Clonal Diversity and Stability of Subgingival Eikenella corrodens

Abstract

Eikenella corrodens is a commensal subgingival bacterium commonly found in both periodontally nondiseased and diseased subjects. The present study examined the clonal diversity and stability of subgingival E. corrodens over time. Ninety-five subjects were enrolled at the baseline examination, including 44 periodontally nondiseased subjects and 51 subjects with aggressive periodontitis. Twenty-two nondiseased subjects and 11 subjects with aggressive periodontitis were subsequently reexamined after an average interval of 14 months. Two subgingival plaque samples were obtained from each subject to determine the total cultivable bacteria. In addition, multiple E. corrodens isolates from each sample were recovered for clonal analysis by arbitrarily primed PCR. The mean numbers of distinct E. corrodens clones harbored by nondiseased subjects and subjects with aggressive periodontitis were 1.3 and 3.0, respectively. Thirty-nine percent of the nondiseased subjects and 63% of the subjects with aggressive periodontitis harbored multiple clones of E. corrodens. The numbers of distinct E. corrodens clones increased significantly (Mann-Whitney ranking test, P < 0.05) in sites from patients with aggressive periodontitis, in sites with pocket depths of 4 mm or greater, in sites with a clinical attachment loss of 2 mm or greater, and in sites coinfected with Porphyromonas gingivalis. Comparison of E. corrodens clones recovered at the baseline and those recovered at the follow-up examination showed that E. corrodens colonization was not stable. Thirty-eight of the 66 follow-up samples (58%) showed a complete change (including de novo colonization of the sites or complete elimination of the organism from the sites) of the subgingival E. corrodens clonal types between the baseline and the follow-up examinations. Our results suggest a complexity of subgingival microbiota not seen previously.

Eikenella corrodens is a gram-negative, facultative anaerobe found in the human oral cavity and upper respiratory tract (7, 10). This bacterium has been associated with a variety of nonoral human infections (3, 7, 12) as well as with human periodontitis (7). E. corrodens is not a frank pathogen but is a commensal and an occasional periodontal pathogen of humans. The bacterium rarely rises to become the dominant subgingival bacterial species, even in sites with severe periodontitis.

A typical clinical microbiological study of periodontitis can be likened to taking a snapshot of the bacterial community in gingival crevices at the time of the study. While the bacterial components are identified at a single time point, little is known about the stability of the subgingival bacterial community. Moreover, the content of the subgingival bacterial community is usually determined to the species level, with the genetic heterogeneity of individual strains within species being ignored. This may even hamper our understanding of the etiologic role of bacteria in periodontal disease, because not all strains within a periodontal pathogenic species possess the same degree of virulence. Some basic questions about the clonal diversity and the clonal stability of subgingival bacterial species over time may need to be addressed. It is from this perspective that we initiated this clinical microbiological study of E. corrodens.

While E. corrodens may no longer be the focus of research in periodontal disease pathogenesis, it is an excellent bacterium from which to gain information about the subgingival bacterial community in general. E. corrodens is relatively easy to identify from subgingival plaque by culture, is likely to be present in all subjects irrespective of their periodontal disease status, and is not likely to be eliminated after periodontal treatment. Furthermore, the species E. corrodens comprises genetically diverse strains that may be easily distinguished by a variety of DNA fingerprinting methods (5, 6, 8, 9). Taken together, we may examine the clonal diversity of subgingival E. corrodens and, over time, track the presence of subgingival E. corrodens to determine the stability of the colonization. The objectives of this study were to (i) examine the clonal diversity of subgingival E. corrodens in periodontally nondiseased and diseased subjects, (ii) determine the clonal stability of subgingival E. corrodens within individuals over time, and (iii) identify factors associated with the clonal diversity and clonal stability of subgingival E. corrodens.

MATERIALS AND METHODS

Study design and subjects.

The subjects were recruited from patients and students seeking dental care at the University of Southern California School of Dentistry. Three study subject groups were targeted: periodontally nondiseased subjects, subjects with localized aggressive periodontitis (LAP), and subjects with generalized aggressive periodontitis (GAP). All subjects were screened, accepted to the study, examined clinically, and sampled by two of the investigators (C.C. and S.R.). Demographic information was obtained by a questionnaire. The nondiseased subjects were 30 years old or younger with a minimum of 26 teeth, showed no radiographic evidence of alveolar bone loss, and had no periodontal sites with probing pocket depths (PDs) ≥4 mm. LAP and GAP are two forms of aggressive periodontitis recognized by the 1999 International Workshop for the Classification of Periodontal Diseases and Conditions (1). In this study, subjects with LAP were 30 years old or younger with at least two teeth, either central incisors or first molars, displaying ≥2 mm of clinical attachment loss (AL). In addition, fewer than four teeth (other than central incisors or first molars) exhibited clinical AL due to periodontitis. Subjects with GAP were 35 years old or younger with at least two teeth with ≥2 mm of AL due to periodontitis in each of the four quadrants. All subjects were systemically healthy. Subjects with the following conditions or behaviors were excluded from the study: immunodeficiency, diabetes, smoking, and medical conditions requiring premedication with antibiotics for any dental procedures. Subjects were not allowed to have received antibiotic therapy for 3 months prior to the baseline examination or to have received any periodontal therapy, including scaling and root planing, for 6 months prior to the baseline.

At the baseline, the subjects received a routine oral and periodontal examination that included determinations of PD and AL (determined by measuring the distance between the cemento-enamel junction and the base of the pocket). Subgingival plaque samples were collected from selected sites for analysis (see “Bacterial sampling” for more details). The follow-up examination was conducted at least 9 months after the baseline and included a subpopulation of the periodontally nondiseased subjects and diseased subjects (from both the LAP and the GAP groups). The subjects were examined and again sampled as they were at the baseline. No attempt was made to determine the disease activity (progression) for subjects with aggressive periodontitis.

Bacterial sampling.

Two subgingival plaque samples were obtained from each subject. The sample sites were selected on the basis of the existing periodontal probing information that was obtained at the initial examination. The chance of cross-contamination by transferring bacteria from the non-sample sites to the sample sites was minimized by performing little or no probing immediately prior to sampling. One sample was obtained from the periodontal site with the deepest PD in the mouth. The other sample was a pooled sample from three subgingival sites, each of which was selected from the deepest pocket in each of the remaining three quadrants. Subgingival plaque samples were collected with sterile periodontal curettes, placed into vials containing transport medium VMGAIII, and immediately processed in the laboratory. The same sample sites were used for the subjects who participated in both the baseline and the follow-up examinations.

Culture identification of subgingival bacteria.

The subgingival bacteria were processed by established methods (2). Briefly, samples in VMGAIII were warmed to 35°C, dispersed with a Vortex mixer at the maximal setting for 60 s, and serially diluted; and 0.1-ml aliquots were plated onto 4.3% brucella agar base (BBL Microbiology Systems, Cockeysville, Md.) supplemented with 0.3% Bacto Agar, 5% defibrinated sheep blood, 0.2% hemolyzed sheep red blood cells, 0.2% yeast extract, 0.0005% hemin, and 0.00005% menadione for determination of total viable counts and the proportions of Porphyromonas gingivalis, Tannerella forsythensis, Campylobacter spp., Eubacterium spp., Fusobacterium spp., Prevotella melaninogenicus, and members of the Prevotella intermedia-Prevotella nigrescens group. Primary incubation of the brucella blood agar medium took place at 37°C for 5 to 7 days in an anaerobic chamber (Coy Laboratory Products, Ann Arbor, Mich.) containing 85% N₂, 10% H₂, and 5% CO₂. The subgingival plaque samples were also plated onto tryptic soy-serum-bacitracin-vancomycin (TSBV) medium selective for Actinobacillus actinomycetemcomitans (18). The TSBV medium was incubated at 37°C for 3 days in 5% CO₂-95% air. The presumptively identified A. actinomycetemcomitans isolates were further analyzed by 16S rRNA PCR analysis (2).

Culture and PCR identification of E. corrodens.

Aliquots of 0.1 ml of the serially diluted subgingival samples were plated onto the selective medium (nonselective brucella blood agar supplemented with 1 mg of clindamycin per liter) and incubated at 37°C in air with 5% CO₂ for 4 days (10). E. corrodens was initially identified on the plates by its characteristic colonial morphology and was then subcultured for further identification (10). To increase the probability of including different clonal types of E. corrodens within a sample for analysis, as many as 24 colonies were presumptively identified from each sample. The colonies were subcultured once to ensure purity and were further identified by an E. corrodens-specific 16S rRNA-based PCR analysis (2). The isolates confirmed to be E. corrodens were then subjected to clonal analysis by arbitrarily primed PCR (AP-PCR) (see below). The proportional recovery of E. corrodens was calculated from the total cultivable cell count in nonselective medium.

Clonal analysis of E. corrodens by AP-PCR.

The AP-PCR method for clonal analysis of E. corrodens described previously was used (5). Briefly, E. corrodens was grown on agar plates for 2 to 3 days. Two to 10 colonies from the agar plates were collected and suspended in 100 μl of lysing buffer (0.45% Nonidet P-40, 0.45% Tween 20, 10 mM Tris-HCl [pH 9.0], 50 mM KCl, 0.1% Triton X-100, 1 mM MgCl₂, 0.06 mg of proteinase K per ml), incubated at 50°C for 1 h, and boiled for 10 min and chilled on ice. The samples were centrifuged at 10,000 × g for 5 min in a microcentrifuge to remove large debris and unbroken cells. The supernatant was immediately used as the template in AP-PCR.

Three 10-base oligonucleotide primers (primers OPA-2, OPA-4, and OPA-10; Operon Technologies, Alameda, Calif.) were used for PCR. The base sequences (5′ to 3′) were TGCCGAGCTG, AATCGGGCTG, and GTGATCGCAG, respectively. The amplification mixture (final volume, 30 μl) consisted of 6 μl of template; each deoxynucleotide triphosphate (Pharmacia LKB Biotechnology, Piscataway, N.J.) at a concentration of 0.2 mM; 2.5 U of Taq DNA polymerase (Promega, Madison, Wis.); and 0.4 μM primer in 10 mM Tris-HCl (pH 9.0)-50 mM KCl-0.1% Triton X-100-4 mM MgCl₂. Amplification was performed in a thermal cycler (PTC-100; MJ Research, Watertown, Mass.). The temperature profile was 94°C for 5 min for initial denaturation and 36 cycles of 94°C for 1 min, 32°C for 2 min, and 72°C for 2 min. After the last cycle, the amplified products were extended for another 10 min at 72°C. Twenty microliters of the amplified products was separated by agarose gel electrophoresis (on a 1.5% agarose gel containing 0.5 μg of ethidium bromide per ml). After electrophoresis, the DNA banding patterns were visualized under UV illumination. One microgram of a 1-kb DNA ladder (Life Technologies, Gaithersburg, Md.) was used as the molecular size marker.

Statistical analysis.

The statistical significance of the levels of cultivable subgingival bacteria between subject groups was evaluated by the Mann-Whitney ranking test. The chi-square test or Fisher's exact test was used to determine the statistical significance of the differences of the prevalence of subgingival bacteria between subject groups. The associations of demographic, clinical, or microbiological variables and the numbers of distinct E. corrodens clones in individuals and in samples were analyzed by the Mann-Whitney ranking test. The Wilcoxon signed-rank test was used for statistical analysis of the differences in the cultivable levels of subgingival bacteria between the baseline and the follow-up examinations.

RESULTS

Demographic and clinical characteristics of the study subjects.

We noted no significant differences between the LAP and the GAP groups from the results of microbial analyses (data not shown). Therefore, the two groups with disease, the LAP and the GAP groups, were combined into a single subject group designated “subjects with aggressive periodontitis.” A total of 95 subjects were accepted into the baseline study, and this total included 44 periodontally nondiseased subjects and 51 subjects with aggressive periodontitis. The demographics of the study subjects and the clinical characteristics of the sample sites at the baseline are shown in Table 1. Subsequently, 22 nondiseased subjects and 11 subjects with aggressive periodontitis participated in the follow-up examination after an average interval of 14 months (range, 9 to 31 months) from the baseline examination. At the time of the follow-up examination, all 22 nondiseased subjects had received at least one prophylaxis. Among the 11 subjects with aggressive periodontitis, two had completed periodontal treatment, which included scaling and root planing, systemic antibiotic therapy, and osseous surgery of the affected sites. Five subjects with aggressive periodontitis had received scaling and root planing only, while three had received scaling and root planing plus systemic antibiotics. One subject with aggressive periodontitis did not receive any treatment.

TABLE 1.

Demographic and clinical characteristics of subject groups

Characteristic	Nondiseased subjects (n = 44)	Subjects with aggressive periodontitis (n = 51)
Mean (SD, range) age (yr)	25.2 (3.9, 11-30)	24.6 (4.6, 14-32)
Race or ethnicitya (no. of subjects)
African-American	0	12
Asian-American	23	17
Caucasian	12	3
Hispanic	4	19
Unidentified	5	0
Gender (no. of subjects)
Male	28	19
Female	16	32
Mean (SD, range) PD of sample site (mm)
Single	3.0 (0.4, 1-4)	7.1 (1.9, 4-13)
Pooled	2.7 (0.3, 2-4)	5.4 (1.3, 3-8)
Mean (SD, range) AL of sample site (mm)
Single	0	5.3 (2.4, 2-13)
Pooled	0	3.7 (1.8, 1-9)

^aSelf-identified.

Subgingival microbial profile at the baseline.

The results of microbial culture were similar for the single-site and the pooled samples for the two study subject groups. For the sake of simplicity, the results for the single-site and the pooled samples are analyzed and presented together unless specified otherwise.

Table 2 illustrates the prevalence and the cultivable level of subgingival bacterial species in subgingival samples from the nondiseased and aggressive periodontitis groups. Not surprisingly, the prevalence and/or the cultivable levels of the majority of the candidate periodontopathic bacteria were significantly elevated in the diseased subjects.

TABLE 2.

Cultivable bacteria in subgingival plaque samples from nondiseased subjects and subjects with aggressive periodontitis

Species	Prevalence (%) in samples from:		% of cultivable bacteria in samples from:
	Nondiseased groupa	Aggressive periodontitis groupb	Nondiseased group	Aggressive periodontitis group
Fusobacterium spp.	60.2	83c	8.2 ± 18.1	5.6 ± 7.0
E. corrodens	40.9	64.7c	0.4 ± 1.4	0.5 ± 2.3d
P. intermedia-P. nigrescens	29.5	61.8c	2.5 ± 5.3	9.4 ± 17.9d
P. melaninogenicus	12.5	39.2c	0.4 ± 1.6	2.2 ± 5.9d
Eubacterium spp.	12.5	37.3c	0.3 ± 1.0	1.2 ± 5.2
Campylobacter spp.	8.0	42.2c	0.3 ± 1.1	2.8 ± 6.2d
A. actinomycetemcomitans	2.3	18.6c	0.0 ± 0.3	1.6 ± 7.5
T. forsythensis	2.3	26.5c	0.1 ± 0.7	1.2 ± 3.3d
P. gingivalis	1.1	41.2c	0.0 ± 0.0	5.0 ± 10.2d

^aEighty-eight single-site and pooled samples from 44 nondiseased subjects were examined.

^bOne hundred two single-site and pooled samples from 51 diseased subjects were examined.

^cA significant difference was detected between nondiseased subjects and subjects with aggressive periodontitis by the chi-square test or Fisher's exact probability test (P < 0.05).

^dA significant difference (P < 0.05) between nondiseased subjects and subjects with aggressive periodontitis was detected by the Mann-Whitney ranking test.

Forty-one percent of the samples from nondiseased subjects and 65% of the samples from subjects with aggressive periodontitis harbored cultivable E. corrodens at the baseline. The organism constituted, on average, 0.5% or less of the total cultivable subgingival bacteria in plaque samples. No synergistic relationship was detected between the presence or the cultivable levels of E. corrodens and those of any other study periodontal pathogens, including A. actinomycetemcomitans (data not shown).

The clonal diversity of E. corrodens at the baseline.

Previous studies (5) suggested that multiple distinct clones of E. corrodens may be found in individual periodontal pockets. To ensure the inclusion of as many different E. corrodens clonal types as possible, multiple E. corrodens colonies were recovered from primary cultures for species identification by 16S rRNA PCR analysis. The resultant 1,021 confirmed E. corrodens colonies from 103 E. corrodens-positive plaque samples were then subjected to AP-PCR clonal analysis. Figure 1 illustrates an example of the results of clonal analysis for six E. corrodens clinical isolates by AP-PCR with two different random primers (primers OPA-2 and OPA-4 for lanes 2 to 7 and lanes 9 to 14, respectively). The results of the AP-PCR analysis distinguished the six E. corrodens clinical isolates into three clonal types.

FIG. 1.

Gel electrophoresis patterns of amplicons of six E. corrodens clinical isolates obtained by AP-PCR with OPA-2 (lanes 2 to 7) and OPA-4 (lanes 9 to 14) as the random primers. The same six clinical isolates were used in the same order in lanes 2 to 7 and lanes 9 to 14, respectively. Lanes 1 and 8, DNA size markers. The six isolates were grouped into three distinct clonal types (lanes 2 and 3, lanes 4 and 5, and lanes 6 and 7, respectively) with OPA-2, and the results were confirmed with OPA-4 (lanes 9 and 10, lanes 11 and 12, and lanes 13 and 14, respectively).

The distribution patterns of the distinct E. corrodens clones within individuals (combined from both the single-site and the pooled samples for each individual) at the baseline is illustrated in Fig. 2. Thirty-nine percent of the nondiseased subjects and 63% of the subjects with aggressive periodontitis harbored more than one clone of E. corrodens. The numbers of within-individual E. corrodens clones were higher in the subjects with aggressive periodontitis than in nondiseased subjects (Mann-Whitney ranking test, P < 0.05).

FIG. 2.

Pattern of E. corrodens clonal diversity within individuals in the nondiseased and the aggressive periodontitis groups at the baseline. The number of distinct E. corrodens clones represented all distinct clones identified from both the single-site and the pooled samples for each individual.

Table 3 shows the variables associated with an increased clonal diversity of E. corrodens on the basis of the sample sites. The clonal diversity of E. corrodens increased in the subgingival sites in subjects with aggressive periodontitis, in sites with PDs of 4 mm or greater, in sites with ALs of 2 mm or greater, and in subgingival sites colonized by P. gingivalis (Mann-Whitney ranking test, P < 0.05). There was a trend (but not statistically significant) for an increased clonal diversity of E. corrodens in sites colonized by other periodontal pathogens listed in Table 2 (data not shown).

TABLE 3.

Characteristics associated with an increased genotype diversity of subgingival E. corrodens at the baseline^a

Characteristic (no. of subjects)	No. of E. corrodens clonal typeb
Sample source
Nondiseased (88)	0.8 ± 1.1
Aggressive periodontitis (102)	1.7 ± 1.7
PDs of sample sites
<4 mm (91)	0.8 ± 1.2
≥4 mm (99)	1.6 ± 1.7
ALs sample sites
<2 mm (96)	0.9 ± 1.2
≥2 mm (94)	1.6 ± 1.7
P. gingivalis coinfection of sample sites
Positive (20)	0.9 ± 1.3
Negative (75)	2.0 ± 1.7

^aBoth single-site and pooled samples were used for the analyses.

^bFor each characteristic, a significant difference (P < 0.05) between the two groups of subjects was detected by the Mann-Whitney ranking test.

The demographic factors were examined for their possible confounding effects on the clonal diversity of subgingival E. corrodens. No statistically significant differences were found in the clonal diversity of the subgingival E. corrodens isolates between samples from males and females in either the nondiseased or the aggressive periodontitis subject group (Mann-Whitney ranking test, P = 0.54 and 0.70 for the nondiseased and aggressive periodontitis groups, respectively). Age was also not a factor in the clonal diversity of the subgingival E. corrodens isolates (data not shown).

Due to the limited subject sample size, the effect of race or ethnicity was examined for nondiseased subjects in the Asian-American and Caucasian groups and for the subjects with aggressive periodontitis in the African-American, Asian-American, and Hispanic groups. None of the races or ethnicities was found to affect the clonal diversity of the subgingival E. corrodens isolates (Mann-Whitney ranking test, P > 0.05). However, we noted a trend for a higher clonal diversity among subgingival E. corrodens isolates from subjects with aggressive periodontitis in the Hispanic group than among isolates from subjects with aggressive periodontitis in the Asian-American group (Mann-Whitney ranking test, P = 0.055).

Stability of the cultivable subgingival microbiota.

Table 4 and Table 5 compare the microbiological findings between the baseline and the follow-up examinations for nondiseased subjects and subjects with aggressive periodontitis, respectively. In the nondiseased group, the subgingival microbiota was relatively stable. In contrast, the subgingival microbiota in the subjects with aggressive periodontitis appeared to be less stable. The prevalence of E. corrodens, P. intermedia-P. nigrescens, Fusobacterium spp., and P. melaninogenicus was significantly lower (chi-square test or Fisher's exact test, P < 0.05) at the follow-up examination. The cultivable levels of subgingival T. forsythensis, Eubacterium spp., and P. melaninogenicus also decreased significantly at the follow-up examination (Wilcoxon signed-rank test, P < 0.05).

TABLE 4.

Comparison of microbial compositions between baseline and follow-up samples from nondiseased subjects^a

Species	Prevalence (%) in samples from:		% of cultivable bacteria in samples from:
	Baseline	Follow-up	Baseline	Follow-up
E. corrodens	38.6	36.4	0.5 ± 1.7	0.5 ± 1.4
A. actinomycetemcomitans	0	0	0 ± 0	0 ± 0
P. gingivalis	2.3	0	0.0 ± 0.1	0 ± 0
P. intermedia-P. nigrescens	25	29.5	1.5 ± 3.6	2.4 ± 5.7
T. forsythensis	2.3	9.1	0.0 ± 0.1	0.2 ± 0.9
Campylobacter spp.	4.5	9.1	0.1 ± 0.6	0.5 ± 1.7
Eubacterium spp.	11.4	6.8	0.3 ± 1.0	0.3 ± 1.3
Fusobacterium spp.	36.4	61.4b	6.6 ± 19.1	7.2 ± 13.1
P. melaninogenicus	6.8	18.2	0.1 ± 0.3	1.0 ± 5.0

^aForty-four single-site and pooled samples from 22 nondiseased subjects were examined.

^bA significant difference (P < 0.05) between baseline and follow-up samples was detected by the chi-square test.

TABLE 5.

Comparison of microbial composition between baseline and follow-up samples from subjects with aggressive periodontitis^a

Species	Prevalence (%) in samples from:		% of cultivable bacteria in samples from:
	Baseline	Follow-up	Baseline	Follow-up
E. corrodens	63.6	27.3b	0.1 ± 0.2	3.3 ± 12.5
A. actinomycetemcomitans	13.6	0	0.0 ± 0.0	0 ± 0
P. gingivalis	45.5	22.7	5.9 ± 10.3	1.5 ± 3.8
P. intermedia-P. nigrescens	81.8	45.5b	14.4 ± 18.7	9.2 ± 24.7
T. forsythensis	45.5	22.7	3.4 ± 5.3	0.1 ± 0.2c
Campylobacter spp.	40.9	27.3	1.5 ± 2.9	1.1 ± 2.6
Eubacterium spp.	50	18.2	1.4 ± 1.9	0.2 ± 0.6c
Fusobacterium spp.	95.5	68.2b	6.7 ± 5.3	4.8 ± 5.8
P. melaninogenicus	45.5	13.6b	3.6 ± 9.6	1.0 ± 4.0c

^aTwenty-two single-site and pooled samples from 11 subjects with aggressive periodontitis were examined.

^bA significant difference (P < 0.05) between baseline and follow-up samples was detected by the chi-square test and Fisher's exact probability test.

^cA significant difference (P < 0.05) between baseline and follow-up samples was detected by the Wilcoxon signed-rank test.

Clonal stability of subgingival E. corrodens over time.

A total of 234 E. corrodens isolates from 22 E. corrodens-positive plaque samples (from 22 nondiseased and 11 subjects with aggressive periodontitis) were recovered and subjected to AP-PCR analysis. The clonal typing results were compared to the E. corrodens clonal types found at the baseline. In the simplest terms, clonal stability was indicated by the presence of identical E. corrodens clones at both the baseline and the follow-up examinations, while clonal instability was exemplified by the emergence or the disappearance of distinct E. corrodens clonal types in the follow-up examination.

Table 6 summarizes the results of the clonal stability analysis. We first differentiated eight possible patterns of change in clonal stability, of which seven were identified in this study. The patterns of change can be further grouped into three categories, each of which represents a different degree of clonal stability: stable (category A), unstable (category B), and partially stable (category C) (Table 6). We did not note any significant difference in the patterns of clonal stability between the nondiseased subjects and the subjects with aggressive periodontitis. Twenty-six samples (single-site and pooled samples from both subject groups) exhibited the category A patterns (25 samples with pattern 0→0 and 1 sample with pattern X→X). Thirty-eight samples exhibited category B patterns (10 samples with pattern 0→X, 19 samples with pattern X→0, and 9 samples with pattern X→Y). The remaining two samples exhibited clonal changes of category C (patterns X→XY and XY→XZ). We did not find any variables (demographic, clinical, or microbiological) to be associated with specific patterns of clonal stability in the sample sites.

TABLE 6.

Patterns of E. corrodens clonal stability in nondiseased subjects and subjects with aggressive periodontitis

Pattern of clonal stability and change	No. of clones in:
	Nondiseased group		Group with aggressive periodontitis
	Single-site sample	Pooled sample	Single-site sample	Pooled sample
A (stable)
0→0a	10	9	3	3
X→Xb	1	0	0	0
B (unstable)
0→Xc	4	4	1	1
X→0d	4	5	4	6
X→Ye	3	2	3	1
C. Partially stablef
X→XY	0	1	0	0
XY→XZ	0	1	0	0
XY→X	0	0	0	0

^aNo E. corrodens isolates were found at the baseline or at the follow-up examination.

^bIdentical E. corrodens genotypes were found at the baseline and at the follow-up examination.

^cE. corrodens isolates were not found at the baseline but were found at the follow-up examination.

^dE. corrodens isolates were found at the baseline but were not found at the follow-up examination

^eE. corrodens isolates were found at both the baseline and the follow-up examination, but the isolates were genetically distinct.

^fE. corrodens isolates were found at both the baseline and the follow-up examination; at least one genetically distinct isolate persisted in both examinations, while other distinct clones appeared or disappeared at the follow-up examination.

DISCUSSION

The present study examined the subgingival microbial compositions of periodontally nondiseased subjects and subjects with aggressive periodontitis and focused on the colonization patterns of genetically distinct E. corrodens isolates within individuals. AP-PCR clonal analysis was a convenient and sensitive method for distinguishing E. corrodens clones rapidly. In general, the AP-PCR generated amplicons ranging from 0.5 to 2 kb. After gel electrophoresis of the amplicons obtained by AP-PCR, the gel patterns were usually quite distinct among different E. corrodens strains and could be interpreted unambiguously. The results of the AP-PCR clonal analyses with the different random primers were always consistent. To avoid difficulties in the interpretation of AP-PCR gel patterns, E. corrodens isolates from the same subjects were always compared in the same gel and electrophoresis was repeated twice or more for each primer used.

E. corrodens is a naturally transformable species (20) and may acquire new genotypes through recombination. However, the AP-PCR patterns among the E. corrodens clones were quite distinct and suggested broad genetic differences among the clones. It seems unlikely that in vivo recombination between E. corrodens strains will result in genetic differences that generate such distinct AP-PCR patterns among different clones.

At the baseline, 17 of the 44 (39%) nondiseased subjects and 32 of the 51 (63%) subjects with aggressive periodontitis were found to harbor more than one clone of E. corrodens. In general, the clonal diversity of the subgingival E. corrodens isolates was increased in sites with periodontitis and also in sites coinfected with P. gingivalis. No other demographic or microbiological variables were found to be significantly associated with the subgingival clonal diversity of E. corrodens. However, we noted a trend (but not statistically significant) for a higher number of subgingival E. corrodens clones in diseased Hispanic subjects than in diseased Asian-American subjects. The association between the clonal diversity of subgingival E. corrodens and race or ethnicity remains to be determined.

There may be a simple explanation for the multiclonal infection with E. corrodens, in contrast to the predominantly monoclonal infection with bacterial species such as A. actinomycetemcomitans. E. corrodens is a commensal bacterium in gingival crevices. All humans may have harbored subgingival E. corrodens some time in their lives. The high prevalence of E. corrodens may lead to a greater chance for individuals to be exposed to exogenous E. corrodens and a greater accumulation of distinct E. corrodens clones in individuals over their lifetimes. In contrast, the predominantly monoclonal infection with A. actinomycetemcomitans may be a consequence of the relatively low prevalence of this bacterium in the general population.

The variable efficiency of colonization may further affect the likelihood of multi- or monoclonal infection with oral bacteria. The result of colonization with bacteria that can easily colonize the oral cavity may be multiclonal infections, and vice versa. Perhaps E. corrodens can colonize subgingival sites more efficiently than species such as A. actinomycetemcomitans. The ease of de novo colonization may be influenced by the presence of periodontal disease or other subgingival bacterial species. In this regard, Mandell and colleagues (15, 16) reported a synergistic relationship between E. corrodens and A. actinomycetemcomitans. We noted an association between the clonal diversity of E. corrodens and coinfection with P. gingivalis at the same sites. However, we did not detect any association between the cultivable level or the clonal diversity of E. corrodens and coinfection with A. actinomycetemcomitans in subgingival sites.

Twenty-two of the 44 nondiseased subjects and 11 of the 51 subjects with aggressive periodontitis were followed up and resampled. We did not attempt to control the treatment received by the nondiseased subjects or the subjects with aggressive periodontitis. By the time of the follow-up examination, all nondiseased subjects had received at least one professional prophylaxis. The treatments received by the subjects with aggressive periodontitis, while highly variable, were generally more disruptive to the subgingival microbiota than those received by the nondiseased subjects. The clonal types of E. corrodens from the baseline and the follow-up examinations were compared for each individual to evaluate the stability of the subgingival colonization. Our results suggested that a clonal shift of the subgingival E. corrodens isolates was very common. Thirty-eight of the 66 follow-up samples (58%) showed a complete change of the E. corrodens clonal types from those detected at the baseline examination. Among the 26 follow-up samples that showed clonal stability, 25 were devoid of E. corrodens at either the baseline or the follow-up examination. Perhaps these sites were inhospitable to E. corrodens. When the samples from the sites where colonization by E. corrodens may not be possible were excluded, 38 of the 41 samples (93%) were categorized as unstable. Such a degree of instability was not discernible when the subgingival bacterial content was analyzed to the species level (for example, compare the cultivable microbial species at the baseline and the follow-up examinations in Table 4).

The clonal stability of E. corrodens may be compared to those from other studies of persistent colonization of bacteria in humans. Suchett-Kaye et al. (19) examined the clonal stability of subgingival Fusobacterium nucleatum in healthy young adults. Among eight dental students harboring F. nucleatum at the baseline, none was found to harbor the same clonal types of F. nucleatum 16 months later. The high turnover of F. nulceatum in their study appeared to be similar to the frequent clonal shift of E. corrodens in this study. Saarela et al. (17) and Ehmke et al. (11) examined the persistence of A. actinomycetemcomitans in subjects with or without periodontal interventions. The results showed that while A. actinomycetemcomitans may be eliminated by periodontal treatment, the recurrent strains invariably belonged to the same clonal types, indicating a capacity of this bacterial species for persistent infection. Lamell et al. (14) showed that colonization with A. actinomycetemcomitans and P. gingivalis appeared to be transient in nondiseased children and young adults. In comparison, subgingival colonization with E. corrodens appears to be less stable than that with A. actinomycetemcomitans in adult subjects.

The pattern of clonal shift of subgingival E. corrodens was also similar to that exhibited by Escherichia coli in the gastrointestinal tract. The stability of E. coli in the gastrointestinal tract was examined by Caugant et al. (4). The electrophoretic types (ETs) of the E. coli isolates recovered from fecal samples from a human host were determined over an 11-month period. Most of the ETs appeared for a short duration and were termed transient. The complete turnover of transient ETs was quite common in the fecal samples. However, two ETs persisted many times over long periods of times and were termed residents. Hartley et al. (13) examined fecal samples from nine adults and four children over a period of 24 to 265 days. The fecal samples were obtained twice a week. The O antigens of the E. coli isolates recovered were determined. The results showed a wide spectrum of stability of the E. coli types in the fecal samples of individuals. Some of the E. coli types persisted for a short period of time, disappeared, but then reemerged later. Other E. coli types appeared for a short duration and then disappeared completely. Still other E. coli types seemed to persist through the greater part of the experimental period.

In summary, it was common to find clonally diverse E. corrodens isolates in single subgingival sites. The diversity of the E. corrodens isolates increased in sites with periodontitis and sites coinfected with P. gingivalis. Examination of the E. corrodens isolates at the follow-up examination showed several patterns of clonal shift. Some sites were not colonized with E. corrodens at both time points. A few sites harbored the same E. corrodens clones over the period examined. However, a significant proportion of the sites examined showed the acquisition of E. corrodens clones, the loss of new E. corrodens clones, or the complete replacement of the original clones by new clones at the follow-up examination. E. corrodens colonization did not appear to result in a pattern of persistent infection with the same clones. Our results suggest a complexity of subgingival microbiota not seen previously.