INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

The microbial composition of dental plaque is highly diverse, with as many as 300 genera of bacteria commonly isolated. This microflora contains bacteria with widely different requirements in terms of nutrition, environmental parameters, and reaction to the presence or absence of oxygen and other gases. For example, although the oral cavity is overtly aerobic, a large proportion of the oral microflora bacteria are obligately anaerobic (19). The mechanisms by which these anaerobes survive may depend on oxygen consumption by aerotolerant species, thereby generating gradients within plaque biofilms.

Experiments using a two-stage chemostat model system, in which the second-stage chemostat vessel was aerated, showed that anaerobes were able to persist and grow in both the planktonic and biofilm phases of growth (2). A subsequent experiment showed that they were unable to persist in such a system in the absence of facultative species, even in biofilms (4). These studies suggested that intimate interspecies interactions mediated anaerobe survival under these conditions.

Coaggregation has been shown to be a highly specific mechanism by which dental plaque bacteria may interact physically. Most authors have described the potential role of coaggregation in terms of the formation of dental plaque biofilms and, in particular, in accretion of secondary colonizers to the pioneer species in plaque, and thus the development of a spatially organized community (14). It is, however, also possible that coaggregation may provide some metabolic advantage (e.g., cross-feeding, enzyme complementation) to neighboring cells by facilitating close physical juxtaposition of partner cells, as has been shown for glucose metabolism of mixtures of Actinomyces and streptococci (10, 11).

The aim of the present study was initially to examine the pair-wise coaggregation patterns of the 10 bacterial species included in the chemostat system and, in particular, the interactions between obligately anaerobic and oxygen-tolerant species. Following on from this, the ability of Fusobacterium nucleatum and other coaggregating strains to act as bridges between otherwise-noncoaggregating obligately anaerobic-oxygen-tolerant species pairs was examined. Finally, the survival of anaerobes in planktonic and biofilm communities in which F. nucleatum was omitted was assessed to determine whether a relationship existed between coaggregation mediated by this organism and survival in hostile conditions.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Bacterial strains. The following bacteria were used in the study: Streptococcus mutans R9, Streptococcus sanguis 209, Streptococcus oralis EF186, Lactobacillus rhamnosus AC413, Actinomyces naeslundii WVU627, Neisseria subflava A1078, Veillonella dispar ATCC 17745, Porphyromonas gingivalis W50, Prevotella nigrescens T588, and F. nucleatum ATCC 10953. The strains were maintained in 10% glycerol in BM broth in the gas phase above liquid nitrogen (at approximately 130°C).

Coaggregation assay. Each species was inoculated into 1 liter of BM medium (20) and grown at 37°C. N. subflava was grown in a static culture in air for 3 days; S. sanguis, S. oralis, S. mutans, A. naeslundii, and L. rhamnosus were grown in static cultures in an anaerobic atmosphere (80% N₂, 10% CO₂, 10% H₂) for 4 days; and V. dispar, F. nucleatum, P. nigrescens, and P. gingivalis were grown in static cultures under an anaerobic atmosphere for 7 days. Bacteria were harvested by centrifugation at 5,000 × g for 20 min and resuspended in coaggregation buffer (7). This buffer contains 0.1 mM CaCl₂, 0.1 mM MgCl₂, 0.15 M NaCl, and 3.1 mM NaN₃, dissolved in 1 mM Tris adjusted to pH 8. The strains were washed twice in coaggregation buffer and finally resuspended to give an optical density of 2.0 at 660 nm. Equal volumes of 2 ml of each bacterial suspension were mixed together by vortex mixing, with each possible pair of species, in test tubes. Individual bacterial suspensions were dispensed alone as controls. Coaggregation was scored on the following scale after 90 min, as described by Cisar et al. (7): 0, no coaggregation visible; 1+, small coaggregates, remaining in solution; 2+, larger coaggregates, not immediately falling out of solution; 3+, large coaggregates, immediately clearing, slightly turbid suspension; 4+, large coaggregates, settling immediately.

Continuous-culture growth conditions and inoculation. The continuous-culture system consisted of a two-stage system, set up as described previously (2). Briefly, the first-stage vessel was operated anaerobically (gas phase, 5% CO₂ in N₂) and was linked to a second-stage vessel, which was supplied with a gas phase of 200 ml of 5% CO₂ in air · min¹. The first-stage vessel was a conventional fermentor vessel, but the second-stage vessel consisted of a wide-top vessel with a top plate modified to allow the aseptic insertion and removal of hydroxyapatite discs. The growth medium was BMHGM medium (a diluted basal peptone medium supplemented with porcine gastric mucin) (5), supplied to the first-stage culture at 50 ml · h¹, to give a dilution rate of 0.1 h¹ (corresponding to a mean generation time of 6.9 h). The overflow weir of the first-stage vessel was connected to the second-stage vessel, thus giving a 50-ml · h¹ flow rate of culture into the second-stage vessel. In addition, fresh medium was supplied to the second-stage vessel at a rate of 75 ml · h¹. Culture pH was maintained in each vessel at 7.0 ± 0.1 (set point ± range) by the automatic addition of 2 M NaOH, and the temperature was maintained at 37 ± 0.1°C.

Statistical analysis. Log₁₀-transformed counts of total numbers of bacteria and of individual species in the two aerated cultures (i.e., with and without F. nucleatum) were compared by using Student's t test, with significance assumed at a P of <0.05.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Pair-wise coaggregation. The aerobe N. subflava coaggregated only poorly with the obligate anaerobes P. gingivalis, P. nigrescens, and V. dispar (Table 1). All strains coaggregated with F. nucleatum, and this species also auto-aggregated strongly (Table 1). Coaggregations between F. nucleatum and A. naeslundii, F. nucleatum and S. mutans, and F. nucleatum and S. oralis gave unequivocal 4+ results. S. sanguis coaggregated strongly with A. naeslundii, L. rhamnosus, and N. subflava; L. rhamnosus also coaggregated strongly with S. mutans, S. oralis, and N. subflava and, to some extent, with S. sanguis, P. gingivalis, and P. nigrescens.

Three-species coaggregation. F. nucleatum was able to act as a bridge between all of the otherwise-noncoaggregating (or weakly coaggregating) pairs of aerobe-anaerobe species, in all cases tested (Table 2). Despite the strong auto-aggregation of F. nucleatum, the 3+ scores for the three-species coaggregates (large coaggregates, immediately clearing, leaving a slightly turbid suspension), in comparison with generally very low scores for the pairs (0 to 1+, corresponding either to no coaggregation at all or to small coaggregates, remaining in solution), indicate that the majority of the noncoaggregating bacteria were precipitated by the addition of this species.

Chemostat community composition in the first-stage (anaerobic) chemostat. The compositions of the steady-state communities that developed in the first-stage chemostat, in the presence and absence of F. nucleatum, are shown in detail in Table 3. The numbers of the majority of species in the two communities were not markedly different, although the viable counts of streptococci and N. subflava were increased significantly in the absence of F. nucleatum.

Community development in the aerated second-stage chemostat. (i) Environmental conditions. The environmental conditions in the aerated second-stage chemostats were broadly similar in the parallel experiments with and without F. nucleatum. The redox potential (E_h) was 259 ± 14 mV in the presence of F. nucleatum and 280 ± 65 mV in the absence of F. nucleatum. These values were not significantly different (t test, P = 0.55). Dissolved oxygen (dO₂) was only intermittently detectable in the culture at 0 to 20% saturation with the same gas mixture (5% CO₂ in air).

(ii) Planktonic communities. Data on the composition of steady-state, aerated planktonic cultures in the presence and absence of F. nucleatum are presented in Table 4. The total viable count of the community was slightly increased in the absence of F. nucleatum. N. subflava predominated in the planktonic phase of both of the aerated cultures, comprising 46 and 69% of the total CFU in the 9- and 10-species communities, respectively. Anaerobes were able to persist and grow in high numbers in the planktonic phase when F. nucleatum was present. However, in the absence of F. nucleatum, the proportions of all of the anaerobes were reduced, with the log-transformed numbers of the black-pigmented species (P. nigrescens and P. gingivalis) reduced significantly (t test, P < 0.001). In contrast, the numbers of other facultative species in the community (the three streptococci and two gram-positive rods) were significantly increased in the absence of F. nucleatum.

(iii) Biofilm communities. The data on the composition of 4-day biofilms are shown in Table 5, and these largely mirrored the composition of the planktonic phase. The total viable counts in the biofilms developed in the absence of F. nucleatum were increased, but the proportions of black-pigmented anaerobes were lower, and their log₁₀ viable counts were significantly reduced (t test, P  0.001). This decrease contrasted with a large and significant increase in the numbers of facultatively anaerobic species and also with an increase in the numbers of V. dispar.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Bacteria are found in many natural environments which, at least superficially, appear to be hostile. For example, obligately anaerobic bacteria may be recovered from a wide variety of habitats which are highly aerated (e.g., the oral cavity, the skin, and seawater). The bulk of natural bacterial populations exist in biofilms, and these may afford protection to the constituent organisms (8, 9). The majority of natural ecosystems support the growth of diverse mixed populations of bacteria, which often appear to have conflicting environmental requirements (e.g., aerobes and obligate anaerobes coexisting at a site) (15). A wide spectrum of physical, biochemical, and genetic interactions between different species enable bacteria in nature to exist as true communities, as suggested by Wimpenny (23). In multispecies biofilms, distinct activity domains may develop, depending on the spatial organization of bacteria and on their metabolic activities. As a result of these factors, spatial heterogeneity develops; this can allow sharp gradients of nutrients, pH, and oxygen, etc., to develop over relatively short distances. Thus, species with radically differing requirements can coexist in biofilms.

Chemostats have been used to model oral microbial communities (5, 18, 20), and surfaces have been introduced into these systems to allow biofilm development (3). Recently, we have examined the effects of aeration on the development of these oral bacterial communities. Initial experiments examined the importance of the "aerobe" N. subflava in the reaction of the community to oxygen. The N. subflava population was increased in aerated cultures, and the oxygen added was rapidly depleted. The numbers of anaerobes remained high (>10⁷ ml¹). In the absence of N. subflava, oxygen remained in higher concentrations in the culture and yet the anaerobes were still able to persist and grow in both the planktonic and biofilm phases of growth (2). A later study showed that anaerobes were able to survive only when facultative or aerobic species were present. In the absence of such species, the anaerobes died quickly, and even biofilms could not provide a haven for the anaerobes under these circumstances (4).

A number of previous studies have suggested an important role for F. nucleatum in the microbial ecology of the oral cavity. This species has been proposed as a key organism in bridging between early colonizers (such as streptococci) and bacteria (especially obligate anaerobes) more usually associated with mature dental plaque (14). This phenomenon has thus usually been considered in terms of facilitating an ordered microbial succession during the formation of dental plaque and in the development of a structured "climax plaque." In the present study, aerobes and facultative species were unable to coaggregate with any of the anaerobes except F. nucleatum. Thus, the protection afforded to these anaerobes by growth in a mixed culture, which was apparent in our previous studies (2, 4), could not be mediated by direct coaggregation. The coaggregation patterns of F. nucleatum suggested a possible role for this organism in this protective effect. Indeed, this study has shown that F. nucleatum could act as a bridge between otherwise-noncoaggregating pairs to form three-species aggregates.

A striking finding from these studies was the persistence of obligate anaerobes in aerated environments, not only in biofilms but also in planktonic cultures. When F. nucleatum was omitted from the inoculum, however, the composition of the aerated community was significantly perturbed. In particular, the viable counts of the black-pigmented anaerobes P. gingivalis and P. nigrescens were significantly reduced (1,000- and 100-fold, respectively). The fact that these effects were observed in both biofilm and planktonic phases of growth suggests that ordered, metabolically organized aggregates were responsible for the persistence of the anaerobes. The data suggest that gradients in oxygen must be generated within these aggregates over relatively short distances. Metabolism of aerobic and/or oxygen-tolerant species may reduce the concentration of oxygen to levels that the defenses of obligate anaerobes are able to detoxify (15). These microbial community effects result from intimate physical contact, spatial organization, and efficient metabolic coupling, as was reported with glucose metabolism by coaggregating Streptococcus-Actinomyces pairs (10, 11). This study indicates that microbial community effects may be as important as, if not more important than, the beneficial effects often ascribed to the biofilm mode of growth for oral bacteria.

The findings of this study have shown that coaggregation may provide additional benefits to the interacting species, beyond promoting adherence and facilitating bacterial succession. In addition to allowing close coupling of obligate anaerobes with oxygen-tolerant species, coaggregation may encourage other important interactions. Mixed consortia are required for the efficient degradation of complex host proteins and glycoproteins (1, 12, 13, 21, 22). It may be that coaggregation is also a strategy for increasing the probability of interactions between species with complementary metabolic capabilities. This helps to explain the powerful homeostatic properties of oral microbial communities, despite regular perturbations (16). It also follows that interfering with community interactions may be a promising route towards an ecological approach to plaque control (17).