Dent Mater 8:16-20, January, 1992
Inhibition of microbial adherence and growth by various glass ionomers in vitro C.J. Palenik ~, M.J. Behnen 1, J.C. Setcos 2,3,C.H. Miller 1 Department of Oral Microbiology, 2Department of Dental Materials Indiana Univ. School of Dentistry, Indianapolis IN USA 3Now at Department of Restorative Dentistry, University of California, San Francisco, School of Dentistry San Francisco CA USA
Abstract. This study measured the in vitro inhibition of growth and adherence of five oral bacteria by glass-ionomer materials. Disks were prepared from two cavity liners and four restorative class materials, by use of Teflon plates with circular wells, five mm wide and two mm deep. The bacterial species tested included: A. viscosus, S. mitis, S. mutans, L. casei, and S. sanguis. Growth inhibition studies were performed by the spreading of 0.1 mL of standardized inocula over agar plates produced with selective media, followed by the direct application of glass-ionomer disks onto the agar. On other plates, disks were placed onto uninoculated agars for 48 h, followed by bacterial inoculation. All agar plates were incubated under optimal growth conditions for each bacterial species. The four restorative materials were also placed aseptically into sterilized bovine incisors and placed into sucrose containing broth media, inoculated with S. mutans for three days. Adhering materials were disclosed and scored. An ion-exchange electrode was used to measure fluoride release over a seven-day period for all six glass ionomers. The two cavity liners and two of the restorative materials produced the largest growth inhibition zones by direct contact. No growth inhibition occurred when the specimens were allowed to come into contact with the agars prior to inoculation. All four restorative materials reduced bacterial accumulations on enamel surfaces by over 80%. Elevations in short-term fluoride release levels were positively correlated with growth inhibition. Recurrent human dental caries has been associated with the deterioration of dental restorative materials. Breakdown in marginal areas between cavity preparations and restorative materials can provide potential pathways for re-infection. Cariogenic micro-organisms present in the normal human flora could easily penetrate underlying dentin through such defects (Br~innstrSm, 1984; Br~innstrSm and Nordenvall, 1978; Br~innstrSm and Vojinovic, 1976; Lavelle, 1976). Reducing or preferably preventing such marginal breakdown could reduce the chances of recurrent caries. Obviously, physically superior restorative materials are important in the prevention of recurrent caries. However, it would also be advantageous if such materials possessed frank antimicrobial potentials. Published data (Barkhordar et al., 1989; DeSchepper et al., 1989; McComb and Ericson, 1987; Scherer et al., 1989; Schwartzman et al., 1980) have indicated that some commercially available glass-ionomer restorative materials and cements have demonstrable in vitro antimicrobial abilities. Such abilities have been related to the release of high concentrations of fluoride ions (DeSchepper et al., 1989; Forsten, 1977; Maldonado et al., 1978; Mount, 1984) or to a n
16 Palenik et aL/In vitro GI inhibition of microbial adherence
initially low material pH (DeSchepper et al., 1989; Scherer et al., 1989; McComb and Ericson, 1987). However, information in this area is limited, both in the number of materials evaluated and in the variety of oral micro-organisms tested. In addition, secondary caries around glass-ionomer restorations has been reported by in vitro studies (Kidd, 1978; Derand and Johansson, 1984). The purpose of this study was to observe the in vitro effects of various glass-ionomer restorative materials and cements on the growth and plaque-forming abilities oforal bacteria believed to be responsible for recurrent caries in humans. Antimicrobial activity, when present, is expected to be related to the amount of fluoride present in a material and to the amount these ions are released into the growth environment. MATERIALS AND METHODS The six glass ionomers used in this study are shown in Table 1. Two were cavity liners, while the remaining four are considered restorative materials. All materials were commercially-available formulations and were prepared in strict compliance with their manufacturers' recommendations. Specimens were prepared with the use of Teflon templates containing circular holes 5.0 mm wide and 2.0 mm deep. A glass slab was placed under a template, and mixed ionomer material was placed into the holes. Glass cover-slips were depressed onto the materials so that a surface flush with the top of the template would be achieved. A curing light was then applied to cure materials when indicated. Set discs were pushed into empty Petri plates with a glass rod. Sterile technique was practiced throughout the experimental protocol where possible, with templates, glass slabs, Petri dishes, glass rods, and other inert supplies sterilized in a steam autoclave between uses. The five bacteria used in this study were obtained from the American Type Culture Collection (Rockville, MD). Each is considered to be a bacterial species involved with both primary and secondary caries formation or plaque formation in humans. The bacteria included Actinomyces viscosus ATCC 19246, Streptococcus mitis ATCC 9811, Streptococcus mutans ATCC 27351, Lactobacillus casei ATCC 7469, and Streptococcus sanguis ATCC 10556. Micro-organisms were grown aerobically from frozen stock cultures in trypticase soy broth supplemented with 0.25% (w/ v) glucose. In all cases, 18-20-hour cultures were used. Cells were harvested by centrifugation (8000 rpm at 4°C) and washed twice with sterile saline. Inocula were prepared by the resuspension of washed cells to pre-determined optical densities which relate to known concentrations of 100,000 CPU per mL.
TABLE 1: GLASS-IONOMERMATERIALS
Materials
Type
Manufacturer
Batch Number
Fuji Cap II
Restorative, encapsulated, machine-mixed
G-C
241181
Ketac Fil
Restorative, encapsulated, machine-mixed
Espe
R329 MD 112488
Ketac Silver
Restorative, encapsulated, silver cermet, machine-mixed
Espe
S025 MD 012589
Fuji Miracle-Mix
Restorative, cement + alloy, hand-mixed
G-C
300191
Vitrabond
Cavity Liner, hand-mixed, light- and chemical-activated
3M
751(]
XR-Glass Ionomer
Cavity Liner, hand-mixed, light- and chemical-activated
Kerr
060889
Because there was a chance of environmental contamination, only selective growth agars were used. Mitis salivarius agar (Gold et al., 1973) was used to culture the streptococci, while CFAT (trypticase soy supplemented with cadmium, fluoride, acriflavin, tellurite, and defibrinated sheep blood) agar (Zylber and Jordan, 1982) was used forA. viscosus and Rogosa SL agar (Rogosa et al., 1951) for L. casei. All plates were incubated anaerobically (85% N°-10% H°-5% CO2) at 37°C for 48 h. S. mutans was cultured for an additional 24 h at 21°C in air. Four sets of experiments were performed, including: direct inhibition of microbial growth, indirect inhibition of growth, measurement of fluoride ion release, and inhibition of bacterial adherence to enamel. For direct inhibition tests, 0.1-mL aliquots ofthe standardized bacterial cultures were spread-plated on each of five plates of the appropriate agars. Then, two specimens of a material were applied to each plate. After incubation, the agar plates were examined for bacterial growth inhibition. When present, the widths of the zones of inhibition were measured in millimeters with a dial caliper. Ten specimens per material type were also used in the indirect growth inhibition tests. Two specimens per plate were applied symmetrically on the surfaces of each type of agar with a sterile forceps. Samples were not placed within 12 mm of a plate's side. Plates were then incubated anaerobically at 37°C for 48 h. The test specimens were removed and discarded. The plates were then inoculated using a spread plating technique with the appropriate test micro'organisms. Agar types, test micro-organisms, incubation parameters, and scoring of inhi-
bition zones were the same as described previously. An attempt was made to correlate growth and/or plaque inhibition with the release of inorganic ions from test specimens. Five test specimens were added individually to tubes containing 2.0 mL of sterile saline (0.85% w/v, pH 7.0). Tubes were then gently oscillated in a 37°C incubator for periods of 1, 8, 24, 72, and 168 h. One mL was removed and mixed with 1.0 mL of total ionic strength adjustor buffer, pH = 5.3 (Orion Research, 1982). The fluoride content of the tubes was then measured with the use of a combination fluoride electrode (Orion Research, Cambridge, MA). The pH of the remaining solution was also determined. Bovine central incisors were used in the bacterial adherence inhibition tests. After the root tip was cleaned, 90% of it was removed, and a small hole was drilled facial-lingually through the remaining root structure. Two cavity preparations, each 5.0 mm in diameter and approximately 2.0 mm deep, were made in the facial enamel surface of each tooth. Preparations were placed on the midlines of the teeth, with one being at least 3 mm from the incisal edge, while the other was at least 3 mm from the cemento-enamel junction. Some teeth were not restored and served as controls. Prepared and control teeth were suspended from rubber stoppers with the use of stainless steel orthodontic wires. The wires were passed through holes drilled in the roots. The tooth-stopper apparatus were then packaged in Nyclave plastic bags (Lorvic Corporation, St. Louis, MO) and steam-sterilized. After sterilization, the teeth were restored with one of the four restorative glass ionomers under aseptic conditions. Five teeth per material (a total of 10
TABLE 2: DIRECT INHIBITION OF BACTERIAL GROWTH Diameter of Inhibition Zone with Each Bacterial Species Materials*
S. mutans
Ketac-Silver
0.00'*
Ketac-Fil
0.00
#
S. mitis
S. sanguis
L. case±
A. viscosus
0.00
0.00
0.00
0.72 ±0.24
0.00
0.00
0.62 _+0.23
0.81 ±0.31
0.19 ±0.25
0.97 ±0.27
0.00 ##
Fuji Cap II
0.89 ±0.40
1.02 ±0.36
XR-Ionomer
1.10 ±0.40
1.20 ±0.07
0.60 _+0.31
0.05 ±0.11
1.14 _+0.24
Vitrabond
1.11 ±0.36
1.48 ±0.27
1.20 ±0.34
0.58 ±0.35
1.26 ±0.25
Miracle Mix
1.55 ±0.40
1.29 ±0.53
1.52 ±0.65
0.85 ±0.42
1.77 ±0.46
* Each group consisted of 10 specimens. ** Mean ±SEM Meanvalues expressed in millimeters. '~ Values connected by verticle lines are not significantly different (p > 0.05). "Vitrabond is significantly different from Fuji Cap II and XR-Ionomer (p < 0.05).
Dental Materials~February 1992 17
TABLE 3: FLUORIDEION RELEASE Fluoride Release(ppm)** Material*
1h
8 hr
24 hr
72 hr
168 hr
Ketac-Silver
6.4 +1.17
11.2 +0.49
14.0 +1.63
30.0 +2.61
46.0 +3.16
Fuji Cap II
42.4 +3.19
62.0 +7.48
64.4 +7.69
85.2 +5.82
119.6 +6.76
Vitrabond
38.4 _+1.47
65.2 +6.50
71.2 + 1.51
139.6 +2.40
156.4 +3.27
Ketac-Fil
23.2_+1.85
24.4 +0.75
28.4 +3.19
171.8 +8.88 I
178.2 +8.84 I
Miracle Mix
149.2 + 3.71
154.0 + 12.3
156.4 + 8.45
162.0 + 9.03
XR-Ionomer
175.0 _+6.23
195.2 +3.83
211.2 _+0.80
216.0 _+7.48
#
I
178.2 + 8.48
I
227.4 _+8.18
* Each group consisted of five specimens. ** Mean +SEM. # Values connected by verticle lines are not significantlydifferent (p > 0.05).
specimens) were prepared. The tooth-stoppers were then individually placed into tubes containing 10 mL of trypticase soy broth supplemented with 2.0% (w/v) sucrose. Each tube was inoculated with 0.1 mL of a 24-hour culture ofS. mutans and incubated aerobically for 24 h at 37°C. The apparatus were then transferred to tubes containing freshly inoculated media. The process was repeated twice. The teeth were then rinsed gently with saline and the adherent material disclosed with a 0.075% (w/v) basic fuchsin solution. The teeth were then rinsed again and scored by use of a modification of a method described by Park and Katz (1974). Coverage scores were "0" (0% coverage), "1" (1-25%), "2" (26-50%), "3" (51-75%), and "4" (76100%). Plaque thickness (stain intensity) was also evaluated. These scores were "0" (none), "1" (thin or light), "2" (medium), and "3" (heavy). The final score for each restoration was determined by multiplication of the coverage score by the intensity score. All quantitative results were subjected to an analysis of variance. If the variance ofthe data was found to be homogeneous by Box's test, the one-way analysis (ANOVA) was performed (Sokal and Rohlf, 1981). Where the variance was not homogeneous, the Welch test was used (Welch, 1951). Determination of significant difference among the groups was made by the Newman-Keuls procedure (Snodgrass, 1977).
be delayed in fluoride release, levels were consistent with the other materials after 72 h. The other three materials produced their highest concentrations after 24 h. Fluoride release levels over 50 ppm were achieved within eight hours by Fuji Cap II, Miracle Mix, Vitrabond, and XR-Ionomer. The four restorative glass ionomers markedly inhibited adherence byS. mutans to the bovine incisors (Table 4). While a comparison of mean adherence between the test materials was not statistically different, all differed from the means of the unrestored controls (p < 0.01).
RESULTS The results of the direct bacterial growth inhibition tests are shown in Table 2. Miracle Mix exhibited the greatest overall antibacterial ability, producing the largest inhibitory zones for four of the bacteria after 48-72 h of exposure. Fuji Cap II, XRIonomer, and Vitrabond exhibited similar inhibitory abilities for all but S. sanguis. Ketac-Fil and Ketac-Silver demonstrated the least inhibition. They inhibited only the growth ofA. viscosus. L. casei was the most resistant bacterial species, whileA, viscosus was inhibited to some extent by all six test materials. Placement of material specimens onto uninoculated plates for 48 h and then removal prior to bacterial plating produced no outwardly extending zones of inhibition. Growth occurred even in areas over which the specimens had been placed. Fluoride release data from all six materials are presented in Table 3. XR-Ionomer had the most rapid ion release and produced the largest amounts of fluoride in each of the five time intervals. Miracle Mix also released fluoride very quickly, but amounts leveled offafter one hour. Ketac-Silver elaborated the least amounts of fluoride. While Ketac-Fil initially appeared to
TABLE 4: INHIBITIONOF Streptococcus mutans BACTERIAL ADHERENCE
18 Palenik et aL/In vitro GI inhibition of microbial adherence
DISCUSSION Laboratory studies have demonstrated that appreciable amounts offluoride can be released from glass ionomers (Forsten, 1990; Swartz et al., 1984; Wilson and Kuhn, 1985; Thornton et al., 1986). Release has been related to the infrequent appearance of recurrent caries when glass ionomers have been used as restorative materials or as cements (Christensen, 1990; Phillips, 1990; Barkhordar et al., 1989; Scherer et al., 1989). Most studies have measured the release of fluoride into deionized water. E1Mallakh and Sarkar (1988) indicated that the use of de-ionized water was not representative of the oral
TO BOVINE ENAMEL Inhibition Values Materials** Adherence Score
% of Control Score % of MaximumScore*
#
11.0
6.7
1.2 +0.20
16.7
10.0
Fuji Cap II
1.6 +0.39
22.2
13.3
Ketac-Silver
2.0 +0.61
27.8
16.7
Control##
7.2 +0.73
100.0
60.0
Ketac-Fil
0.8 +0.25
Miracle Mix
* Maximum plaque score possiblewas 12. ** Restorativegroups contained 10 specimens. # Mean+SEM. Valuesconnected by verticlelinesare not significantlydifferent (p > 0.05) ##Control group containedfive specimens.
condition. They contended that significantly lower amounts of fluoride are released when specimens are placed into saline or artificial saliva. It has also been reported that lowering environmental pH increases fluoride release from glass ionomers (Hamilton and Bowden, 1988; Forsten, 1990; Walls et al., 1988). Fluoride release in this study varied markedly between and among materials. Values were expressed as amounts released over a specific time period, rather than as ppm fluoride released per hour or day. The initial pH of the saline solutions was 7.0. The pH dropped to approximately 6.0 for all materials after 72 h. The pH values remained at 6.0 for the 168-hour sampling. The release levels for individual materials differed from those described in other reports. In most cases, the amounts were less. These differences could be explained by lot variation, by the use of newer material formulations, and by the fact that the decline in the terminal pH was modest. Also, since the saline solutions were not changed, some released fluoride could have re-entered the specimens. XR-Ionomer rapidly released large amounts of fluoride. Vitrabond, which is also considered to be a fluoride-rich cavity liner, demonstrated a slower release pattern. Ketac Silver gave offless than 50 total ppm fluoride. Since the inhibition-of-growth and bacterial-accumulations tests were completed within 72 h, rapidly-releasing materials (such as XR-Ionomer and Miracle Mix) should have been most active. A two-year fluoride release study (Forsten, 1990) indicated great variability among products as to their initial (up to seven days) release levels. However, the differences between materials diminished over time. By eight months, the amounts being released were similar. The clinical effect of an initially large fluoride release is not fully known. Forsten (1990), however, suggests that the presence of large amounts of fluoride immediately after placement of a glass-ionomer restoration could be bacteriostatic or even bactericidal and thus could stop the caries process in dentin unintentionally left behind (e.g., "tunnel" preparations). Forsten (1990) also suggests that the fluoride could help remineralize unaffected inner dentin and demineralized enamel. Fast fluoride-releasers, XR-Ionomer and Miracle Mix, inhibited the growth of all test bacteria. Miracle Mix was most effective for four of the five organisms. However, Fuji Cap II and Vitrabond also performed well. The amount of fluoride required to inhibit growth varies per organism (Hamilton and Bowden, 1988). Streptococci at neutral pH are inhibited by concentrations of 150-200 ppm fluoride. Decreasing environmental pH to 5.5-6.5 lowers the minimum inhibitory concentration of fluoride required by 20-80%. Initially, the pH of the bacterial growth media was between 6.9 and 7.2. Although final agar pH values were not determined, a drop of at least 1.01.5 pH units might be expected at the agar surface. This is because all the bacteria used are acidogenic and because ample amounts of fermentable carbohydrates were present in the media. This would help explain the credible amounts of growth inhibition exhibited by Fuji Cap II and Vitrabond. Either more fluoride was released into the agar because of acidic conditions or the fluoride level required to inhibit growth was less than that produced by Miracle Mix and XR-Ionomer. L. casei was the most fluoride-resistant bacterium. This supports the results of previous studies (Hamilton and Bowden, 1988; Milnes et al., 1985; Maltz and Emilson, 1982). There was no growth inhibition after indirect contact by any of the materials. It could be expected that fluoride would be released from the specimens into the agars over the 48-hour exposure period. However, the inoculation of the plates follow-
ing removal of the specimens did not result in any inhibition zones. Two possible explanations seem appropriate. The pH of the agar was initially near neutrality. No microbial acids were present to allow for extra fluoride release from the specimens. Another possibility is that the fluoride released diffused widely throughout the agars, thus reducing significantly the effective concentration available for inhibitory activity. All four restorative glass ionomers were placed into enamel specimens which were added into a system known to generate significant amounts of S. mutans accumulations. All four ionomers significantly (p < 0.01) reduced bacterial deposition when compared with unrestored controls. There were no significant differences in reduction values when the materials were compared with each other. This result could possibly be explained by one or more of the following factors: First, the terminal pH of the growth tubes approached 5.2 after 24 h of microbial growth. Such an acidic pH would increase fluoride release and could locally affect bacterial growth. Even though Ketac-Silver produced low fluoride release into saline, low environmental pH could enhance release to growth inhibition levels. Second, the concentration of fluoride necessary to inhibit deposition could be less than the higher amounts released into saline by Miracle Mix and XR-Ionomer. Other factors may have also been involved with decreased bacterial accumulation. Ionomers have low pH values for periods of up to seven days after being placed. This could have a negative effect on local bacterial growth. Also, release of silver from Ketac-Silver could have been sufficient to be antibacterial. In summary, the glass ionomers tested all demonstrated an inhibitory effect on growth and/or adherence of oral bacteria thought to be often involved with recurrent dental caries. The extent of bacterial inhibition differed between and among the test materials and among the different bacteria used. Inhibition of bacterial growth could be related to the amounts of fluoride ion released. Reduction in bacterial adherence could be due to the release of fluoride or silver or to an initially low material pH. This investigation was supported in part by USPHS Research Grant DE0712506 (Professional Student Short-term Research Training Program) from the National Institute of Dental Research, Bethesda, MD 20892.
Received August 6, 1990/Accepted September 6, 1991 Address correspondence and reprint requests to: C. J. Palenik Department of Oral Microbiology Indiana University School of Dentistry 1121 W. Michigan St. Indianapolis, IN 46202 USA
REFERENCES Barkhordar RA, Kempler D, Pelzner RRB, Stark MM (1989). Technical note: antimicrobial action ofglass-ionomer lining cement onS. sanguis andS. mutans. DentMater 5:281-282. Br~innstr6m M (1984). Smear layer: pathological and treatment considerations. Oper Dent (Suppl) 3:35-42. Br~nnstr6m M, Nordenvall KJ (1978). Bacterial penetration, pulpal reaction and the inner surface of concise enamel bond composite fillings in etched and unetched cavities. J Dent Res 57:3-10. Br~innstr6m M, Vojinovic O (1976). Response of the dental pulp to invasion of bacteria around three filling materials. JDent Child 43:83-89.
Dental Materials~February 1992 19
Christensen GJ (1990). Glass ionomer as a luting material. J Am Dent Assoc 120:59-62. DeSchepper EJ, White RR, von der Lehr W (1989). Antibacterial effects of glass ionomers. Am JDent 2:51-56. D~rand T, Johansson B (1984). Experimental secondary caries around restorations in roots. Caries Res 18:548-554. E1 Mallakh B, Sarkar NK (1988). Fluoride ion release from glass ionomer cements in deionized water and artificial saliva (abstract). JDent Res 67:197. Forsten L (1977). Fluoride release from a glass ionomer cement. Scand J Dent Res 85:503-504. Forsten L (1990). Short- and long-term fluoride release from glass ionomers and other fluoride-containing filling materials in vitro. Scand J Dent Res 98:179-185. Gold OG, Jordan HV, van Houte JV (1973). A selective medium for Streptococcus mutans. Arch Oral Biol 18:1357-1364. Hamilton I, Bowden G (1988). Effect of fluoride on oral microorganisms. In: Ekstrand J, Fejerskov O, Silverstone LM, editors. Fluoride in dentistry. Copenhagen: Munksgaard, 77-103. Kidd EAM (1978). Cavity sealing ability of composite and glass ionomer cement restorations. BrDent J 144:139-142. Lavelle CLB (1976). A cross-sectional longitudinal study into the durability of amalgam restorations. JDent Res 55:139143. Maldonado A, Swartz ML, Phillips RW (1978). An in vitro study of certain properties of a glass ionomer cement. J A m Dent Assoc 96:785-791. Maltz M, Emilson CG (1982). Susceptibility of oral bacteria to various fluoride salts. J Dent Res 61:786-790. McComb D, Ericson D (1987). Antimicrobial action of new, proprietary lining cements. JDentRes 66:1025-1028. Milnes AR, Bowden GH, Hamilton IR (1985). Effect ofNaF and pH on the growth and glycolytic rate of recently isolated strains of oral lactobacillus species. JDentRes 64:401-404. Mount GJ (1984). Glass ionomer cements: clinical considerations. Clin Dent 4:4-5.
20 Palenik et aL/In vitro GI inhibition of microbial adherence
Orion Research (1982). Instrument manual for fluoride electrodes. Cambridge, MA: Orion Research Corporation, 2-4. Park KC, Katz S (1974). In vivo plaque measurement in the rat. J Dent Res 53:1346-1349. Phillips RW (1990). The glass ionomer cement. J Am Dent Assoc 120:19. Rogosa M, Mitchell JA, Weiseman RF (1951). A selective medium for the isolation and enumeration of oral and fecal lactobacilli. JBacteriol 62:132-133. Scherer W, Lippman N, Kaim J (1989). Antimicrobial properties of glass-ionomer cements and other restorative materials. Oper Dent 14:77-81. Schwartzman B, Caputo AA, Schein B (1980). Antimicrobial action of dental cements. J Prosthet Dent 43:309-312. Snodgrass JO (1977). The number game. Statistics for psychology. Baltimore: The Williams and Wilkins Co., 318-324. Sokal RR, Rohlf FJ (1981). Biometry. The principles and practice of statistics in biological research. 2nd ed. New York: W.H. Freeman and Co., 402-412. Swartz ML, Phillips RW, Clark HE (1984). Long-term F release from glass ionomer cements. JDentRes 63:158-160. Thornton JB, Retief DH, Bradley EL (1986). Fluoride release from and tensile bond strength of ketac-fil and ketac-silver to enamel and dentin. Dent Mater 2:241-245. Walls AWG, McCabe JF, Murray JJ (1988). The effect of the variation in pH of the eroding solution upon the erosion resistance of glass polyalkenoate (ionomer) cements. Dent Mater 4:141-144. Welch BL (1951). On the comparison of sequentially measured values: an alternative approach. Biometrika 38:330-336. Wilson AD, Kuhn AT (1985). The release of fluoride and other chemical species from a glass-ionomer cement. Biomaterials 6:431-433. Zylber LJ, Jordan NV (1982). Development of a selective medium for detection and enumerations of Actinomyces viscosus andActinomyces naeslundii in dental plaque. J Clin Microbiol 15:253-259.
Inhibition of microbial adherence and growth by various glass ionomers in vitro C.J. Palenik ~, M.J. Behnen 1, J.C. Setcos 2,3,C.H. Miller 1 Department of Oral Microbiology, 2Department of Dental Materials Indiana Univ. School of Dentistry, Indianapolis IN USA 3Now at Department of Restorative Dentistry, University of California, San Francisco, School of Dentistry San Francisco CA USA
Abstract. This study measured the in vitro inhibition of growth and adherence of five oral bacteria by glass-ionomer materials. Disks were prepared from two cavity liners and four restorative class materials, by use of Teflon plates with circular wells, five mm wide and two mm deep. The bacterial species tested included: A. viscosus, S. mitis, S. mutans, L. casei, and S. sanguis. Growth inhibition studies were performed by the spreading of 0.1 mL of standardized inocula over agar plates produced with selective media, followed by the direct application of glass-ionomer disks onto the agar. On other plates, disks were placed onto uninoculated agars for 48 h, followed by bacterial inoculation. All agar plates were incubated under optimal growth conditions for each bacterial species. The four restorative materials were also placed aseptically into sterilized bovine incisors and placed into sucrose containing broth media, inoculated with S. mutans for three days. Adhering materials were disclosed and scored. An ion-exchange electrode was used to measure fluoride release over a seven-day period for all six glass ionomers. The two cavity liners and two of the restorative materials produced the largest growth inhibition zones by direct contact. No growth inhibition occurred when the specimens were allowed to come into contact with the agars prior to inoculation. All four restorative materials reduced bacterial accumulations on enamel surfaces by over 80%. Elevations in short-term fluoride release levels were positively correlated with growth inhibition. Recurrent human dental caries has been associated with the deterioration of dental restorative materials. Breakdown in marginal areas between cavity preparations and restorative materials can provide potential pathways for re-infection. Cariogenic micro-organisms present in the normal human flora could easily penetrate underlying dentin through such defects (Br~innstrSm, 1984; Br~innstrSm and Nordenvall, 1978; Br~innstrSm and Vojinovic, 1976; Lavelle, 1976). Reducing or preferably preventing such marginal breakdown could reduce the chances of recurrent caries. Obviously, physically superior restorative materials are important in the prevention of recurrent caries. However, it would also be advantageous if such materials possessed frank antimicrobial potentials. Published data (Barkhordar et al., 1989; DeSchepper et al., 1989; McComb and Ericson, 1987; Scherer et al., 1989; Schwartzman et al., 1980) have indicated that some commercially available glass-ionomer restorative materials and cements have demonstrable in vitro antimicrobial abilities. Such abilities have been related to the release of high concentrations of fluoride ions (DeSchepper et al., 1989; Forsten, 1977; Maldonado et al., 1978; Mount, 1984) or to a n
16 Palenik et aL/In vitro GI inhibition of microbial adherence
initially low material pH (DeSchepper et al., 1989; Scherer et al., 1989; McComb and Ericson, 1987). However, information in this area is limited, both in the number of materials evaluated and in the variety of oral micro-organisms tested. In addition, secondary caries around glass-ionomer restorations has been reported by in vitro studies (Kidd, 1978; Derand and Johansson, 1984). The purpose of this study was to observe the in vitro effects of various glass-ionomer restorative materials and cements on the growth and plaque-forming abilities oforal bacteria believed to be responsible for recurrent caries in humans. Antimicrobial activity, when present, is expected to be related to the amount of fluoride present in a material and to the amount these ions are released into the growth environment. MATERIALS AND METHODS The six glass ionomers used in this study are shown in Table 1. Two were cavity liners, while the remaining four are considered restorative materials. All materials were commercially-available formulations and were prepared in strict compliance with their manufacturers' recommendations. Specimens were prepared with the use of Teflon templates containing circular holes 5.0 mm wide and 2.0 mm deep. A glass slab was placed under a template, and mixed ionomer material was placed into the holes. Glass cover-slips were depressed onto the materials so that a surface flush with the top of the template would be achieved. A curing light was then applied to cure materials when indicated. Set discs were pushed into empty Petri plates with a glass rod. Sterile technique was practiced throughout the experimental protocol where possible, with templates, glass slabs, Petri dishes, glass rods, and other inert supplies sterilized in a steam autoclave between uses. The five bacteria used in this study were obtained from the American Type Culture Collection (Rockville, MD). Each is considered to be a bacterial species involved with both primary and secondary caries formation or plaque formation in humans. The bacteria included Actinomyces viscosus ATCC 19246, Streptococcus mitis ATCC 9811, Streptococcus mutans ATCC 27351, Lactobacillus casei ATCC 7469, and Streptococcus sanguis ATCC 10556. Micro-organisms were grown aerobically from frozen stock cultures in trypticase soy broth supplemented with 0.25% (w/ v) glucose. In all cases, 18-20-hour cultures were used. Cells were harvested by centrifugation (8000 rpm at 4°C) and washed twice with sterile saline. Inocula were prepared by the resuspension of washed cells to pre-determined optical densities which relate to known concentrations of 100,000 CPU per mL.
TABLE 1: GLASS-IONOMERMATERIALS
Materials
Type
Manufacturer
Batch Number
Fuji Cap II
Restorative, encapsulated, machine-mixed
G-C
241181
Ketac Fil
Restorative, encapsulated, machine-mixed
Espe
R329 MD 112488
Ketac Silver
Restorative, encapsulated, silver cermet, machine-mixed
Espe
S025 MD 012589
Fuji Miracle-Mix
Restorative, cement + alloy, hand-mixed
G-C
300191
Vitrabond
Cavity Liner, hand-mixed, light- and chemical-activated
3M
751(]
XR-Glass Ionomer
Cavity Liner, hand-mixed, light- and chemical-activated
Kerr
060889
Because there was a chance of environmental contamination, only selective growth agars were used. Mitis salivarius agar (Gold et al., 1973) was used to culture the streptococci, while CFAT (trypticase soy supplemented with cadmium, fluoride, acriflavin, tellurite, and defibrinated sheep blood) agar (Zylber and Jordan, 1982) was used forA. viscosus and Rogosa SL agar (Rogosa et al., 1951) for L. casei. All plates were incubated anaerobically (85% N°-10% H°-5% CO2) at 37°C for 48 h. S. mutans was cultured for an additional 24 h at 21°C in air. Four sets of experiments were performed, including: direct inhibition of microbial growth, indirect inhibition of growth, measurement of fluoride ion release, and inhibition of bacterial adherence to enamel. For direct inhibition tests, 0.1-mL aliquots ofthe standardized bacterial cultures were spread-plated on each of five plates of the appropriate agars. Then, two specimens of a material were applied to each plate. After incubation, the agar plates were examined for bacterial growth inhibition. When present, the widths of the zones of inhibition were measured in millimeters with a dial caliper. Ten specimens per material type were also used in the indirect growth inhibition tests. Two specimens per plate were applied symmetrically on the surfaces of each type of agar with a sterile forceps. Samples were not placed within 12 mm of a plate's side. Plates were then incubated anaerobically at 37°C for 48 h. The test specimens were removed and discarded. The plates were then inoculated using a spread plating technique with the appropriate test micro'organisms. Agar types, test micro-organisms, incubation parameters, and scoring of inhi-
bition zones were the same as described previously. An attempt was made to correlate growth and/or plaque inhibition with the release of inorganic ions from test specimens. Five test specimens were added individually to tubes containing 2.0 mL of sterile saline (0.85% w/v, pH 7.0). Tubes were then gently oscillated in a 37°C incubator for periods of 1, 8, 24, 72, and 168 h. One mL was removed and mixed with 1.0 mL of total ionic strength adjustor buffer, pH = 5.3 (Orion Research, 1982). The fluoride content of the tubes was then measured with the use of a combination fluoride electrode (Orion Research, Cambridge, MA). The pH of the remaining solution was also determined. Bovine central incisors were used in the bacterial adherence inhibition tests. After the root tip was cleaned, 90% of it was removed, and a small hole was drilled facial-lingually through the remaining root structure. Two cavity preparations, each 5.0 mm in diameter and approximately 2.0 mm deep, were made in the facial enamel surface of each tooth. Preparations were placed on the midlines of the teeth, with one being at least 3 mm from the incisal edge, while the other was at least 3 mm from the cemento-enamel junction. Some teeth were not restored and served as controls. Prepared and control teeth were suspended from rubber stoppers with the use of stainless steel orthodontic wires. The wires were passed through holes drilled in the roots. The tooth-stopper apparatus were then packaged in Nyclave plastic bags (Lorvic Corporation, St. Louis, MO) and steam-sterilized. After sterilization, the teeth were restored with one of the four restorative glass ionomers under aseptic conditions. Five teeth per material (a total of 10
TABLE 2: DIRECT INHIBITION OF BACTERIAL GROWTH Diameter of Inhibition Zone with Each Bacterial Species Materials*
S. mutans
Ketac-Silver
0.00'*
Ketac-Fil
0.00
#
S. mitis
S. sanguis
L. case±
A. viscosus
0.00
0.00
0.00
0.72 ±0.24
0.00
0.00
0.62 _+0.23
0.81 ±0.31
0.19 ±0.25
0.97 ±0.27
0.00 ##
Fuji Cap II
0.89 ±0.40
1.02 ±0.36
XR-Ionomer
1.10 ±0.40
1.20 ±0.07
0.60 _+0.31
0.05 ±0.11
1.14 _+0.24
Vitrabond
1.11 ±0.36
1.48 ±0.27
1.20 ±0.34
0.58 ±0.35
1.26 ±0.25
Miracle Mix
1.55 ±0.40
1.29 ±0.53
1.52 ±0.65
0.85 ±0.42
1.77 ±0.46
* Each group consisted of 10 specimens. ** Mean ±SEM Meanvalues expressed in millimeters. '~ Values connected by verticle lines are not significantly different (p > 0.05). "Vitrabond is significantly different from Fuji Cap II and XR-Ionomer (p < 0.05).
Dental Materials~February 1992 17
TABLE 3: FLUORIDEION RELEASE Fluoride Release(ppm)** Material*
1h
8 hr
24 hr
72 hr
168 hr
Ketac-Silver
6.4 +1.17
11.2 +0.49
14.0 +1.63
30.0 +2.61
46.0 +3.16
Fuji Cap II
42.4 +3.19
62.0 +7.48
64.4 +7.69
85.2 +5.82
119.6 +6.76
Vitrabond
38.4 _+1.47
65.2 +6.50
71.2 + 1.51
139.6 +2.40
156.4 +3.27
Ketac-Fil
23.2_+1.85
24.4 +0.75
28.4 +3.19
171.8 +8.88 I
178.2 +8.84 I
Miracle Mix
149.2 + 3.71
154.0 + 12.3
156.4 + 8.45
162.0 + 9.03
XR-Ionomer
175.0 _+6.23
195.2 +3.83
211.2 _+0.80
216.0 _+7.48
#
I
178.2 + 8.48
I
227.4 _+8.18
* Each group consisted of five specimens. ** Mean +SEM. # Values connected by verticle lines are not significantlydifferent (p > 0.05).
specimens) were prepared. The tooth-stoppers were then individually placed into tubes containing 10 mL of trypticase soy broth supplemented with 2.0% (w/v) sucrose. Each tube was inoculated with 0.1 mL of a 24-hour culture ofS. mutans and incubated aerobically for 24 h at 37°C. The apparatus were then transferred to tubes containing freshly inoculated media. The process was repeated twice. The teeth were then rinsed gently with saline and the adherent material disclosed with a 0.075% (w/v) basic fuchsin solution. The teeth were then rinsed again and scored by use of a modification of a method described by Park and Katz (1974). Coverage scores were "0" (0% coverage), "1" (1-25%), "2" (26-50%), "3" (51-75%), and "4" (76100%). Plaque thickness (stain intensity) was also evaluated. These scores were "0" (none), "1" (thin or light), "2" (medium), and "3" (heavy). The final score for each restoration was determined by multiplication of the coverage score by the intensity score. All quantitative results were subjected to an analysis of variance. If the variance ofthe data was found to be homogeneous by Box's test, the one-way analysis (ANOVA) was performed (Sokal and Rohlf, 1981). Where the variance was not homogeneous, the Welch test was used (Welch, 1951). Determination of significant difference among the groups was made by the Newman-Keuls procedure (Snodgrass, 1977).
be delayed in fluoride release, levels were consistent with the other materials after 72 h. The other three materials produced their highest concentrations after 24 h. Fluoride release levels over 50 ppm were achieved within eight hours by Fuji Cap II, Miracle Mix, Vitrabond, and XR-Ionomer. The four restorative glass ionomers markedly inhibited adherence byS. mutans to the bovine incisors (Table 4). While a comparison of mean adherence between the test materials was not statistically different, all differed from the means of the unrestored controls (p < 0.01).
RESULTS The results of the direct bacterial growth inhibition tests are shown in Table 2. Miracle Mix exhibited the greatest overall antibacterial ability, producing the largest inhibitory zones for four of the bacteria after 48-72 h of exposure. Fuji Cap II, XRIonomer, and Vitrabond exhibited similar inhibitory abilities for all but S. sanguis. Ketac-Fil and Ketac-Silver demonstrated the least inhibition. They inhibited only the growth ofA. viscosus. L. casei was the most resistant bacterial species, whileA, viscosus was inhibited to some extent by all six test materials. Placement of material specimens onto uninoculated plates for 48 h and then removal prior to bacterial plating produced no outwardly extending zones of inhibition. Growth occurred even in areas over which the specimens had been placed. Fluoride release data from all six materials are presented in Table 3. XR-Ionomer had the most rapid ion release and produced the largest amounts of fluoride in each of the five time intervals. Miracle Mix also released fluoride very quickly, but amounts leveled offafter one hour. Ketac-Silver elaborated the least amounts of fluoride. While Ketac-Fil initially appeared to
TABLE 4: INHIBITIONOF Streptococcus mutans BACTERIAL ADHERENCE
18 Palenik et aL/In vitro GI inhibition of microbial adherence
DISCUSSION Laboratory studies have demonstrated that appreciable amounts offluoride can be released from glass ionomers (Forsten, 1990; Swartz et al., 1984; Wilson and Kuhn, 1985; Thornton et al., 1986). Release has been related to the infrequent appearance of recurrent caries when glass ionomers have been used as restorative materials or as cements (Christensen, 1990; Phillips, 1990; Barkhordar et al., 1989; Scherer et al., 1989). Most studies have measured the release of fluoride into deionized water. E1Mallakh and Sarkar (1988) indicated that the use of de-ionized water was not representative of the oral
TO BOVINE ENAMEL Inhibition Values Materials** Adherence Score
% of Control Score % of MaximumScore*
#
11.0
6.7
1.2 +0.20
16.7
10.0
Fuji Cap II
1.6 +0.39
22.2
13.3
Ketac-Silver
2.0 +0.61
27.8
16.7
Control##
7.2 +0.73
100.0
60.0
Ketac-Fil
0.8 +0.25
Miracle Mix
* Maximum plaque score possiblewas 12. ** Restorativegroups contained 10 specimens. # Mean+SEM. Valuesconnected by verticlelinesare not significantlydifferent (p > 0.05) ##Control group containedfive specimens.
condition. They contended that significantly lower amounts of fluoride are released when specimens are placed into saline or artificial saliva. It has also been reported that lowering environmental pH increases fluoride release from glass ionomers (Hamilton and Bowden, 1988; Forsten, 1990; Walls et al., 1988). Fluoride release in this study varied markedly between and among materials. Values were expressed as amounts released over a specific time period, rather than as ppm fluoride released per hour or day. The initial pH of the saline solutions was 7.0. The pH dropped to approximately 6.0 for all materials after 72 h. The pH values remained at 6.0 for the 168-hour sampling. The release levels for individual materials differed from those described in other reports. In most cases, the amounts were less. These differences could be explained by lot variation, by the use of newer material formulations, and by the fact that the decline in the terminal pH was modest. Also, since the saline solutions were not changed, some released fluoride could have re-entered the specimens. XR-Ionomer rapidly released large amounts of fluoride. Vitrabond, which is also considered to be a fluoride-rich cavity liner, demonstrated a slower release pattern. Ketac Silver gave offless than 50 total ppm fluoride. Since the inhibition-of-growth and bacterial-accumulations tests were completed within 72 h, rapidly-releasing materials (such as XR-Ionomer and Miracle Mix) should have been most active. A two-year fluoride release study (Forsten, 1990) indicated great variability among products as to their initial (up to seven days) release levels. However, the differences between materials diminished over time. By eight months, the amounts being released were similar. The clinical effect of an initially large fluoride release is not fully known. Forsten (1990), however, suggests that the presence of large amounts of fluoride immediately after placement of a glass-ionomer restoration could be bacteriostatic or even bactericidal and thus could stop the caries process in dentin unintentionally left behind (e.g., "tunnel" preparations). Forsten (1990) also suggests that the fluoride could help remineralize unaffected inner dentin and demineralized enamel. Fast fluoride-releasers, XR-Ionomer and Miracle Mix, inhibited the growth of all test bacteria. Miracle Mix was most effective for four of the five organisms. However, Fuji Cap II and Vitrabond also performed well. The amount of fluoride required to inhibit growth varies per organism (Hamilton and Bowden, 1988). Streptococci at neutral pH are inhibited by concentrations of 150-200 ppm fluoride. Decreasing environmental pH to 5.5-6.5 lowers the minimum inhibitory concentration of fluoride required by 20-80%. Initially, the pH of the bacterial growth media was between 6.9 and 7.2. Although final agar pH values were not determined, a drop of at least 1.01.5 pH units might be expected at the agar surface. This is because all the bacteria used are acidogenic and because ample amounts of fermentable carbohydrates were present in the media. This would help explain the credible amounts of growth inhibition exhibited by Fuji Cap II and Vitrabond. Either more fluoride was released into the agar because of acidic conditions or the fluoride level required to inhibit growth was less than that produced by Miracle Mix and XR-Ionomer. L. casei was the most fluoride-resistant bacterium. This supports the results of previous studies (Hamilton and Bowden, 1988; Milnes et al., 1985; Maltz and Emilson, 1982). There was no growth inhibition after indirect contact by any of the materials. It could be expected that fluoride would be released from the specimens into the agars over the 48-hour exposure period. However, the inoculation of the plates follow-
ing removal of the specimens did not result in any inhibition zones. Two possible explanations seem appropriate. The pH of the agar was initially near neutrality. No microbial acids were present to allow for extra fluoride release from the specimens. Another possibility is that the fluoride released diffused widely throughout the agars, thus reducing significantly the effective concentration available for inhibitory activity. All four restorative glass ionomers were placed into enamel specimens which were added into a system known to generate significant amounts of S. mutans accumulations. All four ionomers significantly (p < 0.01) reduced bacterial deposition when compared with unrestored controls. There were no significant differences in reduction values when the materials were compared with each other. This result could possibly be explained by one or more of the following factors: First, the terminal pH of the growth tubes approached 5.2 after 24 h of microbial growth. Such an acidic pH would increase fluoride release and could locally affect bacterial growth. Even though Ketac-Silver produced low fluoride release into saline, low environmental pH could enhance release to growth inhibition levels. Second, the concentration of fluoride necessary to inhibit deposition could be less than the higher amounts released into saline by Miracle Mix and XR-Ionomer. Other factors may have also been involved with decreased bacterial accumulation. Ionomers have low pH values for periods of up to seven days after being placed. This could have a negative effect on local bacterial growth. Also, release of silver from Ketac-Silver could have been sufficient to be antibacterial. In summary, the glass ionomers tested all demonstrated an inhibitory effect on growth and/or adherence of oral bacteria thought to be often involved with recurrent dental caries. The extent of bacterial inhibition differed between and among the test materials and among the different bacteria used. Inhibition of bacterial growth could be related to the amounts of fluoride ion released. Reduction in bacterial adherence could be due to the release of fluoride or silver or to an initially low material pH. This investigation was supported in part by USPHS Research Grant DE0712506 (Professional Student Short-term Research Training Program) from the National Institute of Dental Research, Bethesda, MD 20892.
Received August 6, 1990/Accepted September 6, 1991 Address correspondence and reprint requests to: C. J. Palenik Department of Oral Microbiology Indiana University School of Dentistry 1121 W. Michigan St. Indianapolis, IN 46202 USA
REFERENCES Barkhordar RA, Kempler D, Pelzner RRB, Stark MM (1989). Technical note: antimicrobial action ofglass-ionomer lining cement onS. sanguis andS. mutans. DentMater 5:281-282. Br~innstr6m M (1984). Smear layer: pathological and treatment considerations. Oper Dent (Suppl) 3:35-42. Br~nnstr6m M, Nordenvall KJ (1978). Bacterial penetration, pulpal reaction and the inner surface of concise enamel bond composite fillings in etched and unetched cavities. J Dent Res 57:3-10. Br~innstr6m M, Vojinovic O (1976). Response of the dental pulp to invasion of bacteria around three filling materials. JDent Child 43:83-89.
Dental Materials~February 1992 19
Christensen GJ (1990). Glass ionomer as a luting material. J Am Dent Assoc 120:59-62. DeSchepper EJ, White RR, von der Lehr W (1989). Antibacterial effects of glass ionomers. Am JDent 2:51-56. D~rand T, Johansson B (1984). Experimental secondary caries around restorations in roots. Caries Res 18:548-554. E1 Mallakh B, Sarkar NK (1988). Fluoride ion release from glass ionomer cements in deionized water and artificial saliva (abstract). JDent Res 67:197. Forsten L (1977). Fluoride release from a glass ionomer cement. Scand J Dent Res 85:503-504. Forsten L (1990). Short- and long-term fluoride release from glass ionomers and other fluoride-containing filling materials in vitro. Scand J Dent Res 98:179-185. Gold OG, Jordan HV, van Houte JV (1973). A selective medium for Streptococcus mutans. Arch Oral Biol 18:1357-1364. Hamilton I, Bowden G (1988). Effect of fluoride on oral microorganisms. In: Ekstrand J, Fejerskov O, Silverstone LM, editors. Fluoride in dentistry. Copenhagen: Munksgaard, 77-103. Kidd EAM (1978). Cavity sealing ability of composite and glass ionomer cement restorations. BrDent J 144:139-142. Lavelle CLB (1976). A cross-sectional longitudinal study into the durability of amalgam restorations. JDent Res 55:139143. Maldonado A, Swartz ML, Phillips RW (1978). An in vitro study of certain properties of a glass ionomer cement. J A m Dent Assoc 96:785-791. Maltz M, Emilson CG (1982). Susceptibility of oral bacteria to various fluoride salts. J Dent Res 61:786-790. McComb D, Ericson D (1987). Antimicrobial action of new, proprietary lining cements. JDentRes 66:1025-1028. Milnes AR, Bowden GH, Hamilton IR (1985). Effect ofNaF and pH on the growth and glycolytic rate of recently isolated strains of oral lactobacillus species. JDentRes 64:401-404. Mount GJ (1984). Glass ionomer cements: clinical considerations. Clin Dent 4:4-5.
20 Palenik et aL/In vitro GI inhibition of microbial adherence
Orion Research (1982). Instrument manual for fluoride electrodes. Cambridge, MA: Orion Research Corporation, 2-4. Park KC, Katz S (1974). In vivo plaque measurement in the rat. J Dent Res 53:1346-1349. Phillips RW (1990). The glass ionomer cement. J Am Dent Assoc 120:19. Rogosa M, Mitchell JA, Weiseman RF (1951). A selective medium for the isolation and enumeration of oral and fecal lactobacilli. JBacteriol 62:132-133. Scherer W, Lippman N, Kaim J (1989). Antimicrobial properties of glass-ionomer cements and other restorative materials. Oper Dent 14:77-81. Schwartzman B, Caputo AA, Schein B (1980). Antimicrobial action of dental cements. J Prosthet Dent 43:309-312. Snodgrass JO (1977). The number game. Statistics for psychology. Baltimore: The Williams and Wilkins Co., 318-324. Sokal RR, Rohlf FJ (1981). Biometry. The principles and practice of statistics in biological research. 2nd ed. New York: W.H. Freeman and Co., 402-412. Swartz ML, Phillips RW, Clark HE (1984). Long-term F release from glass ionomer cements. JDentRes 63:158-160. Thornton JB, Retief DH, Bradley EL (1986). Fluoride release from and tensile bond strength of ketac-fil and ketac-silver to enamel and dentin. Dent Mater 2:241-245. Walls AWG, McCabe JF, Murray JJ (1988). The effect of the variation in pH of the eroding solution upon the erosion resistance of glass polyalkenoate (ionomer) cements. Dent Mater 4:141-144. Welch BL (1951). On the comparison of sequentially measured values: an alternative approach. Biometrika 38:330-336. Wilson AD, Kuhn AT (1985). The release of fluoride and other chemical species from a glass-ionomer cement. Biomaterials 6:431-433. Zylber LJ, Jordan NV (1982). Development of a selective medium for detection and enumerations of Actinomyces viscosus andActinomyces naeslundii in dental plaque. J Clin Microbiol 15:253-259.
