface-free energy on supra- and subgingival plaque formation in man. A review of the literature. J Clin Periodontol. 1995;22:1. 11. Rundegren J, Simonsson T, Petersson L, Hansson E. Effect of Delmopinol on the cohesion of glucan-containing plaque formed by Streptococcus mutans in a ow cell system. J Dent Res. 1992;71:1792.

2004 Blackwell Publishing Ltd, Journal of Oral Rehabilitation 31; 278285

ADHERENCE OF S. MUTANS TO MATERIALS

12. Satou J, Fukunaga A, Morikawa A, Matsumae I, Satou N, Shintani H. Streptococcal adherence to uncoated and salivacoated restoratives. J Oral Rehabil. 1991;18:421. 13. Gibbons RJ, Hay DI. Adsorbed salivary acidic proline-rich proteins contribute to the adhesion of Streptococcus mutans to apatite surfaces. J Dent Res. 1989;68:1303. 14. Shahal Y, Steinberg D, Hirschfeld Z, Bronshteyn M, Kopolovic K. In vitro bacterial adherence onto pellicle-coated aesthetic restorative materials. J Oral Rehabil. 1998;25:52. 15. Rimondini L, Fare S, Brambilla E, Felloni A, Consonni C, Brossa F, Carassi A. The effect of surface roughness on early in vivo plaque colonization on titanium. J Periodontol. 1997; 68:556. 16. Kawai K, Takaoka T. Inhibition of bacterial and glucan adherence to various light-cured uoride-releasing restorative materials. J Dent. 2000;29:119. 17. Carlen A, Nikdel K, Wennerberg A, Holmberg K, Olsson J. Surface characteristics and in biolm formation on glass ionomer and composite resin. Biomaterials. 2001;22:481. 18. Netuschil L, Brecx M, Vohrer KG, Riethe P. Vital uorescence to assess in vitro and in vivo the antibacterial effects of amalgams. Acta Stomatol Belg. 1996;93:129. 19. Orstavik D. Antibacterial properties of and element release from some dental amalgams. Acta Odontol Scand. 1985;43: 231. 20. Scherer W, Lippmann N, Kaimes J. Antimicrobial properties of glassionomer cements and other restorative materials. Operat Dent. 1989;14:77. 21. Wataha JC, Rueggeberg FA, Lapp CA, Lewis JB, Lockwood PE, Ergie JW, Mettenburg DJ. In vitro cytotoxicity of resincontaining restorative materials after aging in articial saliva. Clin Oral Invest. 1999;3:144. 22. Friedl KH, Schmalz G, Hiller KA. Liquid culture tests of the effect of dental material on bacterial growth. Deutsche Zahna rztliche Zeitschrift. 1992;47:826. 23. Palenik CJ, Behnen MJ, Setcos JC, Miller CH. Inhibition of microbial adherence and growth by various glass ionomers in vitro. Dent Mater. 1992;8:16. 24. Glockmann G, Gehroldt C, Triemer K. Fluoride release from different glass ionomer cements. Deutsche Zahna rztliche Zeitschrift. 1997;52:668. 25. Shu M, Wong L, Miller JH, Sissons CH. Development of multispecies consortia biolms of oral bacteria as an enamel and root caries model system. Arch Oral Biol. 2000;45:27.

285

Correspondence: Dr Sigrun Eick, Department of Medical Microbiology, University Hospital of Jena, Semmelweisstr. 4, 07740 Jena, Germany. E-mail: sigrun.eick@med.uni-jena.de

2004 Blackwell Publishing Ltd, Journal of Oral Rehabilitation 31; 278285