Characterization of Actinomyces with Genomic DNA Fingerprints and rRNA Gene Probes
G. BOWDEN', J. JOHNSON, and C. SCHACHTELE2
Department of Oral Sciences and Clinical Research Center for Periodontal Diseases, School of Dentistry, University
of Minnesota, Minneapolis, Minnesota 55455; and 'Department of Oral Biology, Faculty of Dentistry, University of
Manitoba, Winnipeg, Manitoba, Canada R3E OW2
Cellular DNA from 25 Actinomyces naeslundii and Acti-
nomvces viscosus strains belonging to the 7 taxonomic
clusters of Fillery et al. (1978) and several unclustered
strains was obtained by enzymatic and N-lauroylsarcosine/
guanidine isothiocyanate treatment of whole cells, fol-
lowed by extraction of the nucleic acid. The DNA samples
were digested with restriction endonucleases BamHl or
PvuII, and agarose gel electrophoresis was used to obtain
DNA fingerprints. The DNA fragments were subjected to
Southern blot hybridization with a digoxigenin-labeled
cDNA probe transcribed from Escherichia coli 16S and
23S rRNA. The patterns of bands from genomic (DNA
fingerprints) and rDNA fingerprints (ribotypes) were used
for comparison between the taxonomic cluster strains and
strains within clusters. Representative strains from each
taxonomic cluster provided different BamHI DNA finger-
prints and ribotype patterns with 3 to 9 distinct bands.
Some strains within a cluster showed identical ribotype
patterns with both endonucleases (A. naeslundii B120
and A. naeslundii B102 from cluster 3), while others
showed the same pattern with BamHI but a different
pattern with PvuII (A. naeslundii ATCC 12104 and 398A
from cluster 5). A. viscosus ATCC 15987 (cluster 7) and its
parent strain T6 yielded identical fingerprint and ribotype
patterns. The genomic diversity revealed by DNA finger-
printing and ribotyping demonstrates that these tech-
niques, which do not require phenotypic expression, are
suited for study of the oral ecology ofthe Actinomyces, and
for epidemiological tracking of specificActinomyces strains
associated with caries lesions and sites of periodontal
destruction.
J Dent Res 72(8):1171-1179, August, 1993
Introduction.
Actinomyces, particularly strains designatedA. naeslundii
andA. uisCosus, make up a major portion of the indigenous
human oral microflora (Socransky and Manganiello, 1971;
Bowden et al., 1975; Ellen, 1976; Socransky et al., 1977;
Sanyal and Russell, 1978). These bacteria are of special
interest, due to their suggested involvement in several
diseases (Jordan, 1982), including periodontitis (Loesche
and Syed, 1978; Moore et al., 1982, 1984; Johnson et al.,
1990) and possibly root-surface caries (Syed et al., 1975;
Received for publication October 29, 1992
Accepted for publication January 21, 1993
rTo whom correspondence and reprint requests should be ad-
dressed
This research was supported in part by USPHS Grant #1 P50
DE98489 from the National Institute of Dental Research, National
tnstitutes of Health, Bethesda, MD 20892, and Grant MT-7611 from
the Medical Research Council of Canada, Ottawa. Canada K1A0W9.
Downloaded from jdr.sagepub.com at UNIV OF MICHIGAN on January 5, 2015 For personal use only. No other uses without permission.
Ellen et al., 1985; Bowden et al., 1990). In addition, these
species are unique to the mouth and associated tissues,
comprise a significant component of dental plaque, and
also colonize mucosal surfaces. Any understanding of the
ecology of Actinomyces is, therefore, important in appre-
ciating the interaction of the human host with members of
their resident oral flora (Ellen, 1976).
Much effort has been directed toward the taxonomy of
the Actinomyces (Schaal, 1986), including various physi-
ological tests, carbohydrate fermentation (Schofield and
Schaal, 1980; Kalfas and Edwardsson, 1990), cellular and
colonial morphology (Slack and Gerencser, 1975), cell wall
analysis (Cummins and Harris, 1958), polyacrylamide gel
electrophoresis of whole-cell protein preparations
(McCormick et al., 1985), and DNA hybridization
(Coykendall and Munzenmaier, 1979; Johnson et al.,
1990). Serological techniques have also been used for
identification of species and serotypes (Bowden et al.,
1976), with initial emphasis on gel diffusion (King and
Meyer, 1963; Georg et al., 1964; Lambert et al., 1967;
Slack et al., 1969) and fluorescent antibody staining
(Brock and Georg, 1969; Slack and Gerencser, 1970;
Holmberg and Forsum, 1973; Gerencser and Slack, 1976;
Gerencser, 1979). Fillery et al. (1978) used 65 tests on 43
strains of A. viscosus and A. naeslundii from human and
animal sources and applied numerical analysis to divide
these bacteria into 7 taxonomic groupings (i.e., clusters).
More recently, monoclonal antibodies (Firtel and Fillery,
1988) have been used for identification of antigenic deter-
minants on strains representative ofthe clusters, with the
conclusion that the surfaces ofA. viscosus andA. naeslundii
present a complicated mosaic of epitopes responsible for
cross-reactions between isolates. In all ofthese approaches
to characterization and identification ofA. viscosus andA.
naeslundii, high degrees of phenotypic and serological
similarity among strains have been observed, although
serotyping is possible (Gerencser, 1979). At this time,
there is no completely satisfactory series of phenotypic
tests that will distinguish between these two species of
Actinomyces from humans (Johnson et al., 1990). This is
obviously a major deterrent to studies attempting to
relate specific types ofA. naeslundii andA. viscosus to oral
sites and/or disease. Johnson et al. (1990) have proposed,
based on a detailed study ofthe DNA relatedness between
strains of Actinomyces, that isolates of A. naeslundii and
A. viscosus from humans be collected into two genospecies:
A. naeslundii genospecies 1 to include A. naeslundii sero-
type I and A. naeslundii genospecies 2 to include A.
viscosus serotype II andA. naeslundii serotypes II and III.
A. viscosus serotype I strains from animals would be
retained as A. viscosus.
Recent developments in DNA technology have pro-
vided new approaches to bacterial taxonomy. DNA finger-
printing has recently been used to distinguish between
1171
1172 BOWDEN et al. J Dent Res August 1993

different strains of a number of bacterial species (Owen, All of the strains either were from the culture collec-
1989). Genco and Loos (1991) have reviewed the use of tions of the University of Minnesota and the University of
DNA fingerprinting for studies on oral bacteria, including Manitoba, or were purchased from the American Type
Streptococcus mutans (Caufield and Walker, 1989; Culture Collection (Rockville, MD). Frozen (-80°C) or
Kulkarni et al., 1989), Haemophilus influenzae (Loos et lyophilized stocks were grown on supplemented blood
al., 1989), Eikenella corrodens (Chen et al., 1990), agar plates (Takahashi and Schachtele, 1990) at 37°C in
Porphyromonas gingivalis (Loos et al, 1990; Loos et al., candle jars and identification confirmed by colony mor-
1992), Fusobacterium nucleatum (Dzink et al., 1990), and phology, Gram stain, sugar fermentation pattern (John-
Actinobacillus actinomycetemcomitans (Zambon et al., son et al., 1990), and by whole-cell agglutination analysis
1990; Han et al., 1991). DNA fingerprinting is providing with rabbit antisera raised against strains representative
important new information on the relatedness of oral of clusters (i.e., taxonomic groupings) 1 through 6 of
bacterial strains and will, when used appropriately, pro- Fillery et al. (1978). Actinomyces strains from clusters 1,
vide insight into the intra-oral and inter-oral transmis- 2, 3, 4, and 6 were placed intoA. naeslundii genospecies 2
sion of these bacteria. and cluster 5 strains into A. naeslundii genospecies 1,
One problem with bacterial DNA fingerprinting is the based on a comparison with agglutination results with
large number of bands (. 50) found in the agarose gels strains W1053 and ATCC12104, respectively (Putnins
(Owen, 1989). Fingerprinting can be successfully used to and Bowden, 1991). Strains W1053 and ATCC 12104 were
identify patterns that are identical, but it is difficult forutilized in the studies in which Johnson et al. (1990)
this technology to be used to build a strain identification defined two groups of oral Actinomyces designated A.
scheme. One partial solution to this problem is utilization naeslundii genospecies 1 and 2. In this paper, strains are
of cDNA probes which allow for the identification of designated as A. naeslundii genospecies 1 or 2, or A.
fingerprint bands containing specific DNA sequences. viscosus, following the proposal of Johnson et al. (1990).
Specific probes have been developed for the DNA of oral DNA extraction. -Total cellular DNA was extracted by
bacteria, including the Actinomyces (Stackebrandt and modifications of several methods (Barsotti et al., 1988;
Charfreitag, 1990; Yeung, 1992), A. actinomycetem- Pitcher et al.,1989; Smith et al., 1989a) to obtain consis-
comitans (DiRienzo et al., 1990), and various other tent cell lysis and DNA recovery. The method was ad-
subgingival bacteria (Smith et al., 1989b; Dix et al., 1990).equate for obtaining DNA from all of the strains tested.
Ribotyping is a general and widely used probing sys- Cells were grown at 37°C on supplemented blood agar
tem based on restriction fragment length polymorphisms plates in a candle jar for a minimum of 48 h. A sterile
in the chromosomal DNA containing rRNA genes (Grimont inoculating loop was used to obtain cells (2-3 loops full, ca.
and Grimont, 1986; Stull et al., 1988; Owen et al., 1988; 80 mg wet weight) which were placed into 1.0 mL of 10
Baloga and Harlander, 1991). This technique involves the mmol/L Tris (pH 8.0) with 1 mmol/L EDTA (TE buffer).
probing ofDNA fingerprints, after transfer to nylon mem- The suspension was briefly sonicated for 30 s and the cells
branes, with labeled cDNA transcribed from Escherichia collected by centrifugation. A mixture of chicken egg-
coli 16S and 23S rRNA. Since rRNA is ubiquitous in white lysozyme (15 mg/mL, L6876, Sigma, St. Louis, MO)
bacteria, the molecule's ribonucleotide sequences are con- and Achromobacter tyticus achromopeptidase (5 mg/mL,
served (Woese, 1987), and bacterial rRNA cistrons are A3547, Sigma) in TE buffer was added (150 ML) and the
present in multiple copies (Nomura and Post, 1980). cells suspended by being vortexed. The mixture was
Therefore, ribotyping is applicable to molecular epidemio- incubated at 370C for 60 min, and then 10 pL of Pronase
logical investigations on a diversity of bacteria (Grimont E (Type XXV, Sigma, 20 mg/mL TE buffer) was added, and
and Grimont, 1986). after 30 min at 370C the cells were lysed by the addition
In this communication, we describe a technique for of 750 tL of guanidine isothiocyanate [5535ua, Bethesda
effective cell lysis and extraction of chromosomal DNA Research Laboratories, Gaithersburg, MD; 60 g/20 mL of
from strains ofA. naeslundii andA. viscosus, and demon- 0.5 mol/L EDTA, pH 8.0; 20 mL H20; 5 mL 10% N-
strate that the taxonomic clusters of these bacteria yield lauroylsarcosine (Sigma), heated until dissolved and fil-
different DNA fingerprints, and that a cDNA probe for tered through a 0.45-rtm membrane]. During 30 min of
rRNA genes can be used to distinguish strains from incubation at 37°C, the mixture was briefly vortexed
different clusters and, in several instances, strains within several times. After centrifugation for removal of any
a specific cluster. The genetic heterogeneity of the Actino- debris, the viscous supernatant was removed, treated
myces, as revealed by DNA fingerprinting and ribotyping, with 370 pL of cold (-20°C) 7.5 mol/L ammonium acetate,
provides an opportunity to evaluate both the relatedness and kept on ice for 10 min. Chloroform:isoamyl alcohol
of large numbers of strains from human clinical studies (24:1) was added (500 gL), and after being mixed, the
and inter-oral and intra-oral transfer of these bacteria. aqueous phase was separated. The DNA was precipitated
with 600 pL of cold isopropanol and stored at -20°C for 10
Materials and methods. min. After centrifugation, the DNA pellet was washed
with 70% alcohol. The pellet was dissolved in 250 ,L of
Bacteria and growth conditions.-The bacterial strains H20 by incubation at 60°C; sodium chloride (1.4 mol/L,
used in this study are presented in Tables 1 and 2, and we 250 pL) was added, followed by cetyl trimethyl ammo-
have retained the nomenclature of the strains as they nium bromide (10%, 50 pL, Sigma), and the mixture
were received in the laboratory. In the text and Figs., incubated for 30 min at 60°C. The mixture was cooled,
strains have been identified by their cluster and as A. vortexed with 400 pL of chloroform: isoamyl alcohol, and
naeslundii genospecies 1 or 2 or as A. viscosus following centrifuged. After centrifugation, the supernatant was
Johnson et al. (1990) (see below). Reference to Table 1 will removed and mixed with 1 mL of cold ethanol. After
give the previous nomenclature of strains. precipitation, the DNA was washed 3 times with 70%
Downloaded from jdr.sagepub.com at UNIV OF MICHIGAN on January 5, 2015 For personal use only. No other uses without permission.
Vol i ' ANo ' DNA FINGERPRINTING AND RIBOTYPING OF ACTINOMYCES 1173

TABLE 1
ORIGIN OF STRAINS USED IN Actinomyces DNA FINGERPRINTING AND RIBOTYPING
Strains Origin Source
A. viscosuls serotype II
WVVJa 627 human mouth Dr. M.A. Gerencser
WVUI 371 (ATCCb 19246) human lesion Dr. M.A. Gerencser
W1053 (WVU474/ATCC27044) human sputum Dr. L.K. Georg
11 B2 plaque from New Guinea tribesman Dr. R.G. Schamschula
8AO 6 plaque from New Guinea tribesman Dr. R.G. Schamschula
Be 32 carious dentin Dr. S. Edwardsson
Be 64 carious dentin Dr. S. Edwardsson
H 103 child's dental plaque Dr. G.H. Bowden
B 25 child's dental plaque Dr. G.H. Bowden
B236 child's dental plaque Dr. G.H. Bowden
AN 6 -- - - - - -- plaque
human dental
---- -- - - jc- - - -I- - Dr. K. Holmberg
--- --- - - C7,
,A. naeslundii serotype I
ATCC 12104 (WVU45/W826) sinus after tooth extraction ATCC
W\TU 45 (ATCC 12104/W826) sinus after tooth extraction Dr. M.A. Gerencser
WVU 398A human dental calculus Dr. M.A. Gerencser
A. naes/u ndii
WV 752 wound abscess Dr. L.K. Georg
B 120 child's dental plaque Dr. G.H. Bowden
B 102 child's dental plaque Dr. G.H. Bowden
134 child's dental plaque Dr. G.H. Bowden
VTPI 3428A human mouth Dr. C.S. Cummins
WV1048 (ATCC 27038) human lesion Dr. L.K. Georg
X 600 (ATCC 19039) human dental plaque Dr. L.K. Georg
A. ViSC(SlIS serotype 1

ATCC 15987 (HT6/WVU 745) hamster mouth ATCC

T6 'ATCC 15987/WVU 754) hamster mouth Dr. P. Keyes
WXTVU 440 (HS2/X 601) hamster mouth Dr. M.A. Gerencser
NY I rat mouth Dr. J.S. van der Hoeven
aWest Virginia University.
'American Type Culture Collection.
cVirginia Polytechnic Institute.

ethanol and freeze-dried. Prior to use, the DNA was phoresis.-A survey of a wide range of restriction endo-
dissolved in 40 p.L of H20, and yield and quality were nucleases revealed that BamHI (Boehringer Mannheim
determined by measurement ofthe A260 and A280 (UV-1201 Biochemicals, Indianapolis, IN) caused complete diges-
UV-Vis Spectrophotometer, DNA/Protein data program, tion ofthe cellular DNA from all of theActinomyces cluster
Shimadzu, Columbia, MD). strains tested and provided clear DNA fingerprinting
Restriction enzyme digestion and agarose gel electro- patterns. PvuII (Boehringer Mannheim Biochemicals)
Downloaded from jdr.sagepub.com at UNIV OF MICHIGAN on January 5, 2015 For personal use only. No other uses without permission.
1174 BOWDEN et al. J Dent Res August 1993

was used in several experiments under conditions defined
by the manufacturer. Approximately 8 gg of DNA was
digested with 10 U of BamHI according to the
manufacturer's recommendations. A 1.0-kb DNA ladder
(Bethe'sda Research Laboratories) was used as a molecu-
lar-weight standard. DNA fragments were separated by
horizontal agarose gel electrophoresis (gel bed dimen-
sions, 16 cm wide by 15 cm long). Gels were run at constant
voltage (40 V) for 19 h at room temperature in Tris-borate
buffer (89 mmol/L Tris-base, 89 mmol/L boric acid, 2.4
mmol/L EDTA, pH 8.3). Following staining with ethidium
bromide, fragments were visualized with short-wave UY
light and photographed on Polaroid 57 Professional film
(Polaroid Corp., Cambridge, MA).
Synthesis ofprobes.-The larger rRNA subunits from
Escherichia coli (16S and 23S) obtained from Boehringer
Mannheim Biochemicals were used for preparation of
cDNA probes by a nonradioactive labeling method. The
reverse transcription reaction mixture contained: 1.0 jig
of 16S and 23S RNA; 0.2 mmol/L each of dATP, dCTP, and
dGTP; 0.1 mmol/L dTTP (Promega Research Labs, Madi-
son, WI); 0.05 mmol/L digoxigenin-11-dUTP (Boehringer
Mannheim Biochemicals); 0.2 jigrandom oligodeoxynucleo-
tide primers (Promega Research Labs); 400 U RNase H-
reverse transcriptase (Gibco BRL, Grand Island, NY); 1 X
reaction buffer (250 mmol/L Tris-HCl [pH 8.3], 375 mmoV
L KCl, 15 mmol/L MgCl2); and 2 ,uL of 0.1 mol/L
dithiothreitol. After incubation at 42°C for 6 h, the reac-
tion was stopped by the addition of 2 jiL of 0.5 molIL
EDTA. Standard molecular-weight DNA (K DNA cleaved
with HindIII molecular-weight marker II, digoxigenin-
labeled) was obtained from Boehringer Mannheim
Biochemicals. Labeled cDNA was precipitated by addi-
tion of glycogen (20 mg/mL, 2 jtL, Boehringer Mannheim
Biochemicals), 4 jL of 4 molIL lithium chloride, and 120
jiL cold ethanol.
Southern blot hybridization.-The DNA in the gel was
depurinated with 0.25 molIL HCl for 10 min, denatured in
0.5 mol/L NaOH-1.5 molIL NaCl for 30 min, and neutral-
ized in 0.5 mol/L Tris-HCl-1.5 mol/L NaCl (pH 7.6) for at
least 30 min. DNA was transferred to MagnaCharge
nylon transfer membranes (Micron Separations, Inc.,
Westboro, MA) with 20X SSC (SSC: 0.5 mol/L NaCl, 0.05
mol/L sodium citrate, pH 7.0) under vacuum (Model 785
Vacuum Blotter, BioRad, Richmond, CA). The membranes
were washed once in 6X SSC and baked at 70°C for 1 h.
Membranes were prehybridized at 70°C for 6 h in 5X SSC-
10% N-lauroylsarcosine-10% sodium dodecyl sulfate-2%
blocking agent (Genius Kit; Boehringer Mannheim
Biochemicals). The membranes were hybridized at 70°C

TABLE 2
Actinomyces STRAINS AND ESTIMATED SIZES OF rDNA-CONTAINING BANDS
IN BamHI DIGESTS OF CHROMOSOMAL DNA
Strain Fillery Cluster Band Size (kilobases ± standard deviation)
A. viscosus WVU627 1 11.4 ±0.49 8.2 + 0.2, 6.2 + 0.3, 4.8+ 0.7, 3.5+ 1.2, 3.3+ 1.3, 1.7+ 1.0
A. naeslundii W752 1 18.6 ± 1.3, 11.1 + 0.8, 9.6 + 0.5
A. viscosus H103 1 + 1.31. 3.7
23.2 ±0.6, 7.0+ 0.7, 6.0+ 0.6, 4.9 1.5, 3.4+ 1.5, 2.3+ 1.3
A. viscosus B236 2 17.9 ± 4.0, 13.4+ 2.9, 9.2 + 1.4, 5.4 + 0.8, 3.44+ 1.3, 3.33+ 1.4, 1.3± 1.0
A. naeslundii B120 3 23.0 ± 5.7, 18.1 + 1.0, 8.4 + 0.3
A. naeslundii B102 3 22.7, 18.7 + 0.1, 8.5 ± 0.2
A. naeslundii Be32 4 8.1 +0.2,6.0+0.7,5.3 +0.9, 4.8+1.2, 3.6+1.4, 3.5 ± 1.4
A. naeslundii WVU398A 5 22.1 ± 0.8, 13.8 + 1.59 9.0+ 0.4, 8.0+ 1.0, 6.0+ 0.7, 5.0 + 1.1, 4.0+ 1.3, 3.4+1.5,
2.2 ± 1.3
A. naeslundii X600 5 6.0 + 1.2, 5.0 + 1.4, 4.0 + 1.7, 3.5 + 1.7, 2.2 + 1.3
A. naeslundii 12104 5 5.9 + 1.3, 4.9 + 1.6, 4.0 + 1.8, 3.4 + 1.7, 2.2 + 1.4
A. viscosus W1053 6 12.5 ± 2.8, 7.3 + 1.6, 5.5 + 1.0, 4.1 + 1.1, 3.8 + 0.9
A. viscosus ATCC 15987 7 11.9 ± 1.4,9.3 +0.3, 8.1 +0.2, 7.9 +1.5, 4.9 +0.8, 4.2 +0.8, 4.0 ± 0.9, 3.4+ 0.9
A. viscosus WVU440 7 8.7 + 2.2, 5.3 + 0.4, 4.5 + 0.0, 4.3 + 0.1, 3.4 ± 0.6
A. viscosusNY1 7 13.2±0.3,8.4+0.1,8.822.3,7.3+1.6,5.4+0.4,4.5+0.0,4.3+0.0,3.3+0.7
A. viscosusT6 7 13.4±0.19.8+0.3,80.3,7.2±0.1,5.00+.1,4.5+0.1,4.3+0.04,3.8+0.0
A. viscosus WVU371 NC 21.8 ±0.1,7.4 +0.2,6.3 +0.06.5 ± 2.0, 5.1 + 1.2 4.4 +0.3, 4.2+ 0.1, 2.5± 0.8
A. naeslundii VPI3428A NC 17.3 ± 1.5, 8.5 + 0.2, 6.4 + 0.1,4.3+0.1, 4.0 ± 0.2, 1.9 ± 0.8
NC = no cluster assigned.
Downloaded from jdr.sagepub.com at UNIV OF MICHIGAN on January 5, 2015 For personal use only. No other uses without permission.
Vol. 72 No. 8 DNA FINGERPRINTING AND RIBOTYPING OF ACTINOMYCES 1175

x C1 C2 C3 C4 C5 C6 C7 x A B C D E F G H I J X

kb
23.1 4
23.1 . 9.4 * -
9.4. 6.6 . ..--e- .
_- - -
6.6* 4.4 *
---
- -_

4.4 *

2.3 .
2.0 .

Fig. 2-Ribotype patterns generated from Actinomnyces strains

belonging to clusters C1-C5. X, HindIII fragments ofX DNA labeled
with digoxigenin; Lane A, A. naeslundii genospecies 2 WVW627;
Lane B, A. naeslundii genospecies 2 W752; Lane C, A. naeslundii
genospecies 2 H103; Lane D,A. naeslundii genospecies 2 B236; Lane
E,A. naeslundii genospecies 2 B120; Lane F,A. naeslundii genospecies
2 B102; Lane G, A. naeslundii genospecies 2 Be32; Lane H, A.
Fig. 1-DNA fingerprints of Actinomyces cluster strains after naeslundii genospecies 1 WVU398A; Lane I,A. naeslundii genospecies
treatment of chromosomal DNA with BamHI. (X) Hindlll fragments 1 X600; and Lane J, A. naeslundii genospecies 1 ATCC 12104.
of X DNA ladder; (Cl) Cluster 1 strain A. naeslundi genospecies 2
WVU627; (C2) Cluster 2 strain A. naeslundii genospecies 2 B236;
for 3 to 5 fingerprinting experiments.
(C3) Cluster 3 strainA. naeslundii genospecies 2 B 120; (C4) Cluster
The BamHI DNA fingerprints obtained from strains
4 strain A. naeslundii genospecies 2 Be32; (CS Cluster 5 strain A.
naesluodii genospecies 1 W;VU398A; (C6) Cluster 6 strain A. representing each of the 7 Actinomyces clusters (Fig. 1)
naeslundii genospecies 2 W1053; (C7) Cluster 7 strain A. uiscosus
demonstrated that each ofthe tested strains had a distinct
ATCC 15987.
pattern, with numerous bands ranging from approxi-
mately 23 kb to less than 2 kb. The DNA was extracted
for 20 h in prehybridization solution containing freshly from each of these strains from 5 to 7 times, and in each
denatured digoxigenin-labeled DNA. After hybridization, case, restriction and fingerprinting provided patterns
the membranes were washed, and probed DNA was visu- which could not be distinguished by visual inspection.
alized colorimetrically as described in the Genius Kit. When the fingerprints from strains within clusters were
Analysis.-A Visage 110 Image Analyzer (Bio Image/ compared, it appeared from visual inspection that some
Kodak) was used for determination of the molecular patterns were very similar, while others were clearly
weights of the major bands obtained by ribotyping. Occa- different (e.g., cluster 1 strains A. naeslundii WVU627,
sionally, "ghost" bands were generated ifthe DNA concen- W752, and H103) (data not shown).
tration was too high or if the DNA was incompletely Figs. 2 and 3 present the ribotyping patterns obtained
digested; therefore, ribotype patterns were confirmed on with selected cluster strains, and Table 2 summarizes the
at least 5 independent gels, and the bands listed in Table results of ribotyping obtained from each strain after
1 were the major bands observed in every case. repeated experiments. Consistent with the DNA finger-
print patterns, the three cluster 1 strains (Fig. 2, lanes
Results. A,B,C) showed distinctly different patterns. Also consis-
tent with fingerprinting was the observation that the two
So that cellular DNA could be obtained from a wide range cluster 3 strains (A. naeslundii genospecies 2 strains B 120
of Actinomyces, a method was developed which combined and B102; Fig. 2, lanes E and F) provided identical
various enzymatic treatments with N-lauroylsarcosine/ patterns with 3 bands. The band at 18.1-18.7 kb in Fig. 2
guanidine isothiocyanate for cell lysis, the cells and ex- is faint and can be seen better in Fig. 4 (lanes A and B).
traction with appropriate salts and solvents. We have Interestingly, two cluster 5 strains gave identical ribotypes
tested this method on over 200 laboratory strains and (A. naeslundii genospecies 1 X600, ATCC 12104; Fig. 2,
fresh isolates and have not found any strain ofA. naeslundii lanes I and J), while a third strain, WVU398A, Fig. 2, lane
or A. viscosus which did not provide DNA which could be H, showed a similar pattern but with a group of additional
cut with BamHI. The DNAs obtained by the procedure high-molecular-mass bands. Close inspection of Fig. 2
hadaA 20 toA280 ratio of 1.8 ± 0.2 (mean + SD, n = 129), and (lanes I and J) reveals the presence of "ghost" bands in the
the small quantity of cells used provided DNA adequate higher-molecular-mass regions of the blot. Inspection of
Downloaded from jdr.sagepub.com at UNIV OF MICHIGAN on January 5, 2015 For personal use only. No other uses without permission.
1176 BOWDEN et al.
the DNA fingerprints of these strains indicated that they
had identical patterns (data not shown). Three different
cluster 7A. viscosus strains showed similar patterns with
5 bands in common but also with additional bands, mak-
ing each one unique (Fig. 3, lanes C,D,E). Interestingly,A.
viscosus T6 (Fig. 3, lane J) showed a ribotype pattern
identical with that ofA. viscosus ATCC15987 (Fig. 3, lane
C), which is consistent with T6 being the parent of
ATCC15987. Two unclustered strains (Fig. 3, lanes G and
H) showed ribotype patterns which were distinctly differ-
ent from any of the cluster strain patterns.
Comparison of strains (A. naeslundii genospecies 2
B120 and B102) with identical BamHI ribotyping pat-
terns (Fig. 4, lanes A and B) with PvuII used as the
restriction enzyme demonstrated identical ribotype pat-
terns (Fig. 4, lanes C and D), which adds support to the
close relationship between these two strains. However, in
contrast to strains B120 and B102, two A. naeslundii
genospecies 1 strains ATCC 12104 and WVU 398A, which
gave the same patterns on ribotyping with BamH I digests
(Fig. 5, lanes A and B), were not identical with PvuII
digests (Fig. 5, lanes C and D). Visual inspection of the
DNA fingerprints from each of the digests was consistent
with this observation, since the fingerprints ofA. naeslundii
genospecies 1 strains ATCC 12104 and WVU 398A with
BamHl appeared identical, while the patterns with PvuuI
were clearly different (data not shown).
Discussion.
The identification of a role for specific species or strains of
Actinomyces in oral diseases requires accurate identifica-
tion at the species and sub-species levels. Johnson et al.
(1990) have confirmed that phenotypic differentiation
x A B C D E F G H I
23.1 4
9.4 4
6.6 *
4.4 4
2.3 .
2.0 4
Fig. 3-Ribotype patterns generated from Actinomyces strains
belonging to clusters C6-C7 and several nonclustered strains. X
HindIII fragments of X DNA labeled with digoxigenin; Lane A, A.
naeslundii genospecies 2 W1053; Lane C, A. viscosus ATCC 15897;
Lane D, A. viscosus WVU440; Lane E, A. viscosus NY1; Lane G, A.
naeslundii genospecies 2 WVU37 1; Lane H,A. naeslundii VP13428A;
and Lane J, A. viscosus T6.
Downloaded from jdr.sagepub.com at UNIV OF MICHIGAN on January 5, 2015 For personal use only. No other uses without permission.
J Dent Res August 1993
between strains of A. naeslundii and A. viscosus-now
designated A. naeslundii genospecies 1 and 2 from hu-
mans-is difficult. Apparently, species not previously
identified, such asActinomycesgeorgeii, are present in the
human mouth, and these can be separated from the A.
naeslundii genospecies. It is in the differentiation be-
tween strains within the genospecies, at a level that will
be useful in ecological studies, that problems arise. Differ-
entiation between the genospecies is possible by aggluti-
nation (Putnins and Bowden, 1991), and serotypes can be
identified with fluorescent-labeled antibody. However,
serology suffers from the non-availability of standard
antisera and also the known variation in antisera from
different animals. Johnson et al. (1990) have noted varia-
tion in strain reactions with some of their absorbed fluo-
rescent-labeled antisera. Although Fillery et al. (1978)
used a large number of tests to place these bacteria into 7
x A B C D
23.1 -
9.4 -.
6.6 -+
J X
2.3 -.
1.9 -.
Fig. 4-Ribotype patterns generated from Actinomyces strains
within cluster 3. X HindIII fragments of X DNA labeled with
digoxigenin; Lane A, BamHI digest ofstrainA. naeslundii genospecies
2 B102; Lane B, BamHI digest of strain B102; Lane C, Pvull digest
of strainA. naeslundii genospecies 2 B102; and Lane D,PvuII digest
of strain B102.
Vol. 72 No. 8 DNA FINGERPRINTING AND RIBOTYPING OF ACTINOMYCES 1177

A ter. Our results demonstrate that some strains within a

B C D
given cluster did conform to this expectation, with similar
fingerprints and ribotypes (clusters 3, 5, 7), while in other
clusters, e.g., cluster 1, the strains were very different.
The diversity among the examined strains may relate
23.1 -. to their different origins (Table 1). It may be significant
- _ _ l__
I_ XIthat strains B102 and B120 originated from children in
England, while other strains were isolated from New
-* Guinea natives. The similarity between the animal strains
9.4 -4 _ .may also relate to their origins from a fairly restricted
- -habitat, the mouths of rats and hamsters living in captiv-
ity. Other studies of epidemiologically unrelated strains
of bacteria have also shown considerable diversity. De
__ _______
- Buyser et al. (1989) found 44 distinct HindIII and EcoRI
6.6 -+ _____ - B SrRNA gene restriction patterns ofStaphylococcus. Strains
within S. aureus showed a variety of patterns, where the
patterns of both endonucleases matched, e.g., pattern HI
-n , and E2, for five strains from humans. However, human
strains did not always give the same patterns with the
individual endonucleases; one strain gave patterns HI
4.4 -4 _____ iand E2, while a variety of pattern pairs was seen in
HindIII pattern 7. These patterns were: H7-E7, one strain;
H7-E8, two strains; and H7-E9, one strain. These results
are equivalent to our finding with ATCC 12104 and WVU
398A and suggest that these strains are different from
each other, compared with B120 andB102.
In this study, hybridization of digoxigenin-labeled E.
coli 16S + 23S rDNA probes, prepared from commercially
available E. coli rRNA, to DNA fingerprints ofActinomy-
ces naeslundii genospecies 1 and 2 and A. viscosus was
used successfully to differentiate strains within and among
the 7 taxonomic clusters of Fillery et al. (1978). These
methods may have application in tracing strains of A.
2.3 -4 naeslundii genospecies in ecological studies of the micro-
biology of the human resident flora and also in showing
the association ofspecific strains or ribotypes in restricted
1.9 -+ habitats such as caries lesions or areas of periodontal
destruction.

Baloga AO, Harlander SK (1991). Comparison of methods for dis-

Fig. 5-Ribotype patterns generated fromActinomyces strains in
cluster 5. X HindIII fragments of X DNA labeled with digoxigenin;
Lane A, BamHI digest of strain A. naeslundii genospecies 1 ATCC
12104; Lane B, BamHI digest of strain WVU398A; Lane C, PuuJI
digest of strain A. naeslundii genospecies 1 ATCC 12104; and Lane
D, PuuII digest of strain WVU398A.
clusters, there is still no really efficient and useful identi-
fication scheme for these bacteria (Johnson et al., 1990).
We have attempted to circumvent the use of phenotypic
tests or serology for specific characterization of strains of
A. naeslundii genospecies 1 and 2 and A. viscosus by
looking directly at their genomic DNA. The site-specific
cleavage of DNA by various endonucleases, DNA finger-
printing methods, and the availability of appropriate
probes which identify rRNA sequences has facilitated the
development of ribotyping methods, which are ideally
suited for strain characterization and identification. Prior
to this investigation, we expected that strains comprising
the individual clusters described by Fillery et al. (1978)
would be unique with regard to DNA fingerprint and
ribotype pattern. However, we had hoped that there
would be some commonality among strains within a clus-
Downloaded from jdr.sagepub.com at UNIV OF MICHIGAN on January 5, 2015 For personal use only. No other uses without permission.
crimination between strains of Listeria monocytogenes from epi-
demiological surveys. Appl En uiron Microbiol 57:2324-2331.
Barsotti 0, Morrier JJ, Freney J, Renaud F, Benay G, Decoret D, et
al. (1988). Achromopeptidase for rapid lysis of oral anaerobic
Gram-positive rods. Oral Microbiol Immunol 3:86-88.
Bowden G, Ekstrand J, McNaughton B, Challacombe SJ (1990). The
association of selected bacteria with the lesions of root-surface
caries. Oral Microbiol Immunol 5:346-351.
Bowden G, Hardie J, Fillery E (1976). Antigens from Actinomyces
species and their value in identification. J Dent Res 55(Spec Iss
A):192-204.
Bowden G, Hardie J, Slack G (1975). Microbial variations in
approximal dental plaque. Caries Res 9:253-277.
Brock DW, Georg LK (1969). Determination and analysis of Actino-
myces israelii serotypes by fluorescent antibody procedures. J
Bacteriol 97:581-588.
Caufield PW, Walker TM (1989). Genetic diversity within Strepto-
coccus mutans evident from chromosomal DNA restriction frag-
ment polymorphisms. J Clin Microbiol 27:274-278.
Chen C-KC, Sunday GJ, Zambon JJ, Wilson ME (1990). Restriction
endonuclease analysis of Eikenella corrodens. J Clin Microbiol
28:1265-1270.
Coykendall AL, Munzenmaier AJ (1979). Deoxyribonucleic acid
1178 BOWDEN et al.
hybridization among strains ofActinomyces viscosus andActino-
myces naeslundii. J System Bacteriol 29:234-240.
Cummins CS, Harris H (1958). Studies on the cell-wall composition
and taxonomy of Actinomycetales and related groups. J Gen
Microbiol 18:173-189.
De Buyser M, Morvan A, Grimont F, ElSolh N (1989). Characteriza-
tion of Staphylococcus species by ribosomal RNA gene restriction
patterns. J Gen Microbiol 135:989-999.
DiRienzo JM, Cornell S, Kazoroski L, Slots J (1990). Probe-specific
DNA fingerprinting applied to the epidemiology of localized
juvenile periodontitis. Oral Microbiol Immunol 5:49-56.
Dix K, Watanabe SM, McArdle S, Lee DI, Randolph C, Moncla B,
Schwartz DE (1990). Species-specific oligodeoxynucleotide probes
for the identification of periodontal bacteria. J Clin Microbiol
28:319-323.
Dzink JL, Martin LM, Socransky SS (1990). Typing ofFusobacterium
nucleatum strains byrestriction endonuclease analysis (abstract).
J Dent Res 69:354.
Ellen R (1976). Establishment and distribution ofActinomyces viscosus
and Actinomyces naeslundii in the human oral cavity. Infect
Immun 14:1119-1124.
Ellen RP, Banting DW, Fillery ED (1985). Streptococcus mutans and
Lactobacillus detection in the assessment of dental root-surface
caries risk. J Dent Res 64:1245-1249.
Fillery ED, Bowden GH, Hardie JM (1978). A comparison of strains
of bacteria designated Actinomyces viscosus and Actinomyces
naeslundii. Caries Res 12:299-312.
Firtel M, Fillery E (1988). Distribution of antigenic determinants
between Actinomyces viscosus and Actinomyces naeslundii. J
Dent Res 67:15-20.
Genco RJ, Loos BG (1991). The use of genomic DNA fingerprinting
in studies of the epidemiology of bacteria in periodontitis. J Clin
Periodontol 18:396-405.
Georg LK, Robertstad GW, Brinkman SA (1964). Identification of
species of Actinomyces. J Bacteriol 88:477-490.
Gerencser MA (1979). The application of fluorescent antibody tech-
niques to the identification of Actinomyces and Arachnia. In:
Bergan T, Norris JR, editors. Methods in microbiology. Vol. 13.
London: Academic Press, 187-221.
Gerencser MA, Slack JM (1976). Serological identification ofActino-
myces using fluorescent antibody techniques. JDent Res 55 (Spec
Iss):A184-A191.
Grimont F, Grimont PAD (1986). Ribosomal ribonucleic acid gene
restriction patterns as potential taxonomic tools. Ann Inst Pasteur
Microbiol 137B:165-175.
Han N, Hoover CI, Winkler JR, Ng CY, Armitage GC (1991).
Identification of genomic clonal types of Actinobacillus
actinomycetemcomitans by restriction endonuclease analysis. J
Clin Microbiol 29:1574-1578.
Holmberg K, Forsum U (1973). Identification of Actinomyces,
Arachnia, Bacterionema, Rothia and Propionibacterium species
by defined immunofluorescence. Appl Microbiol 25:834-843.
Johnson JL, Moore LVH, Kaneko BW, Moore EC (1990).Actinomyces
georgiae sp. nov., Actinomyces gerencseriae sp. nov., designation
of two genospecies ofActinomyces naeslundii, and inclusion ofA.
naeslundii serotypes II and III and Actinomyces viscosus sero-
type II in A. naeslundii genospecies 2. Int J System Bacteriol
40:273-286.
Jordan HV (1982). Pathogenicity of Actinomyces species. In: Genco
RJ, Mergenhagen SE, editors. Host-parasite interactions in peri-
odontal disease. Washington (DC): American Society for Microbi-
ology, 169-177.
Kalfas S, Edwardsson S (1990). Identification procedures for oral
Actinomyces species. Oral Microbiol Immunol 5:39-42.
Downloaded from jdr.sagepub.com at UNIV OF MICHIGAN on January 5, 2015 For personal use only. No other uses without permission.
J Dent Res August 1993
King S, Meyer E (1963). Gel diffusion technique in antigen-antibody
reactions of Actinomyces species and anaerobic diphtheroids. J
Bacteriol 85:186-190.
Kulkarni GV, Chan KH, Sandham HJ (1989). An investigation into
the use of restriction endonuclease analysis for the study of
transmission of mutans streptococci. J Dent Res 68:1155-1161.
Lambert FW, Brown JM, Georg LK (1967). Identification of Acti-
nomyces israelii and Actinomyces naeslundii by fluorescent
antibody and agar gel diffusion techniques. JBacteriol 94:1287-
1295.
Loesche WJ, Syed SA (1978). Bacteriology of human experimental
gingivitis: Effect of plaque and gingivitis score. Infect Immun
21:830-839.
Loos BG, Bernstein JM, Dryja DM, Murphy TF, Dickinson DP
(1989). Determination of the epidemiology and transmission of
nontypeable Haemophilus influenzae in children with otitis me-
dia by comparison oftotal genomic DNA restriction fingerprints.
Infect Immun 57:2751-2757.
Loos BG, Mayrand D, Genco RJ, Dickinson DP (1990). Genetic
heterogeneity of Porphyromonas (Bacteroides) gingivalis by
genomic DNA fingerprinting. J Dent Res 69:1488-1493.
Loos BG, Van Winkelhoff AJ, Dunford RG, Genco RJ, De Graaff J,
Dickinson DP, et al. (1992). A statistical approach to the ecology
of Porphyromonas gingivalis. J Dent Res 71:353-358.
McCormick SS, Mengoli HF, Gerencser MA (1985). Polyacrylamide
gel electrophoresis ofwhole-cell preparations ofActinomyces spp.
Int J System Bacteriol 35:429-433.
Moore WEC, Holdeman LV, Smibert RM, Cato EP, Burmeister JA,
Palcanis KG, Ranney RR (1984). Bacteriology of experimental
gingivitis in children. Infect Immun 46:1-6.
Moore WEC, Holdeman LV, Smibert RM, Good IJ, Burmeister JA,
Palcanis KG, Ranney RR (1982). Bacteriology of experimental
gingivitis in young adult humans. Infect Immun 38:651-667.
Nomura M, Post LE (1980). Organization of ribosomal genes and
regulation of their expression in Escherichia coli. In: Chambliss
G, Graven GR, Davis K, Cahan L, Nomura M, editors. Ribosomes:
structure function and genetics. Baltimore (MD): University
Park Press, 671-691.
Owen RJ (1989). Chromosomal DNA fingerprinting-a new method
of species and strain identification applicable to microbial patho-
gens. J Med Microbiol 30:89-99.
Owen RJ, Beck A, Dayal PA, Dawson C (1988). Detection of genomic
variation in Providencia stuartii clinical isolates of DNA restric-
tion fragment length polymorphisms containing rRNA cistrons.
J Clin Microbiol 26:2161-2166.
Pitcher DG, Saunders NA, Owen RJ (1989). Rapid extraction of
bacterial genomic DNA with guanidium thiocyanate. Lett Appl
Microbiol 8:151-156.
Putnins E, Bowden G (1991). Antigenic crossreactivity between
Actinomyces viscosus and Actinomyces naeslundii (abstract). J
Dent Res 70:317.
Sanyal B, Russell C (1978). Nonsporing, anaerobic, Gram-positive
rods in saliva and the gingival crevice of humans. Appl Environ
Microbiol 35:670-678.
Schaal KP (1986). Genus Actinomyces. In: Sneath PHA, editor.
Bergey's manual of systematic bacteriology. Vol. 2. Baltimore:
Williams and Wilkins, 1383-1418.
Schofield GM, Schaal KP (1980). Carbohydrate fermentation pat-
terns of facultative anaerobic Actinomycetes using micrometh-
ods. FEMS Microbiol Lett 8:67-69.
Slack JM, Gerencser MA (1970). Two new serological groups of
Actinomyces. J Bacteriol 103:266-267.
Slack JM, Gerencser MA (1975). Actinomyces, filamentous bacteria.
Biology and pathology. Minneapolis (MN): Burgess Publishing Co.
Vol. 72 No. 1' DNA FINGERPRINTING AND RIBOTYPING OF ACTINOMYCES
Slack JM, Landfried S, Gerencser MA (1969). Morphological, bio-
chemical and serological studies on 64 strains of Actinomyces
israelii. J Bacteriol 97:873-884.
Smith GLF, Sansone C, Socransky SS (1989a). Comparison of two
methods for the small-scale extraction of DNA from subgingival
microorganisms. Oral Microbiol Immunol 4:135-140.
Smith GLF, Socransky SS, Smith CM (1989b). Non-isotopic DNA
probes for the identification of subgingival microorganisms. Oral
Microbiol Immunol 4:41-46.
Socranskv S, Manganiello S (1971). The oral microbiota of man from
birth to senility. J Periodontol 42:485-496.
Socransky S, Manganiello A, Propas D, Oram V, Van Houte J (1977).
Bacteriological studies ofdeveloping supragingival dental plaque.
J Periodont Res 12:90-106.
Stackebrandt E, Charfreitag 0 (1990). Partial 16S rRNA primary
structure of five Actinomyces species: phylogenetic implications
and development of an Actinomyces israelii-specific oligonucle-
Downloaded from jdr.sagepub.com at UNIV OF MICHIGAN on January 5, 2015 For personal use only. No other uses without permission.
1179
otide probe. J Gen Microbiol 136:37-43.
Stull TL, LiPuma JJ, Edling TD ( 1988). A broad-spectrum probe for
molecular epidemiology of bacteria: ribosomal RNA. J Infect Dis
157:280-286.
Syed SA, Loesche WJ, Pape HL, Grenier E (1975). Predominant
cultivable flora isolated from human root-surface caries plaque.
Infect Immun 11:727-731.
Takahashi N, Schachtele CF (1990). Effect of pH on the growth and
proteolytic activity ofPorphyromonas gingitalis and Bacteroides
intermedius. J Dent Res 69:1266-1269.
Woese CR (1987). Bacterial evolution. Microbiol Rev 51:221-271.
Yeung MK (1992). Conservation of an Actinomyces tiscosus T14V
Type 1 fimbrial subunit homolog among divergent groups of
Actinomyces spp. Infect Immun 60:1047-1054.
Zambon JJ, Sunday GJ, Smutko JS (1990). Molecular genetic analy-
sis of Actinobacillus actinornycetemcoinitans epidemiologv. J
Periodontol 61:75-80.