JOURNAL OF CLINICAL MICROBIOLOGY, Apr. 1982, p. 606-609
Vol. 15, No. 4
Selective Medium for Isolation of Actinobacillus
Department ofOral Biology and Periodontal Disease Clinical Research Center, State University ofNew York
at Buffalo, Buffalo, New York 14226
Received 3 November 1981/Accepted 16 December 1981
A selective medium, TSBV (tryptic soy-serum-bacitracin-vancomycin) agar,
was developed for the isolation ofActinobacillus actinomycetemcomitans. TSBV
agar contained (per liter) 40 g of tryptic soy agar, 1 g of yeast extract, 100 ml of
horse serum, 75 mg of bacitracin, and 5 mg of vancomycin. The TSBV medium
suppressed most oral species and permitted significantly higher recovery of A.
actinomycetemcomitans than nonselective blood agar medium. The distinct
colonial morphology and positive catalase reaction ofA. actinomycetemcomitans
easily distinguished this bacterium from Haemophilus aphrophilus, Capnocyto-
phaga species, and a few other contaminating organisms. With the TSBV
medium, even modestly equipped laboratories will be able to isolate and identify
A. actinomycetemcomitans from clinical specimens.
Isolation and identification of Actinobacillus
actinomycetemcomitans has become increasingly
important because of this organism's suspected
role in certain types of human periodontal dis-
ease. Patients with localized juvenile periodonti-
tis frequently harbor high numbers ofA. actino-
mycetemcomitans in the subgingival plaque (6)
and demonstrate high levels of serum antibodies
against A. actinomycetemcomitans (2; P. Ham-
mond, J. Slots, and R. J. Genco, J. Dent. Res.
60A:498, 1981). Toxic substances from this orga-
nism including a potent endotoxin (4) and leuko-
cidin (1, 7, 8) may play a role in pathogenesis.
Development of a selective medium for the
specific recovery of A. actinomycetemcomitans
would be of particular value because the orga-
nism forms small, translucent colonies which
will be readily overlooked if other organisms
outnumber it significantly. Nevertheless, even if
A. actinomycetemcomitans is present only in
small proportions, it may be of etiological signif-
icance due to its virulence. Furthermore, the
growth of A. actinomycetemcomitans can be
inhibited in vitro by common oral streptococcal
species (J. D. Hillman and S. S. Socransky, J.
Dent. Res. 60A:603, 1981; Y. Yamamoto, P. A.
Mashimo, H. Reynolds, and R. J. Genco, Abstr.
Annu. Meet. Am. Soc. Microbiol. 1981, D41, p.
50). A bacterial medium which suppresses spe-
cies that are inhibitory to A. actinomycetemco-
mitans should facilitate the recovery of the
This paper describes and evaluates an im-
proved medium for selective culturing of A.
actinomycetemcomitans. This medium, which is
designated TSBV, contains tryptic soy agar
(Difco Laboratories, Detroit, Mich.), 10% se-
rum, and 0.1% yeast extract, and it includes the
selective agents bacitracin (75 ,ug/ml) and vanco-
mycin (5 p.g/ml). A. actinomycetemcomitans,
unlike most oral bacterial species,
resistant to bacitracin and vancomycin (J. Slots,
Arch. Microbiol., in press).
MATERIALS AND METHODS
Bacterial strains. The A. actinomycetemcomitans
strains used were: ATCC 29522, ATCC 29523, and
ATCC 29524, obtained from the American Type Cul-
ture Collection, Rockville, Md.; NCTC 9709 and
NCTC 9710, obtained from the National Collection of
Type Cultures, London, England; Y4, obtained from
S. S. Socransky, Boston, Mass.; and fresh clinical
isolates 1, 15, 67, 90, 107, and 125, from our labora-
tory. The strains were maintained by weekly subcul-
ture on tryptic soy agar supplemented with 5% sheep
blood and 0.1% yeast extract (BBL Microbiology
Systems, Cockeysville, Md.) (TB medium) in an atmo-
sphere of 90% air-10% CO2.
Medium. TSBV was prepared with tryptic soy agar
to which was added 1.0 g of yeast extract per liter. The
pH was adjusted to 7.2, and the medium was auto-
claved for 15 min at 121°C. The medium was cooled to
50°C, and horse serum and filter-sterilized bacitracin
(Sigma Chemical Co., St. Louis, Mo.) and vancomy-
cin (Sigma) were added to give final concentrations of
10%, 75 ,ug/ml, and 5 ,ug/ml, respectively. Preliminary
studies with pure cultures of A. actinomycetemcomi-
tans and dental plaque samples showed that these
concentrations of serum, bacitracin, and vancomycin
were optimal with respect to supporting growth of A.
actinomycetemcomitans and suppressing growth of
other oral species. The plates were stored aerobically
at 5°C and used within 7 days of preparation.
Recovery efficiency of pure cultures ofA. actinomyce-
temcomitans. Test strains of A. actinomycetemcomi-
MEDIUM FOR ISOLATING A. ACTINOMYCETEMCOMITANS
TABLE 1. Comparative growth on nonselective TB medium and on TSBV medium of 12 pure cultures of A.
Geometric mean growth +
SD in % of the TB
Median growth in %
of the TB medium
85% N2-10%o H2-5% CO2
85% N2-10%o H2-5% CO2
90% air-10% CO2
99 ± 14
96 ± 43
85 ± 23
tans were inoculated into prereduced, anaerobically
sterilized brain heart infusion broth (BBL). Overnight
cultures were dispersed by mixing with a Vortex mixer
at the maximal setting for 60 s and were serially 10-fold
diluted in anaerobic Ringer solution. Using a bent-
glass rod, 0.1-ml portions of 10-5 and 10-6 dilutions
were spread on TB and TSBV plates. TB plates were
incubated for 72 h at 37°C in a Coy anaerobic chamber
(Coy Manufacturing Co., Ann Arbor, Mich.) contain-
ing 85% Nz-109o H2-5% CO2. TSBV plates were
incubated for 72 h at 37°C in a Coy anaerobic chamber,
in a Torbal jar (The Torsion Balance Co., Clifton,
N.J.) with no catalyst and containing 90o air-10%
C02, and in a candle jar.
Clinical specimens. Subjects were adolescents who
were admitted to our dental clinic for treatment of
rapidly advancing periodontal disease on molars and
incisors (localized juvenile periodontitis). Samples
from deep periodontal pockets were obtained by using
paper points as previously described (6). These were
transferred to 9 ml of anaerobic Ringer solution, the
bacteria were dispersed by mixing with a Vortex mixer
at the maximal setting for 60 s, and the bacterial
suspension was serially diluted in 10-fold steps in
anaerobic Ringer solution. Using a bent-glass rod, 0.1-
ml portions of appropriate dilutions were plated on TB
and TSBV. After incubation for 72 h at 37°C in a Coy
anaerobic chamber, the plates were examined for A.
actinomycetemcomitans. Randomly selected isolates
suspected of being A. actinomycetemcomitans were
subcultured and confirmed as A. actinomycetemcomi-
tans if they were gram-negative, capnophilic (exhibit-
ed scant growth in air but grew well in 10o C02),
fermentative, catalase-positive coccobacilli which did
not require X (hemin) and V (NAD+) factors for
Table 1 shows the recovery efficacy of 12
strains of A. actinomycetemcomitans on TB
after anaerobic chamber incubation and on
TSBV after anaerobic chamber, Torbal jar with
90% air-10% CO2, and candle jar incubation.
For each strain tested, the viable count on TB in
the anaerobic chamber (nonselective culturing)
was designated 100%, and the viable counts on
the TSBV plates were compared with this. Un-
der anaerobic incubation, growth on TSBV was
similar to that on TB. Some inhibition tended to
occur when incubation took place in 90% air-
10% CO2 or in a candle jar; however, the
observed differences in recovery rates were not
statistically significant at the 5% significance
level as analyzed by the Wilcoxon signed-ranks
test for matched pairs.
Fourteen specimens ofadvanced juvenile per-
iodontitis lesions were obtained. Table 2 shows
the number of A. actinomycetemcomitans re-
TABLE 2. Recovery of A. actinomycetemcomitans from dental plaque on TSBV medium and on TB medium
after 72 h of anaerobic incubation
A. actinomycetemocomitans recovery
528 x 102
294 x 103
12772 x 102
14 14 x 102
% Increase in
88 x 102
27 x 102
42 x 103
106 x 102
206 x 103
19 X 103
32 x 102
140 x 103
152 x 102
aData were obtained by subtracting the value for TB medium from the value for TSBV medium, dividingthe
resulting number by the value for TB medium, and thenmultiplying by 100.
VOL. 15, 1982
J. CLIN. MICROBIOL.
FIG. 1. Characteristic colony of A. actinomyce-
temcomitans after growth on TSBV for 72 h. Note
starlike inner structure. Photography performed with
light transmitted through the medium.
covered on TB and TSBV media. Significantly,
eight samples yielded
counts of A. actinomycetemcomitans on TSBV
than on TB. Also, in three samples in which A.
actinomycetemcomitans occurred in low num-
bers, the organism could be recovered only on
the TSBV medium.
Colonies of A. actinomycetemcomitans
cent, glistening, and 0.5 to 1.0 mm in diameter
with slightly irregular edges after incubation for
3 days. On primary
strongly adhered to the agar surface and com-
monly exhibited a starlike inner structure (Fig.
1). Continued subculture resulted in loss of the
adherence and the starlike structure.
Organisms other than A. actinomycetemcomi-
tans which grew on TSBV in 90% air-10% CO2
or candle jar included mainly Haemophilus aph-
rophilus and Capnocytophaga species and occa-
sionally Neisseria species, Staphylococcus-Mi-
crococcus species, and yeasts. When incubated
anaerobically, TSBV also supported growth of
strains offusobacteria and gram-negative anaer-
obic motile rods. The ability of A. actinomyce-
temcomitans to vigorously decompose 3% H202
applied on the primary isolation plate could
distinguish this organism from the morphologi-
cally similar catalase-negative or weakly cata-
aphrophilus. The remaining
readily be distinguished
from A. actinomycetemcomitans on the basis of
major differences in colonial morphology.
to 10-times higher
as circular, convex, translu-
Most existing data on A. actinomycetemcomi-
tans in clinical infections have been obtained by
using nonselective culture techniques. Since A.
actinomycetemcomitans produces a small colo-
ny which readily can be overlooked and since
various streptococcal species can inhibit the
organism's growth in vitro, it seems appropriate
to use a special selective procedure for the
isolation of A. actinomycetemcomitans.
Kilian and Schi0tt (5) recovered A. actinomy-
cetemcomitans from dental plaque by using
chocolate agar supplemented with 300 pLg of
bacitracin per ml, a medium developed by Hovig
and Aandahl (3) for the selective recovery of
haemophili. The isolation and identification of
A. actinomycetemcomitans on this medium can
be hampered by the growth of streptococci,
several Haemophilus species, Eikenella corro-
dens, and neisseriae (5).
Mandell and Socransky (R. L. Mandell and
S. S. Socransky, J. Dent. Res. 59A:512, 1980)
isolated A. actinomnycetemcomitans on Trypti-
case soy agar (BBL Microbiology Systems) sup-
plemented with 128 ptg of bacitracin per ml, 8 p.g
of malachite green per ml, and 5% sheep blood.
This medium suppressed the growth of pure
cultures of A. actinomycetemcomitans as much
as 20%. The malachite green, which at relatively
low concentrations can be inhibitory to A. acti-
nomycetemcomitans (Slots, in press), may be
responsible for this suppression. Mandell and
Socransky's medium also may grow several
contaminating Haemophilus species because of
the blood supplement.
The TSBV medium overcomes several of
these problems. The enrichment with 10% horse
serum instead of with blood did not change the
recovery rate ofA. actinomycetemcomitans, but
it suppressed hemin-requiring Haemophilus
strains and allowed direct application of H202 to
the primary isolation plate to verify the A.
was added because many oral species are partic-
ularly susceptible to this antibiotic (Slots, un-
published data). The vancomycin content of 5
pg/ml suppressed the growth of potentially in-
hibitory streptococcal strains and other gram-
positive species. Our finding of lower recovery
rates ofA. actinomycetemcomitans from clinical
specimens on nonselective blood agar medium
than on TSBV suggests that in vitro bacterial
antagonism against A. actinomycetemcomitans
can be of a significant magnitude. Additional
suppression of contaminants could be achieved
by incubating in 90% air-10% CO2 or in a candle
jar instead of anaerobically. The observation
that the candle jar system provided good growth
for A. actinomycetemcomitans may be of value
in a clinical setting or in epidemiological studies
where microbiological facilities are limited.
In conclusion, the TSBV medium described in
this report can considerably aid the microbiolog-
ical monitoring ofA. actinomycetemcomitans; it
is relevant in the diagnosis and treatment of A.
MEDIUM FOR ISOLATING A. ACTINOMYCETEMCOMITANS
This investigation was supported in part by Public Health
Grant DE04898 from the National Institute of Dental Re-
search. The assistance of H. S. Reynolds is highly appreciat-
1. Baehni, P., C.-C. Tsai, W. P. McArthur, B. F. Hammond,
and N. S. Taichman. 1979. Interaction of inflammatory
cells and oral microorganisms. VIII. Detection of leuko-
toxic activity of a plaque-derived gram-negative microor-
ganism. Infect. Immun. 24:233-243.
2. Genco, R. J., J. Slots, C. Mouton, and P. Murray. 1980.
Systemic immune responses to oral anaerobic organisms,
p. 277-293. In D. W. Lambe, Jr., R. J. Genco, and K. J.
topics. Plenum Press, New York and London.
3. Hovig, B., and E. H. Aandahl. 1969. A selective method
for the isolation of Haemophilus in material from the
respiratory tract. Acta Pathol. Microbiol. Scand. 77:676-
4. Kiley, P., and S. C. Holt. 1980. Characterization of the
lipopolysaccharide from Actinobacillus actinomycetem-
comitans Y4 and N27. Infect. Immun. 30:862-873.
5. Kilian, M., and C. R. Schisltt. 1975. Haemophili and
related bacteria in the human oral cavity. Arch. Oral Biol.
6. Slots, J., H. S. Reynolds, and R. J. Genco. 1980. Actinoba-
cillus actinomycetemcomitans in human periodontal dis-
ease: a cross-sectional microbiological investigation. In-
fect. Immun. 29:1013-1020.
7. Taichman, N. S., R. T. Dean, and C. J. Sanderson. 1980.
Biochemical and morphological characterization of the
killing of human monocytes by a leukotoxin derived from
Actinobacillus actinomycetemcomitans. Infect. Immun.
8. Tsai, C.-C., W. P. McArthur, P. C. Baehni, B. F. Ham-
mond, and N. S. Taichman. 1979. Extraction and partial
characterization of a leukotoxin from a plaque-derived
gram-negative microorganism. Infect. Immun. 25:427-
VOL. 15, 1982