distribution of selected bacterial species on intraoral surfaces

Distribution of selected bacterialspecies on intraoral surfaces

Donna L. Mager,Laurie Ann Ximenez-Fyvie,Anne D. Haffajee andSigmund S. SocranskyDepartment of Periodontology, The Forsyth

Institute, Boston, MA, USA

Mager DL, Ximenez-Fyvie LA, Haffajee AD, Socransky SS: Distribution of selectedbacterial species on intraoral surfaces. J Clin Periodontol 2003; 30: 644–654.r Blackwell Munksgaard, 2003.

AbstractBackground/aim: To examine the proportions of 40 bacterial species in samples from8 oral soft tissue surfaces and saliva in systemically healthy adult subjects and tocompare these microbiotas with those of supra- and subgingival plaque.

Methods: Microbial samples were taken from 8 oral soft tissue surfaces of 225systemically healthy subjects using a ‘‘buccal brush’’. Saliva was taken byexpectoration. Forty-four of these subjects provided additional supra- and subgingivalplaque samples. Samples were individually evaluated for their content of 40 bacterialspecies using checkerboard DNA–DNA hybridization. The percentage of total DNAprobe count was determined for each species, at each sample location and averagedacross subjects. The significance of differences among the proportions of the 40 testspecies at different sample locations was sought in the 225 and 44 subjects separatelyusing the Quade test and adjusted for multiple comparisons. Cluster analysis wasperformed using the proportions of the 40 species at the different sample locationsusing the minimum similarity coefficient and an average unweighted linkage sort. Theproportions of each species were averaged across subjects in the resulting clustergroups and the significance of differences was tested using the t-test and ANOVA.

Results: Microbial profiles differed markedly among sample locations in the 225subjects, with 34 of 40 species differing significantly. Proportions of Veillonellaparvula and Prevotella melaninogenica were higher in saliva and on the lateral anddorsal surfaces of the tongue, while Streptococcus mitis and S. oralis were insignificantly lower proportions in saliva and on the tongue dorsum. Cluster analysisresulted in the formation of 2 clusters with 485% similarity. Cluster 1 comprisedsaliva, lateral and dorsal tongue surfaces, while Cluster 2 comprised the remaining softtissue locations. V. parvula, P. melaninogenica, Eikenella corrodens, Neisseriamucosa, Actinomyces odontolyticus, Fusobacterium periodonticum, F. nucleatum ssvincentii and Porphyromonas gingivalis were in significantly higher proportions inCluster 1 and S. mitis, S. oralis and S. noxia were significantly higher in Cluster 2.These findings were confirmed using data from the 44 subjects providing plaquesamples. The microbial profiles of supra- and subgingival plaque differed from theother sample locations, particularly in the increased proportions of the Actinomycesspecies. Species of different genera exhibited different proportions on the variousintraoral surfaces, but even within the genus Streptococcus, there were differences incolonization patterns. S. oralis, S. mitis and S. constellatus colonized the soft tissuesand saliva in higher proportions than the samples from the teeth, while the other 4streptococcal species examined colonized the dental surfaces in proportionscomparable to the soft tissue locations and saliva.

Conclusions: Proportions of bacterial species differed markedly on different intraoralsurfaces. The microbiota of saliva was most similar to that of the dorsal and lateralsurfaces of the tongue. The microbiotas of the soft tissues resembled each other morethan the microbiotas that colonized the teeth both above and below the gingivalmargin.

Key words: soft tissue microbiota; saliva;periodontal disease; supra- and subgingivalplaque; systemically healthy

Accepted for publication: 30 September 2002

The mean surface area of the adultoral cavity is approximately 215 cm2

(Collins & Dawes 1987). The teeth,keratinized and nonkeratinized soft

tissues comprise about 20%, 30%and 50% of this surface area respec-tively. While a great deal is knownabout the microbial composition of

hard tissue biofilms, surprisingly littleis known about the microbiotas thatcolonize approximately 80% of thesurface area of the oral cavity. Most

J Clin Periodontol 2003; 30: 644–654 Copyright r Blackwell Munksgaard 2003Printed in Denmark. All rights reserved

studies that examined the microbiotaon oral mucous membranes haveemployed relatively time-consumingcultural techniques. For this reason, thecomposition of the oral soft tissuemicrobiota is less well understoodthan that of dental plaque, wheremore studies, subjects, samples andspecies have been examined often bymeans of more rapid microbiologicaltechniques.

Studies that have sought the presenceof specific bacterial species on differentoral soft tissue surfaces have beenmotivated, primarily, because of possi-ble associations of soft tissue residentswith various clinical conditions includ-ing oral malodor, dental caries andperiodontal diseases. A number ofstudies have shown that soft tissuesurfaces harbor periodontal pathogens(van Winkelhoff et al. 1986, Dahlenet al. 1992, Loos et al. 1992, Bosy et al.1994, De Boever & Loesche 1995,Muller et al. 1995, 1997, Quirynenet al. 1995, 1998, Matto et al. 1996a,b,1998, Lee et al. 1999). For example,van Winkelhoff et al. (1986) character-ized the microbiotas of the tongue andtonsils using culture and phase contrastmicroscopy. They showed that spiro-chetes, motile organisms and black-pigmented species such as Prevotellaintermedia colonized the tongue.Further, there was an associationbetween periodontal breakdown andthe presence of these organisms onthe tongue, suggesting that the tonguecould serve as a significant ecologicalhabitat for periodontal pathogens. Dah-len et al. (1992) examined the tongue ofperiodontally diseased and nondiseasedyoung adult Kenyan subjects for thepresence of 7 putative periodontalpathogens. All test species includingPorphyromonas gingivalis, P. interme-dia, Campylobacter rectus and Actino-bacillus actinomycetemcomitans couldbe detected in samples from both groupsof subjects, although P. gingivalis wasdetected significantly more frequentlyin the tongue samples from the period-ontally diseased subjects. In anotherstudy of P. gingivalis, Loos et al.(1992) analyzed saliva and microbialsamples from the tongue dorsum,tonsil and attached gingiva of 8 adultsubjects with untreated periodontitisand 1 patient with an untreated infectedroot canal. Their data indicated thatP. gingivalis could colonize multipleecological sites and that subjectscould be infected with more than one

clonal type. Danser et al. (1996) studiedthe effect of SRP and periodontalsurgery on the levels of A. actinomyce-temcomitans, P. gingivalis and P. inter-media on the oral mucous membranesusing indirect immunofluorescence.Improvement in clinical parametersafter treatment was accompanied by asignificant decrease in the subgingivalprevalence of the 3 putative periodontalpathogens, but essentially no reductionin the prevalence of the test bacteriaon the oral mucous membranes.The studies of Muller et al. (1995,1997) and Asikainen et al. (1991)demonstrated that A. actinomycetemco-mitans could be detected on the tonguedorsum, buccal mucosa and in saliva.Although no difference in the detectionrate of A. actinomycetemcomitans wasfound between periodontally healthyand diseased subjects, A. actinomyce-temcomitans was recovered moreoften on soft tissues and in saliva thanin plaque samples (Asikainen et al.1991).

Studies of oral malodor have con-tributed to our understanding of micro-bial colonization of oral soft tissues,particularly the dorsum of the tongue.These studies indicated that a widerange of species including Treponemadenticola, P. gingivalis, Bacteroidesforsythus, Prevotella melaninogenica,P. intermedia, Fusobacterium, Strepto-coccus, Lactobacillus, Rothia, Capno-cytophaga, and Actinomyces spp. couldbe found in samples taken from thetongue (Bosy et al. 1994, De Boever &Loesche 1995). Further, sulfate-redu-cing bacteria, which may contribute tooral malodor, were found in samplestaken from the tongue as well as thehard palate and vestibulum (Langendijket al. 1999).

The above studies suggest that softtissues may serve as reservoirs forinfection or reinfection of the period-ontium and may deserve therapeuticattention (Quirynen et al. 2001). How-ever, to date there have been fewinvestigations of a broad range ofbacterial species in multiple hardand soft tissue samples taken from thesame subjects. Thus, the purpose ofthe present investigation was to exam-ine the proportions of 40 bacterialspecies in samples from 8 oral softtissue surfaces and saliva in systemi-cally healthy adult subjects. In addition,the microbiotas from the soft tissuesurfaces were compared with samples ofsupra- and subgingival plaque.

Material and Methods

Subject population

Two hundred and twenty-five systemi-cally healthy subjects 418 years of agewith at least 20 teeth were selected forthe study. Periodontally healthy subjectshad no pocket depth or attachment levelmeasurements 43 mm, but could havegingival inflammation. Periodontallydiseased subjects had at least 4 siteswith pocket depths 44 mm and 4 siteswith attachment level measurements44 mm. Subjects were excluded if theyhad received antibiotic therapy withinthe previous 3 months, had any systemiccondition that could affect the host’speriodontal status (e.g. diabetes, AIDS),had any condition that would requireantibiotics for monitoring or treatmentprocedures (e.g. heart conditions, jointreplacements), or had mucosal lesions,previous chemotherapy, radiation ther-apy or medications that cause xerosto-mia. Subjects were chosen from thepatient pool at The Forsyth Institute.The purpose and nature of the studyincluding the types of clinical measure-ments and sample collection were ex-plained to all potential subjects. Afterreading and signing the consent form,the subjects were entered into the study.

Clinical monitoring

After initial screening for suitability andobtaining informed consent, subjectswere clinically and microbiologicallymonitored. Clinical measurements weretaken at 6 sites per tooth (mesiobuccal,buccal, distobuccal, distolingual, lingualand mesiolingual) at all teeth excludingthird molars (a maximum of 168 sitesper subject) at a baseline visit (Haffajeeet al. 1983). The clinical parametersmeasured were gingival redness (0 or1), pocket depth (mm), bleeding onprobing (0 or 1), suppuration (0 or 1),attachment level (mm) and plaqueaccumulation (0 or 1). Saliva, oralmucosa and supra- and subgingivalplaque samples for microbiologicalassessment were taken prior to theclinical measurements. The mean base-line clinical parameters for the 225systemically healthy subjects are shownin Table 1. Two hundred and twenty-five subjects provided soft tissue sam-ples and 44 of these subjects alsoprovided samples of supra- and sub-gingival plaque taken separately fromeach tooth.

Microbiotas of intraoral surfaces 645

Collection of samples

Subjects expectorated whole saliva intosterile Eppendorf microcentrifuge tubes.A 0.2 ml sample of whole saliva wasvortexed with 0.15 ml sterile, filteredTris EDTA buffer (TE; 10 mM Tris-HCl, 1 mM EDTA, pH 7.6). A 0.2 mlsample of this mixture was takenand 0.1 ml of 0.5 M NaOH added.Subsequently, microbial samples of 8soft tissue sites were taken from eachsubject. Each mucosal sample wascollected using a MasterAmpt buccalswab brush (Epicentre Technologies,Madison, WI, USA). The soft tissuesamples were obtained by gently strok-ing each site in an area large enough toyield sufficient numbers of microorgan-isms to examine. The size of the areaand length of sampling time werelocation dependent. Soft tissue sitesexamined were 3 areas of the tongue,dorsum, lateral, and ventral surfaces, thefloor of the mouth, buccal mucosa, hardpalate, anterior vestibule and mucosa ofthe maxillary and mandibular lips, andthe maxillary anterior attached gingiva.Sampling areas and times were asfollows: 1 cm2 of the center of thedorsum tongue for 5 s, full length ofboth sides of the lateral tongue for 5 seach, and the entire ventral tongue for10 s. The right and left sides of the floorof the mouth required 10 s each. Theright and left buccal mucosas requiredsampling the entire area of both sidesfor 10 s each, without touching theteeth. The entire hard palate wassampled for 10 s. The vestibule/lip wassampled 10 s each for the maxillary andmandibular areas. Finally, the attachedgingiva of the maxillary anterior wassampled for 10 s. Soft tissue sampleswere placed into individual tubes con-taining 150ml of TE buffer to which100 ml of 0.5 M NaOH was added.

Samples of supra- and subgingivalplaque were obtained from up to 28teeth in each of the subset of 44subjects. After drying and isolation withcotton rolls, supragingival plaque wassampled from the mesiobuccal aspect ofeach tooth using sterile Gracey curettes.Each plaque sample was placed inindividual tubes containing 150ml ofTE buffer. After removal of the supra-gingival sample and any remainingsupragingival plaque, subgingival pla-que samples were taken from the samesites (i.e., the mesiobuccal aspect ofeach tooth) using sterile Gracey curettesand placed in similar individual tubesalso containing 150ml of TE buffer. Allsamples from all subjects were indivi-dually analyzed for their content of 40bacterial species using the checkerboardDNA–DNA hybridization technique.The 40 bacterial species studied arelisted in Fig. 1.

Microbiological assessment

Samples were evaluated using a mod-ification (Haffajee et al. 1997) of thecheckerboard DNA–DNA hybridizationtechnique (Socransky et al. 1994). Thesamples were lysed and the DNA placedin lanes on a nylon membrane using aMinislot device (Immunetics, Cam-bridge, MA, USA). After fixation ofthe DNA to the membrane, the mem-brane was placed in a Miniblotter 45(Immunetics) with the lanes of DNA at901 to the lanes of the device. Digox-igenin-labeled whole genomic DNAprobes to 40 bacterial species werehybridized in individual lanes of theMiniblotter. After hybridization, themembranes were washed at high strin-gency and the DNA probes detectedusing antibody to digoxigenin conju-gated with alkaline phosphatase and

chemifluorescence detection. Signalswere detected using AttoPhos substrate(Amersham Life Science, ArlingtonHeights, IL, USA) and were read usinga Storm Fluorimager (Molecular Dy-namics, Sunnyvale, CA, USA), a com-puter-linked instrument that read theintensity of the fluorescent signalsresulting from the probe–target hybridi-zation. Two lanes in each run containedstandards at concentrations of 105 and106 cells of each species. The sensitivityof the assay was adjusted to permit thedetection of 104 cells of a given speciesby adjusting the concentration of eachDNA probe. Signals evaluated using theStorm Fluorimager were converted toabsolute counts by comparison with thestandards on the same membrane. Fail-ure to detect a signal was recorded aszero.

Data analysis

Data available for all subjects were thecounts of 40 bacterial species in salivaand from 8 different oral mucousmembrane locations. In addition, sam-ples of supragingival and separatelysubgingival plaque from all teeth, ex-cluding third molars, were availablefrom 44 subjects. For the plaque sam-ples, the total DNA probe count for eachspecies was computed at each sampledsite in each subject and averaged toprovide a single supra- and subgingivalvalue for each species for each subject.The proportion (percentage of the DNAprobe count) that each species com-prised of the total DNA probe count ateach sample location was computed.The significance of differences amongthe proportions of the 40 test speciesin saliva and at the 8 mucous membranelocations in the 225 subjects weresought using the Quade test andadjusted for multiple comparisons(Socransky et al. 1991). A similaranalysis was performed comparingsaliva, supra- and subgingival plaqueand the 8 oral soft tissue locationsin samples from the 44 subjects whoprovided plaque samples.

Cluster analysis was performed onthe proportions of the 40 species at thedifferent sample locations in the 225subjects. Similarities were computedusing the minimum similarity coeffi-cient (Socransky et al. 1982) and sortedusing an average unweighted linkagesort (Sneath & Sokal 1973). Theproportions of each species were aver-aged across subjects in the different

Table 1. Mean (7SD) baseline clinical characteristics of the 225 systemically healthy subjectsand the subset of 44 subjects providing plaque samples

Soft tissue samples only Soft and hard tissue samples

N 225 44age (years) 40716 43716number of missing teeth 2.072.3 1.972.2% males 42 41mean pocket depth (mm) 2.770.7 2.770.5mean attachment level (mm) 2.671.1 2.671.0

% of sites withplaque accumulation 65743 87761gingival redness 53731 62738bleeding on probing 23721 29722

646 Mager et al.

cluster groups and the significance ofdifferences was tested using the t-testand ANOVA. The analysis was repeatedusing the data from the 44 subjects whoprovided plaque samples.

ResultsComparison of soft tissue samples and

saliva

Fig. 1 presents the mean percentage ofthe DNA probe count of samples fromthe 8 intraoral mucous membrane loca-tions and saliva in the 225 subjects. Themicrobial profiles differed markedlyamong sample locations, with 34 of 40species differing significantly amongsample locations even after adjusting for40 comparisons. In particular, proportionsof Veillonella parvula and P. melanino-genica were higher in saliva and on thelateral and dorsal surfaces of the tongue,while Streptococcus mitis and S. oraliswere in significantly lower proportionsin saliva and on the tongue dorsumcompared with the other sampled sites.

Cluster analysis was performed toseek similarities in microbial profilesamong the 9 sample locations (Fig. 2).

Two clusters were formed with 485%similarity. One cluster comprised saliva,and the lateral and dorsal tonguesurfaces, while a second cluster wasmade up of the remaining soft tissuelocations. The species that differentiated

the 2 cluster groups are shown in Fig. 3.V. parvula, P. melaninogenica, Eikenel-la corrodens, Neisseria mucosa, Acti-nomyces odontolyticus, Fusobacteriumperiodonticum, F. nucleatum ss vincen-tii and P. gingivalis were in significantly

Fig. 1. Mean percentage DNA probe count (7SEM) for samples from the 8 oral soft tissue locations and saliva taken from 225 systemicallyhealthy adult subjects. Species are ordered according to their mean proportions in saliva. Significance of differences among sample locationswas determined using the Quade test and adjusted for multiple comparisons (Socransky et al. 1991). The shaded bars represent species thatshowed particularly marked differences among sample locations.

Fig. 2. Dendrogram of a cluster analysis of the mean species proportions from the 9 samplelocations. A minimum similarity coefficient was employed and an average unweightedlinkage sort. Two clusters were formed at 485% similarity.


higher proportions in Cluster 1 (saliva–lateral–dorsal tongue) compared withCluster 2 (the remaining soft tissuelocations). In contrast, the mean propor-tions of S. mitis, S. oralis and Seleno-monas noxia were significantly higherin Cluster 2 than Cluster 1.

Comparison of soft tissue samples, saliva

and samples of supra- and subgingival

plaque

The mean percentage of the DNA probecount of samples from the 8 intraoralmucous membrane locations, saliva andsupra- and subgingival plaque in asubset of 44 of the 225 subjects ispresented in Fig. 4. Once again therewere marked differences in microbialprofiles among the different samplelocations, with 35 of 40 test speciesdiffering significantly among groups. Inparticular, supra- and subgingival pla-que samples harbored significantlyhigher proportions of Actinomyces spe-cies compared with saliva or the softtissue locations. Saliva and the dorsumof the tongue exhibited the highestproportions of P. melaninogenica.These sample sites as well as the lateralsurface of the tongue harbored highproportions of V. parvula. As shown inFig. 1, the other soft tissue locationstended to exhibit high proportions ofstreptococcal species as well as Gemella

morbillorum. Cluster analysis of the 11sample locations reaffirmed the 2 clus-ters observed in Fig. 2 and indicated thatsamples of supra- and subgingivalplaque differed markedly in composi-tion from the other locations (Fig. 5).The species that differed among clustergroups are presented in Fig. 6. Similarto the individual locations, the hardtissue biofilm samples (Cluster A) weredominated by Actinomyces species,Cluster B (saliva–lateral–dorsal tongue)by P. melaninogenica and V. parvula,and Cluster C (remaining soft tissuelocations) by S. mitis, S. oralis, Strepto-coccus constellatus, Capnocytophagagingivalis and G. morbillorum.

Proportions of specific species at

different intraoral locations

Several of the oral species examinedexhibited different proportions on dif-ferent intraoral surfaces (Fig. 7). Forexample, A. naeslundii genospecies 2was found in high proportions in supra-and subgingival plaque samples but inlow proportions in the soft tissuesamples and saliva. Eubacterium sabur-reum and to some extent S. mitisshowed the opposite pattern of coloni-zation. These species were found in lowproportions on the hard tissues but inhigh proportions on the soft tissuesurfaces. P. melaninogenica was found

in high proportions on the lateral anddorsal surfaces of the tongue and in lowproportions on the hard tissues and hardpalate. The periodontal pathogen B.forsythus was found in highest propor-tions in subgingival plaque, whileLeptotrichia buccalis was found insimilar proportions in all sample loca-tions. Even within a genus such asStreptococcus, there were differences incolonization patterns (Fig. 8). S. oralis,S. mitis and S. constellatus colonizedthe soft tissues and saliva in higherproportions than the samples from theteeth. The other 4 streptoccocal speciescolonized the dental surfaces in propor-tions comparable to the soft tissuelocations and saliva.

Discussion

It has been shown that oral bacteriademonstrate specific tropisms towardthe different biological surfaces inthe oral cavity such as the teeth,mucosa and other bacteria (Gibbons1989). The nonshedding surfaces ofthe teeth offer a far different habitatthan the continually shedding surfacesof the oral mucosa. Owing to therepeated shedding of epithelial cells,there is less time for a complex bio-film to develop on soft tissue surfaces;thus, a premium is placed on potentmechanisms of adhesion. Clearly, the

Fig. 3. Mean microbial proportions (7SEM) of the 2 clusters in Fig. 2. The left panel indicates mean percentage of the 40 bacterial species ofthe saliva and lateral surface and dorsum of the tongue cluster (Cluster 1), while the right panel presents the mean proportions of the 40bacterial species at the 6 remaining sampled surfaces (Cluster 2). Significance of differences between cluster groups was determined using thet-test: npo0.05; nnpo0.01; nnnpo0.001. The shaded bars represent the species that were significantly higher in proportion in a given cluster.

648 Mager et al.

differences in habitat and colonizationfor the 40 bacterial species studiedsuggest that different intraoral surfacesand bacterial species have differentreceptors and adhesion molecules thatdictate the development of biofilms ondifferent oral surfaces.

The purpose of the present investiga-tion was to compare the proportions of40 bacterial species in samples from8 intraoral soft tissue locations, salivaand supra- and subgingival plaque.The results indicated that the majordifference in colonization patterns was

between samples from the hard tissuesand the various intraoral soft tissuelocations and saliva. Actinomycesspecies colonized hard tissues at farhigher proportions than soft tissues,while P. melaninogenica, V. parvulaand S. mitis were found in higherproportions on soft tissue surfaces.Supra- and subgingival plaque mostclosely resembled each other in micro-bial composition and the microbiota ofsaliva resembled that of the dorsum andlateral surfaces of the tongue. However,clear differences were found in themicrobiotas that colonized the differentintraoral soft tissues. Indeed, sites thatmight be expected to be nearly identicalin microbial profile were found to bedistinctly different. For example, themicrobiota of the keratinized attachedgingiva might be expected to closelyresemble that of the keratinized hardpalate, yet it was more similar to that ofthe nonkeratinized floor of the mouth.

The present investigation demon-strated that Actinomyces spp. (especiallyA. naeslundii genospecies 1 and 2)colonized teeth at far greater propor-tions than the soft tissues. Membersof this species are known to be a

Fig. 4. Mean percentage DNA probe count (7SEM) for samples from the 11 intraoral locations in the subset of 44 of the 225 subjects whoprovided samples of supra- and subgingival dental plaque. Species are ordered according to their mean proportions in supragingival plaque.Significance of differences among sample locations was determined using the Quade test and adjusted for multiple comparisons (Socranskyet al. 1991). The shaded bars represent species that showed particularly marked differences among sample locations.

Fig. 5. Dendrogram of a cluster analysis of the mean species proportions from the 11 samplelocations. A minimum similarity coefficient was employed and an average unweightedlinkage sort. Three clusters were formed at 475% similarity. Cluster B was more similar toCluster C than Cluster A.


Fig. 6. Mean microbial proportions (7SEM) of the species in the 3 clusters in Fig. 5. The left panel (Cluster A) indicates mean percentage ofthe 40 bacterial species in supra- and subgingival plaque samples; the middle panel presents the mean microbial proportions of species inCluster B (saliva, lateral surface and dorsum of the tongue), while the right panel presents the mean proportions of the 40 bacterial species atthe 6 remaining sampled surfaces (Cluster C). Significance of differences among cluster groups was determined using ANOVA: npo0.05;nnpo0.01; nnnpo0.001. The shaded bars represent the species that were significantly higher in proportion in a given cluster.

Fig. 7. Mean proportion (7SEM) of 6 selected species in samples from the 11 intraoral locations. Significance of differences among samplelocations was determined using the Quade test and adjusted for multiple comparisons (Socransky et al. 1991). The bars are shaded accordingto the cluster groups in Fig. 5.

650 Mager et al.

heterogeneous group, which can beprominent in dental plaque. Liu et al.(1991) showed that some strains of A.naeslundii bind to collagen, a moleculepresent in the matrices of both cemen-tum and dentin. This ability to bind tocollagen may be an important factor inactinomyces colonization at gingivaland subgingival sites. Strains of Actino-myces spp. differ in their abilities toattach to salivary components (Gibbons1989). Stromberg & Boren (1992)suggested that receptor specificities ofbacterial cell-surface adhesins mightdetermine the abilities of differentActinomyces strains to colonize differ-ent oral sites. Hallberg et al. (1998)found that 102 strains of Actinomycesisolated from teeth, buccal mucosa andtongue in 8 individuals could be classi-fied into 3 major groups based onbinding and coaggregation properties.A. naeslundii genospecies 1 was foundto be prevalent on teeth, A. naeslundiigenospecies 2 was the dominant Acti-nomyces species on both teeth andbuccal mucosa, while A. odontolyticuswas the dominant species on the tongue.In the present investigation the findingswere similar, with the exception thatA. naeslundii genospecies 1 was moredominant on the buccal mucosa than A.naeslundii genospecies 2. Hallberg et al.(1991) found that the colonizationpatterns of the 3 species correlated wellwith the binding specificities of eachspecies to beta-linked galactosamine,

acidic proline-rich protein structuresand beta-linked galactosamine-inhibita-ble coaggregation with selected strainsof Streptococcus species. The authorssuggested that surface-associated adhe-sion molecules, as well as surfacestructures (namely fimbriae), may med-iate the intraoral colonization and dis-tribution of Actinomyces species.

There were even greater differencesin colonization patterns within thegenus Streptococcus. Other investiga-tors have observed differences in colo-nization patterns of species in thisgenus. van Houte et al. (1971) examinedthe adherence of labeled Streptococcusspecies to dental plaque and oralepithelial surfaces in vivo. The propor-tions of labeled S. sanguis recoveredfrom clean teeth or preformed dentalplaques were much higher than those oflabeled Streptococcus salivarius andsimilar to those observed for S. sanguisin naturally occurring oral biofilms.Proportions of labeled S. salivariuswere higher than those of S. sanguison the dorsum of the tongue but loweron the vestibular mucosa. LabeledStreptococcus mutans was found inlower proportions than S. sanguis andS. salivarius on the soft tissue surfaces.The authors indicated that the relativeadherence of the Streptococcus speciesto the tongue and vestibular mucosacorrelated with their proportions foundnaturally in these sites. Frandsen et al.(1991) investigated the colonization of

oral and pharyngeal surfaces by viridansstreptococci. Samples taken from thebuccal mucosa, tongue dorsum, oro-pharynx, supra- and subgingival plaquewere evaluated culturally for 7 speciesof streptococci. Different species wereassociated with specific oral surfaces.For example, IgA1 protease activity wasassociated almost exclusively withstreptococcal species colonizing acleaned tooth or buccal mucosal sur-face. S. mitis biovar 1 and S. sanguisproduced IgA1 protease that mightfacilitate colonization of sites such asbuccal mucosa or the pellicle of acleaned tooth. In contrast, habitats inwhich protease activity was unnecessaryfor colonization (mature plaque ortongue dorsum) were colonized by S.mitis biovar 2, Streptococcus gordoniiand S. oralis, species that do notproduce IgA1 protease. Our findingswere similar to those of Frandsen et al.(1991), in that the most prominentstreptococci detected on the buccalmucosa was S. mitis, although wedid not distinguish between biovars1 and 2.

Neeser et al. (1995) showed that S.sanguis OMZ 9 bound to human buccalepithelial cells in a sialic acid-sensitivemanner, suggesting that S. sanguis maycolonize both hard and soft tissues bybinding to salivary glycoproteins withsialic acid residues. In accord with thesefindings, the results of the presentinvestigation indicated that S. sanguis

Fig. 8. Mean proportion (7SEM) of 7 Streptococcus species in samples from the 11 intraoral locations. Significance of differences amongsample locations was determined using the Quade test and adjusted for multiple comparisons (Socransky et al. 1991). The bars are shadedaccording to the cluster groups in Fig. 5.


colonized both teeth and soft tissues incomparable proportions.

Sklavounou & Germaine (1980) sug-gested that keratinization of epithelialcells was likely to be a significant factorin the adherence of oral streptococci.The results of the present investigationwere in contrast to those findings.Cluster analysis using a minimumsimilarity coefficient found that kerati-nization of epithelial cells did notappear to be a significant factor in theadherence of oral bacteria, includingstreptococci, to the oral mucosa. Forexample, S. mitis colonized the hardpalate and buccal mucosa at similarproportions, while S. oralis was found atsimilar proportions on the hard palateand floor of the mouth, and S. con-stellatus was at similar proportions onthe attached gingiva and ventral tongue.In fact, the microbial profile of kerati-nized attached gingiva was more similarto that of the buccal mucosa than to thatof the hard palate.

In accord with many studies in theliterature (Krasse 1953, 1954, Gibbonset al. 1964, Beighton et al. 1987), thecomposition of the microbiota in salivawas most closely related to that of thedorsum of the tongue. Krasse (1953,1954) examined the levels of S. salivar-ius in plaque, saliva, tongue, vestibuleand buccal mucosa samples. S. salivar-ius was found at higher levels insamples of saliva and the tongue thanin samples from the vestibule, buccalmucosa and teeth. Gibbons et al. (1964)found that S. salivarius and P. melani-nogenica preferentially colonized thetongue and saliva compared with theteeth and the buccal mucosa. Otherstudies have found similarities betweenthe microbiota recovered from thetongue and saliva samples. For example,Beighton et al. (1987) found that S.mutans levels in tongue samples weresignificantly correlated with levels ofthis species in saliva samples.

This investigation supports the find-ings of Asikainen et al. (1991) that A.actinomycetemcomitans may be foundas often or more often on soft tissuesand in saliva than in dental plaque.Indeed, several investigators have sug-gested that oral colonization of a subjectby A. actinomycetemcomitans could besuccessfully determined by using oralmucosa or saliva samples (Muller et al.1995, Eger et al. 1996, Zoller et al.1996). Our data indicate that saliva orthe tongue dorsum may be the best sitesto sample the presence of A. actinomy-

cetemcomitans, while subgingival sam-ples may be superior for the detection ofB. forsythus. Further, in accord withUmeda et al. (1998), our data suggestthat saliva samples may be a gooddiagnostic indicator of the presence ofperiodontal pathogens including P. gin-givalis, P. intermedius, P. nigrescensand T. denticola.

In the present investigation, themicrobiota of the tongue was found tobe colonized predominantly by Gram-negative species including P. melanino-genica, V. parvula and C. gingivalis,rather than streptococci as reported byGordon & Gibbons (1966). However,the present study confirms numerousreports in the literature that the softtissues, especially the tongue, mayharbor periodontal pathogens includingP. gingivalis, T. denticola, A. actinomy-cetemcomitans and P. intermedia (vanWinkelhoff et al. 1986, Asikainen et al.1991, Dahlen et al. 1992, Bosy et al.1994, De Boever & Loesche 1995).These data suggest that the soft tissuesmay be an important reservoir forperiodontal pathogens and could be amajor factor in the recolonization oftooth surfaces after periodontal therapy.

There were limitations to the currentinvestigation. The 40 bacterial speciesthat were examined were those thoughtto be important in dental plaque. Thesespecies have been for many years instudies examining the composition ofdental plaque as well as in studiescomparing the therapeutic effects ofdifferent periodontal treatments. Thus,these species were useful for comparingthe microbiotas of oral hard and softtissues. This battery did not includespecies that may be important membersof the soft tissue microbiota such as S.salivarius and Streptococcus vestibu-laris. Nonetheless, a large number ofsubjects, samples and species wereexamined providing a unique perspec-tive on intraoral patterns of coloniza-tion. While differences in thecomposition of samples obtained fromhard and soft tissues were expected, themagnitude of the effect of samplelocation on taxa such as the Actino-myces or streptococci as well as thedifferences in the microbial profilesobtained from different soft tissuesurfaces were not anticipated. Althoughour results showed that the proportionsof 34 of 40 bacterial species differedsignificantly, some of the species werefound at such low proportions that theymay have no clinical significance. For

this reason, cluster analyses were per-formed to reveal similarities and dis-similarities in the colonization patternsof oral soft tissues. The similaritybetween microbiotas of saliva and thetongue was expected. However, thesimilarities between some of the tissuesin Cluster 2 were surprising.

The present investigation demon-strated that all of the tested speciescould be found, on average, on all of thesampled surfaces. The major differenceswere in the proportions that colonizedthe different surfaces, suggesting thatreceptors, coaggregation or local habitatdifferences play major roles in definingcommunity structure. Nevertheless, cer-tain locations showed greater microbialsimilarities than others. The microbialcomposition of saliva was most similarto that found on the lateral and dorsalsurfaces of the tongue, suggesting thatthese surfaces may be the major sourceof salivary bacteria. The microbiotascolonizing the remaining surfacesshowed greater similarities to eachother, but differences could be detectedamong surface locations. This wasinteresting given that both keratinizedand nonkeratinized surfaces were repre-sented in this group. The biofilmscolonizing the teeth were somewhatsimilar to each other but quite differentfrom the microbiotas on the oral softtissue surfaces and in saliva. However,as pointed out above, tooth-colonizingspecies could be detected on the softtissues. Thus, as discussed by manyinvestigators, the soft tissues may act asreservoirs for tooth-borne pathogensand may require therapeutic attention.Soft tissues and saliva may also provideconvenient sites for monitoring thera-peutic interventions.

Acknowledgments

This work was supported in part by NIHGrants K23 DE-00453, DE-12861, DE-10977, DE-12108.

Zusammenfassung

Verteilung ausgewahlter Bakterienarten aufintrooralen OberflachenZielsetzung: Untersuchung der Anteile von40 Bakterienarten in Proben von 8 oralenWeichgewebsoberflachen und aus dem Speichelvon allgemein gesunden Erwachsenen undVergleich mit der Verteilung dieser Mikroorga-nismen in supra- und subgingivaler Plaque.Methoden: Mikrobiologische Proben wurdenvon 8 oralen Oberflachen bei 225 allgemein

652 Mager et al.

gesunden Personen mittels einer Bukkalburstegewonnen. Speichel wurde durch Spuckengewonnen. Von 44 dieser Personen wurdenzusatzlich supra- und subgingivale Plaquepro-ben entnommen. Die Proben wurden einzelnauf ihren Gehalt von 40 Bakterienarten mittelsder Schachbrett-DNS-DNS-Hybridisierung un-tersucht. Der prozentuale Anteil an der GesamtDNS-Sondenzahl wurde fur jede Art an jederStelle bestimmt und pro Person gemittelt.Signifikante Unterschiede zwischen den Ante-ilen der 40 Testspezies an den unterschiedlichenEntnahmestellen wurde fur die 225 und 44Personen separat mittels des Quade-Tests nachKorrektur fur multiple Testung bestimmt. EineClusteranalyse wurde mit den Anteilen der 40Bakterienarten an den verschiedenen Entnah-mestellen mittels eines Minimum-Ahnlichkeits-Koeffizienten und einer durchschnittlichen un-gewichteten Verbindungsortierung durchge-fuhrt. Die Anteile jeder Bakterienart wurdenfur jede Person in den sich ergebendenClustergruppen gemittelt und die Signifikanzder Unterschiede mittels t-Test und ANOVAbestimmt.Ergebnisse: Die mikrobiologischen Profileunterschieden sich deutlich zwischen den un-terschiedlichen Entnahmestellen bei den 225Personen, wobei sich fur 34 von 40 Bakter-ienarten signifikante Unterschiede ergaben. DieAnteile von V. parvula und P. melaninogenicawaren im Speichel und an den lateralen unddorsalen Flachen der Zunge hoher, wahrend S.mitis und S. oralis in signifikant geringerenAnteilen im Speichel und am Zungenruckengefunden wurden. Die Clusteranalyse ergab 2Cluster mit 4 85% Ahnlichkeit; Cluster 1bestand aus Speichel, seitlichen und dorsalenZungenflachen, wahrend Cluster 2 die ubrigenGewebeoberflachen umfasste. V. parvula, P.melaninogenica, E. corrodens, N. mucosa, A.odontolyticus, F. periodonticum, F. nucleatumss vincentii und P. gingivalis wurden zusignifikant hoheren Anteilen in Cluster 1 undS. mitis, S. oralis und S. noxia in Cluster 2gefunden. Diese Ergebnisse wurden durchdie Daten der 44 Patienten, von denen Plaque-proben genommen wurden, bestatigt. Diemikrobiologischen Profile der supra- und sub-gingivalen Plaqueproben unterschieden sichinsbesondere hinsichtlich der erhohten Anteilean Actinomyces-Arten von den anderenEntnahmestellen. Arten unterschiedlicher Gen-era zeigten unterschiedliche Anteile aufden verschiedenen intraoralen Oberflachen,aber selbst innerhalb des Genus Streptococcusbestanden Unterschiede hinsichtlich desBesiedlungsmusters. S. oralis, S. mitis und S.constellatus besiedelten die Weichgewebe undden Speichel zu hoheren Anteilen als dieProben von Zahnen, wahrend die anderen 4untersuchten Streptokokkenarten die Zahnober-flachen zu ahnlichen Anteilen besiedelten wiedie Weichgewebe und den Speichel.Schlussfolgerungen: Die Anteile der Bakter-ienarten unterschieden sich deutlich auf denverschiedenen intraoralen Oberflachen. DieMikroorganismen des Speichels waren denender dorsalen und lateralen Flachen der Zungeam ahnlichsten. Die Mikroorganismen derWeichgewebsoberflachen ahnelten einandermehr als die Bakterien, die die Zahne ober-und unterhalb des Gingivarandes besiedeln.

Resume

Repartition d’especes bacteriennes selection-nees sur les surfaces intrabuccales

Le but de cet examen a ete de calculer lesproportions de 40 especes bacteriennes prove-nant d’echantillons de huit surfaces de tissumou et de la salive chez des adultes sains et decomparer ces microbiotypes a ceux de la plaquesus- et sous-gingivale. Des echantillons micro-biens ont ete preleves de huit surfaces buccalesde tissu mou chez 225 patients sains au point devue systemique utilisant une ‘‘brosse buccale’’.La salive a ete prelevee par expectoration.Quarante-quatre des sujets ont egalement donnede la plaque sus- et sous-gingivale. Lesechantillons ont ete evalues individuellementpour leur teneur en 40 especes bacteriennesutilisant l’hybridisation ADN-ADN par damier.Le pourcentage du comptage par sonde ADNtotal a ete determine pour chaque espece, achaque localisation et leur moyenne a etecalculee parmi les patients. La significationdes differences parmi les proportions des 40especes testees a differents endroits a eteexaminee chez les 225 et 44 sujets separementutilisant le test de Quade et ajustee pour lescomparaisons multiples. L’analyse par groupe aete effectuee a l’aide des proportions des 40especes aux differentes localisations des echan-tillons avec le coefficient de similarite minimaleet une relation de moyenne non-ponderee. Lamoyenne des proportions de chaque espece aete calculee parmi les sujets dans les groupesresultant, et la signification des differencestestees via le test-t et ANOVA. Les profilsmicrobiens differaient extremement parmi leslocalisations de l’echantillonnage chez les 225sujets avec 34 des 40 especes differantsignificativement. Les proportions de V. parvu-la et P. melaninogenica etaient plus importantesdans la salive et sur les surfaces laterales etdorsales de la langue tandis que S. mitis et S.oralis se retrouvaient dans des proportionssignificativement inferieures dans la salive etsur le dos de la langue. L’analyse par grouperesultait dans la formation de deux groupes avec485% de similarite : le groupe 1 comprenait lasalive et les surfaces laterales et dorsales de lalangue tandis que le groupe 2 comprenait leslocalisations des tissus mous restantes. V.parvula, P. melaninogenica, E. corrodens, N.mucosa, A. odontolyticus, F. periodonticum, F.nucleatum ss vincentii et P. gingivalis etaient enproportions significativement plus importantesdans le groupe 1 et S. mitis, S. oralis et S. noxiaetaient significativement plus importants dans legroupe 2. Ces decouvertes ont ete confirmeespar les donnees provenant des 44 sujets dont laplaque dentaire avait ete prelevee. Les profilsmicrobiens de la plaque sus- et sous-gingivaledifferaient des autres localisations d’echantil-lonnage particulierement pour les proportionsaugmentees des especes Actinomyces. Desespeces de differents genres montraient desproportions diferentes sur les surfaces intra-buccales distinctes mais meme parmi le genreStreptococcus, il y avait des differences dans lesmodeles de colonisation. S. oralis, S. mitis et S.constellatus colonisaient les surfaces des tissusmous et la salive dans des proportions plusimportantes que les echantillons provenant

des dents tandis que les quatre autres especesde streptocoques examinees colonisaient lessurfaces dentaires dans des proportions compar-ables aux localisations des tissus mous et de lasalive. Les proportions d’especes bacteriennesdifferaient enormement sur les sites intrabuc-caux respectifs. Le microbiote de la salive etaittres semblable a celui des surfaces dorsales etlaterales de la langue. Les microbiotes destissus mous se ressemblaient les uns aux autres,davantage que les microbiotes qui colonisaientles dents tant en sus- qu’en sous-gingival.

References

Asikainen, S., Alaluusua, S. & Saxen, L. (1991)

Recovery of A. actinomycetemcomitans from

teeth, tongue, and saliva. Journal of Period-

ontology 62, 203–206.

Beighton, D., Rippon, H. R. & Thomas, H. E.

(1987) The distribution of Streptococcus

mutans serotypes and dental caries in a

group of 5- to 8- year-old Hampshire

schoolchildren. British Dental Journal 162,

103–106.

Bosy, A., Kulkarni, G. V., Rosenberg, M. &

McCullock, C. A. G. (1994) Relationship of

oral malodor to periodontitis: evidence of

independence in discrete subpopulations.

Journal of Periodontology 65, 37–46.

Collins, L. M. C. & Dawes, C. (1987) The

surface area of the adult human mouth and

thickness of the salivary film covering the

teeth and oral mucosa. Journal of Dental

Research 66, 1300–1303.

Dahlen, G., Manji, F., Baelum, V. & Fejerskov,

O. (1992) Putative periodontopathogens in

‘‘diseased’’ and ‘‘non-diseased’’ persons

exhibiting poor oral hygiene. Journal of

Clinical Periodontology 19, 35–42.

Danser, M. M., Timmerman, M. F., van

Winkelhoff, A. J. & van der Velden, U.

(1996) The effect of periodontal treatment on

periodontal bacteria on the oral mucous

membranes. Journal of Periodontology 67,

478–485.

De Boever, E. H. & Loesche, W. J. (1995)

Assessing the contribution of anaerobic

microflora of the tongue to oral malodor.

Journal of the American Dental Association

126, 1384–1393.

Eger, T., Zoller, L., Muller, H-P., Hoffmann, S.

& Lobinsky, D. (1996) Potential diagnostic

value of sampling oral mucosal surfaces for

Actinobacillus actinomycetemcomitans in

young adults. European Journal of Oral

Science 104, 112–117.

Frandsen, E. V. G., Pedrazzoli, V. & Kilian, M.

(1991) Ecology of viridans streptococci in

the oral cavity and pharynx. Oral Microbiol-

ogy and Immunology 6, 129–133.

Gibbons, R. J., Kapsimalis, B. & Socransky, S.

S. (1964) The source of salivary bacteria.

Archives of Oral Biology 9, 101–103.

Gibbons, R. J. (1989) Bacterial adhesion to oral

tissues: a model for infectious diseases.

Journal of Dental Research 68, 750–760.

Gordon, D. F. Jr. & Gibbons, R. J.

(1966) Studies of the predominant cultivable


micro-organisms from the human tongue.


Haffajee, A. D., Cugini, M. A., Dibart, S.,

Smith, C., Kent, R. L. Jr. & Socransky, S. S.

(1997) The effect of SRP on the clinical and

microbiological parameters of periodontal

diseases. Journal of Clinical Periodontology

24, 324–334.

Haffajee, A. D., Socransky, S. S. & Goodson, J.

M. (1983) Comparison of different data

analyses for detecting changes in attachment

level. Journal of Clinical Periodontology 10,

298–310.

Hallberg, K., Hammarstrom, K. -J., Falsen, E.,

Dahlen, G., Gibbons, R., Hay, D. I. &

Stromberg, N. (1998) Actinomyces naeslundii

genospecies 1 and 2 express different binding

specificities to N-acetyl-B-D-galactosamine,

whereas Actinomyces odontolyticus expresses

a different binding specificity in colonizing

the human mouth. Oral Microbiology and

Immunology 13, 327–336.

van Houte, J., Gibbons, R. J. & Pulkkinen, A. J.

(1971) Adherence as an ecological determi-

nant for streptococci in the human mouth.


Krasse, B. (1953) The proportional distribution

of different types of Streptococci in saliva

and plaque material. Odontologisk Revy 4,

304–312.

Krasse, B. (1954) The proportional distribution

of Streptococcus salivarius and other Strep-

tococci in various parts of the mouth.

Odontologisk Revy 5, 203–211.

Langendijk, P. S., Hagemann, J. & van der

Hoeven, J. S. (1999) Sulfate-reducing bacter-

ia in periodontal pockets and in healthy oral

sites. Journal of Clinical Periodontology 26,

596–599.

Lee, K. H., Tanner, A. C. R., Maiden, M. F. J.

& Weber, H. P. (1999) Pre- and post-

implantation microbiota of the tongue, teeth,

and newly-placed implants. Journal of Clin-

ical Periodontology 26, 822–832.

Liu, T., Gibbons, R. J., Hay, D. I. & Skobe, Z.

(1991) Binding of Actinomyces viscosus to

collagen: association with type 1 fimbrial

adhesin. Oral Microbiology and Immunology

6, 1–5.

Loos, B. G., van Winkelhoff, A. J., Dunford, R.

G., Genco, R. J., De Graaff, J., Dickinson, D.

P. & Dyer, D. W. (1992) A statistical

approach to the ecology of Porphyromonas

gingivalis. Journal of Dental Research 71,

353–358.

Matto, J., Saarela, M., Alaluusua, S. &

Jousimies-Somer, H. (1996a) Similarity

of salivary and subgingival Prevotella

intermedia and Prevotella nigrescens isolates

by arbitrarily primed polymerase chain reac-

tion. Oral Microbiology and Immunology 11,

395–401.

Matto, J., Saarela, M., von Troil-Linden, B.,

Kononen, E., Jousimies-Somer, H., Torkko,

H., Alaluusua, S. & Asikainen, S. (1996b)

Distribution and genetic analysis of oral

Prevotella intermedia and Prevotella nigres-

cens. Oral Microbiology and Immunology 11,

96–102.

Matto, J., Saarela, M., Alaluusua, S., Oja, V.,

Jousimies-Somer, H. & Asikainen, S. (1998)

Detection of Porphyromonas gingivalis from

saliva by PCR by using a simple sample-

processing method. Journal of Clinical

Microbiology 36, 157–160.

Muller, H. -P., Eickholz, P., Heinecke, A., Pohl,

S., Muller, R. F. & Lange, D. E. (1995)

Simultaneous isolation of Actinobacillus

actinomycetemcomitans from subgingival

and extracrevicular locations of the mouth.

Journal of Clinical Periodontology 22,

413–419.

Muller, H. -P., Heinecke, A., Borneff, M.,

Knopf, A., Kiencke, C. & Pohl, S. (1997)

Microbial ecology of Actinobacillus

actinomycetemcomitans, Eikenella corrodens

and Capnocytophaga spp. in adult period-

ontitis. Journal of Periodontal Research 32,

530–542.

Neeser, J. R., Grafstrom, R. C., Woltz, A.,

Brassart, D., Fryder, V. & Guggenheim, B.

(1995) A 23 kDa membrane glycoprotein

bearing NeuNAc alpha 2-3Gal beta 1-3Gal-

NAc O-linked carbohydrate chains as a

receptor for Streptococcus sanguis OMZ

9 on human buccal epithelial cells. Glyco-

biology 5, 97–104.

Quirynen, M., Bollen, C. M. L., Vandekerc-

khove, B. N. A., Dekeyser, C., Papaioannou,

W. & Eyssen, H. (1995) Full- vs. partial-

mouth disinfection in the treatment of period-

ontal infections: short-term clinical and

microbiological observations. Journal of

Dental Research 74, 1459–1467.

Quirynen, M., De Soete, M., Dierickx, K. & van

Steenberghe, D. (2001) The intra-oral trans-

location of periodontopathogens jeopardises

the outcome of periodontal therapy. A review

of the literature. Journal of Clinical Period-

ontology 28, 499–507.

Quirynen, M., Mongardini, C. & van Steen-

berghe, D. (1998) The effect of a 1-stage

full-mouth disinfection on oral malodor and

microbial colonization of the tongue in

periodontitis patients. A pilot study. Journal

of Periodontology 69, 374–382.

Sklavounou, A. & Germaine, G. R.

(1980) Adherence of oral streptococci to

keratinized and nonkeratinized human oral

epithelial cells. Infection and Immunity 27,

686–689.

Socransky, S. S., Haffajee, A. D., Smith, C. &

Dibart, S. (1991) Relation of counts of

microbial species to clinical status at the

site. Journal of Clinical Periodontology 18,

766–775.

Socransky, S. S., Smith, C., Martin, L., Paster,

B. J., Dewhirst, F. E. & Levin, A. E. (1994)

Checkerboard DNA–DNA hybridization.

Biotechniques 17, 788–792.

Socransky, S. S., Tanner, A. C. R., Goodson, J.

M., Haffajee, A. D., Walker, C. B.,

Ebersole, J. L. & Sornberger, G. C.

(1982) An approach to the definition of

periodontal disease syndromes by cluster

analysis. Journal of Clinical Periodontology

9, 460–471.

Sneath, P. H. A. & Sokal, R. R. (1973)

Numerical taxonomy. The principles and

practice of numerical classification. San

Francisco: W.H. Freeman & Co.

Stromberg, N. & Boren, T. (1992) Actinomyces

tissue specificity may depend on differences

in receptor specificity for GalNAc-containing

glycoconjugates. Infection and Immunity 60,

3268–3277.

Umeda, M., Contreras, A., Chen, C., Bakker, I.

& Slots, J. (1998) The utility of whole saliva

to detect the oral presence of periodonto-

pathic bacteria. Journal of Periodontology

69, 828–33.

van Winkelhoff, A. J., van der Velden, U.,

Winkel, E. G. & De Graaff, J. (1986)

Black-pigmented Bacteroides and motile

organisms on oral mucosal surfaces in

individuals with and without periodontal

breakdown. Journal of Periodontal Research

21, 434–439.

Zoller, E. T., Muller, H. -P., Hoffmann, S. &

Lobinsky, D. (1996) Potential diagnostic

value for sampling oral mucosal surfaces

for Actinobacillus actinomycetemcomitans in

young adults. European Journal of Oral

Science 104, 112–117.

Address:

Donna L. Mager

Department of Periodontology and

Molecular Genetics

The Forsyth Institute

140 The Fenway

Boston, MA 02115

USA

654 Mager et al.

distribution of selected bacterial species on intraoral surfaces

Documents