Naturally Transmitted between Rats and Consequent Caries Inhibition by Streptococcus salivarius TOVE-R Infection

Transcription

1 INFECTION AND IMMUNITY, JUlY 1985, p /85/ $02.00/0 Copyright 1985, American Society for Microbiology Vol. 49, No. 1 Inhibition of Ecological Emergence of Mutans Streptococci Naturally Transmitted between Rats and Consequent Caries Inhibition by Streptococcus salivarius TOVE-R Infection JASON M. TANZER, 2* ANDREA B. KURASZ,' AND JONATHAN CLIVE3 Departments of Oral Diagnosis,' Laboratory Medicine,2 and Behavioral Science and Community Health,3 University of Connecticut Health Center, Farmington, Connecticut Received 28 January 1985/Accepted 12 April 1985 The ability of Streptococcus salivarius strain TOVE-R to inhibit the ecological emergence of virulent representatives of the most prevalent human mutans streptococci on the teeth of specific pathogen-free Osborne-Mendel rats was studied. Rats which were infected by TOVE-R, or either S. mutans 10449S or S. sobrinus WT, or uninfected were transiently co-caged so as to allow natural fecal transfer of organisms due to coprophagy. The infectants were differentially recovered from swabs of the teeth over the time course of the experiments apd from sonified teeth at termination. Data were expressed on both relative (percentage) and absolute (CFU) bases. Initial oral colonization of rats by TOVE-R inhibited the ecological emergence of fecally transmitted S. mutans and S. sobninus WT. There was a generally inverse relationship between the percentages and absolute numbers of TOVE-R and the mutans streptococci on the teeth, which strongly suggested their competition for tooth sites. Absolute numbers of total recoverable flora from the teeth upon sonification were correlated with caries scores, thus suggesting that total recoverable flora counts substantially reflect cavitation status. TOVE-R itself induced no apparent caries activity and its transmission to rats already infected by 10449S or its colonization of rats before 10449S infection inhibited caries induction by this S. mutans strain; similar anticaries effects were not statistically significant for TOVE-R against WT in these experiments. These data on the inhibition of the ecological emergence of the mutans streptococci supplement the already reported ability of TOVE-R to preempt initial colonization of teeth and partially displace the colonization of teeth by the mutans streptococci. Streptococcus salivarius strain TOVE-R, like the mutans group of streptococci, synthesizes alkali-soluble, cell surfaceassociated glucans from sucrose and sticks to solid surfaces in vitro and teeth in vivo (13, 27, 29, 32, 34). Unlike the mutans streptococci, however, it is not detectably cariogenic in rats (29, 34). TOVE-R-colonized rats are difficult to colonize later by virulent S. mutans strain 10449S (29). Remarkably, when rats are first colonized by 10449S or the virulent S. sobrinus strain WT, TOVE-R inoculations partially displace those mutans streptococci and reduce caries (34). We report here that the initial colonization of rats by TOVE-R inhibits the ecological emergence of natural, fecally transmitted strains 10449S and WT from other rats. Initial colonization of rats by TOVE-R thereby inhibits caries associated with mutans infection. MATERIALS AND METHODS Microorganisms. Streptomycin-resistant, rough-colony S. salivarius strain TOVE-R (29, 34), S. mutans NCTC 10449S (27, 33), and S. sobrinus strain WT (formerly S. mutans serotype dig [4, 6, 23, 27, 32]) were studied. Strains were maintained in a lyophilized state until tested in laboratory animals, and inocula were cultured in fluid thioglycolate medium (Difco Laboratories, Detroit, Mich.) supplemented with 20% (vol/vol) meat extract (Difco) and excess CaCO3. The high virulence of strains 10449S and WT in * Corresponding author. 76 rodents has been described (27, 29, 32-34); both were originally isolated from humans and represent the mutans streptococcal species most frequently isolated from humans (4, 23, 27). Animals, diets, inoculations, and experimental design. SPFOM (specific pathogen-free Osborne-Mendel) rat breeders colonized by normal rodent gut flora as previously described (27, 31, 32) were maintained in a laminar flow hood with bacteriological air filters (Portable Containment System, PCS-80; Hazelton Systems, Inc., Aberdeen, Md.) and provided with autoclaved chow no (Ralston Purina Co., St. Louis, Mo.) and sterile demineralized water. The colony was maintained free of mutans and salivarius streptococcal infection. After weaning at 20 days of age, rats were removed from the laminar flow hood, and several litters were pooled in a large stainless steel cage and provided with caries test diet 2000 (Zeigler Brothers, Inc., Gardeners, Pa.) containing 56% sucrose (16) and sterile demineralized water ad libitum. One day later the rats were randomly distributed, one per cage, and inoculated with test organisms at various times as described in the protocols below. The design of the experiments is given in Fig. 1. Rats were simultaneously inoculated orally with a micropipette; each animal was either inoculated two times with about 6 x 108 cells (per inoculation) of the mutans streptococci (10449S or WT) or TOVE-R grown in the thioglycolate medium or left uninoculated. After suitable cultures were taken (see below) to confirm that the rat teeth were heavily colonized by the inoculants, mutans-infected rats were transiently co-caged with TOVE- R-infected rats, uninfected rats, or other mutans-infected

2 VOL. 49, 1985 INHIBITION OF MUTANS STREPTOCOCCI BY TOVE-R S EXPERIMENT o 0 Co0 50 ( c 09 0 & 10 cc co o0 C cn 0 M0 a l CO CO Uninfected Uninfected/Uninfected Uninfected/mutans mutansauninfected mutans Streptococcus Tove-R mutans/mutans mutans/tove-r Tove-R/mutans Tove-RlTove-R 1 Ii 5 6 l 0 _ o o? C ir c S +0 a Iit Co CD 1.C 3:1 Q: WT EXPERIMENT FIG. 1. Experimental design for the study of effects of TOVE-R on the ecological emergence of S. mutans 10449S and S. sobrinus WT after natural fecal transmission of infection between rats. The calendars at the top (10449S trial) and bottom ( WT trial) of the figure begin with the date of initial inoculation by the mutans streptococci and TOVE-R. The arrows indicate the subsequent dates of events, including dates oral cultures were taken, dates of co-caging and separation of animals, and dates of sacrifice by the procedures described in the text. The seven animal groups, based on initial inoculation, transient co-caging, and separation of cage-mates, are depicted by the shading and labeling of horizontal bars. rats. Similarly, TOVE-R-infected rats were transiently cocaged with mutans-infected rats or other TOVE-R-infected rats; uninfected rats were co-caged with other uninfected rats or with mutans-infected rats. By this method seven groups of rats (with 9 or 10 animals per group) were formed and given the following designations which describe the initial colonizing organism of the rat and the colonizing organism of the transient cage-mate: mutans-infected/ TOVE-R-infected, mutans-infected/uninfected, mutansinfected/mutans-infected, TOVE-R-infected/mutansinfected, TOVE-R-infected/TOVE-R-infected, uninfected/ uninfected, and uninfected/mutans-infected. We chose not to study uninfected/tove-r-infected and TOVE-Rinfected/uninfected groups because the number of rats of exactly the same age was limited and because we knew that uninfected rats challenged by TOVE-R become colonized by this strain but develop no additional (though sometimes slightly fewer) carious lesions (29, 34). It is well known that rats consume their own feces and that of their cage-mate's, thereby transmitting the mutans streptococci between cage-mates and recycling their gut flora by oral reinoculation of fecally cleared bacteria (22, 28). We determined in preliminary experiments that newly co-caged rats 21 to 35 days of age do not reliably share their feces before being co-caged for at least 24 h. This was determined by measuring the oral recovery of streptomycin-tagged, fecally excreted TOVE-R and mutans streptococci colonizing the transient cage-mate. To limit the magnitude of the natural transmission challenge that was incident to co-caging, animals were separated after 48 h and housed singly in freshly autoclaved cages. We know of no way to inhibit the recycling of flora by an animal's ingestion of its own feces without adversely affecting the health of the rats (22). Recovery of microorganisms. The teeth of all animals were swabbed at intervals after inoculation and after separation from cage-mates. The cotton swabs were immediately placed in buffered yeast extract, vortex mixed, and plated on appropriate media (see below) within 1 h by using a spiral plater (Spiral Systems, Cincinnati, Ohio). The problems of sampling plaque from the teeth of small animals and of plate counts of streptococci are well known (27, 28, 32, 34). Previous data had shown that neither TOVE-R nor the mutans streptococci colonize the tongues of rats (34) and that the percentage recoveries of S. mutans 10449S monitored by careful swabbing of teeth were not statistically different from the percentage recoveries monitored by directly scraping plaque from the teeth (31). Nonetheless, at the termination of experiments, three molar tooth crowns from one hemimandible of each animal were harvested with a small rongeur as described previously. The plaque on these teeth and bacteria lodged in their fissures were dislodged, and clumps and chains were disrupted by sonification under conditions which gave maximal numbers of total recoverable flora (34).

3 78 TANZER ET AL. TABLE 1. Percentage of S. mutans 10449S in competition trial between TOVE-R and 10449Sa S. mutans 10449S (%) at day: Animal/cage-mate Uninfected/10449S 18 ± 7b 28 ± 9C 34 ± 13c 10449S/uninfected 43 ± 4 43 ± 9b 42 ± gb 57 ± llb TOVE-R/10449S 0.6 ± ± 6 8 ± S/TOVE-R 43 ± 4 35 ± ± 12c 45 ± 12b 10449S/10449S 43 ± c 37 ± 9b 53 ± 10j a The animal/cage-mate designation indicates the condition of infection of the animal/the condition of infection of the transient cage-mate. The values are for the "animal" of the animal/cage-mate pairs. They are given as the mean ± standard error of the mean for the percentage of the total recoverable flora which was S. mutans 10449S at swab-sampling intervals after initial inoculation. b p < 0.01 versus the TOVE-R/10449S group. c P < 0.05 versus the TOVE-R/10449S group. As previously described (34), both swabbed and sonicated cultures were plated on mitis salivarius agar containing Chapman tellurite (MS; Difco), MS supplemented with 200,ug of streptomycin per ml (MSS), and Trypticase soy agar supplemented with 5% sheep blood (BA; BBL Microbiology Systems, Cockeysville, Md.). The very large TOVE-R colonies grew rapidly on all of the agars, both in the presence and absence of CO2 atmospheric supplementation. The relatively small colonies of S. mutans 10449S and S. sobrinus WT grew well, but only with CO2 atmospheric supplementation, and they had typical mutans morphology on MS and MSS. CFUs on BA were used to estimate the total recoverable facultative flora; CFUs on MSS allowed quantitation of the mutans streptococci and TOVE-R. Thus, recoveries of the streptomycin-labeled TOVE-R, 10449S, and WT infectants in swab samples could be differently quantitated and expressed as percentages of total flora; but, because swab samples vary in size, absolute counts could not be computed. However, sonicated tooth sample counts could be expressed in both relative and absolute terms because the entire tooth crowns were sampled. The presence of possible non-streptomycin-resistant streptococcal infectants could be detected on MS. Periodically, the identities of selected reisolates were confirmed by wellestablished biochemical-physiological and morphological techniques (30). Statistical analyses of recoveries were carried out with percentage data arcsine transformed to meet the requirements of the parametric analysis of variance (ANOVA) and t test (26). Statistically significant contrasts resulting from ANOVA were identified by the least-significant-difference procedure (26). Data were also analyzed by the nonparametric Dixon-Mood sign test (25). Absolute bacterial CFUs were analyzed statistically without transformation. Plaque and caries evaluation. Excised, coded hemimandibles with teeth in situ were viewed microscopically for the presence of plaque after being stained with 0.25% aqueous safranin. They and the similarly coded maxillas were defleshed by dermestid beetles and scored for carious enamel areas by the method of Keyes (15) as modified by Larson (19). Because carious lesions are symmetrically represented bilaterally (14), the caries score for the molars of a single hemimandible were doubled before adding the score for each animal to the score for its maxillary molar teeth. Statistical evaluations of mean caries scores were carried out by ANOVA with isolation of significant contrasts by the least-significant-difference method (26). TABLE 2. Percentage of S. sobrinus WT in competition trial between TOVE-R and WTa S. sobrinus WT (%) at day: Animal/cage-mate Uninfected/ WT 14 ± 8b 30 ± 12C 54 ± WT/uninfected 11 ± 1 14 ± 4c 55 ± 12c 72 ± 5c TOVE-R/ WT 0.44 ± ± ± WT/TOVE-R 11 ± c 33 ± 8c 52 ± 7b WT/ WT 11 ± 1 23 ± 8c 25 ± 8c c a For an explanation of data, see Table 1, footnote a. b p < 0.05 versus the TOVE-R/ WT group. c P < 0.01 versus the TOVE-R/ WT group. RESULTS INFECT. IMMUN. In each experiment, seven groups of singly caged rats were studied (Fig. 1): (i) uninoculated, transiently co-caged with other uninoculated rats or (ii) with mutans- (either 10449S or WT) inoculated rats; (iii) TOVE-Rinoculated, transiently co-caged with other TOVE-Rinoculated rats or (iv) with mutans-inoculated rats; and (v) mutans-inoculated, transiently co-caged with mutansinoculated, (vi) TOVE-R-inoculated, or (vii) uninoculated rats. Never was (i) an uninoculated rat not co-caged with a mutans-infected rat observed to harbor either streptomycinsusceptible or streptomycin-resistant mutans streptococci or S. salivarius or (ii) a rat not exposed by design to either TOVE-R or a mutans Streptococcus sp. observed to harbor either of them. Thus, there was no evidence of extraneous or unwanted cross-infection of rats or of reversion of the streptomycin-resistant phenotypes to streptomycin-sensitive ones. There were no statistically significant differences among mean body weight gains of animal groups; thus, there was no evidence of effects of inoculations other than those described below. Recoveries of TOVE-R and mutans streptococci by tooth swabbing. Uninfected animals had essentially no plaque on their teeth at the time of sacrifice, whereas animals which had been infected by either of the mutans streptococci strains or TOVE-R had generally moderate plaque quantities. However, no differences in quantities of plaque could be photographically demonstrated among the infected rat groups. In both co-cage competition experiments with strains 10449S (Table 1) and WT (Table 2), prior colonization of rats by TOVE-R significantly inhibited the emergence of the mutans streptococci compared with the ecological emergence of the mutans streptococci in previously uninoculated rats (uninfected/mutans versus TOVE- R/mutans). These differences are evident by multiple t-test comparisons at different bacterial recovery sampling dates and by ANOVA (see table footnotes for statistical significance data). In addition, the nonparametric probability of observing these consistent directional differences for the 10449S and WT trials is P < Thus, colonization by TOVE-R inhibited the emergence of the mutans streptococci in the oral ecology. By contrast, the relative recoveries (percentages) of strain 10449S from uninfected rats by 10 days after co-caging (20 days after initial inoculation) with 10449S-infected rats (uninfected/10449s) were lower (P < 0.05) than from 10449S/uninfected and 10449S/10449S rats. However, by 25 days after co-caging the relative recoveries were not statistically different from those percentages recovered from rats

4 VOL. 49, 1985 INHIBITION OF MUTANS STREPTOCOCCI BY TOVE-R 79 TABLE 3. Percentage of S. salivarius TOVE-R in competition trial between TOVE-R and 10449S' S. salivarius TOVE-R (%) at day: Animal/cage-mate TOVE-R/TOVE-R 42 ± 5 21 ± 4b 21 ± 8 28 ± 7C TOVE-R/10449S 42±5 18±7c 17±7 24± S/TOVE-R 4±3 7±4 10±5 a For an explanation of data, see Table 1, footnote a. b p < 0.01 versus the 10449S/TOVE-R group. P p < 0.05 versus the 10449S/TOVE-R group. rl-. 0 x 3 SONIFIED TEETH initially colonized by 10449S (i.e., 10449S/uninfected, 10449S/ TOVE-R, and 10449S/10449S), and this lack of difference persisted (Table 1). Analogously, the percentages of strain WT recovered from uninfected rats by 9 days after co-caging (15 days after initial inoculation) with WTinfected rats (uninfected/ wt) were not statistically different from those percentages recovered from rats initially colonized by WT (i.e., WT/uninfected, WT/TOVE-R, and WT/ WT), and this lack of difference also persisted throughout the experiment. The lower percentage of 10449S and WT in animals first colonized by TOVE-R, therefore, cannot be explained as the result of the shortened duration of exposure to these strains. Thus, mutans streptococcal infection after TOVE-R colonization inhibited the relative emergence of both S. mutans 10449S and S. sobrinus WT in the oral flora. The mean percentage of TOVE-R among the total recoverable flora was lower in animals first colonized by strain 10449S (Table 3) or WT (Table 4). The nonparametric probability of observing such directional differences is P < 0.02 for these experiments. Thus, although TOVE-R, like the mutans streptococci, was transmitted by co-caging, it achieved lower relative levels if animals had first been colonized by the mutans streptococci, suggesting that TOVE- R colonization of the teeth occurred in competition with mutans streptococci already colonizing them. Recoveries of TOVE-R, mutans streptococci, and total recoverable flora by sonification of teeth. Although many comparisons could be described statistically for the two illustrated experiments, because of their complexity only certain key contrasts are presented here. For both the 10449S and WT trials, terminal recoveries (on days 41 and 50, respectively) of the mutans streptococci and TOVE-R upon sonification of the crowns of three mandibular molars (Fig. 2 and 3) paralleled the percentage of recoveries obtained by swabbing the teeth on that same date (Tables 1 through 4). Recoveries of the two mutans streptococci from rats initially infected by them were TABLE 4. Percentage of S. salivarius TOVE-R in competition trial between TOVE-R and WTa S. salivarius TOVE-R (%) at day: Animal/cage-mate TOVE-R/TOVE-R ± 2b 49 ± 10b TOVE-R/ WT 22 ± 6 28 ± 8c 3 ± 1c 46 ± 11c WT/TOVE-R 6 ± ± ± 6 a For an explanation of data, see Table 1, footnote a. b p < 0.05 versus the WT/TOVE-R group. C P < 0.01 versus the WT/TOVE-R group. C-) CL) I- 0 z w L-) w TOVE- R 10449S NON-mutons, NON-solivarius FLORA FIG. 2. Histogram of the absolute CFU recoveries for 10449S, TOVE-R, and non-mutans, non-salivarius flora and the relative (percentage) recoveries, both plotted as mean + standard error of the mean, at day 41 of the 10449S experiment. The absolute CFU recoveries are above the horizontal axis, and the relative recoveries are below. The data are given for the three sonicated (sonified) mandibular molars of the rats of the seven experimental groups. very high, significantly higher on both absolute and relative bases than recoveries from rats first infected by TOVE-R (P < 0.05; determined by ANOVA). The nonparametric probability of this consistent directional difference was P < 0.02 for both experiments. Some additional specific contrasts should be noted. For the 10449S study, the absolute and relative numbers of 10449S recovered from uninfected/10449s, 10449S/ uninfected, 10449S/TOVE-R, and 10449S/10449S rats were not different from one another, although they were higher than for TOVE-R/10449S rats (P < 0.05). The absolute and relative numbers of TOVE-R recovered from sonified teeth were higher from rats not previously inoculated with 10449S than from rats so inoculated (P < 0.05). It is interesting that the total CFUs recovered from the teeth of rats infected by 10449S, where caries scores were highest (see below), were greater (P < 0.05) than the total CFUs recovered from the uninfected/uninfected, TOVE- R/TOVE-R, and TOVE-R/10449S groups. In addition, there was no difference in total CFUs among the groups first

5 80 TANZER ET AL. C-0 Lo- J U.U 0 L w 0 TOVE-R 3 SONIFIED TEETH Om WT NON-mutans, NON-solivarius FLORA FIG. 3. Histogram of the absolute CFU recoveries for WT, TOVE-R, and non-mutans, non-salivarius flora and the relative (percentage) recoveries, both plotted as mean ± standard error of the mean, at day 50 of the WT experiment. The absolute CFU recoveries are above the horizontal axis, and the relative recoveries are below. The data are given for the three sonicated (sonified) mandibular molars of the rats of the seven experimental groups. infected by 10449S and the uninfected/10449s groups or between the uninfected/uninfected and TOVE-R/TOVE-R groups. For the WT study, the absolute and relative numbers of WT recovered from uninfected/ wt, WT/uninfected, WT/TOVE-R, and WT/ WT rats were not different from one another, but most were higher than for TOVE-R/ WT rats (P < 0.05; determined by ANOVA), and they were always higher (P < 0.02) than for TOVE-R/ WT rats as determined by the nonparametric test. The absolute and relative numbers of TOVE-R recovered from sonified teeth were higher (P < 0.05) from rats previously not inoculated by WT than from those so inoculated. The total recoverable CFUs for all of the groups were lower than those for the WT/TOVE-R group (P < 0.05) and not significantly different from one another. Caries scores. The caries scores for the two experiments are plotted (Fig. 4 and 5) so as to discriminate between sulcal (fissure) and smooth-surface (buccal, lingual, approximal, and morsal) tooth surfaces (24, 27). Tables 5 and 6 sum- INFECT. IMMUN. marize the statistically significant differences (P < 0.05) between pairs of group means for the 10449S and WT experiments, respectively. Caries scores for 10449S/10449S rats were the same as for 10449S/uninfected rats; but when 10449S-infected rats were co-caged transiently with TOVE-R-colonized rats, caries scores were lower (Fig. 4, Table 5). Scores for TOVE- R/TOVE-R, uninfected/uninfected, TOVE-R/10449S, and uninfected/10449s rats were lower than those for all animals initially infected by 10449S and were not statistically different from one another. Although scores for TOVE- R/10449S animals appeared lower than for uninfected/10449s animals, this difference was not quite statistically significant (P < 0.10). These data indicate, therefore, (i) that initial and longer-duration infection by strain 10449S induces higher caries scores than does the absence of such an infection or a shorter-duration infection, (ii) that infection by TOVE-R after infection by 10449S significantly inhibits caries, and (iii) that all animal groups first infected by TOVE-R or left uninfected have caries scores not statistically distinct from one another. Caries scores for all animal groups initially colonized by strain WT were not different from one another (Fig. 5, Table 6). Also, caries scores for the group initially uninfected but later co-caged with WT-infected rats (uninfected/ wt) were lower than those for WT/uninfected and WT/ WT rats. However, caries scores of the rat group initially colonized by TOVE-R and later co-caged with WT-infected animals (TOVE-R WT) were lower than those of the three groups of rats initially colonized by WT but not significantly lower than those of the uninfected/ wt rats. Scores for uninfected/uninfected and TOVE-R/TOVE- R rats were not different from each other. Therefore, initial and longer-duration infection by WT induced higher caries scores than if infection by WT had occurred later and for a shorter duration. Prior colonization by TOVE- R was not detectably protective against caries associated with later colonization by WT. DISCUSSION These data demonstrate that initial oral colonization of rats by S. salivarius TOVE-R inhibits the ecological emergence of fecally transmitted S. mutans 10449S and S. sobrinus WT; conversely, prior colonization by these mutans streptococci probably inhibits the ecological emergence of fecally transmitted TOVE-R. There is a generally inverse relationship between the percentages of TOVE-R and the mutans streptococci on the teeth. The data support previously published evidence that these bacteria compete for the same tooth-surface ecological sphere (29, 34). Other studies have shown that TOVE-R, like the mutans streptococci, colonizes the teeth, but not the tongues, of rats (34). Previous studies of preemptive colonization of teeth by TOVE-R and of displacement of mutans streptococci from teeth by TOVE-R have used oral inoculations of known numbers of cells (29, 34). The present study, employing natural, fecal inoculation resulting from co-caging, provides an intense, although difficult to quantitate, infectious challenge with fecal impaction into the fissures of the teeth probably occurring several times over the course of 2 days. This technique thus tests the effect of TOVE-R on the ecological emergence of mutans streptococci in a model in which transmission of microorganisms is virtually assured. Oral inoculation with a pipette of both TOVE-R (29) and weakly virulent S. mutans mutant 805 (31) and their coloni-

6 VOL. 49, 1985 INHIBITION OF MUTANS STREPTOCOCCI BY TOVE-R S/ 10449S S/TOVE-R TOVE-R/ 10449S TOVE-R/TOVE-R FE =JMAJOR SULCAL BUCCO- LINGUAL APPROXIMAL MORSAL I i : EMMMMM... 'I"W I*-:1mmw.- _ *. 1=4.-q 0bllm.. H S/UNINFECTED UNINFECTED/ 10449S UNINFECTED/UNINFECTED FE i EIZ 10 I - I I Average Number Carious Enamel Areas FIG. 4. Average number of carious enamel areas of S. mutans 10449S-infected, S. salivarius TOVE-R-infected, or uninfected SPFOM rats. The designations of the bars indicate the condition of infection of the rat groups versus the condition of infection of their transient cage-mates. The mean and standard error of the mean are plotted for each indicated type of tooth surface. A detailed statistical comparison of the total caries score for each of the seven animal groups is given in Table 5. zation of the teeth inhibits subsequent implantation of carefully quantitated numbers of inoculated 10449S. However, from this preemptive-inoculation model one cannot ascertain possible anticaries effects; the periods of cariogenic challenge by strain TOVE-R or 805, by strain 10449S, and by the indigenous non-mutans, non-salivarius flora of the rats are all different, making groups incomparable as to consequent caries induction. The present co-caging studies substantially overcome this problem because rats of the same age with the same indigenous flora are simultaneously inoculated with either TOVE-R or mutans streptococci shortly after the cariogenic diet is presented. In addition, co-caging and the consequent coprophagous sharing of flora produces additional animal groups challenged for a shorter time with either the mutans streptococci or TOVE-R and allows tests of whether TOVE-R inhibits the ecological emergence of the mutans streptococci. It must be recognized that the inception of most carious lesions in rats, as in humans, occurs in the early period after the teeth erupt and the bacteriologicaldietary challenge to develop decay is imposed (5, 10). Thus, C MAJOR-SULCAL BUCCO-LINGUAL M APPROXIMAL E3 MORSAL WT/ WT Si WT/TOVE-R TOVE-R/ WT TOVE- R/TOVE-R WT/UNINFECTED F-H UNINFECTED/ WT UNINFECTED/ UNINFECTED Average Number rou of Carious Enamel Areas FIG. 5. Average number of carious enamel areas of S. sobrinus WT-infected, S. salivarius TOVE-R-infected or uninfected SPFOM rats. The designations of the bars indicate the condition of infection of the rat groups versus the condition of infection of their transient cage-mates. The mean and standard error of the mean are plotted for each indicated type of tooth surface. A detailed statistical comparison of the total caries score for each of the seven animal groups is given in Table 6. 50

7 82 TANZER ET AL. TABLE 5. Multiple statistical comparisons of total caries scores for each of seven rat groups in the co-cage competition trial of TOVE-R and 10449S depicted in Fig. 4" T/T U/U T/10449 U/ /T 10449/ /U T/T U/U T/10449 U/ /T * * * 10449/10449 * * * * * 10449/U * * * * * a Statistical significance is indicated by * at the P < 0.05 level, using the least-significant-difference method to isolate differences revealed by AN- OVA. Although there were significant differences at P < 0.01, they are not reported in this table. Abbreviations: T, TOVE-R-infected; U, uninfected; 10449, 10449S-infected. Column and row designations represent the condition of infection of the rats/the condition of infection of the transient cage-mates. The groups are arrayed, from top to bottom and from left to right, in the order of increasing total caries scores. sucrose-eating rats infected by the mutans streptococci soon after weaning develop the highest caries scores, and later infection or sucrose provision induces fewer carious lesions, as seen here. However, rats infected by TOVE-R develop no higher caries scores than animals harboring their normal mutans- and salivarius-free floras. Thus, TOVE-R again (29, 34) was observed not to be detectably cariogenic. Rats infected first by S. mutans 10449S and then by naturally transmitted TOVE-R developed fewer carious lesions than if they had not been exposed to TOVE-R. This caries-protective effect was not evident in the WT experiment reported here, a failure perhaps due to two factors. In the WT trial, animals ate diet 2000 for 51 days and were heavily infected by the highly virulent S. sobrinus strain, in addition to harboring their indigenous fermentative flora (27), for 50 days before sacrifice. Also the Keyes' caries-scoring technique requires that damaged tooth surfaces be scored as carious. Thus, a smooth tooth surface which evidences a cusp fracture or carious involvement due to extension of sulcal caries is nonetheless scored as carious. One cannot be sure whether the caries process originated in the sulcus or on the smooth surface; one can only observe that such a surface is missing or carious (and thus score it as carious). Because of this scoring artifact and the duration of TABLE 6. Multiple statistical comparisons of total caries scores for each of seven rat groups in the co-cage competition trial of TOVE-R and WT depicted in Fig. 5a U/U T/T T/6715 U/ /T 6715/U 6715/ U/U T/T T/6715 * U/6715 * * 6715/T * * * 6715/U * * * * 6715/6715 * * * * " Statistical significance is indicated by * at the P < 0.05 level, using the least-significant-difference method to isolate differences revealed by AN- OVA. Although there were significant differences at P < 0.01, they are not reported in this table. Abbreviations: U, Uninfected; T, TOVE-R-infected; 6715, WT-infected. Column and row designations represent the condition of infection of the rats/the condition of infection of the transient cagemates. The groups are arrayed, from top to bottom and from left to right, in the order of increasing total caries scores. INFECT. IMMUN. this particular WT trial, base-line fissure and smoothsurface caries scores were unusually high, even for uninfected and TOVE-R-infected animals. This perhaps obscured the caries-protective effects of TOVE-R against WT that were observed in the 10449S experiment. In fact, TOVE-R protection against caries by displacement of WT and 10449S was previously reported (34). There has been uncertainty in the literature about whether to report percentage recoveries of bacteria from teeth or absolute numbers. The latter, of course, requires extraction of the teeth or jaws and thus the termination of experiments. However, total bacterial recoveries reported in absolute CFUs were suggested to be a reflection of the state of cavitation of the teeth, with the cavities mechanically accommodating more bacteria as though they were additional tooth fissures (34). For our 10449S experiment, the correlation between caries score and total recoverable flora at the time of sonification of teeth was r = 0.489, P < 0.005; for the WT experiment it was r = 0.340; P < Thus, total recoverable flora counts are correlated with the caries score, independent of mutans streptococcal infection. It is also clear, however, that the groups of animals with the highest caries scores are those first infected by the mutans streptococci (10449S or WT). It is not surprising that absolute counts of strain 10449S or WT recovered from teeth at termination of the experiments do not correlate with caries scores for those same mutans-infected animals because most lesions begin to develop soon after the bacteriological-dietary cariogenic challenge (5, 10). Also, because caries scores for mutansinfected animals and bacterial recoveries from them are rather similar within an experiment, they do not allow spread of the data, thus diminishing the likelihood of detecting correlations of statistical significance. Therefore, counts at the time of sacrifice may not so much reflect causation of disease as they do the eventual dimension of the lesions. The principal cause of human coronal caries appears to be infection by the mutans streptococci (7, 9, 11, 12, 20), the transmission of which frequently occurs from female to child after its first teeth erupt into the oral cavity (1, 3, 17, 18, 21). No practical ways have yet been demonstrated to suppress or cure such infections (8, 18, 35-37). The presently reported inhibition by TOVE-R of the ecological emergence in rats of the two mutans streptococcal species most frequently isolated from humans, the previously reported successful preemption of S. mutans infection by TOVE-R colonization (29), and the previously reported displacement of virulent S. mutans and S. sobrinus strains from rats and consequent caries inhibition (34) suggest that TOVE-R or similar microorganisms may have important human clinical therapeutic utility. Initial human studies along these lines are under way. Studies of the mechanistic bases for the lack of cariogenicity of TOVE-R and its ability to successfully compete with the mutans streptococci have also begun. There has been a long and newly revived interest in the inhibition of various bacterial infections by analogous bacterial competitive means (2). ACKNOWLEDGMENT This study was supported by Public Health Service grant DE from the National Institute of Dental Research. The authors thank John Richi and Lynn Laskowski for excellent technical support. LITERATURE CITED 1. Alaluusua, S., and 0. V. Renkonen Streptococcus mutans establishment and dental caries experience in children from 2 to

8 VOL. 49, 1985 INHIBITION OF MUTANS STREPTOCOCCI BY TOVE-R 83 4 years old. Scand. J. Dent. Res. 91: Aly, R., and H. R. Shinefield (ed.) Bacterial interference, p CRC Press, Inc., Boca Raton, Fla. 3. Berkowitz, R. J., and H. V. Jordan Similarity of bacteriocins of Streptococcus mutans from mother and infant. Arch. Oral Biol. 20: Bratthali, D Demonstration of five serological groups of streptococcal strains resembling Streptococcus mutans. Odontol. Revy 21: Carlos, J. P., and A. M. Gittelsohn Longitudinal studies of the natural history of caries. Arch. Oral Biol. 10: Coykendall, A. L Streptococcus sobrinus nom. rev. and Streptococcus ferus nom. rev.: habitat of these and other mutans streptococci. Int. J. Syst. Bacteriol. 33: de Stoppelaar, J. D., J. van Houte, and 0. Backer-Dirks The relationship between extracellular polysaccharide-producing streptococci and smooth surface caries in 13-year-old children. Caries Res. 3: de Stoppelaar, J. D., J. van Houte, and 0. Backer-Dirks The effect of carbohydrate restriction on the presence of S. mutans, S. sanguis and iodophilic polysaccharide-producing bacteria in human dental plaque. Caries Res. 4: Duchin, S., and J. van Houte Relationship of Streptococcus mutans and lactobacilli to incipient smooth surface dental caries in man. Arch. Oral Biol. 23: Fitzgerald, R. J., and R. H. Larson Age and caries susceptibility in gnotobiotic rats. Helv. Odontol. Acta 11: Ikeda, T., and H. J. Sandham Prevalence of Streptococcus mutans on various tooth surfaces in Negro children. Arch. Oral Biol. 16: Ikeda, T., H. J. Sandham, and E. L. Bradley, Jr Changes in Streptococcus mutans and lactobacilli in plaque in relation to the initiation of dental caries in Negro children. Arch. Oral Biol. 18: Kelstrup, J Extracellular polysaccharides of smooth and rough variants of Streptococcus salivarius. Scand. J. Dent. Res. 89: Keyes, P. H Dental caries in the molar teeth of rats. I. Distribution of lesions induced by high-carbohydrate low-fat diets. J. Dent. Res. 37: Keyes, P. H Dental caries in the molar teeth of rats. II. A method for diagnosing and scoring several types of lesions simultaneously. J. Dent. Res. 37: Keyes, P. H., and H. V. Jordan Periodontal lesions in the Syrian hamster. III. Findings related to an infectious and transmissible component. Arch. Oral Biol. 9: Kohler, B., and D. Bratthall Intra-familial levels of Streptococcus mutans and some aspects of the bacterial transmission. Scand. J. Dent. Res. 86: Kohler, B., D. Bratthall, and B. Krasse Preventive measures in mothers influence in the establishment of Streptococcus mutans in their infants. Arch. Oral Biol. 28: Larson, R. H Merits and modifications of scoring rat dental caries by Keyes' method, p In J. M. Tanzer (ed.), Animal models in cariology. Information Retrieval, Inc., Washington, D.C. 20. Littleton, N. W., S. Kakehashi, and R. J. Fitzgerald Recovery of specific "caries-inducing" streptococci from carious lesions in the teeth of children. Arch. Oral Biol. 15: Masuda, N., N. Tsutsumi, S. Subue, and S. Hamada Longitudinal survey of the distribution of various serotypes of Streptococcus mutans in infants. J. Clin. Microbiol. 10: Olsson, J., and W. Bowen Effect of diet on the colonization of the mouth by Actinomyces viscosus (T-6) in Osborne- Mendel rats. Arch. Oral Biol. 27: Perch, B., E. Kjems, and T. Ravn Biochemical and serological properties of Streptococcus mutans from various human and animal sources. Acta Pathol. Microbiol. Scand. Sect. B 82: Rinehimer, L. A., and J. M. Tanzer Statistical relationship of morsal to bucco-lingual, approximal and sulcal caries in rats. J. Dent. Res. 59: Sachs, L Applied statistics: a handbook of techniques, p Springer Verlag, New York. 26. Snedecor, G., and W. Cochran Statistical methods, 6th ed., p Iowa State University Press, Ames, Iowa. 27. Tanzer, J. M Essential dependence of smooth surface caries on, and augmentation of fissure caries by, sucrose and Streptococcus mutans infection. Infect. Immun. 25: Tanzer, J. M. (ed.) Animal models in cariology, p Information Retrieval, Inc., Washington, D.C. 29. Tanzer, J. M., and J. Fisher Competition of a rough non-cariogenic Streptococcus salivarius-like strain with Streptococcus mutans in dental plaque, p In S. E. Holm and P. Christiansen (ed.), Basic concepts of streptococci and streptococcal diseases. Reedbooks, Ltd., Chertsey, England. 30. Tanzer, J. M., and M. L. Freedman Genetic alterations of Streptococcus mutans virulence. Adv. Exp. Med. Biol. 107: Tanzer, J. M., M. L. Freedman, and J. Fisher Preemption of Streptococcus mutans 10449S colonization by its mutant 805. Infect. Immun. 35: Tanzer, J. M., M. L. Freedman, R. J. Fitzgerald, and R. H. Larson Diminished virulence of glucan synthesis-defective mutants of Streptococcus mutans. Infect. Immun. 10: Tanzer, J. M., M. L. Freedman, F. N. Woodiel, R. L. Eifert, and L. A. Rinehimer Association of Streptococcus mutans virulence with synthesis of intracellular polysaccharide, p In H. M. Stiles, W. J. Loesche, and T. C. O'Brien (ed.), Microbial aspects of dental caries. Information Retrieval, Ltd., New York. 34. Tanzer, J. M., A. B. Kurasz, and J. Clive Competitive displacement of mutans streptococci and inhibition of tooth decay by Streptococcus salivarius TOVE-R. Infect. Immun. 48: van Houte, J., and S. Duchin Streptococcus mutans in the mouths of children with congenital sucrase deficiency. Arch. Oral Biol. 20: Zickert, I., C.-G. Emilson, and B. Krasse Streptococcus mutans, lactobacilli and dental health in year-old Swedish children. Community Dent. Oral Epidemiol. 10: Zickert, I., C.-G. Emilson, and B. Krasse Effect of caries preventive measures in children highly infected with the bacterium Streptococcus mutans. Arch. Oral Biol. 27: