Journal of Applied Bacteriology 1991,71, 343-346 ADONIS 0021884791001310

A selective medium for Fusobacterium spp.

J.S. Brazier, Diane M. Citron' and E.J.C. Goldstein'
Anaerobe Reference Unit, Public Health Laboratory, Luton, UK and 'R.M.Alden Research Laboratory, Santa
Monica Hospital and Medical Center, CA 90404, USA
3557/12/90:accepted 28 March 1991

J.S. BRAZIER, D . M . CITRON AND E.J.C. GOLDSTEIN. 1991. A new selective medium (JVN) for
the isolation of Fusobacterium spp. from clinical material is described. T h e medium
incorporates josamycin, vancomycin and norfloxacin (at 3 , 4 a n d 1 pg/ml, respectively) as the
selective agents, plus 5% defibrinated horse blood in Fastidious Anaerobe Agar Base ( L a b
M). T h i s formula allowed luxuriant growth of all 82 strains (eight recognized species) of
fusobacteria tested, while significantly inhibiting 5 1/5 1 (100%) strains of facultative anaerobes
and 45/51 (88%) strains of other obligate anaerobes. JVN medium allowed the successful
isolation of strains of Fusobacteriurn naviforme, F . nucleatum a n d F . necrophorum from t h e
gingivae of 9/16 healthy volunteers, a n d strains o f F . varium a n d F . mortferum from faecal
suspensions seeded with these organisms.

INTRODUCTION T h e present study sought to develop a selective medium

Members of the genus Fusobacterium form part of the
normal anaerobic oral and gastrointestinal flora of man. In
common with many other components of these normal
floras, fusobacteria are opportunistic pathogens and have
been reported in various pneumonias, soft-tissue abscesses,
intra-abdominal infections (Finegold 1989) and in cases of
necrobacillosis (Moore-Gillon et al. 1984). The isolation of
fusobacteria from clinical specimens requires optimal con-
ditions of anaerobiosis and a basal medium capable of sup-
plying the nutritional requirements of these fastidious
anaerobes. Fastidious Anaerobe Agar Base (FAA; Lab M)
has been described as supporting luxuriant growth of fuso-
bacteria (Brazier 1986). The addition of selective agents to
an agar medium is a recognized method for improving the
isolation of specific organisms from mixed culture. Agents
such as neomycin sulphate may be used for the selection of
clinically significant obligate anaerobes from mixed floras
because aminoglycosides have little activity against obligate
anaerobes. Such agents, however, while successfully inhi-
biting many facultative anaerobes, allow the growth of
most, but not all, other obligately anaerobic genera. T h e
medium described by Morgenstein et al. (1981) for the iso-
lation of fusobacteria incorporated neomycin, vancomycin
and josamycin at 100, 5 and 3 pg/ml respectively, a com-
bination which did not satisfactorily suppress all facultative
Gram-negative bacilli.
Correspondence to :Dr J.S. Brazrer, Anaerobe Refirenre Unit, Public Health
Loboratory, Lewsey Road, Luton L U 4 ODZ, U K .
for fusobacteria based upon antimicrobial susceptibility
data of recent clinical strains of fusobacteria. I t was hoped
that this medium would inhibit not only facultative anaer-
obes but also other obligate anaerobes, and thus facilitate
easy isolation of these important pathogens from mixed
flora.
MATERIALS AND METHODS
Susceptibility studies
Susceptibility studies against erythromycin (Eli Lilley),
josamycin (Yamanouchi Pharmaceuticals), norfloxacin
(Merck Sharp and Dohme) and vancomycin (Eli Lilley)
were performed on 82 strains of fusobacteria from the com-
bined culture collections of the R.M. Alden Research
Laboratory (Santa Monica, USA) and the Public Health
Laboratory Service Anaerobe Reference Unit (Luton, UK).
All strains were recently isolated from clinical or veterinary
sources and had been identified by the criteria of Hold-
eman et al. (1977). The species represented and the number
of strains included were : Fusobacterium gonidiaformans (3),
F. mortiferum (9), F. naviforme (9), F. necrophorum (12), F.
nucleatum (17), F. russii (3), F . ulcerans (9), F. varium (11)
and nine unidentified species. Susceptibility testing was
performed by a multi-point agar dilution technique (Brazier
et al. 1990). Each agent was tested at doubling dilutions
ranging from either 32 or 16 to 0.03 ,ug/rnl. Strains were
prepared for inoculation by picking several colonies from a
344 J . S . B R A Z I E R ET A L .
blood agar plate and making suspensions in Brucella Broth
(Difco) made to a visual turbidity of McFarland 0.5 stan-
dard (1 x 10' cfu/ml).
inhibition and growth of facultative and obligate
anaerobes
From the results obtained from the susceptibility studies, a
selective medium incorporating 3 pg/ml josamycin, 4 ,ug/ml
vancomycin and 1 ,ug/ml norfloxacin in FAA plus 5% de-
fibrinated horse blood was formulated UVN agar).
T o test the inhibition of facultative anaerobes on JVN
agar a range of stock strains of facultative organisms was
selected. These were : Escherichia coli (3), Enterococcus fae-
calis (S), Enterobacter aerogenes (3), Ent. cloacae (3), Kleb-
siella pneumoniae (3), methicillin-resistant Staphylococcus
aureus (MRSA) (3), Staph. epidermidis (4), Staph. aureus
( S ) , Morganella morganii (4), Proteus mirabilis (3), P . stuartii
(3), Pseudomonas aeruginosa (3), Streptococcus mitis (l),
viridans streptococci (2) and Strep. pyogenes (1). Each
organism was emulsified in Brucella Broth (Difco) to a tur-
bidity of 0.5 McFarland standard and 0.01 ml of each bac-
terial suspension was plated on JVN agar to obtain single
colonies. Control plates of unselective FAA blood agar were
similarly inoculated, and all plates were incubated in an
anaerobic chamber at 37°C. Plates were examined after
overnight incubation, and again after 48 h.
T o determine the ability of JVN agar to inhibit the
growth of obligate anaerobes a range of stock strains of
Clostridium spp. ( S ) , Bacteroides spp. (29 including 18 of
the Bacteroides fragilis group), Peptostreptococcus spp. (8),
Veillonella spp. (4) and miscellaneous anaerobic non-spore-
forming Gram-positive bacilli (5) were inoculated on the
selective medium. Inocula were prepared as described for
the fusobacteria and incubated similarly. After incubation
for 48 h the plates were examined for growth of each strain.
Effect of JVN and blood supplement on fusobacteria
T o test the ability of the selective medium to grow pure
cultures of Fusobacterium spp. all 82 strains of fusobacteria
used in the susceptibility studies were emulsified in Bru-
cella broth to 0.5 McFarland standard, inoculated on JVN
blood agar, and incubated anaerobically at 37°C for 48 h
before examination. In a smaller study designed to measure
any restriction of fusobacterial colonies on JVN compared
with that on FAA blood agar without the selective agents
both media were inoculated with 27 strains of eight species
of fusobacteria. In addition, 5% sheep blood was substi-
tuted for horse blood in FAA base to determine if the type
of blood used affected colony size. All plates were incu-
bated in an anaerobic chamber for 48 h. Colony sizes were
measured with a plate microscope fitted with a micrometer
eyepiece and graticule calibrated in millimetres. The mean
colony size for each strain on each medium was determined
by measuring the diameter of five representative discrete
colonies from the streaked-out inoculum.
Recovery of Fusobacterlum spp. from clinical
samples
T o test the efficacy of JVN agar in the recovery of fusobac-
teria from human gingiva with concurrent inhibition of
other components of the normal oral flora, gingival material
was obtained from 16 healthy volunteers. T h e material was
collected by a fine gauged cotton-tipped wire ear-nose-
throat swab applied to volunteers' gingival crevices. The
swabs were plated directly on JVN agar and unselective
FAA blood agar to obtain single colonies. T h e efficacy of
JVN agar in the recovery of gastrointestinal species of fuso-
bacteria and its ability to inhibit other components of the
normal faecal flora was tested by making suspensions of
faecal material from four volunteers into 5 ml volumes of
Brucella broth. T h e faecal suspensions were seeded with
suspensions of F . varium and F . mortijerum made to 0.5
McFarland standard (1 x 10' cfu/ml) to give a final con-
centration of 2 x lo5 cfu/ml. These seeded faecal suspen-
sions were plated on JVN agar and unselective FAA blood
agar to obtain single colonies. All plates were incubated
anaerobically for 48 h and the resulting growth recorded.
RESULTS
The minimum inhibitory concentration (MIC) results of
the various species of fusobacteria for the four antimicrobial
agents tested are shown in Table 1.
Of the 51 strains of facultative organisms tested, 47
showed no growth on JVN agar. The remainder (one strain
of Serratia marcesans, two of Ent. cloacae and one of
MRSA) were significantly inhibited, yielding less than
seven colonies from an inoculum of 0.01 ml of a McFarland
0.5 suspension.
Table 2 lists the obligate anaerobes inhibited on JVN
agar. The only non-fusobacteria to grow significantly were
two strains each of Bacteroides fragilis, B . distasonis and B .
ovatus. The colonies of five other fragilis group strains were
reduced to pin-point colonies of less than 0.5 mm diameter,
and the remaining four were completely inhibited.
I n the studies on gingival material, JVN was successful
in isolating fusobacteria from 9/16 specimens. Strains were
identified according to the criteria of Holdeman et al.
(1977). Six were identified as F. naviforme, two as F . nude-
atum and one as F. necrophorum. The most common accom-
panying strain on the selective medium was Leptotrichia
buccalis, which was present in 12/16 of the gingival samples.
 SELECTIVE M E D I U M FOR FUSOBACTERIA 345

Table 1 Susceptibility results of 82 strains of Fusobacterium against josamycin, vancomycin, norfloxacin and erythromycin

Organism Strains Agent Range MIC 50% MIC 90%

Fusobacterium ulcerans 9 Erythromycin 32- > 32 > 32 > 32

Josamycin 8->32 > 32 > 32
Vancomycin > 16 > 16 > 16
Norfloxacin > 32 > 32 > 32
mortrferum 9 Erythromycin > 32 > 32 > 32
Josamycin > 32 > 32 > 32
Vancomycin > 16 > 16 > 16
Norfloxacin 8-32 16 32
varium 11 Erythromycin > 32 > 32 > 32
Josamycin > 32 > 32 > 32
Vancomycin > 16 > 16 > 16
Norfloxacin 16>32 32 > 32
nucleatum 17 Erythromycin 1->32 16 > 32
Josamycin 4->32 16 > 32
Vancomycin > 16 > 16 > 16
Norfloxacin 4-16 8 16
necrophorum 12 Erythromycin 1-32 2 8
Josamycin 2-32 16 32
Vancom ycin &> 16 > 16 > 16
Norfloxacin 2-16 8 8
navaJorme Erythromycin 4->32 32 32
Josamycin 8->32 8 > 32
Vancomycin > 16 > 16 > 16
Norfloxacin 1632 32 32
russii Erythromycin > 32 > 32 > 32
Josamycin > 32 > 32 > 32
Vancom ycin > 16 > 16 > 16
Norfloxacin 32 32 32
gonidiaformans Erythromycin > 32 > 32 > 32
Josamycin 8-32 16 32
Vancom ycin > 16 > 16 > 16
Norfloxacin 32 32 32
Fusobacterium spp. Erythromycin 1-32 2 > 16
Josam ycin 1-32 4 32
Vancomycin 1 6 > 16 > 16 > 16
Norfloxacin 4-32 32 32

MIC, Minimum inhibitory concentration in pg/ml.

Table 2 Obligate anaerobes inhibited on

JVN agar Gram-positive (n) Gram-negative

Peptostreptococcus spp. (8) Bacteroides fragilis (2)

Clostridium perfringens (1) distasonis (1)
bij2rmentans (1) thetaiotaomicron (2)
septicum (1) vulgatus (2)
Clostridium spp. (1) stercoris (1)
Propionibacterium acnes (3) caccae (1)
Miscellaneous non- bivius (2)
spore-forming bacilli (5) gracilis (1)
melaninogenicus group (7)
Bacteroides spp. (2)
Veillonella spp. (4)
n, Number of strains.
346 J . S . B R A Z I E R ET A L .
The unselective FAA blood agar inoculated simultaneously
with JVN also enabled fusobacteria to be isolated from 9/16
volunteers, but recognition was difficult because of the
heavy growth of other anaerobic and facultative com-
mensals. The seeded faecal suspensions on JVN agar
yielded a heavy ( +++ +) growth of F. m o r t f e r u m and F.
varium with a light ( + ) growth of gut commensals, com-
pared with a heavy ( + ++
+) mixed growth of faecal flora
and fusobacteria on the non-selective medium.
In terms of the mean colony sizes of 27 representative
strains of Fusobacterium spp. there was no significant differ-
ence between media supplemented with sheep or horse
blood. Colonies on JVN selective medium were only mar-
ginally smaller (mean 0.8 mm) than on non-selective FAA
blood agar.
DISCUSSION
The susceptibility results showed that of the four anti-
microbial agents tested, erythromycin was of the least value
as a selective agent since the M I C range was lower against
4/8 species of fusobacteria than for any of the other agents.
In particular, F. necrophorum strains had an MIC5O to
erythromycin of between 1 and 2 pg/ml. At this low
threshold, erythromycin would probably be of limited
selective value, while at higher concentrations it would
jeopardize the isolation of 50% of F. necrophorum strains.
The closely related macrolide, josamycin, however, had less
activity against F. necrophorum with an MICSo of between
8 and 16 pg/ml. Only one strain of F. necrophorum was
inhibited at 2 pg/ml; the remaining 11 strains had an M I C
of > 4 pg/ml. For vancomycin, apart from one strain of F.
necrophorum with an M I C of 8 pg/ml, all strains of fusobac-
teria had an M I C of > 16 pg/ml.
Norfloxacin was chosen for inclusion in the selective
medium in the light of the review by Goldstein (1987), who
showed that most facultative Gram-negative bacilli had
MICs of < 1 pg/ml, and it had very little activity on anaer-
obes. The present study confirmed its promise as a selec-
tive agent for fusobacteria at this concentration, since the
MIC9' was either 32 or >32 &ml for 7/8 recognized
species. Only one species, F. necrophorum, had lower M I C
values (range 2-16 pg/ml). The incorporation of either
horse or sheep blood in FAA base had no discernible effect
on the growth of fusobacteria. This has implications in the
U K and the USA where preferences for blood supplemen-
tation differ.
T h e performance of JVN in the inhibition of facultative
organisms was impressive. Significant or complete inhibi-
tion was seen with 100% of the stock strains of facultative
organisms tested. Inhibition of obligate anaerobes was also
significant, with 88% of strains failing to grow on JVN.
These figures were reinforced by the results of the clinical
samples which showed significant inhibition of the majority
of anaerobic and facultative commensals, while allowing
growth of fusobacteria. Thus, JVN agar fulfils its purpose.
T h e results of the above evaluation indicate that JVN
agar supplemented with either horse or sheep blood will be
of value in the isolation of Fusobacterium spp. from clincial
material.
REFERENCES
B R A Z I E RJ, . S . (1986) Yellow fluorescence of fusobacteria.
Letters in Applied Microbiology 2, 125-126.
B R A Z I E RJ,. S . , G O L D S T E I NE, . J . C . , C I T R O N D , .M. &
OSTOVARM I , . I . (1990) Fastidious anaerobe agar compared
with Wilkins-Chalgen agar, Brain Heart Infusion agar and
Brucella agar for susceptibility testing of Fusobacterium species.
Antimicrobial Agents and Chemotherapy 34, 228s2282.
FINEGOLD, S . M . (1989) General aspects of anaerobic infection.
In Anaerobic Infections in Humans. ed. Finegold, S.M. &
George, W.L. pp. 137-153. San Diego: Academic Press.
G O L D S T E I NE, . J . C . (1987) Norfloxacin, a fluoroquinolone
antibacterial agent. Classification, mechanism of action, and in
vitro activity. American Journal of Medicine 82 (suppl. 6B),
3-17.
HOLDEMAN L ,. V . , C A T O ,E . P . & M O O R EW , . E . C . (1977)
Anaerobe Laboratory Manual, 5th edn. Blacksberg, Virginia :
Southern Printing.
M O O R E - G I L L O JN.,, L E E , T . H . , E Y K E N S , .J. & PHIL-
L I P S , I . (1984) Necrobacillosis : a forgotten disease. British
Medical Journal 288, 15261527.
M O R G E N S T E I NA,. A . , C I T R O N ,D . M . & F I N E G O L D ,
S . M . (1981) New medium selective for Fusobacterium species
and differential for Fusobacterium necrophorum. Journal of Clini-
cal Microbiology 13, 4. 666669.