INFECTION AND IMMUNITY, Nov. 1976, p. 1119-1124Copyright X) 1976 American Society for Microbiology

Vol. 14, No. 5Printed in U.S.A.

Establishment and Distribution of Actinomyces viscosus andActinomyces naeslundii in the Human Oral Cavity

RICHARD P. ELLENFaculty ofDentistry, University of Toronto, Toronto, Ontario, Canada M5G IG6

Received for publication 2 July 1976

The intraoral establishment and proportional distribution of suspected perio-dontal pathogens Actinomyces viscosus and Actinomyces naeslundii were stud-ied using a recently developed differential plating medium, CNAC-20. Salivaand dental plaque samples were collected from 108 subjects ranging in age frominfants to young adults; tongue and buccal mucosa samples were collected fromonly the adult subjects. Catalase-negative A. naeslundii was isolated from 40%of the predentate infants' and almost all other subjects' saliva samples. Itpredominated among CNAC-20 isolates in the saliva of subjects of all agegroups, in the plaques of young children, and in the adult tongue samples. Incontrast, catalase-positive A. viscosus was not isolated from predentate infantsamples, and its frequency of isolation increased slowly with age (>50% detec-tion by age 7). A. viscosus was isolated in highest relative proportions fromdental plaque and buccal mucosa samples. The two closely related species A.viscosus and A. naeslundii apparently differ in respect to factors determiningthe host age at which they colonize and their relative intraoral distribution inhumans.

Actinomyces viscosus and Actinomyces naes-lundii have been implicated in the etiology ofperiodontal disease primarily because of theirassociation with naturally transmissible andexperimentally induced periodontal lesions inlaboratory animals (13, 15-18, 21). However,little is known about their ecological relation-ship with the human host and their impact onthe course of naturally occurring human perio-dontal diseases. Presumably, this lack of infor-mation reflects previously encountered difficul-ties in rapidly enumerating these species inclinical samples and in differentiating themfrom other facultatively anaerobic, gram-posi-tive, pleomorphic rods and filaments that alsocolonize the mouth. Such identification prob-lems were evident in a previous attempt totrace the establishment ofActinomyces speciesin the developing oral flora of newborn infants(20). The recent development of a partially se-lective, differential medium for detecting facul-tatively anaerobic Actinomyces colonies (6) hasfacilitated the processing of numerous clinicalsamples, making the present investigation fea-sible. This report describes investigations intothe host age at which A. viscosus and A. naes-lundii infect humans and the distribution ofthe two species in the mouth. Data are pre-sented that indicate that the closely relatedspecies A. viscosus and A. naeslundii differmarkedly in their patterns of human intraoralcolonization.

MATERIALS AND METHODSPopulations under investigation. The 108 sub-

jects for this investigation ranged in age from 3months to 30 years. Thirty-six healthy infants wereselected from the population at Edgewood ManorDay Nursery and the Family Practice Unit of St.Michael's Hospital, Toronto, Canada. Fifty-sevenchildren from 3 to 16 years of age were sampled atthe Faculty of Dentistry clinics, University of To-ronto. The adult population included 15 subjects, 20to 30 years of age, who were selected from the staffand students of the Faculty of Dentistry and werefound to be relatively free of clinical signs of perio-dontal disease.

Sample collection and cultural condi-tions. Samples of whole saliva were collected fromall the subjects. Unstimulated whole saliva was col-lected in sterile vials from the children and adults.Infant saliva samples were collected by absorbingthe saliva bathing their mucosal surfaces with cal-cium alginate swabs (Calgiswabs, Wilson Diagnos-tics, Inc., Glenwood, Ill.). Pooled smooth-surfaceplaque samples were obtained from all dentuloussubjects by swabbing the facial and lingual surfacesof their maxillary anterior teeth. Additional swabsamples were collected from the dorsum of thetongue and from the buccal mucosa of the adultsubjects. The calcium alginate swabs were partiallydissolved in 2.0 ml of modified Ringer solution withhexametaphosphate (23) by vibrating the sampletubes on a Vortex mixer for 20 s.

Tenfold serial dilutions of each sample were pre-pared in 0.05% yeast extract broth (Difco), and ap-propriate dilutions were spread in duplicate onplates of tryptic soy agar (Difco) containing 5.0%

1119

on June 26, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

1120 ELLEN

sheep blood and Columbia CNA agar (Difco) con-taining 20.0 ,ug of 3CdSO4 8H2O per ml (CNAC-20)(6). Undiluted swab samples of plaque and infantsaliva were also inoculated directly onto CNAC-20.For the whole saliva samples, the lowest dilutionplated on CNAC-20 was 10-3. The blood plates wereincubated at 35°C for 4 days in an atmosphere of 95%N2 and 5% CO2. The CNAC-20 plates were incubatedfor 2 days in 90% air and 10% CO2 to encourage thegrowth ofActinornyces which could grow aerobically(6).

Collection and analysis of data. The number ofcolony-forming units (CFU) per milliliter of samplewas determined by calculating the average colonycount on duplicate plates of dilutions yielding be-tween 20 and 200 colonies. The number of faculta-tively anaerobic Actinornyces was determined bycounting only the large, white, smooth, opaqueCNAC-20 colonies previously shown to contain bac-teria resembling either A. viscosus or A. naeslundii(6). Questionable colony types were isolated, andpure cultures thereof were identified by culturalmethods (6, 14). Catalase activity to differentiatebetween catalase-positive A. viscosus-like and cata-lase-negative A. naeslundii-like CNAC-20 isolateswas tested by applying 3.0% H202 directly to thecolonies. The proportions of catalase-positive andcatalase-negative isolates in the samples were de-termined on the same plates used for the colonycount. If no catalase-positive colonies were detected,plates from successively lower dilutions were cho-sen.The frequency of A. viscosus and A. naeslundii

isolation was calculated and expressed as the per-centage of subjects in each age group whose samplesyielded detectable CNAC-20 Actinomyces isolates.The infant group was composed of 15 predentatesand 21 infants with teeth. The group of preschool-aged children (3 to 4 years) included 9 subjects.Among the school-aged children, 10 were 5 to 6 yearsof age; 14 were 7 to 8 years; 11 were 9 to 10 years; 7were 11 to 12 years; and 6 were 13 to 16 years of age.

For each sample, the proportions of A. viscosusand A. naeslundii (CNAC-20 plate counts) amongthe total CFU recovered (blood plates) were calcu-lated. An estimate of the efficiency of CNAC-20 inrecovering Actinomyces from the samples was ob-tained by comparing the number of ActinomycesCFU detected on CNAC-20 with the number ofCFUrecovered on blood agar that could be identifiedsubsequently as A. viscosus or A. naeslundii. Thisestimate was made on five of the clinical samples.Every colony on blood agar plates containing 50 to100 colonies was isolated. Pure cultures of the iso-lates were Gram stained. Gram-positive rods, pleo-morphic rods, filaments, and coccobacilli were sub-mitted to several cultural tests (6, 14) to determinetheir resemblance to A. viscosus and A. naeslundii.An index was calculated to determine the relative

preferences ofA. viscosus and A. naeslundii for colo-nizing specific intraoral surfaces. The index con-sisted of the ratio of catalase-positive to catalase-negative CFU in samples from only those subjectsfound to be infected with both catalase-positive andcatalase-negative Actinomyces. Thus, for any in-

INFECT. IMMUN.

traoral site, a mean ratio significantly greater than1.0 indicated a predominance of catalase-positiveCFU among the facultatively anaerobic Actino-myces isolated; a mean ratio of less than 1.0 indi-cated a predominance of catalase-negative isolates.

RESULTSEstablishment of facultatively anaerobic

Actinomyces in the human mouth. The fre-quency of isolation of bacteria resembling A.viscosus and A. naeslundii is illustrated in Fig.1. In general, the frequency of facultativelyanaerobic Actinomyces isolation increased withage, similar to findings for anaerobic Actino-myces reported by others (20, 22). Catalase-negative Actinomyces were detected in 40% ofthe predentate infants and 82% of the infantswith teeth. The saliva samples of all other sub-jects contained catalase-negative isolates. Al-though the plaque samples from most subjectsalso contained catalase-negative CNAC-20CFU, there were some samples in which theywere not detected. In contrast, the establish-ment of catalase-positive Actinomyces in theoral cavity was delayed. They were not detectedin the samples from predentate infants. Thefrequency of their detection in saliva andplaque samples increased slowly, reaching anisolation frequency greater than 50% by age 7.Neither saliva nor plaque samples from any ofthe age groups yielded a 100% detection fre-quency for catalase-positive Actinomyces.

Intraoral distribution of facultatively an-aerobic Actinomyces. Data reflecting the pro-portions of Actinomyces-like isolates in the to-tal cultivated flora should be considered accu-

100

90CD

E 80

10

-C460

C)50

-F

, 30

K 20

10

Ai 1/

_ / ,

I ,8,~~~~0

- /22 CatalaseSaliva O____ +

o- Smooth surface Aplaque*----- +

a I I I I Age Years,X_ , 3-4

_c c

T?, C: Cs

5-6 7-8 9-10 11-12 13-16 20-30

FIG. 1. Frequency of isolation of facultatively an-aerobic Actinomyces from saliva and dental plaquecollected from subjects of various ages.

on June 26, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

A. VISCOSUS AND A. NAESLUNDII ORAL COLONIZATION 1121

TABLE 1. Recovery offacultatively anaerobic Actinomyces from saliva and smooth surface dental plaques ofchildren

Saliva Dental plaque

Subject age Mean % of cultivated Actino- Mean % of cultivated Actino-(years) Mean % of culti- mycesa Mean % of culti- myces

vated flora Catalase vated flora CatalaseCatalase positive negative Catalase positive negative

3-4 7.1 + 3.9 3.8 ± 3.5 96.2b 1.4 + 1.0 10.1 ± 9.5 89.9b(0.2-36.7)c (<0.1-9.6)

5-6 6.1 + 1.3 3.5 + 1.6 96.5 3.6 ± 1.7 18.0 ± 11.1 82.0(2.1-14.5) (<0.1-13.3)

7-8 5.2 ± 1.0 10.0 ± 4.1 90.0 0.6 ± 0.2 36.0 + 10.5 64.0(<0.1-13.5) (<0.1-2.0)

9-10 11.0 ± 3.0 7.6 ± 2.5 92.4 2.6 ± 1.2 55.2 + 13.7 44.8(2.9-34.4) (<0.1-8.7)

11-12 6.0 ± 1.1 5.0 ± 4.2 95.0 0.7 ± 0.3 23.1 ± 12.7 76.9(1.9-9.7) (<0.1-1.8)

13-16 9.9 ± 4.9 33.2 + 13.6 66.8 0.7 ± 0.5 87.3 ± 4.7 12.7(1.7-33.2) (<0.1-2.9)

a Facultatively anaerobic Actinomyces isolated on CNAC-20.b Standard error same as for catalase-positive proportions.C Mean ± standard error of mean (range).

rate only to the limits of the methods used toenumerate them. We have found previouslythat some stock strains of both A. viscosus andA. naeslundii either fail to grow on CNAC-20or lose the ability to grow on it after extendedlaboratory culture (6; unpublished observa-tions). However, pure cultures of stock andfreshly isolated strains that can grow onCNAC-20 yield equivalent numbers of CFU onCNAC-20 and nonselective media (6). In thesamples from this study chosen to estimate theefficiency of CNAC-20 in recovering faculta-tively anaerobic Actinomyces, the number ofActinomyces CFU on CNAC-20 represented70.0 + 12.0% (mean + standard error; range,36.8 to 90.4) of the predominant CFU recoveredon tryptic soy blood agar that could be identi-fied subsequently as A. viscosus or A. naeslun-dii.Among the children and adults, the average

percentage of the salivary flora recovered onblood agar that could be accounted for byCNAC-20 Actinomyces isolates ranged from 5.2+ 1.0 to 11.0 + 3.0% (Tables 1 and 3). No trendsof increasing or decreasing proportions withage were noted. Catalase-negative isolates pre-dominated among CNAC-20 CFU recoveredfrom saliva of subjects in all age groups. Themean percentage of catalase-negative isolatesamong salivary Actinomyces isolates wasgreater than 90% at all ages younger than theteenage and adult groups. Under the conditionsused, the adult saliva samples contained anaverage of8.9 x 10" catalase-negative and 2.9 x10" catalase-positive Actinomyces CFU/ml.

The mean proportions of the cultivated den-tal plaque flora recovered as Actinomyces CFUon CNAC-20 were quite low in the children ofall age groups, ranging from 0.6 + 0.2 to 3.0 ±1.7% (Table 1). Among these, catalase-negativeisolates predominated in the plaques from theyoung children. The relative proportions of cat-alase-positive isolates generally increased withage. Catalase-positive isolates predominated inplaque samples from the teenage and adultgroups (Tables 1 and 3).

Facultatively anaerobic Actinomyces ac-counted for 7.0 and 4.5% of the cultivated florafrom the adult tongue and buccal mucosal sam-ples, respectively. The frequency of isolationand the relative proportions ofcatalase-positiveand catalase-negative Actinomyces isolates dif-fered in samples from the two sites (Table 3).Catalase-negative CFU were detected in 100%of both tongue and buccal mucosa samples. Incontrast, catalase-positive CFU were isolatedfrom only 40% of the tongue and 93% of thecheek samples. Tongue samples contained anaverage of greater than fourfold more catalase-negative than catalase-positive isolates. Theiraverage relative proportions in the buccal mu-cosa samples were almost equal.

Differences in the preferences ofcatalase-pos-itive and catalase-negativeActinomyces for col-onizing various sites in the human mouth weresuggested by the mean relative proportions de-scribed above. These preferences became moreevident when the ratio of their proportions wascalculated for samples from only those subjectsfrom whom both catalase-positive and -nega-

VOL. 14, 1976

on June 26, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

1122 ELLEN

tive Actinornyces were isolated (Tables 2 and3). The mean ratios for the saliva samples dif-fered markedly from those of the plaque sam-ples at all ages. With one exception (salivasamples from the teenage group), the ratio indi-cated an overall predominance of catalase-neg-ative isolates in saliva samples and catalase-positive isolates in plaque. Differences in theratio were also observed between the samplesobtained from the tongue and buccal mucosa inadults (Table 3). The ratio indicated a predomi-nance of catalase-negative isolates in thetongue samples, similar to that detected in sa-liva. In contrast, catalase-positive isolates pre-dominated in buccal mucosa samples fromthose subjects with detectable levels of bothcatalase-positive and -negative CFU.

DISCUSSIONInformation regarding the host age at which

oral pathogens colonize, their distribution inthe mouth, and the conditions necessary forsuccessful implantation should suggest somemeans to prevent their establishment, and

TABLE 2. Ratio of catalase-positive to catalase-negative Actinomyces-like isolates in samples from

children a

Mean ratio (catalase-posi-

Subject age No. of sub- tive/negative CFU)(years) jects Dental

Saliva plaque

3-4 2 1/lb 0.23 3.105-6 4/10 0.10 10.277-8 12/14 0.19 7.969-10 7/11 0.14 24.2311-12 4/7 0.12 2.5013-16 4/6 1.32 14.24

a Only children from whom both catalase-positiveand catalase-negative Actinomyces were isolated.

b Number of children sampled.

thereby lower the pathogenic potential of theflora. Data presented herein indicate that thesuspected periodontal pathogens A. viscosusand A. naeslundii differ in both the host age atwhich they establish in the human mouth andtheir preferences for colonizing various in-traoral sites. Catalase-negative isolates resem-

blingA. naeslundii were found to colonize mostinfants, including 40% of the predentates, andto maintain predominance among facultativelyanaerobic Actinornyces in saliva and on thetongue. These data agree with earlier findingsthat the proportions of bacterial species in sa-

liva generally reflect the proportions on thetongue dorsum (9, 12, 19). In contrast, the colo-nization of bacteria resembling catalase-posi-tive A. viscosus was delayed until after teethhad erupted, and even then their rise to pre-dominance over catalase-negative Actinornycesin plaque was delayed for several years. Al-though the medium used to detect A. viscosusand A. naeslundii could only recover approxi-mately 70% of their CFU cultivable on bloodagar, it seems unlikely that undetected A. vis-cosus CFU could have comprised a significantproportion of the flora, considering the fact thatinfant saliva samples and plaque samples were

plated undiluted.If oral bacteria are transmitted between hu-

mans primarily via saliva, then the relativeease with which specific bacteria are transmit-ted would likely be influenced in part by theirrelative numbers in saliva. In addition, differ-ences in their preferred intraoral sites of coloni-zation would be expected to affect the probabil-ity of implantation, dependent upon the availa-bility of a specific local environment conduciveto colonization at the time of transmission. Forexample, Streptococcus salivarius, which colo-nizes the tongue and is present in high num-

bers in adult saliva, has been shown by Carls-son and co-workers to establish in the mouths

TABLE 3. Recovery offacultatively anaerobic Actinomyces from various intraoral sites of adults

Mean % of cultivated flora Mean % of cultivated Actinomyces Mean ratioaSite sampled (catalase-positive/

Catalase positive Catalase negative Catalase positive Catalase negative negative CFU)

Saliva 1.5 ± 0.6 4.3 ± 0.8 23.9 ± 7.1 76.1' 0.73(<0.1-6.4)b (0.6 ± 12.9)

Smooth-surface 4.4 + 1.4 2.9 ± 1.3 58.3 ± 11.2 41.7 13.34dental plaque (<0. 1-19.5) (<0. 1-16.9)

Tongue 0.9 + 0.4 5.1 + 1.1 15.7 ± 7.8 84.3 0.67(<0.1-4.9) (0.8-12.8)

Buccal mucosa 2.9 + 1.1 1.6 + 0.7 51.2 ± 9.5 48.8 5.81(<0.1-13.5) (0.2-10.4)

Includes only subjects from whom both catalase-positive and catalase-negative Actinomyces weredetected.

b Mean ± standard error of mean (range).' Standard error same as for catalase positive.

INFECT. IMMUN.

on June 26, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

A. VISCOSUS AND A. NAESLUNDII ORAL COLONIZATION 1123

of neonates (3). Thus, the large surface area ofthe tongue, which is a prime ecological nichefor S. salivarius (10, 11, 19), provides an envi-ronment suitable for colonization by the rela-tively high numbers ofS. salivarius cells trans-mitted via adult saliva. In contrast, the estab-lishment of Streptococcus mutans and Strepto-coccus sanguis, which preferentially colonizeteeth, is delayed until after tooth eruption (1, 4,5). In the case of S. mutans, adult salivarynumbers are relatively low compared withthose of other oral streptococci, and oral coloni-zation is often delayed until after the first yearofage (1, 4, 5, 22). In addition to its favored, andperhaps essential, ecological niche being una-vailable until teeth erupt, the frequency withwhich infants' teeth are exposed to high enoughS. mutans doses to favor implantation may berare.

Evidently, members of the genus Actino-myces also differ in their ease of implantationin humans. A. naeslundii, which colonizes thetongue and is present in relatively high num-bers in saliva, is readily transmitted to theinfant and may establish in the predentatemouth. In contrast, oral colonization ofA. vis-cosus is delayed. This may be explained in partby its preference for colonizing teeth and itsrelatively low salivary numbers available fortransmission. It is probable that other age-re-lated phenomena also influence the ability ofspecific bacteria to colonize. In studies to deter-mine the relationship between the age of ratsand their susceptibility to the implantation ofA. viscosus Ny-1, Brecher and van Houte havefound that the organism established more read-ily in older than in younger rats (Int. Assoc.Dent. Res., abstr. 462, 1976). It is of interestthat the infecting strain was recovered inhigher proportions of the flora from the rats'teeth and buccal mucosae than from theirtongues (S. Brecher and J. van Houte, personalcommunication). Thus, the most heavily colo-nized rat sites were the same as those yieldingthe highest relative proportions of A. viscosusin the present investigation of humans.

Factors influencing the dissimilarities in A.viscosus and A. naeslundii patterns of in-traoral colonization have not yet been eluci-dated. It is doubtful that differences in theirnutrient requirements have a profound effect,since taxonomic studies have revealed few dis-similarities in the two species except for dis-tinct differences in their catalase-like activities(14; E. D. Fillery, G. H. Bowden, and J. M.Hardie, Int. Assoc. Dent. Res., abstr. L-218,1975). Moreover, reported similarities in theirdeoxyribonucleic acid base ratios (A. L. Coy-kendall, T. W. Lee, and A. T. Brown, Int. As-

soc. Dent. Res. abstr. 74, 1974) and cross-reac-tions between some antigens (8) suggest thatwhat are now termed A. viscosus and A. naes-lundii may eventually be classified as catalaseactivity variants of one species. Since severalmembers of the dental plaque flora are knownto produce hydrogen peroxide, it is possible thatthe increased proportions of A. viscosus re-covered from plaque accumulations may haveresulted from the induction of catalase activityamong populations of Actinomyces cells in re-sponse to high local hydrogen peroxide concen-trations. However, the findings in this studythat catalase-negative Actinomyces predomi-nated in plaque collected from young childrenand the findings of Socransky and co-workersthat catalase-positive A. viscosus is one of thefirst species to recolonize cleaned adult teeth(S. S. Socransky, A. D. Manganiello, D. Pro-pas, V. Oram, and J. van Houte, J. Periodon-tal. Res., in press) do not support this conten-tion. If catalase-like activity is of any selectiveadvantage for A. viscosus cells, it does not ap-pear to be a basic ecological determinant affect-ing their initial colonization.A more likely explanation for differences in

the proportional distribution ofA. viscosus andA. naeslundii may involve the selectivity withwhich they adhere to various oral surfaces.Specificity in the extent to which bacteria canattach to surfaces has been substantiated as amajor ecological pressure regulating the indige-nous microbial flora of the mouth and has beenimplicated among factors influencing the tissuetropisms of overt pathogens (7, 10). The factthat A. viscosus colonizes cleaned teeth morereadily than A. naeslundii (Socransky et al., inpress), whereas the number ofA. viscosus cellsin saliva available for attachment is several-fold fewer, suggests that A. viscosus cells havea greater affinity for teeth than cells of A.naeslundii. Such differences in affinity forteeth and salivary cell concentrations havebeen used to predict the probability of coloniza-tion by oral streptococci and lactobacilli onsmooth surfaces of teeth (24).A. viscosus and A. naeslundii may also differ

in the degree to which they can adhere to accu-mulated plaque deposits. Differences in theability of some A. viscosus and A. naeslundiistrains to aggregate selectively with glucose-grown strains of plaque streptococci have beenreported (R. P. Ellen and I. B. Balcerzak-Raczkowski, J. Periodontal Res., in press). Inaddition, Bourgeau and McBride have shownrecently that A. viscosus cells bind readily tosucrose-grown S. mutans and S. sanguis cellsand to glucans elaborated by these streptococci(2). The ability of A. naeslundii to aggregate

VOL. 14, 1976

on June 26, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

1124 ELLEN INFECT. IMMUN.

under similar conditions has not been reported.The full impact of adherence-related phenom-ena and interbacterial affinities on the estab-lishment and proportional distribution of oralActinomyces species is not yet clear. However,it is evident thatA. viscosus and A. naeslundii,which are similar enough physiologically to beconsidered members of the same species, differgreatly in the pattern in which they naturallyinfect the human host.

ACKNOWLEDGMENTSThis investigation was supported by grants TG-2 and

5619 from the Medical Research Council of Canada and agrant from the Atkinson Charitable Foundation.

I thank H. J. Sandham for his helpful suggestions duringthe preparation of the manuscript.

LITERATURE CITED1. Berkowitz, R. J., H. V. Jordon, and G. White. 1975.

The early establishment of Streptococcus mutans inthe mouths of infants. Arch. Oral Biol. 20:171-174.

2. Bourgeau, G., and B. C. McBride. 1976. Dextran-me-diated interbacterial aggregation between dextran-synthesizing streptococci and Actinomyces viscosus.Infect. Immun. 13:1228-1234.

3. Carlsson, J., H. Grahnen, G. Jonsson, and S. Wikner.1970. Early establishment of Streptococcus salivariusin the mouths of infants. J. Dent. Res. 49:415-418.

4. Carlsson, J., H. Grahnen, G. Jonsson, and S. Wikner.1970. Establishment of Streptococcus sanguis in themouths of infants. Arch. Oral B.iol. 15:1143-1148.

5. Catalanotto, F. A., I. L. Shklair, and H. J. Keene. 1975.Prevalence and localization of Streptococcus mutansin infants and children. J. Am. Dent. Assoc. 91:606-609.

6. Ellen, R. P., and I. B. Balcerzak-Raczkowski. 1975.Differential medium for detecting dental plaque bac-teria resembling Actinomyces viscosus and Actino-myces naeslundii. J. Clin. Microbiol. 2:305-310.

7. Ellen, R. P., and R. J. Gibbons. 1974. Parameters af-fecting the adherence and tissue tropisms of Strepto-coccus pyogenes. Infect. Immun. 9:85-91.

8. Gerencser, M. A., and J. M. Slack. 1976. Serologicalindentification ofActinomyces using fluorescent anti-body techniques. J. Dent. Res. 55(special issueA):A184-A191.

9. Gibbons, R. J., B. Kapsimalis, and S. S. Socransky.1964. The source of salivary bacteria. Arch. Oral Biol.9:101-103.

10. Gibbons, R. J., and J. van Houte. 1975. Bacterial adher-ence in oral microbial ecology. Annu. Rev. Microbiol.29:19-44.

11. Gordon, D. F., and R. J. Gibbons. 1966. Studies on thepredominant cultivable microorganisms from the hu-man tongue. Arch. Oral Biol. 11:627-632.

12. Gordon, D. F., and B. B. Jong. 1968. Indigenous florafrom human saliva. Appl. Microbiol. 16:428-429.

13. Guggenheim, B., and H. E. Schroeder. 1974. Reactionsin the periodontium to continuous antigenic stimula-tion in sensitized gnotobiotic rats. Infect. Immun.10:565-577.

14. Holmberg, K., and H. 0. Hallander. 1973. Numericaltaxonomy and laboratory identification of Bacteri-onema matruchotii, Rothia dentocariosa, Actinomycesnaeslundii, Actinomyces viscosus, and some relatedbacteria. J. Gen. Microbiol. 76:43-63.

15. Jordan, H. V. 1971. Rodent model systems in periodon-tal disease research. J. Dent. Res. 50:236-242.

16. Jordan, H. V., and P. H. Keyes. 1964. Aerobic, gram-positive filamentous bacteria as etiologic agents ofexperimental periodontal disease in hamsters. Arch.Oral Biol. 9:401-414.

17. Jordan, H. V., P. H. Keyes, and S. Bellack. 1972.Periodontal lesions in hamsters and gnotobiotic ratsinfected with Actinomyces of human origin. J. Perio-dontal Res. 7:21-28.

18. Keyes, P. H., and H. V. Jordan. 1964. Periodontal le-sions in the Syrian hamster. III. Findings related toan infectious and transmissible component. Arch.Oral Biol. 9:377-400.

19. Krasse, B. 1954. The proportional distribution of Strep-tococcus salivarius and other streptococci in variousparts of the mouth. Odontol. Revy 5:203-211.

20. McCarthy, C., M. L. Snyder, and R. B. Parker. 1965.The indigenous oral flora of man. I. The newborn tothe 1-year-old infant. Arch. Oral Biol. 10:61-70.

21. Socransky, S. S., C. Hubersak, and D. Propas. 1970.Induction of periodontal destruction in gnotobioticrats by a human oral strain of Actinomyces naeslun-dii. Arch. Oral Biol. 15:993-995.

22. Socransky, S. S., and S. D. Manganiello. 1971. The oralmicrobiota of man from birth to senility. J. Periodon-tol. 42:485-494.

23. van Houte, J., R. J. Gibbons, and S. B. Banghart. 1970.Adherence as a determinant of the presence of Strep-tococcus salivarius and Streptococcus sanguis on thehuman tooth surface. Arch. Oral Biol. 15:1025-1034.

24. van Houte, J., and D. B. Green. 1974. Relationshipbetween the concentration of bacteria in saliva andthe colonization of teeth in humans. Infect. Immun.9:624-630.

on June 26, 2020 by guesthttp://iai.asm

.org/D

ownloaded from