Effect of the Environment on Genotypic Diversity of Actinomyces naeslundii and Streptococcus oralis in the Oral Biofilm

Abstract

The genotypic diversity of Actinomyces naeslundii genospecies 2 (424 isolates) and Streptococcus oralis (446 isolates) strains isolated from two sound approximal sites in all subjects who were either caries active (seven subjects) or caries free (seven subjects) was investigated by using the repetitive extragenic palindromic PCR. The plaque from the caries-active subjects harbored significantly greater proportions of mutans streptococci and lactobacilli and a smaller proportion of A. naeslundii organisms than the plaque sampled from the caries-free subjects. These data confirmed that the sites of the two groups of subjects were subjected to different environmental stresses, probably determined by the prevailing or fluctuating acidic pH values. We tested the hypothesis that the microfloras of the sites subjected to greater stresses (the plaque samples from the caries-active subjects) would exhibit reduced genotypic diversity since the sites would be less favorable. We found that the diversity of A. naeslundii strains did not change (χ2 = 0.68; P = 0.41) although the proportional representation of A. naeslundii was significantly reduced (P < 0.05). Conversely, the diversity of the S. oralis strains increased (χ2 = 11.71; P = 0.0006) and the proportional representation of S. oralis did not change. We propose that under these environmental conditions the diversity and number of niches within the oral biofilm that could be exploited by S. oralis increased, resulting in the increased genotypic diversity of this species. Apparently, A. naeslundii was not able to exploit the new niches since the prevailing conditions within the niches may have been deleterious and not supportive of its proliferation. These results suggest that environmental stress may modify a biofilm such that the diversity of the niches is increased and that these niches may be successfully exploited by some, but not necessarily all, members of the microbial community.

Dental plaque bacteria colonize the hard tissues of the oral cavity by ecological succession (4, 29, 30). Early colonizers of the dentition, including Actinomyces naeslundii and Streptococcus oralis, facilitate further bacterial colonization (16, 29) and maintain biofilm integrity (23), protecting the host against colonization by extraoral pathogens (36). S. oralis is an oral commensal organism and is a member of the mitis group of viridans streptococci (41). A. naeslundii is a gram-positive pleomorphic rod which forms a significant component of commensal oral microfloras (7). A. naeslundii strains are assigned to two genospecies (1 and 2) on the basis of DNA homology (19) and may be identified by using genospecies-specific antisera (28).

The ability of bacteria to survive and persist in a given environment will depend, in part, on their inherent genetic plasticity, which determines their ability to respond to fluctuating local environmental conditions or stresses (13). The presence of active carious lesions indicates that the oral cavity is subject to local stresses, including the intake of fermentable carbohydrates and the production of organic acids, resulting in enamel demineralization and carious lesion formation. It has been suggested that acid stress associated with carious lesion formation will result in the reduction of genotypic diversity as strains best suited to survival are selected (5). However, this hypothesis has not been tested in relation to the ecology of the oral biofilm.

Previous studies have shown that oral bacteria, including streptococci (15, 17, 20-22, 24-26, 27, 32, 33) and A. naeslundii (6, 31), are genotypically heterogeneous. Repetitive extragenic palindromic PCR (REP-PCR) (37, 38) is suitable for genotyping oral bacteria, and this method has been used previously to study the epidemiology and relatedness of S. oralis strains in dental plaque (1, 2) and Actinomyces gerencseriae and Actinomyces israelii isolated from root caries lesions (8). In this study REP-PCR was used to investigate the genetic diversity of S. oralis and A. naeslundii populations isolated from sound anterior approximal tooth sites of caries-free and caries-active subjects. These strains were studied to test the hypothesis that the genetic diversity of individual species exhibits increased homogeneity and reduced diversity in environments subjected to increased stress.

MATERIALS AND METHODS

Subjects.

Fourteen adult subjects were enrolled in the study. Seven had active caries with more than two active, cavitated carious lesions designated for operative treatment, while seven subjects were diagnosed as caries free, being clinically and radiographically free of active caries. A standardized approximal plaque sample was taken from each of two sound sites in the same dental arch from each subject. The plaque samples were taken by using sterile unwaxed dental floss (Oral-B, South Becton, Mass.). The local ethics committee approved the study, and all subjects gave informed consent for the collection of plaque samples.

Processing of plaque samples.

The plaque samples were placed individually into 1 ml of sterile PBSTC (1.58 g of K₂HPO₄ · 3H₂O, 0.34 g of KH₂PO₄, 8 g of NaCl, 1.0 g of sodium thioglycolate, and 0.001 g of cetyltrimethylammonium bromide per liter of distilled water). The samples were dispersed by vortexing with sterile glass beads (3.5- to 4.5-mm diameter; BDH, Lutterworth, Leicestershire, United Kingdom) for 30 s and diluted 10-fold in PBSTC; 100 μl of appropriate dilutions was plated onto a range of culture media (1, 8). Streptococci were isolated on mitis salivarius agar (MSA; Becton Dickinson, Cowley, Oxford, United Kingdom), S. mutans was isolated on MSA supplemented with bacitracin and sucrose, and lactobacilli were enumerated on rogosa agar (Oxoid, Basingstoke, United Kingdom). Fastidious anaerobe agar (FAA; LabM, Bury, Lancashire, United Kingdom) supplemented with 5% (vol/vol) horse blood was used for the enumeration of the total number of CFU and A. naeslundii isolates per sample. The FAA plates were incubated anaerobically at 37°C for 7 days, and all other media were incubated anaerobically at 37°C for 3 days.

The total number of CFU per sample was calculated from the FAA plate counts, from which 80 colonies presumptively identified as Actinomyces spp. on the basis of Gram staining were selected at random and subcultured. A. naeslundii colonies were identified from among these isolates (8), further examined using specific antisera, and classified as A. naeslundii genospecies 1 or A. naeslundii genospecies 2 by whole-cell agglutination (28). Nonreacting colonies were not investigated further. All isolates that were to be genotyped were stored at −70°C.

To identify S. oralis in the plaque samples, 48 colonies were picked at random from each sampled site, subcultured from the MSA plates, and identified (41). The isolates identified as S. oralis were stored at −70°C for genotyping. The total numbers of S. mutans organisms and lactobacilli in each approximal plaque sample were enumerated from the appropriate selective media.

Isolation of DNA from A. naeslundii.

A single colony of each isolate was suspended in 20 μl of microLYSIS (Microzone, Ltd., Sussex, United Kingdom) and vortexed for 30 s. The DNA was extracted with an automated thermocycler (Flexigene; Techne, Cambridge, United Kingdom) according to the manufacturer's instructions. The cell and microLYSIS mixture was heated according to the following cycle: 5 min at 65°C, 2 min at 96°C, 4 min at 65°C, 1 min at 96°C, 1 min at 65°C, and 30 s at 96°C. Samples were put through the cycle twice and vortexed between cycles. Supernatant containing the DNA (40 to 50 ng/μl) was carefully removed from the underlying cell debris and transferred to fresh microcentrifuge tubes. The DNA extracts were stored at 4°C for no longer than 7 days until they were used in the REP-PCRs.

Isolation of DNA from S. oralis.

DNA extraction was performed as previously described (2). Briefly, cells were boiled for 10 min and cell debris was removed by centrifugation. DNA extracts were stored at 4°C for no longer than 7 days until they were used in the REP-PCRs.

Genotyping of A. naeslundii and S. oralis.

Individual strains were genotyped by REP-PCR (2, 8). Briefly, DNA extract (3 μl) was added to 25 μl of the REP-PCR mixture, and after amplification, 3 μl of tracking dye (0.25% bromophenol blue, 0.25% xylene cyanol FF, 30% glycerol) was added to 17 μl of REP-PCR product. The amplification products were analyzed with 2% Metaphor agarose (Flowgen, Ashby de la Zouch, Leicestershire, United Kingdom) containing 0.5 μg of ethidium bromide per ml and were separated electrophoretically on 16.5-cm gels at 100 V for 2 h in 450 ml of Tris-borate-EDTA buffer. Molecular size markers (pGEM DNA Markers; Promega, Southampton, United Kingdom) were included in three separate lanes on all gels to facilitate comparisons of tracks between gels. Gels were examined on a UV transilluminator, and images were acquired (FluorChem 8000; Alpha Innotech Ltd., Cannock, United Kingdom). REP-PCR patterns were generated once from all isolates as this technique has been previously shown to be highly reproducible (2).

Computer-assisted analysis of the DNA patterns.

All visible bands for A. naeslundii and S. oralis were compared by applying the Dice coefficient (12) through assigned banding (GelCompar; Applied Maths, Kortrijk, Belgium). Cluster analysis was carried out by the unweighted pair group method using mathematical averages (UPGMA) (35) with a band position tolerance of 2%. The analysis of the patterns was undertaken in accordance with the instructions of the manufacturer. Strains with REP-PCR patterns with >95% similarity were considered to be of the same genotype.

Statistical analysis.

The total number of organisms recovered from each approximal sample and the proportions of the S. mutans, A. naeslundii and S. oralis organisms and the lactobacilli in the plaque samples were calculated, and the data from the caries-free and caries-active subjects were compared with the Mann-Whitney U test (SPSSPC, version 10.1; SPSS, Chicago, Ill.).

To compare the diversity of both A. naeslundii and S. oralis isolates recovered from the caries-free and caries-active subjects, χ² tests were used. We compared (i) the number of genotypes identified among the strains of each species for both the caries-free and caries-active subjects, (ii) the distribution of shared and unique genotypes between the two sample sites in the caries-free and caries-active subjects for both organisms, and (iii) the distribution of the number of genotypes in both the caries-free and caries-active subjects for both organisms. If the χ² test result was not significant (P > 0.05), then there was no difference in the diversity of the species in these comparisons.

RESULTS

In order to establish that the sites in the caries-free and caries-active subjects were subjected to different environmental stresses, we compared the proportion of the flora associated with health (A. naeslundii) and the proportion of the floras associated with dental caries (S. mutans and lactobacilli). There was no significant difference in the total numbers of organisms isolated from the individual sites of the caries-free or caries-active individuals. However, there were significantly (P < 0.05) greater proportions of lactobacilli and S. mutans isolates in the plaque samples from those subjects with active caries, while the caries-free subjects harbored a significantly greater proportion of A. naeslundii isolates than the caries-active subjects. These data demonstrate that subjects with caries harbored plaque that had apparently been subjected to different local stresses, including acid production, and that this difference had resulted in the accumulation of plaque with significantly different microfloras (3, 7, 9). The difference in the proportions of S. oralis isolates in the samples from the caries-free and caries-active subjects was not significant (Table 1).

TABLE 1.

Total number of bacteria and proportions of individual marker organisms recovered from sound approximal sites of caries-free and caries-active subjects

Subjects	No. of microorganisms per samplea	% Of organisms recoveredb
		Lactobacillus spp.	S. mutans	A. naeslundii	S. oralis
Caries free
Mean ± SE	6.07 ± 0.18	0.001 ± 0.001	0.018 ± 0.017	66.4 ± 8.4	19.4 ± 5.7
Median	6.10	<0.001	<0.001	71.8	8.7
Caries active
Mean ± SE	6.39 ± 0.12	0.68 ± 0.59	0.93 ± 0.47	53.8 ± 18.9	5.2 ± 1.6
Median	6.20	0.009	0.003	49.5	3.6

^aResults are given as log₁₀ (CFU per sample + 1).

^bValues are the percentages of total CFU per sample. Values in bold are significantly greater according to the Mann-Whitney U test (P < 0.05).

A total of 454 A. naeslundii isolates were identified from all sites sampled, and among these, A. naeslundii genospecies 2 (n = 424) was the most frequently isolated. A. naeslundii genospecies 1 (n = 30) was recovered from only 5 patients (13 isolates from 3 caries-free subjects and 17 isolates from 2 caries-active subjects), which is a low rate of prevalence in accordance with previous observations, and these isolates will not be considered further.

A. naeslundii genospecies 2 and S. oralis organisms were isolated from 26 of the 28 sites sampled from the 14 patients. The total number of A. naeslundii genospecies 2 and S. oralis strains identified per site varied between 0 and 70 and 0 and 43, respectively. Representative REP-PCR patterns of A. naeslundii genospecies 2 and S. oralis isolates from one patient are shown in Fig. 1 and 2. Isolates 11A21 and 11B6 are examples of identical A. naeslundii genospecies 2 genotypes from different sites in the same subject, while isolates 12A47 and 12A35 are examples of identical S. oralis genotypes from the same site.

FIG. 1.

REP-PCR patterns of A. naeslundii strains isolated from approximal plaque samples from patient 11. Unmarked lanes (1, 9, and 16) contain molecular size markers.

FIG. 2.

REP-PCR patterns of S. oralis strains isolated from approximal plaque samples from patient 12. Unmarked lanes (1, 8, and 15) contain molecular size markers.

The REP-PCR patterns of A. naeslundii genospecies 2 and S. oralis isolates were analyzed and examined as dendrograms in order to determine the similarities among isolates from the same subject (Fig. 3 and 4). For both species a limited number of genotypes was found at both sampling sites in the same subject, but no strains were identified from different subjects which had the same REP-PCR pattern.

FIG. 3.

Dendrogram illustrating genotypic relationships among A. naeslundii strains isolated from the approximal plaque of a single subject. Strains are designated by patient number (11), site (A or B), and strain number. The REP-PCR amplicon patterns for isolates marked with an asterisk are shown in Fig. 1. The scale shows percentages of similarity.

FIG. 4.

Dendrogram illustrating genotypic relationships among S. oralis strains isolated from the approximal plaque of a single subject. Strains are designated by patient number (12), site (A or B), and strain number. The REP-PCR amplicon patterns for isolates marked with an asterisk are shown in Fig. 2. The scale shows percentages of similarity.

The strains of A. naeslundii genospecies 2 were highly heterogeneous, with a maximum of 21 genotypes found in one patient and an average per subject of 9.8 ± 1.9 genotypes. The diversity of the A. naeslundii genospecies 2 isolates in the caries-free and caries-active subjects indicated that this species was equally diverse in both groups. A total of 78 genotypes were identified among the 246 A. naeslundii genospecies 2 strains investigated from the caries-free sites, and 56 were identified among the 208 strains genotyped from the caries-active sites (χ2 = 0.68; P = 0.41). In the caries-free group, 6 of 77 genotypes were common to the two sites, while in the caries-active group 5 of 55 genotypes were shared (χ2 = 0.10; P = 0.756).

S. oralis strains were also heterogeneous, with 21 genotypes identified from one subject and an average per subject of 10.4 ± 1.5 genotypes. There was apparently greater diversity among strains isolated from the caries-active subjects than among strains from the caries-free subjects. The 278 S. oralis strains genotyped from the caries-free subjects represented only 67 genotypes, but 78 genotypes were found among the 168 strains (χ2 = 11.71; P = 0.0006) from the caries-active subjects. Thus, 18 of 67 genotypes were common to the two sites sampled in the caries-free subjects, while in the caries-active subjects only 4 of 78 clones were common to the two sites sampled (χ2 = 9.69; P = 0.002).

A comparison of the diversity of A. naeslundii genospecies 2 and S. oralis strains isolated from the caries-free and caries-active subjects demonstrated that in the caries-free subjects the diversity of the two taxa was not significantly different. Thus, 67 genotypes were identified among 278 S. oralis strains isolated from the caries-free subjects compared to 78 genotypes among 246 A. naeslundii genospecies 2 isolates (χ2 = 2.13; P = 0.144). In the caries-active group A. naeslundii genospecies 2 was less diverse as only 56 genotypes were present among 208 isolates, but 78 genotypes were identified among the 168 S. oralis isolates (χ2 = 7.24; P = 0.007).

DISCUSSION

The microfloras resident in the oral biofilm are subjected to many variable environmental stresses, including the availability of nutrients, acidic pH, and organic acids (9). These stresses would be most apparent when the plaque floras of caries-active and caries-free subjects are compared. To demonstrate that the plaque samples from the caries-free and caries-active subjects were subject to different stresses, we determined the numbers and proportions of bacteria associated with caries (mutans streptococci and lactobacilli) and with health (A. naeslundii) for each sample. We found that the proportions of mutans streptococci and lactobacilli were elevated in the plaque from the caries-active subjects, while A. naeslundii isolates formed a significantly greater proportion of the floras in the samples from the caries-free subjects. These observations support the assertion that the plaque samples from the two subject populations were subjected to different environments and, consequently, given the differences in composition, to different stresses. The exact nature of these stresses is not known but may include acidic stress, exposure to organic acids, and exposure to dietary sugars.

Here we have investigated the effects of these undetermined stresses on the genotypic composition of two prominent members of the oral biofilm, S. oralis and A. naeslundii. These species are commensal bacteria and are predominant in floras during early plaque formation (16, 29). The prominence of both A. naeslundii and S. oralis in the plaque samples from the caries-free and caries-active subjects suggests that the plaque was recently formed. It was unexpected that the two organisms would behave in different ways as it has previously been postulated that environmental stresses in the oral biofilm should lead to the selection of the genotypes best able to proliferate in the particular environment (5). As the environment associated with caries is assumed to be more stressful than that associated with health, we expected that both species would show a reduction in diversity. However, S. oralis strains were more diverse in the caries-active subjects, and the A. naeslundii strains were no more diverse, than in the plaque samples from the caries-free subjects.

In the present study the undefined stresses, associated with the development of plaque floras with the microbial characteristics associated with caries, did not influence the diversity of A. naeslundii but did result in a reduction in the proportion of this species in the oral biofilm. This result suggests that A. naeslundii strains were less able to adapt to whatever stresses were present. These same stresses did not significantly influence the representation of S. oralis isolates in dental plaque but did result in significantly more diverse populations than were found in the caries-free subjects. From previous studies it is known that all individuals harbor distinct genotypes with the ability to grow in acidic conditions and that nonmutans streptococci from subjects with caries are more acidogenic than those from caries-free subjects (1, 34).

The reasons for the different behavior of these two commensal species in response to the same stresses are not known. However, we suggest that in the caries-active subjects, the sites from which plaque was sampled were more diverse, as production of organic acids and the lowering of the pH within the biofilm probably resulted in a more complex site with an increased diversity of distinct niches. This conclusion is substantiated by confocal microscopic studies of artificial oral biofilms which are not homogenous with respect to pH values following exposure to carbohydrate (39). Such studies clearly demonstrate that numerous niches with different pH values are present in the biofilm, and it follows that bacteria that exploit these niches would have to have different phenotypic characteristics. The environmental stress would therefore result in an increase in niche diversity that might be exploited by the colonizing bacteria. Alternatively, the increase in niche diversity might not be exploited by some species for which the new niches were detrimental while other species (here, S. mutans and lactobacilli) could exploit the niches. The present data also indicate that S. oralis was better able than A. naeslundii to exploit these new niches. This ability may be due to the plasticity of the genome of S. oralis (14, 40) and its inherent competence, which may facilitate the selection of novel genotypes better suited to proliferation in these approximal biofilms. However, the origin of these genotypes is not known. We have previously shown that S. oralis populations are not stable and that genotypes are rapidly replaced over an interval as short as 1 month, a situation similar to that reported for the phenotypically related species Streptococcus mitis (18). Escherichia coli shows similar rapid genotype turnover with few persistent genotypes being identified. For this organism, it has been suggested that environmental sources external to the host underpin the diversity of E. coli and the rapid genotypic turnover (10, 11). For S. oralis no sources external to the host are likely to support the persistence of the organism, and person-to-person transmission is uncommon (1); therefore, other mechanisms must be responsible for the observed diversity. Hohwy et al. (18) considered the diversity of S. mitis in the human oral cavity and speculated that “new” clones represented those already present in the mouth and occupying other niches which were transferred intraorally by undetermined mechanisms. The present data neither support nor refute this hypothesis.

That A. naeslundii was less able to respond to the environmental stresses and less able to proliferate is supported by the present data. Thus, it follows that exposure of the biofilm to these stresses resulted in a reduction in the diversity of niches that A. naeslundii was able to occupy compared to the range available and exploitable in the biofilms of the caries-free subjects.

The impact of stress on a biofilm is to select for species and genotypes best able to exploit the changed environment (5). However, it may be that if the stress is not stringent, the complexity of the biofilm may increase and the genotypic diversity of species present in the biofilm may increase to exploit this increased habitat diversity. Here we have found that the genotypic diversity of S. oralis increased, which we propose was due to an increase in the diversity of the biofilm niches that could be exploited by this organism. A. naeslundii was less able to exploit the postulated increase in biofilm complexity as this may have been less favorable to its proliferation.