Content uploaded by Mogens Kilian
Author content
All content in this area was uploaded by Mogens Kilian on Mar 09, 2014
Content may be subject to copyright.
INTERNATIONAL
JOURNAL
OF SYSTEMATIC BACTERIOLOGY,
OCt.
1989,
p.
471484
oO20-7713/89/O40471- 14$02.00/0
Copyright
0
1989,
International Union of Microbiological Societies
Vol.
39,
No.
4
Taxonomic Study
of
Viridans Streptococci: Description
of
Streptococcus gordonii
sp. nov. and Emended Descriptions
of
Streptococcus sanguis
(White and Niven 1946),
Streptococcus oralis
(Bridge and Sneath 1982), and
Streptococcus mitis
(Andrewes and Horder 1906)
MOGENS KILIAN,l* LENA MIKKELSEN,2
AND
JPRGEN HENRICHSEN3
Departments
of
Oral Biology’ and Oral Medicine and Diagnosis,2 Royal Dental College, DK-8000 Aarhus
C,
Denmark,
and Streptococcus Department, Statens Seruminstitut, DK
2300
Copenhagen
S,
Denmark3
We examined a collection of
151
strains of the viridans type of streptococci, which were isolated mainly from
human oral cavities and included several reference strains, in an attempt to revise and improve the taxonomy
of this group of bacteria. Our examinations included determinations of a high number of biochemical and
physiological characteristics and serological reactivity. The resulting data revealed several hitherto unrecog-
nized characters of taxonomic significance, and several of the species can now be more accurately defined.
A
diagnostic key to the taxa is presented. Strains previously identified as
Streptococcus
sanguis
could be divided
into two clearly distinct species,
Streptococcus
sanguis
sensu strict0 (type strain, ATCC 10556) and a new
species,
Streptococcus
gordonii
(type strain, ATCC
10558).
Streptococcus
rnitis
was divided into two biovars,
consisting of strains possessing group
0
antigens and strains possessing group
K
antigen. The group of strains
assigned to
Streptococcus
anginosus
was biochemically and serologically heterogeneous, but the data did not
allow natural subdivisions. Based on the results of this study, emended descriptions of the species
Streptococcus
oralis,
S.
rnitis,
and
S.
sanguk
are provided. The classification resulting from this study is in complete
agreement with previously published genetic data.
The viridans type of streptococci encountered in oral
cavities and pharynges have been curiously refractory to
satisfactory classification. Although several comprehensive
taxonomic studies have been performed
(3,
5,
8, 11, 22, 25,
41, 42, 49),
international consensus on classification and
nomenclature has not been obtained. As
a
consequence,
several synonyms have been applied to the same organisms
(18, 24, 29).
Furthermore, several of the species are geneti-
cally heterogeneous
(12, 13).
However, lack of distinguish-
ing phenotypic traits has left these problems unsolved. There
is no doubt that this situation has hampered
a
clear under-
standing of the ecology of these bacteria and of some of the
molecular mechanisms involved in plaque formation on
teeth.
The Approved Lists of Bacterial Names published in
1980
(43)
included, in addition to some of the “mutans” species,
the following six species of streptococci normally encoun-
tered in oral cavities:
Streptococcus sanguis, Streptococcus
mitis, Streptococcus salivarius, Streptococcus anginosus,
Streptococcus constellatus,
and
Streptococcus intermedius.
Subsequently, Coykendall et
al.
(14)
demonstrated that the
last two names, together with the name
“Streptococcus
rnilleri,”
which is used mainly in European laboratories, are
later synonyms of
S.
anginosus.
Other workers have found
that
S.
constellatus
is sufficiently distinct from
S.
anginosus
to warrant specific recognition
(26).
The name
bbStreptococcus mitior”
has been used mainly
by workers in European laboratories for streptococci that
lack the ability to hydrolyze arginine and esculin and may or
may not produce extracellular polysaccharide. Although an
overwhelming amount of data, from studies of nucleic acid
homology and base composition
(13),
cell wall carbohy-
*
Corresponding author.
drates (lo), biochemical and physiological characteristics
(5,
11, 17),
and serology
(11, 23, 40, 46),
has established that
polysaccharide-producing
strains deserve specific recogni-
tion, these organisms have been regularly referred to, in
some parts of the world, as
“S.
sanguis
11”; other synonyms
are
“S.
sanguis
I:A,”
“S.
sanguis
biotype B,” and
“S.
sanguis
serotype
11.”
In addition, Colman and Williams
(11)
included strains otherwise referred to as
S.
mitis
in
“S.
mitior”
because they considered
S.
mitis
a
later synonym of
“S.
mitior.”
Although extensively used, the name
“S.
mitior”
was not included on the Approved Lists, and no
formal attempt has been made to revive it. Thus, it has no
formal standing. In the meantime, a new species,
Strepto-
coccus oralis,
has been described
(2).
Although this species
appears to be heterogeneous, the type strain is
a
typical
strain of
“S.
mitior,”
and the new species thus covers this
taxon
(29, 30).
The species
S.
mitis
formally exists, yet it is
not clearly defined. The type strain, which was chosen for
the purpose of representing the species on the Approved
Lists, does not fit the authentic description of the species
(1)
and is reminiscent of
S.
sanguis
(29, 47).
Recently, Whiley
and Hardie described a new species,
Streptococcus vestib-
ularis,
which resembles
S.
salivarius
phenotypically but
differs from it as determined by deoxyribonucleic acid
(DNA)-DNA homology and cell wall analyses
(48).
Several extensive serological studies of oral streptococci
have been performed
(7, 11, 22, 23, 39,
40,
46).
However,
partly as a result of the above-mentioned situation, a clear
correlation with the existing classification has not been
achieved. This problem is further amplified by a lack of
international consensus concerning reference strains for
serology, The special problems concerning serological group
H have been extensively discussed and elucidated by Cole et
al.
(7).
471
472
KILIAN
ET
AL.
INT.
J.
SYST.
BACTERIOL.
The taxonomic study described in this paper encompassed
151 viridans streptococci, including reference strains and
freshly isolated strains from oral cavities, from pharynges,
and from cases of subacute bacterial endocarditis. The
strains were subjected to extensive analysis by using both
serology and biochemical tests, many of which were new to
this group of bacteria. Most of the taxa revealed by our
results were clearly defined. On the basis of these results, a
new species is described, and emended descriptions of the
species
S.
sanguis,
S.
oralis,
and
S.
mitis
are presented.
MATERIALS
AND
METHODS
Streptococcal strains.
The designations, sources, taxo-
nomic status, and other pertinent information on the 151
streptococcal strains used in this study are shown in Table
1.
All of the strains were gram-positive, catalase-negative cocci
with a tendency to form chains in broth cultures, and they
lacked soluble hemoly sin. Upon isolation or receipt they
were subcultured on Todd-Hewitt agar (Difco Laboratories,
Detroit, Mich.) plates
5
to 10 times to ensure purity and were
subsequently lyophilized. A code number (beginning with
the prefix
SK)
was assigned to each strain of the collection.
Table
1
lists the strains under the
taxa
to which they were
assigned as a result of this study.
Standard cultural conditions.
Stock agar plate cultures and
cultures for tests performed on agar plates were incubated at
36°C in a
5%
co2-95% N2 atmosphere obtained in anaerobic
jars. All liquid cultures were incubated at 36°C in air. The
standard inoculum used for liquid cultures was
1
drop of a
24-h Todd-Hewitt broth culture.
Colonial and cell morphology.
Strains grown on mitis
salivarius agar (Difco) were examined for colonial features
after incubation for 24 h at 36°C in C0,-N,, followed by
incubation for 24 h at room temperature in air. Overnight
cultures in Todd-Hewitt broth were examined by phase-
contrast microscopy, and the approximate lengths of chains
were recorded.
Action on blood.
Hemolysis was demonstrated on
5%
horse blood agar, and greening was recorded on chocolate
agar prepared with
5%
horse blood.
Susceptibility tests.
Susceptibility to antibiotics was exam-
ined on blood agar plates (Difco) supplemented with
5%
horse blood. The plates were inoculated to give a dense
growth of discrete colonies. Bacitracin disks were prepared
to contain
0.05
U
per disk. In addition, sulfafurazole and
nitrofurazon disks (Neo-Sensitabs; ROSCO, Copenhagen,
Denmark) were used. For bacitracin a zone of inhibition
indicated susceptibility, as described by Facklam (17). The
inhibition zones around sulfafurazole and nitrofurazon disks
were measured; zone diameters of more than 22 and 14 mm,
respectively, were considered indicative of susceptibility
according to the instructions of the manufacturer of the
disks.
The
ability to grow in broth containing 4 or 6.5% NaCl(4)
was recorded after incubation €or 7 days.
Biochemical tests.
Production of acids from carbohydrates
was determined in phenol red broth base (Difco) containing
0.05%
Tween
80
(16) and carbohydrates at concentrations of
1%. Carbohydrates were added to the autoclaved medium
after cooling as filter-sterilized concentrated solutions. The
final pH was recorded with
a
pH meter (Radiometer, Copen-
hagen, Denmark) after incubation for 7 days. The following
carbohydrates were used: glycerol, ribose, galactose,
D-
(+)-glucose,
D-(
-)-fructose, D-(-)-mannose,
L-(
-)-sorbose,
rhamnose, dulcitol, mannitol, sorbitol, N-acetylglucosamine,
amygdalin, arbutin, esculin, salicin,
D-(
+)-cellobiose, malt-
ose, lactose,
D-(
+)-melibiose, sucrose,
D-(
+)-trehalose, in-
ulin,
D-(
+)-melizitose,
D-(
+)-raffinose, and soluble starch.
The carbohydrates were analytical grade and were pur-
chased from BDH, Poole, United Kingdom,
E.
Merck
AG,
Darmstadt, Federal Republic of Germany, Sigma Chemical
Co., St. Louis, Mo., and Koch-Light, Colnbrook, United
Kingdom.
The presence of a number of glycoside hydrolases, pepti-
dases, phosphatases, esterases, and lipases was demon-
strated by using chromogenic substrates included in two
separate kits, an API ZYM kit and an API ZYM osidase kit
(API System, La Balme-Les-Grottes, France). Both sets
of
test preparations were inoculated and read
as
described by
the manufacturer. The bacteria used were harvested from a
24-h Todd-Hewitt broth culture and were suspended (turbid-
ity, between MacFarland standards
5
and 6) in distilled water
for the API ZYM tests and in 0.15 M phosphate buffer (pH
7.5) containing
0.85%
(wthol) NaCl for the API ZYM
osidase tests.
Esculin hydrolysis was detected by the presence of dark
coloration after 7-days of growth in a 1% (wt/vol) Trypticase-
based medium containing
0.5%
(wt/vol) yeast extract,
0.05%
(wt/vol) ferric citrate,
0.5%
(wt/vol) sodium chloride, and
0.1% (wt/vol) esculin (4). D-Arginine hydrolysis was tested
with the Nessler reagent to detect ammonia after
5
days (11).
Starch hydrolysis was detected on Todd-Hewitt broth sup-
plemented with agar and 1% (wthol) starch. After incuba-
tion, the cultures were flooded with an iodine solution. Clear
zones around colonies were considered to be indicative of
starch hydrolysis. Hippurate hydrolysis was demonstrated
as described by Carlsson
(3,
and decarboxylation of orni-
thine and lysine was determined as described by Mgller (35).
Production of the extracellular polysaccharides dextran and
levan from sucrose was tested by precipitation;
1
and 3 parts
of 96% ethanol, respectively, were added to the supernatants
of 7-day cultures in Trypticase medium containing
5%
su-
crose (19).
Urease activity was detected by
a
micromethod (32).
Production of acetoin from glucose (Voges-Proskauer test)
was demonstrated in a medium composed of 1% tryptone,
0.5%
yeast extract,
0.5%
potassium hydrogen phosphate,
and
0.5%
glucose (added as a filter-sterilized solution after
autoclaving) (pH 7.3). A 2-ml portion of the medium was
inoculated, and the culture was incubated for 4 days. Ace-
toin was demonstrated by using a standard procedure (17).
Production of hydrogen peroxide was detected on the me-
dium described by Miiller (36).
Immunoglobulin A1 (IgA1) protease activity was demon-
strated as described previously by using purified human
myeloma IgAl as the substrate (27). Neuraminidase was
detected by using
4-methylumbilliferyl-a-~-N-acetylneuram-
inate as the substrate at a concentration of 0.1 mM dissolved
in 0.1 M sodium cacodylate buffer (pH 6.0)
(44).
One loopful
of bacteria from a Todd-Hewitt agar plate culture was
suspended in
0.2
ml of the substrate solution, and the
preparation was incubated overnight at 36°C. After centrif-
ugation, 0.1 ml of the supernatant was mixed with
1
ml of 0.1
M sodium bicarbonate buffer (pH
9.1).
Release
of
the
fluorescent product was demonstrated with a fluorescence
spectrophotometer (excitation, 365 nm; emission,
450
nm).
Serological examinations.
The streptococcal antisera were
prepared at the Streptococcus Department, Statens Seru-
minstitut, Copenhagen, Denmark. Grouping antisera cover-
ing groups A through V, as well as antisera against
“S.
milleri,”
“S.
mitior,” Streptococcus mutans,
and
S.
salivar-
TABLE
1.
Strains studied, listed according to the classification resulting from this study
G+C
Source, strain designation(s),
content
Refer-
(mol%)b
ence
and taxonomic statusa
Isolated from:
Study reference no.
Received as:
S.
sanguis
biovar
1
SKIT
ATCC
10556T;
type strain
of
S.
sanguis
46
46
S.
sanguis
type
I
S.b.e.
SK36, SK48
SK59
S.
sanguis
Streptococcus
sp.
group
H
S.
sanguis
Dental plaque, human
S.b.e.
27
NCTC
7863
(=
ATCC
10556)
SK72, SK75, SK76, SK77, SK78,
SK104
SKlO8, SK118, SK119
SK85
Dental plaque, human
33
3
Oral cavity, human
Initial dental plaque,
human
Sneath,
PB177
S.
oralis
S.
sanguis
S.
sanguis
biovar
2
SK4
SK115, SK156, SK157, SK158,
SK164
S.
sanguis
biovar
3
SK162, SK163
S.
sanguis
biovar
4
SK45, SK49, SK80
S
K46
SK112
SK6
SK150, SK159, SK160, SK161,
S.
gordonii
sp. nov. biovar
1
Carlsson,
804
Streptococcus
sp.
S.
sanguis
33
Dental plaque, human
Nyvad
37
S.
sanguis
Initial dental plaque,
human
Dental plaque, human
Buccal mucosa, human
Dental plaque, human
27
27
33
S.
sanguis
S.
sanguis
S.
sanguis
Lancefield group H
ATCC
12396
(Lancefield F90A,
Sneath,
PB179 (SK121
is
a
co-
Our isolate
Hare strain Perryer)
lonial variant of
SK120)
42
SK120, SK121
S.
oralis
3
Oral cavity, human
S.b.e.
SK123
SK3=
S.
gordonii
sp. nov. biovar
2
Streptococcus
sp.
ATCC
105BT;
type strain of
S.
gordonii
sp. nov.
NCTC
10231
(Blackburn)
S.
sanguis
type 1-11
41
42
13
SK5
Streptococcus
sp.
group H
S.
sanguis
Streptococcus
sp.
group
H
S.
sanguis
S.
sanguis
Challis
NCTC
7865
(=
ATCC
10558)
41
41
SK7
SK60
S.b.e.
33
SK86
SK183
Dental plaque, human
Burroughs Wellcome Ltd.
Handley,
FT2,
MJ2,
Rosan
M5
(CN2814)
SK184, SK188, SK190
SK8, SK9
SK11, SK12, SK17, SK33, SK83
SK51T
S.
gordonii
sp. nov. biovar
3
S.
sanguis
Dental plaque, human
Oral cavity, human
Oral cavity, human
W.
Liljemark
(S7,
Sl8)
Our isolates
NCTC
3165T;
type strain of
S.
mitis
Handley
,
LGR2
S.
sanguis
S.
sanguis
S.
mitis
39
Dental plaque, human
SK186
SK42, SK44
S.
gordonii
sp. nov. aberrant strains
S.
sanguis
27
S.
sanguis
Dental plaque,
Macaca
Dental plaque, human
fascicularis
SK43
S.
oralis
SK2
SKlO
SK23T
S.
sanguis
S.
sanguis
type I1
S.
sanguis
type I1
S.
oralis
S.b.e.
S.b.e.
Oral cavity, human
ATCC
10557
(=
NCTC
7864)
NCTC
7864
(=
ATCC
10557)
Sneath,
PB182T
(=
NCTC
11427T);
type strain
of
S.
oralis
42
42
41
33
Dental plaque, human
SK92, SK100, SK105, SK107,
SK122
SK133
SK140
SK143
SK144
SK146
SK153, SK155
SK225
SK38, SK39
SK109, SK111, SK134
S.
oralis
aberrant strains
37
5
5
5
5
5
Streptococcus
sp.
Streptococcus
sp. I:A
Streptococcus
sp. I:A
Streptococcus
sp. I:A
Streptococcus
sp. V:A
Streptococcus
sp. V:A
Streptococcus
sp.
S.
oralis
Dental plaque, human
Oral cavity, human
Oral cavity, human
Oral cavity, human
Oral cavity, human
Oral cavity, human
S.b.e.
Nyvad
Carlsson,
BU174
Carlsson,
HPAl
Carlsson,
MDPl
Carlsson,
OTFl
Carlsson,
OP5l
Gutschik
DSM
20066
42
42
42
42
42
39
30
27
Streptococcus
sp.
Oral cavity, human
Continued on following page
473
474
KILIAN ET AL.
INT.
J.
SYST. BACTERIOL.
TABLE
14ontinued
G+C
(mol%)b
Source, strain designation($, content
Refer-
and taxonomic statusa
Study reference no.
Received as: Isolated from:
S.
mitis
biovar
1
SK24
Streptococcus
sp.
SK113
Streptococcus
sp.
SK135
Streptococcus
sp. V:A
SK137, SK138, SK139, SK141
SK142
Streptococcus
sp. V:A
group
0
viridans type
Streptococcus
sp.
NCTC
8029
NCTC
10712
(=
FW75) 40
Carlsson,
FW103 40
Nyvad
Carlsson,
NS51
(=
NCTC
12261); 41
Sputum
Oral cavity, human
Dental plaque, human
Oral cavity, human
Oral cavity, human
37
5
5
30
proposed neotype
of
S.
mitis
Carlsson,
OS51
Hahn (Kiel
C119148)
SK145
Streptococcus
sp.
V:A
40
39
SK226
Streptococcus
sb.
group
0
S.
mitis
biovar
2
SK21
SK34, SK71, SK79, SK90, SK95,
SK132
SK147, SK148, SK149
SK84, SKllO
SK14, SK47
SK16
SK18
SK19
SK22
SK52=
SK96, SK102, SK103
S.
mitis
aberrant strains
S.
anginosus
Streptococcus
sp.
Streptococcus
sp.
SSI
6235164
Our isolates
Oral cavity, human
S.
mitis
Streptococcus
sp.
Ulcerated sore throat
Dental plaque, human
ATCC
903
Nyvad
39
37
Dental plaque, human
33
S.
milleri
S.
anginosus
S.
milleri
S.
anginosus
S.
milleri
S.
anginosus
Our isolates
HF60R
NCTC
10708
NCTC
8037
SSI
1986164
NCTC
10713T;
type strain
of
ATCC
27823T;
type strain
of
ATCC
27335T;
type strain
of
Sneath,
PB183
ATCC
9895
NCTC
10709
S.
anginosus
S.
constellatus
S.
intermedius
38
38
SK53T
S.
constellatus
SK54T
S.
interrnedius
SK55
SK57
SK61
SK63, SK64, SK65, SK66, SK73,
SK81, SK87, SK88, SK89,
SK154
SK98
SK126
S.
oralis
S.
mitis
S.
milleri
Dental plaque, human
33
S.
oralis
Streptococcus
sp.
group
F
Sneath,
PB187
NCTC
10714
S.
salivarius
SK13
SK25
SK26
SK32, SK40
SK56T
S.
salivarius
S.
salivarius
S.
salivarius
S.
salivarius
S.
salivarius
Human tongue
Our isolate
NCTC
8606
Kelstrup, Tove Ru
41
41
Buccal mucosa, human
Cardioarthritis
27
NCTC
8618=;
type strain
of
ATCC
9759
S.
salivarius
SK58
S.
salivarius
SK99, SKlOl
S.
salivarius
SK128
S.
salivarius
SK35
S
K74
S.
mutans
SK27
S.
mutans
SK28T
S.
mutans
S.
salivarius
aberrant strains
Dental plaque, human
Tongue, human
Buccal mucosa, human
Dental plaque, human
Dental plaque, human
Dental plaque, human
Dental plaque, human
Dental plaque, human
33
27
33
Our isolate
NCTC
10449T;
type strain
of
S.
mutans
Gibbons,
GS5
37
SK30
S.
mutans
SK62, SK68, SK69, SK70, SK91,
S.
mutans
SK93, SK97
33
~ ~~~~~~~~
a
ATCC, American Type Culture Collection, Rockville, Md.; Carlsson,
J.
Carlsson, University of Umea, Umea, Sweden; DSM, Deutsche Sammlung von
Mikroorganismen und Zellkulturen GmbH, Braunschweig, Federal Republic of Germany; Gibbons, R.
J.
Gibbons, Forsyth Dental Center, Boston, Mass.;
Gutschik,
E.
Gutschik, Department
of
Clinical Microbiology, Bispebjerg Hopital, Copenhagen, Denmark; Hahn,
G.
Hahn, Intitut for Hygiene, Kiel, Federal
Republic of Germany; Handley,
P. Handley, University of Manchester, Manchester, United Kindgom; Kelstrup,
J.
Kelstrup, Royal Dental College, Aarhus,
Denmark; NCTC, National Collection
of
Type Cultures, Colindale, United Kingdom; Nyvad, B. Nyvad, Royal Dental College, Aarhus, Denmark; Sneath,
P.
H.
A. Sneath, University of Leicester, Leicester, United Kingdom; SSI, Streptococcal Department, Statens Seruminstitut, Copenhagen, Denmark.
G+C values from references
12,
13,
30,
and
48.
VOL.
39, 1989 VIRIDANS STREPTOCOCCI 475
TABLE 2.
Streptococcal strains used
for
the
production
of
antisera
Group,
type,
or
species designation Strain
~~~~ ~ ~
Group
C
Group
F
Group G
Ottens type
I1
Ottens type
I11
Group
H
Group
H
(F90A;
American
group
H)
“S.
mitior”
Group
K
(Levy)
Group
K
Group
0
“S.
milleri”
“S.
miller?’
S.
mutans
type
c
S.
salivarius
S.
salivarius
ATCC 12388
SSI H127
ATCC 12394
H189
MG216
Blackburn
(=
SK5)
ATCC 12396
(=
SK6)
ATCC 10557
(=
SK2)
NCTC 10232
CN477
2/65
Jelinkova
SSI 14132170
SSI MG2 1132
NCTC 10449
ATCC 13419
NCTC 8606
ius
and antisera against the five Ottens-Winkler antigens,
were used to examine each strain
(38).
Those antisera that
gave positive reactions with the strains which we studied are
shown in Table 2; the labels of the sera are listed together
with the designations of the strains used for immunization.
Antisera were prepared by immunizing rabbits with strep-
tococci grown in Todd-Hewitt broth for about 18 h at 36°C.
The cultures were killed by heating them at 60°C for
30
min,
washed three times in saline, and stored in saline at a
concentration of
1:lOO
of the original volume, with
0.5%
Formalin added as preservative. The concentration of bac-
teria used for immunization was either
4
x
lo9
or
8
x
lo9
organisms per ml, and the course of immunization was either
4
weeks or 2 to
3
weeks, respectively. The animals were
given three successive injections of
0.5
ml the first week and
three injections of
1
ml the following weeks. The animals
were not bled until a strong capillary precipitation reaction
had developed. Bleeding generally took place
5
to 6 days
after the last injection.
RESULTS
Cultures.
During the initial phases of this study, it became
evident that many of the cultures examined yielded biochem-
ical reactivity patterns that appeared to be intermediate
between the patterns shown by defined taxa. Repeated
attempts to check the purity of such cultures by serial
subcultivation revealed that two different strains could be
isolated only after up to five subcultures on Todd-Hewitt
agar. Thus, several cultures, including “strains” received
from other laboratories, initially consisted of two diEerent
strains. Consequently, all
of
the strains included in the study
were purified by at least five serial subcultures on Todd-
Hewitt agar.
Even after careful purification, cultures on Todd-Hewitt
agar often showed a mixture of colony types that could not
be separated by subculturing; i.e., they were not stable.
Biochemical examination of individual cultures based on
different colony types revealed identical results.
No
distinct colonial morphology on Todd-Hewitt agar
could be associated with the individual taxa in this study.
However, in most cases, species-associated differences in
colonial morphology on mitis salivarius agar were evident
(Table
3).
Biochemical tests.
In a pilot experiment,
18
strains that
were later assigned to
S.
sanguis
were tested for the ability
to hydrolyze esculin both in the medium described above
and in a medium in which tryptic soy broth was substituted
for Trypticase peptone. While 12 of the strains showed a
positive reaction in the medium containing Trypticase pep-
tone, only
3
strains were positive in the medium containing
tryptic soy broth. In the final examination
of
strains the
medium containing Trypticase peptone was used.
Several glycoside hydrolases were determined by both the
API
ZYM
and API
ZYM
osidase kits. However, as shown in
Table
3,
the two kits did not produce identical results. For
example, the majority of the
S.
sanguis
strains were re-
corded as negative for p-D-glucosidase and p-D-galactosi-
dase activities when they were examined by using the API
ZYM
kit, whereas the majority of strains were positive when
they were examined by using the API
ZYM
osidase kit
(Table
3).
Furthermore, in several cases in which the.API
ZYM
kit yielded weak and often poorly reproducible reac-
tions (e.g., a-D-galactosidase), the reactions obtained with
the API
ZYM
osidase kit were generally strong and pro-
duced more consistent results within the individual taxa.
Generally weak, indistinct, and poorly reproducible reac-
tions were observed in the tests for esterase (API
ZYM
kit),
esterase lipase (API
ZYM
kit), cystine aminopeptidase (API
ZYM
kit), trypsin (API
ZYM
kit), PNP-p-D-galacturonohy-
drolase (API
ZYM
osidase kit), and a-L-arabinosidase (API
ZYM
osidase kit). These reactions are not included in the
tables.
None of the strains included in the study were inhibited by
optochin, had detectable lysine or ornithine decarboxylase
activity, fermented glycerol, dulcitol, L-( -)-sorbose, or
rhamnose, or exhibited a-mannosidase,
a-
or p-xylosidase,
N-acetyl-a-D-glucosaminidase,
a-fucosidase (API
ZYM
kit),
or p-L-fucosidase
(API
ZYM
osidase kit) activity. All
of
the
strains fermented D-(+)-glucose, D-( -)-fructose, and
D-(
-)-
mannose.
The pH values of uninoculated fermentation media were
7.2
to
7.4.
When tests were recorded as positive, the final pH
values ranged from
4.0
to
5.0,
although occasional strains
showed final pH values for single carbohydrates of 5.6 to
5.9.
Exceptions to this pattern were strains assigned to
S.
oralis,
which consistently produced final pH values of 5.6 to 6.2 in
the melibiose medium. Strains assigned to other species
which were recorded as positive in this medium produced
final pH values of
4.3
to 5.2. The final pH values measured in
glucose broth were very consistent within individual species,
with deviations of only 20.2
U
from the calculated mean
(Table
3).
Characteristics
of
the individual taxa.
The biochemical and
serological characteristics of the strains which we classified
are shown in Table
3.
To calculate the data in the table,
strains examined in duplicate that were obtained from two
different sources (American Type Culture Collection, Rock-
ville, Md., and National Collection of Type Cultures, Cen-
tral Public Health Laboratory, London, England) were in-
cluded only once
as
the duplicates always gave identical
results. Strains that we considered aberrant are not included
in Table
3
but are discussed individually below. The results
in Table
3
are presented according to the revised classifica-
tion resulting from this study. Although the names used in
the tables and in the descriptions
of
taxa below to some
extent anticipate the final taxonomic discussion (and one
case is pending the response of the Judicial Commission to
our proposed rejection of the present type strain of
S.
mitis
[28]), it makes the presentation easier to follow.
P
4
m
TABLE
3.
Characteristics of the individual taxa
%
of
strains positive
Characteristic
S.
mitis
S.
sali-
S.
angi-
S.
mu-
S.
sanguis
S.
gordonii
nosus
tans
S. oralis
Biovar
1
Biovar
2
Biovar
3
Biovar
4
Biovar
1
Biovar
2
Biovar
3
(n
=
18)
Biovar
1
Biovar
2
van'us
(n
=
(n
=
13)
(n
=
10)
(n
=
24)
(n
=
10)
5
(n
=
13)"
(n
=
6)
(n
=
6)
(n
=
5)
(n
=
4)
(n
=
9)
(n
=
9)
7
Colonial characteristics
Smooth
High
Firm
Adherent
Diameter
>1
mm
a-Hemolysis
P-He.molysis
Greening
of
chocolate agar
Chain length of
>10
cells per
chain
Inhibition by:
Bacitracin
Nitrofurazon
Sulfafurazole
4%
(wtkol) NaCl
6.5%
(wthol) NaCl
Arginine hydrolysis
Esculin hydrolysis
Hippurate hydrolysis
Starch hydrolysis
Urease
Voges-Proskauer reaction
H202
production
Dextran production
Levan
production
Neuraminidase
IgAl
protease
Galactose
Mannitol
Sorbitol
N-Acetylglucosamine
Amygdalin
Arbutin
Esculin
Salicin
D-(
+)-Cellobiose
Maltose
Lactose
D-(
+
)-Melibiose
Sucrose
D-(
+
)-Trehalose
Inulin
D-(
+
)-Melezitose
D-(
+
)-Raffinose
Starch
Biochemical characteristics
Acid produced fromb:
92
92
100
100
0
100
0
100
0
0
100
0
38
100
100
100
0
100
0
0
100
92
0
0
100
100
0
8
100
0
85"
85
92'
92
100
100
100
100
100
100
0
92"
3gd
100
100
100
100
0
100
0
100
0
0
100
100
100
100
100
0
0
0
0
0
100
100
0
0
100
100
0
0
100
0
100
0
100
33
100
100
0
100
100
100
0
0
0
100
100
100
100
0
100
0
100
0
0
100
0
50
100
100
100
0
83
0
0
100
100
0
0
100
100
0
67
100
0
100
0
100
100
100
100
83
100
100
100
0
83
0
100
40
40
40
0
100
0
100
75
0
100
0
60
100
100
100
0
100
0
0
100
100
0
0
100
100
0
0
100
20
100
80
100
100
100
100
0
100
100
80
0
0
0
100
100
100
100
0
100
0
100
0
0
100
0
100
100
100
100
0
100
0
0
100
100
0
0
0
100
0
0
100
100
100
100
100
100
100
100
100
100
100
100
0
100
0
100
100
100
100
0
100
0
100
0
0
100
0
100
100
100
100
0
89
0
0
100
100
0
0
0
100
0
0
100
100
100
100
100
100
100
100
0
100
100
78
0
0
0
100
67
67
67
0
100
0
100
0
0
100
0
100
100
100
100
0
44
0
0
100
89
0
0
0
100
0
0
100
100
100
100
100
100
100
100
0
100
100
100
0
0
0
100
72
78
78
0
100
0
100
78
0
100
0
100
100
0
0
0
94
0
0
100
78
0
100
100
100
0
0
100
0
0
0
0
6
94
100
lo@
94
17"
0
6
56"
0
100
0
0
0
0
100
0
100
70
0
100
0
10
100
0
0
0
50
0
0
100
0
0
20
20
100
0
0
100
0
0
0
20
50
80
100
40
100
30
0
0
70
0
100
0
0
0
0
92
0
100
31
0
100
0
15
100
100
0
0
92
0
0
100
0
0
62
0
100
0
39
100
0
0
0
39
62
100
100
100
100
8
0
0
100
0
100
20
0
10
100
0
40
20
20
100
0
80
100
0
100
0
90
40
60
0
0
100
0
0
40
0
0
100
90
100
100
100
100
100
50
0
100
30
60
0
0
0
100
0
0
0
0
50
0
67
41
4
100
13
13
100
100
100
4
63
0
71
57
0
0
0
0
100
4
0
100
58
100
83
100
84
100
79
25
100
88
0
0
25
0
M
4
E
!+
40
10
0
70
0
20
60
10
100
0
20
100
0
100
0
90
0
90
20
100
0
0
0
100
100
100
100
70
100
100
100
Y
100
100
100
4
100
4
100
100
3
m
100
y
0
3
100
E
0
r
Enzymes"
Alkaline phosphatase
(z)
Acid phosphatase
(z)
Leucine aminopeptidase
(z)
Valine aminopeptidase
(z)
Chymotrypsin
(z)
Phosphoamidase
(2)
P-D-Glucosidase
(z)
P-D-Glucosidase
(0)
P-D-Galactosidase
(z)
P-D-Galactosidase
(0)
a-D-Glucosidase
(z)
a-D-Glucosidase
(0)
a-D-Galactosidase
(z)
a-D-Galactosidase
(0)
P-Glucosaminidase
(z)
N-acetyl- P-D-glucosamini-
a-Maltosidase
(0)
P-Maltosidase
(0)
P-Mannosidase
(0)
a-L-Fucosidase
(0)
P-D-Fucosidase
(0)
P-D-Glucuronidase
(2)
9-D-Glucuronidase
(0)
P-D-Lactosidase
(0)
a-L-Arabinosidase
(0)
Phospho-P-galactosidase
(0)
Reactions with antisera
Blackburn (group
H)
F90A
(American group H)
ATCC
10557
SSI
14132
("S.
millen'")
CN477
(group
K)
NCTC 8606
(S.
salivanus
ATCC 13419
(S.
salivan'us)
Group
C
Group
G
Group
F
Group
0
Ottens type
I1
Ottens type
I11
MG2
S.
mutans
(serotype
c,
dase
(0)
Levy
(group
K)
group
K
NCTC
10449)
0/23f
100
100
100
5718
0
0
100
100
100
0
0
5413
1'
941of
0
100
100
0
0
0
92
0
0
15138
0
23/62
77
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
100
83
100
0/100
0
100
0
0/100
0
0
0/100
0/100
0
0
100
0
0
0
0
0
0
0
0
0
100
0
0
50
0
0
0
0
0
0
0
0
0
0
0
0
0
0
100
100
100
0
0
33/67
0
0183
0
0
67/17
100
0
0/100
50
0
0
0
0
0
0
0
0
0
100
0
0
100
0
0
0
0
0
0
0
0
0
0
0
0
0
100
100
100
60140
0180
80
100
0
100
0
0
0
0
0
100
100
0
0
0
0
0
0
40160
0
0180
0
0
100
0
0
0
0
0
0
0
0
0
0
0
0
100
100
100
100
100
01100
100
100
100
100
On5
0
75/25
100
100
100
100
0
100
100
0
0
0
100
0
25/75
0
25
75
0
0
0
0
0
0
0
0
0
0
0
0
0
100
100
100
100
8911
1
0
56/22
100
44/44
100
67/33
78/22
0
6710
100
100
100
0
100
100
0
0
0
100
0
88/12
89
0
0
33
0
0
0
0
0
0
0
0
0
0
0
0
100
100
100
100
100
0
44/66
100
11/67
100
22/67
56/44
0
0
100
100
100
0
56/33
100
0
0
0
100
0
78/22
0
22
78
0
0
0
0
0
0
0
0
0
0
0
0
0
89/11
100
100
100
100
0172
0
0
94
100
100
100
17/28
33/33
100
100
9416
0
0
0
39/56
0
0
0
0
0150
0
0
50
0
0
0
0
0
0
0
0
0
0
0
0
0
60
70
100
100
100
0180
0
0
20140
50140
60
60
80
80
0
80120
0
0
0
0
0120
0
0
0
0
0
0
0
0
0
0
0
20
0
0
0
0
90
0
0
0
0
54/31
85
100
100
100
0154
0
92
69/31
100
100
100
77/16
100
100
100
100
0
0
0
3813
1
0
0
46/38
0
62/15
0
0
0
0
46
31
15
0
0
0
0
0
0
0
0
0
60140
80120
100
80120
90110
80/10
80110
90110
50
60
60140
9010
0
0
0
0
100
013
0
0
0
20
0
0
6010
0
0
0
40
0
0
0
0
30
0
0
0
0
0
0
0
0
0
67/33
100
100
83/13
79/21
13/63
25/25
75/17
29/21
33/46
63/25
67/13
13
13
17/29
38/46
54/21
8
3318
8
4
4
4
17/17
8
13/17
0
0
0
4
0
0
0
0
13
17
33
0
8
8
4
0
0
0120
100
20180
70130
0/100
100
100
0
100
100
100
20180
100
0
0
0150
0
0
0
0
0
0
100
0
0140
0
0
0
0
0
0
0
0
0
0
0
0
0
0
100
n
is
the number of strains tested.
The mean terminal
pH
values in glucose medium were as follows:
S.
sanguis (all biovars), 4.7;
S.
gordonii biovar
1,
4.3;
S.
gordonii biovar 2, 4.4;
S.
gordonii biovar
3,
4.3;
S.
oralis,
4.5;
S.
mitis
biovar
1,
The reaction
of
the type strain differed from the reaction of the majority
of
the strains.
The majority of the fermentation reactions were weaker than the reactions with the other carbohydrates
(pH
5.6
to
6.2).
(z),
Determined with an
MI
ZYM
kit;
(o),
determined with an
API
ZYM
osidase kit.
4.6:
S.
mitis
biovar
2,
4.4;
S.
salivarim,
4.1;
S.
anginosus, 4.2;
S.
mutans,
4.0.
f
Percentage of strong positive reactions (values
of
3
to
percentage of
weak positive reactions (values
of
1
or
2).
478
KILIAN
ET
AL.
INT.
J.
SYST.
BACTERIOL.
S.
sanguis.
A total of
31
strains were assigned to
S.
sanguis.
On the basis of biochemical and serological differ-
ences, this species could be subdivided into four biovars,
which are characterized as shown in Table
3.
Biovar
1
included the type strain of
S.
sanguis,
strain ATCC
10556
(=
NCTC
7863).
However, as shown in Table
3,
the character-
istics of the type strain differed in several ways from the
majority reactions of the taxon. Only
1
of the
31
strains
assigned to
S.
sanguis
failed to produce detectable extracel-
lular polysaccharide in
5%
sucrose broth. This strain, strain
SK118,
also differed from other strains assigned to biovar
1
by fermenting sorbitol but not esculin.
With the exception
of
three strains of biovar
1
and five
strains of biovar
4,
all of the strains assigned to
S.
sanguis,
including the aberrant strain, reacted with group
H
(Black-
burn) antiserum. All five biovar
4
strains reacted with the
antiserum against
“S.
mitior”
ATCC
10557.
Besides react-
ing with the group H (Blackburn) antiserum, the majority of
the strains assigned to biovars
2
and
3
reacted with
“S.
milleri”
SSI
14132/70
antiserum.
Streptococcus gordonii
sp.
nov.
The
25
strains assigned to
Streptococcus gordonii
sp. nov. possessed many of the
properties that are considered characteristic of
S.
sanguis,
such as the ability to hydrolyze arginine and esculin and the
ability to produce extracellular polysaccharide. However,
they differed from
S.
sanguis
strains in lacking IgAl protease
activity, in their ability to ferment amygdalin, and in having
P-glucosaminidase, P-mannosidase, a-L-fucosidase, and
strong alkaline phosphatase activities (Table
3).
As dis-
cussed below, the results of DNA base composition and
DNA-DNA homology studies support the point of view that
this group warrants specific taxonomic status.
S.
gordonii
could be subdivided into three biovars on the
basis of the biochemical and serological reactivities of the
strains (Table
3).
One of these biovars (biovar
3)
included the
designated type strain of
S.
mitis,
strain
SK51
(=
NCTC
3165).
However, this strain differed from
all
of the other
strains assigned to the species by lacking extracellular
polysaccharide production and by exhibiting a strong
P-
galactosidase reaction when the API
ZYM
kit was used. The
taxonomic problems raised by this strain are discussed
below.
Eight of the nine strains of biovar
2
reacted with group H
(Blackburn) antiserum, and three showed an additional
reaction with
“S.
milleri”
SSI
14132/70
antiserum. With one
exception, all of the other strains reacted either with the
ATCC
10557
antiserum or with the American group H
(MA) antiserum.
Three strains (strains
SK42, SK43,
and
SK44)
were con-
sidered aberrant strains of
S.
gordonii.
Two of these, strains
SK42
and
SK44,
were isolated from dental plaque of
Macaca
fascicularis
monkeys. These three strains differed
from other strains assigned to
S.
gordonii
by exhibiting IgAl
protease activity (except strain
SK43)
and by lacking
P-
glucosaminidase, P-mannosidase, a-L-fucosidase, and alka-
line phosphatase activities. Several of these properties are
characteristic of
S.
sanguis.
However, strains
SK42, SK43,
and
SK44
differed from strains of
S.
sanguis
by fermenting
amygdalin, by exhibiting p-D-lactosidase activity, and by
lacking
a-
and P-galactosidase and acid phosphatase activi-
ties when the API
ZYM
kit was used. All three strains
reacted in group H (Blackburn) antiserum. Although these
strains appeared to be phenotypically intermediate between
S.
gordonii
and
S.
sanguis,
a
genetic analysis of one of them
(strain
SK42)
revealed that it is closely related to the former
species but not to the latter species
(20).
S.
oralis.
A total of
21
strains were assigned to
S.
oralis.
These strains included the type strain of
S.
oralis
(strain
NCTC
11427)
and several other reference strains previously
designated
“S.
sanguis
11”
or
“S.
mitior”
(Table
1).
The
majority of the strains produced extracellular polysaccha-
ride, and all exhibited both
IgAl
protease and neuraminidase
activities.
A
detailed description of
18
of the strains is
provided in Table
3; 9
of these
18
strains, including the type
strain, reacted in the antiserum raised against strain ATCC
10557.
The remaining strains were negative in all antisera.
Two strains (strains
SK38
and
SK39),
which were consid-
ered aberrant, differed from all of the other strains
of
the
species by lacking P-glucosaminidase, P-galactosidase (API
ZYM
kit), and alkaline phosphatase activities.
S.
mitis.
A total of
25
strains were tentatively assigned to
S.
mitis
(see below). This species was subdivided in two
groups, designated biovar
1
and biovar
2.
None of the strains
produced extracellular polysaccharide. The two biovars
were separated by having different reactions in tests for
arginine hydrolysis, P-glucosaminidase, P-glucosidase, and
a-mannosidase (Table
3).
The detailed biochemical charac-
teristics are shown in Table
3.
Two strains tentatively
assigned to biovar
1
(strains
SK135
and
SK136)
exhibited
both IgAl protease and neuraminidase activities.
All but one of the strains assigned to biovar
1
reacted with
the group
0
antiserum, and two, in addition, reacted with the
antiserum raised against
S.
salivarius
NCTC
8606.
The
majority of the biovar
2
strains reacted with one of the two
group
K
antisera (Levy and
CN477).
One strain, strain
SK132
(=
ATCC
903),
reacted with both of these antisera, as
well as with the NCTC
8606
and Ottens type
I11
antisera, and
one strain reacted with the NCTC
8606
antiserum only.
The four strains assigned to biovar
2,
including strain
SK132
(=
ATCC
903),
which reacted with the
CN477
group
K
antiserum, may warrant recognition as a separate biovar
as they differed from other strains of biovar
2
by lacking
neuraminidase and p-D-fucosidase activities and by ferment-
ing sorbitol.
Likewise, the exact classification of two strains (strains
SK84
and
SK110)
listed as aberrant will require further
study. In contrast to the strains of biovar
2,
these two strains
did not hydrolyze arginine but did hydrolyze esculin and
fermented trehalose. They otherwise possessed the majority
of the reactions of
S.
mitis
biovar
2
(Table
3),
including
neuraminidase activity.
S.
salivarius.
The
10
strains assigned to
S.
salivarius
included the type strain of
S.
salivarius,
strain NCTC
8618
(=
SK56),
and two other reference strains of
S.
salivarius
(Table
1).
This group of strains was relatively homogeneous
and was well separated from the other taxa (Table
3).
The
majority of the strains reacted with antiserum against either
American group
H
(F90A) or
S.
salivarius
NCTC
8606.
Two
additional strains assigned to this species were considered
aberrant. Strain
SK35
differed from other strains in lacking
the ability to ferment N-acetylglucosaminic acid, in ferment-
ing melibiose and railinose, in exhibiting a strong
P-D-
fucosidase reaction, and in lacking alkaline and acid phos-
phatase, valine aminopeptidase, and phosphoamidase
activities. It reacted with the American group
H
(F90A)
antiserum. The other strain (strain
SK74)
was the only strain
in the collection that was capable of hydrolyzing hippurate.
Furthermore, it differed from the other strains by lacking the
ability to ferment N-acetylglucosaminic acid, by exhibiting
P-mannosidase and P-glucuronidase activities, and by lack-
ing a-maltosidase activity. It reacted with ATCC
13419
antiserum.
VOL.
39,
1989
VIRIDANS STREPTOCOCCI
479
S.
anginosus. The
24
strains tentatively assigned to
S.
anginosus
included the type strains of
S.
anginosus,
S.
constellatus,
and
S.
intermedius,
several reference strains of
“S.
milleri,”
S.
mitis
ATCC
9895,
and two strains classified
as
S.
oralis
by Bridge and Sneath
(3)
(Table
1).
All
24
strains
hydrolyzed arginine and esculin, and the majority
(71%)
produced acetoin. However, as revealed by the data in Table
3,
this taxon was
a
biochemically and serologically very
heterogeneous group, and no natural clustering of the
24
strains was found. The strains were separated from strains
assigned to
S.
sanguis
and
S.
gordonii
by their lack of
extracellular polysaccharide formation, by their inability to
ferment inulin, by their colonial morphology on mitis sali-
varius agar, and by their serological reactions. None of the
strains assigned to
S.
anginosus
possessed
a
group H anti-
gen. In contrast, the majority
of
the strains reacted with
group C,
F,
or
G
antiserum and with antiserum to Ottens
type I1 or
I11
(Table
3).
S.
mutans.
The
10
strains assigned to
S.
mutans
formed a
homogeneous group with only minor variations in certain
biochemical reactions (Table
3).
All of the strains reacted in
serotype c antiserum. The strains were clearly separated
from
all
other strains included in this study.
Genetic relatedness between
and
within
the
species.
The
strain collection examined in this study was composed to
include, as far as possible, strains that previously had been
subjected to genetic analyses, such as determinations of
DNA base composition and DNA-DNA homology and mul-
tilocus enzyme analysis.
As
the data from these studies
formed an important background for the taxonomic deci-
sions resulting from our study, they are summarized below
with special emphasis on the species
S.
oralis,
S.
mitis,
S.
sanguis,
and
S.
gordonii.
Six of the strains which we assigned to
S.
oralis
(strains
SK2,
SK133, SK140, SK143,
SK144,
and
SK146)
and three
of the strains assigned to
S.
mitis
(strains
SK135, SK142,
and
SK145)
were examined in DNA-DNA hybridization experi-
ments
(S1
nuclease method) by Coykendall and Munzen-
maier
(12).
Applied to our classification, the results of these
authors showed
65
to
100%
homology within
S.
oralis
and
S.
mitis
and
30
to
38%
homology between the two species.
Values compatible with this level of interspecies relatedness
have been reported for
S.
mitis
biovar
1
(strain
SK226)
and
S.
oralis
(strain
SK23)
by Kilpper-Balz et
al.
(30)
and for
S.
mitis
biovar
2
(strain
SK132)
and
S.
oralis
(strain
SK2)
by
Coykendall and Specht
(13).
DNA homology data demonstrating the intra- and inter-
species relationships of
S.
sanguis
and
S.
gordonii
as defined
in this study have been published for
S.
sanguis
SK1,
SK4,
and
SK59
and
S.
gordonii
SK3,
SK5,
SK6, SK51,
and
SK190
by Coykendall and Specht
(13),
Schmidhuber et al.
(41),
and Welborn et al.
(47).
Applied to our classification,
the data from these studies showed
89
to
90%
genetic
homology within the species and
40
to
60%
homology
between the two species. Furthermore, all three studies
demonstrated
20
to
30%
genetic homology between
S.
san-
guis
and
s.
gordonii
on one side and
s.
oralis
on the other.
These levels of intra- and interspecies relatedness among the
three species are further supported by the results of our own
genetic analysis of
23
strains from this study in which
multilocus enzyme electrophoresis was used
(20).
Information on the intra- and interspecies relationships
and DNA base compositions based on studies of strains
included in this study is summarized in Fig.
1.
DNA base
compositions compiled from previous studies
(12,
13,48)
are
included in Table
1
for individual strains.
S.
mitis
Lysdirect
G+C
40-41
Yo
30-3a%
\
28%
\
I
S.
oralis
Lvsdirect
G+C
41 -42%
S.
sanguis
/
S.
gordonii
-
40-60%
G+C
40-42%
G+C
4g46%
Lys-Ala,,
Lys-Ala,,
FIG.
1.
Genetic relatedness,
DNA
base compositions, and cell
wall peptidoglycan types of selected
Streptococcus
species defined
in this study. The information is compiled from previously published
data for strains included in this study
(12, 13,
30,
41, 47).
DISCUSSION
Several factors undoubtedly contribute to the difficulties
that have been encountered in reaching
a
satisfactory tax-
onomy
of
oral streptococci.
It
is notoriously difficult to
obtain pure cultures of these bacteria, as we experienced in
the initial phases of this study. This has probably led several
authors to consider oral streptococci
a
continuum with
overlapping species and many intermediate forms
(15, 22).
Second, many of the biochemical tests used to examine
streptococci are not sufficiently optimized for these bacteria,
and small alterations in
a
test may give different results. This
is evident when results reported by workers from different
laboratories for identical strains are compared. For example,
the type strain of
S.
sanguis,
strain ATCC
10556
(=
SKl),
is
often reported to be negative for esculin hydrolysis. In our
study this strain was found to be clearly positive in the test
substrate recommended by Carlsson
(4),
but negative, along
with several other
S.
sanguis
strains, if tryptic soy broth was
substituted for Trypticase peptone in the medium. Several
differences between phenotypic data reported by workers in
the laboratory of Schleifer
(30,41),
who used API
20
STREP
(41)
or
“conventional tests”
(30),
also demonstrate the
importance of the methods used to determine specific bio-
chemical activities of streptococci. Thus, the type strain
of
S.
sanguis,
strain ATCC
10556
(=
SKl), has been reported
to be unable to hydrolyze arginine
(30),
whereas the type
strain of
S.
oralis
was found to be positive. Both results are
in conflict with data from several other studies, including our
study. Other examples, which are relevant for the pheno-
typic description of
S.
oralis,
are discussed below.
Similar discrepancies in the reported abilities of strains to
ferment carbohydrates may be due to the use of different pH
indicators. To avoid this problem, we measured the final pH
in every fermentation test culture after
7
days of incubation.
Virtually
all
of the reactions recorded as positive reached
pH
values between
4.0
to
5.0,
and the values were remarkably
consistent within each taxon (Table
3).
However, a few
characteristic exceptions were found. Thus, the final pH
values of
S.
oralis
strains in melibiose and
of
S.
sanguis
biovar
1
strains in starch were consistently between
5.6
and
6.2.
480
KILIAN
ET
AL.
INT.
J.
SYST.
BACTERIOL.
The API kits with chromogenic substrates used to detect
glycoside hydrolases in this study also demonstrate the
significance of test optimization. API ZYM kits have been
used previously in studies of oral and other streptococci (3,
9, 45). However, the reproducibility and diagnostic validity
of the results have not been entirely satisfactory. The same
problems were observed in this study for some of the
reactions. The importance of the substrate used for propa-
gation
of
the inoculum has been demonstrated previously
(9). Another factor is demonstrated by the results obtained in
our study, in which we used two different API kits with
many overlapping tests. Table 3 shows that the results
obtained for identical glycoside hydrolases with the two kits
often were quite different. Generally, results obtained with
the API ZYM osidase kit were more clear-cut, reproducible,
and consistent within individual taxa. The two kits use
different chromogenic substrates (i.e., naphtol and nitrophe-
no1 derivatives, in the API
ZYM
and API ZYM osidase kits,
respectively), which may influence the results. However, it
may
also
be significant that the inocula used for the API
ZYM osidase kit tests were suspended in a buffer at pH 7.5,
which is closer to the pH optimum of the relevant enzymes
(pH 7.5 to
8.0)
(unpublished data) than the pH obtained with
the API ZYM kit tests, which used saline-suspended inoc-
ula.
Two
of
the tests used in this study, the IgAl protease and
neuraminidase tests, have not been used previously in strep-
tococcal taxonomy. Both of these tests proved to be highly
valuable in supporting the classification presented in Table 3.
However, the drawback of the two tests is that they require
specialized equipment and are, at the present time, not easily
applicable to routine bacteriology.
The individual taxonomic groups resulting from this study
are discussed below. The entity previously referred to as
“S.
sanguis
11”
is
a
priori excluded from the discussion of
S.
sanguis
as
“S.
sanguis
11”
strains differ from the type strain
of
S.
sanguis
by multiple biochemical characteristics
(5,
11,
24) (Table 3), by growth requirements (6), by cell wall
composition (10, 34, 39), by serology (11, 23), by DNA base
composition, and by exhibiting only 20 to 40% DNA homol-
ogy (13, 41).
“S.
sanguis
11”
is discussed below under
S.
oralis.
S.
sanguis.
S.
sanguis
is the name given by White and
Niven (49) to the viridans type of streptococci that hydrolyze
arginine, usually hydrolyze esculin, ferment inulin, salicin,
lactose, and trehalose, and produce extracellular dextran
from sucrose (26). However, a certain heterogeneity has
been suggested as a result of both biochemical and serolog-
ical studies (29, 39, 49). Most convincingly, Coykendall and
Specht (13) demonstrated two separate genotypes with only
30 to 60% DNA homology and significantly different DNA
base compositions (guanine-plus-cytosine [G+C] contents,
44
to 46 and 41 to 43 mol%). The names
“S.
sanguis
subsp.
carlssonii”
and
“S.
sanguis
subsp.
sanguis”
were applied to
the two genotypes, but because of a lack of correlating
phenotypic differences this subdivision has never been prac-
tically employed. Our study revealed several phenotypic
traits that clearly differentiate two major clusters, which, as
noted above, contained several strains representing the
genotypes described by Coykendall and Specht. The two
clusters clearly warrant specific recognition (see below). The
cluster that includes the type strain of
S.
sanguis,
strain
ATCC 10556
(=
NCTC 7863) (Table 3), must retain the name
S.
sanguis.
S.
sanguis,
thus defined, correlates with
S.
sanguis
subsp.
carlssonii
of Coykendall and Specht.
S.
sanguis
sensu strict0 was divided into four taxa that
were separated by several biochemical and serological char-
acteristics. We provisionally label these taxa biovars
1
through
4.
The type strain was assigned to biovar
1.
Para-
doxically, this strain was unique among the
S.
sanguis
strains because it lacked the ability to ferment several
carbohydrates typical of the species (Table 3). The majority
of the
S.
sanguis
strains reacted with the group
H
(Black-
burn) antiserum. However, it is interesting that none of the
strains reacted with the American group H (F90A) antise-
rum. The five strains assigned to biovar 4 were exceptional
in reacting with antiserum against
S.
oralis
ATCC 10557
(“S.
mitior”)
but not with any of the group H antisera and by
forming long chains (Table 3). Also, strains of biovar 4 have
been shown to differ from other
S.
sanguis
strains by
possessing a distinct IgAl protease type (Reinholdt et al., in
preparation) and may belong to a separate species.
S.
gordonii
sp.
nov.
The cluster excluded from
S.
sanguis
included strains previously designated
“S.
sanguis
subsp.
sanguis”
by Coykendall and Specht (13). Furthermore, it
contained strain NCTC 3165
(=
SK51), which is the desig-
nated type strain of
S.
mitis
(43). This strain differed from all
other strains of
S.
gordonii
by lacking detectable extracellu-
lar dextran, and we previously demonstrated a genetic
distance of 20% (Gower distance coefficient) between this
strain and other representative strains of
S.
gordonii
based
on multilocus enzyme electrophoresis (20). In our opinion, it
would clearly cause confusion to apply the name
S.
mitis
to
this group of organisms which is characterized by arginine
and esculin hydrolysis, inulin fermentation, and production
of extracellular polysaccharide, and it would be in conflict
with the original description of that species
(1).
Pending the
response of the Judicial Commission to our request to
replace the type strain of
S.
mitis
(28), we propose the name
S.
gordonii
in honor of the British bacteriologist Mervyn H.
Gordon, who apparently was the first researcher to apply
fermentation tests to the classification of viridans strepto-
cocci in a comprehensive way (21). Previously published
DNA-DNA hybridization data on five strains which we
assigned to
S.
gordonii
(see above) show that it is a homo-
geneous taxon. It is clearly distinct from
S.
sanguis
based on
the following data: it exhibits only 40 to 60% DNA homology
with
S.
sanguis
(Fig. 1); it has a significantly lower G+C
content than
S.
sanguis
(42 versus 46 mol%) (13); it ferments
amygdalin; it has alkaline phosphatase and p-glucosamini-
dase activities; and it lacks IgAl protease activity (Table 3).
Further genetic evidence to support the separation of the
two species was recently obtained by a multilocus enzyme
analysis of several representative strains from this study
(20).
The three taxa defined within
S.
gordonii
differ biochem-
ically and serologically. Whereas the majority of strains
assigned to biovars
1
and 3 reacted with the antiserum
against
S.
oralis
ATCC 10557
(“S.
mitior”)
and some
reacted with American group
H
(F90A) antiserum, most of
the strains of biovar 2 reacted with group
H
(Blackburn)
antiserum. Together with some of our findings for group
K
strains (see below), these results confirm and extend previ-
ous
observations
(7,
11)
that streptococcal group antigens
may be shared by separate species. Based on results pub-
lished by Rosan (40) and Cole et al. (7) for strains common to
their studies and our study, it appears that
S.
gordonii
biovar
2 strains contain all of the antigens (antigens
a
through e)
defined by Rosan
(40),
as does the type strain of
S.
sanguis
(strain ATCC 10556).
S.
gordonii
biovar
1
strain SK6
(=
F90A) was previously found to possess antigens a and e,
whereas
S.
gordonii
biovar 3 strain
SK51
(=
NCTC 3165)
VOL.
39,
1989
VIRIDANS STREPTOCOCCI
481
was found to possess antigens a, c, and d
(7).
The latter
finding explains the reaction of
S.
gordonii
biovar 3 strains
with ATCC 10557 antiserum
(S.
ora1isl“S. mitior”)
as strain
ATCC 10557 has been found to possess antigens c and d
(7).
According to Rosan (40), antigen a is the group H antigen.
Thus, it appears that all
S.
gordonii
biovars possess the
group
H
antigen, although only biovar 2 strains reacted with
the Blackburn serum.
S.
oralis.
S.
oralis
is the name applied by Bridge and
Sneath (2) to one of their clusters resulting from a numerical
taxonomic study of streptococci (3). As defined by these
authors, this taxon appears to be heterogeneous because
representative strains from their study were assigned in this
study to widely different species, including
S.
oralis,
S.
sanguis,
S.
gordonii,
and
S.
anginosus.
The taxon resulting
from our study, which included the type strain of
S.
oralis,
was characterized by the production of extracellular
polysaccharide by most strains, by neuraminidase and IgAl
protease activities, and by a lack of arginine and esculin
hydrolysis and inulin fermentation; 9
of
18 strains, including
the type strain, reacted with the antiserum against ATCC
10557, which is consistent with the fact that this strain and
several other strains previously referred to as
“S.
mitior”
or
“S. sanguis
11”
were included in the species. DNA-DNA
hybridization data for six of the strains which we included in
S.
oralis
(see above) confirm that this species
is
genetically
homogeneous and well separated from other species, includ-
ing
S.
mitis
(Fig.
1).
Kilpper-Balz et al.
(30)
have published an emended de-
scription of
S.
oralis
based on data resulting from their own
study and data reported by Bridge and Sneath (2). However,
there are several significant discrepancies between the re-
sults of these authors and our results. Thus, both the type
strain of
S.
oralis,
strain NCTC 11427
(=
SK23),
and
S.
oralis
DSM 20066
(=
SK225) were found to be positive for
arginine hydrolysis and acetoin production, and strain DSM
20066 was
also
found to be positive for esculin hydrolysis by
Kilpper-Balz et al.
(30).
We were unable to confirm these
data, which results in
S.
oralis
being phenotypically more
homogeneous than reported by Kilpper-Balz et al. (30).
Furthermore, we found that none of the strains exhibits
lysine decarboxylase activity as previously reported for
S.
oralis
(3).
On the basis of our results an emended description
of
S.
oralis
is presented below.
S.
rnitis.
The name
S.
mitis
was used by Andrewes and
Horder
(1)
to designate streptococci which are rarely patho-
genic, grow in short chains, and usually fail to ferment inulin
and rafiinose. Because of the lack of sufficient detail in this
description, the definition
of
S.
mitis
is usually based on the
description provided by Sherman et al. (42). Although, these
authors noted that
“S.
mitis
forms a rather heterogeneous
and perhaps complex group,” they found that it was well
separated from other viridans species by the inability of any
culture out of 147 studied to ferment inulin or to synthesize
polysaccharide from sucrose. That the
S.
mitis
group may
contain more than one distinct unit was indicated by the fact
that an appreciable proportion of the strains of this group
hydrolyzed arginine with the production of ammonia (42).
These characteristics concur with the definition of
S.
mitis
used by Carlsson (3, Facklam (17), and workers in several
other laboratories.
The use of the name
S.
mitis
for the two clusters defined in
Table
3
is in complete agreement with the definitions used by
Sherman et al. (42), Carlsson (3, and Facklam (17). How-
ever, the formal acceptability of this name is pending the
response of the Judicial Commission to our request to reject
the present type strain of
S.
mitis,
strain NCTC 3165
(=
SK51)
(28);
we propose to replace this strain with strain
NCTC 12261.
S.
oralis
and
S.
mitis,
as defined here, are related by
having a cell wall composition that is distinct from that of
S.
sanguis.
In contrast to
S.
sanguis,
the cell walls of
S.
oralis
and
S.
mitis
(biovars
1
and 2) lack glycerol teichoic acid and
significant amounts of rhamnose and contain a ribitol tei-
choic acid (10, 11,33,39). Accordingly,
S.
oralis
and
S.
mitis
were grouped together as
“S.
mitior”
by Colman and
Williams (11); however, the subsequent study of Coykendall
and Munzenmaier (12), which included several strains which
we classified as
S.
oralis
and
S.
mitis,
respectively, demon-
strated that these two species exhibit only
40
to
60%
DNA
homology. As demonstrated in Table
3,
the two species are
also well separated by biochemical and serological differ-
ences and furthermore have different ecological properties
(Kilian, manuscript in preparation).
As
the study of Coyk-
endall and Munzenmaier (12) did not contain representatives
of
S.
mitis
biovar 2, further studies will be required to define
definitively the correct affiliation
of
this taxon.
S.
anginosus.
The taxon that contained the type strain of
S.
anginosus
also encompassed the type strains of
S.
interme-
dius
and
S. constellatus
and several strains labeled
“S.
milleri.”
The 24 strains formed a phenotypically heteroge-
neous group, and no natural subdivisions were possible on
the basis of the data. This finding confirms conclusions from
several previous studies that many strains of
S.
anginusus,
S.
constellatus,
and
S.
intermedius
are closely related to
each other genetically in spite of phenotypic heterogeneity
(14,
31, 46). Therefore, we followed the recent proposal by
Coykendall et al. (14) to use the designation
S.
anginosus
for
the whole group of bacteria because of the priority of that
name. In the present study
S.
anginosus
could be separated
from
S.
sanguis
by a lack of extracellular polysaccharide and
IgAl protease activity, by differences in colonial morphol-
ogy, by the inability to ferment inulin, by a positive Voges-
Proskauer reaction in most strains, and by serological reac-
tivity (Table 3). Clearly, further studies will be required to
explain the taxonomically unnatural situation surrounding
this group of bacteria.
S.
sulivurius.
The taxon that contained the type strain of
S.
salivarius
formed a relatively homogeneous group which
was well separated from other species. The
10
strains
possessed the typical characteristics of
S.
salivarius,
includ-
ing extracellular levan production. Thus, none of the strains
appears to be related to the newly described species
S.
vestibularis
(48).
S.
mutans.
The 10 strains assigned to
S.
mutans
included
the type strain and pose no taxonomic problems. Strains of
other mutans species were not included in the study.
Table
4
lists the phenotypic characteristics that are most
useful for distinguishing the streptococcal species and their
biovars as defined here. The characteristics included in this
table are those that were found to be constant and most
reproducible in tests.
Table 5 correlates the classification presented here with
designations used previously in the literature. This table is
based on previously published information concerning
refer-
ence strains common to the individual studies.
Description
of
Streptococcus
gordonii
sp. nov.
Cells are
gram-positive cocci that grow in short chains in serum broth.
They are nonmotile, nonsporing, aerobic and facultatively
anaerobic, fermentative, and catalase negative and show
a-hemolysis on horse blood agar and pronounced greening
on chocolate agar. The cell wall contains glycerol teichoic
482
KILIAN
ET
AL.
TABLE
4.
Differential characteristics
of
the species
INT.
J.
SYST. BACTERIOL.
S.
sanguis
S.
gordonii
S.
mitis
S.
S.
S.
Biovar Biovar Biovar Biovar Biovar Biovar Biovar
s*
Oralis
Biovar Biovar
salivarius anginosus mutans
1
2
3
4
1
2
3
1
2
+
-
+
+
+
+
+
+
-
d d
+
-
-
-
Arginine hydrolysis
+a
+
+
+
+
+ +
Esculin hydrolysis
+
-
+
+ +
+ +
Extracellular poly-
+
+ + +
+ +
d d
saccharide
Acetoin (VP)
Acid produced
- -
-
-
-
-
-
- - - -
- -
- -
from:
+
d
d
-
d
+
+
+
Am ygdalin
- -
Arbutin d
+ + + +
+
+
Melibiose
+
-
+b
d
+
d
+
+
d
Inulin
+
+ +
d
+
d
+
Mannitol
Sorbitol
RaEinose
+
-
-
-
+
+
+
+
+
+
-
d d
+
d
+
-
- -
-
- - -
-
d
-
-
-
-
- - - -
- - -
-
- -
-
-
- -
-
-
- - - -
-
d
+
d
d
- -
+
+
+
+
+
+
+
+
- -
-
d
- -
P-Glucosaminidase“
- -
-
-
a-L-Fucosidased
-
-
-
-
Acid phosphatase“
+
-
-
+
+ + +
d d
+
Sb
d
+b
Alkaline phospha-
db
-
-
-
+
+
+
+
d
d
+
+
IgAl protease
+
+
+ +
- -
-
-
-
-
-
tasec
-
-
-
-
+
d
+
d
-
-
-
-
-
- - - - -
-
- -
-
Neuraminidase
a
+,
90%
or more of the strains are positive;
-,
90%
or more of the strains are negative; d,
11
to
89%
of the strains are positive.
Reactions
are
usually
weak.
Determined with an
API
ZYM
kit.
Determined with an
API
ZYM
osidase kit.
acid and rhamnose
(lo),
and the peptidoglycan type is
Lys-Ala,_, (30,41). Biochemical, physiological, and serolog-
ical characteristics of the species and three biovars are
shown in Table 3. The G+C content of the DNA is 40 to 43
mol% (13).
S.
gordonii
resembles
S.
sanguis
but can be
differentiated from this species by several biochemical char-
acteristics (Tables 3 and 4), by a lack of IgAl protease, and
by a significantly different G+C content. Found in human
oral cavities and pharynges. The type strain is strain ATCC
10558
(=
NCTC 7865). The reactions of the type strain are
the majority reactions for
S.
gordonii
biovar
2
given in Table
3. The G+C content
of
the type strain is 41 mol%.
Emended description
of
Streptococcus
sanguis
White and
Niven
1946,722.
Cells are gram positive and usually grow in
short chains in serum broth. They are nonmotile, nonspor-
ing, aerobic and facultatively anaerobic, fermentative, and
catalase negative and show a-hemolysis on horse blood agar
and pronounced greening on chocolate agar. The cell wall
contains glycerol teichoic acid and rhamnose
(lo),
and the
peptidoglycan type is Lys-Ala,-, (30,
41).
Biochemical,
physiological, and serological characteristics of the species
and its biovars are shown in Table
3.
Found in human
mouths. The type strain is strain ATCC 10556
(=
NCTC
7863). The reactions of the type strain are the majority
reactions for
S.
sanguis
biovar
1
given in Table 3, with the
exceptions indicated. The G+C content
of
the type strain is
46 mol% (13).
Emended description
of
Streptococcus
oralis
Bridge and
Sneath
1982, 414.
Cells are gram-positive cocci and usually
grow in long chains in serum broth. They are nonmotile,
nonsporing, aerobic and facultatively anaerobic, fermenta-
tive, and catalase negative and show a-hemolysis on horse
blood agar and pronounced greening on chocolate agar. The
cell wall contains ribitol teichoic acid and lacks significant
amounts of rhamnose
(10).
The peptidoglycan type is Lys-
direct (30, 41). Biochemical, physiological, and serological
characteristics of the species are shown in Table 3. Contrary
to information provided in previous descriptions
(2,
30),
S.
oralis
does not have lysine decarboxylase activity, does not
hydrolyze arginine or esculin, and is negative in the Voges-
TABLE
5.
Correlation between the classification resulting from this study and designations used previously
Designation used
by:
Hamada Kilpper-Bdz
et al.
(23)
et al.
(30)
Facklam
(17)
in this study Sherman Carlsson Colman and Coykendall and
et
al.
(41)a
(5)
Williams
(11)
Specht
(13)
~~~~~ ~~~ ~ ~
S.
sanguis
Type
1
I:B
S.
sanguis
Genotype
3, “S.
sanguis
S.
sanguis
I Biotype A,
S.
sanguis
S.
gordonii
Type 1-11
S.
sanguis
Genotype
1,
“S.
sanguis
S.
sanguis
I Biotype A,
S.
mitis
S.
oralis
Type 11 I:A “S.
mitior”
Genotype
2,
“S.
mitior”
S.
sanguis
I1 Biotype B,
S.
oralis
S.
mitis
biovar
1
V:A
“S.
mitior”
S.
mitis
S.
mitis
biovar
2
V:B
“S.
mitior”
subsp.
carlssonii’
’
serotype I
subsp.
sanguis”
serotype I11
serotype I1
a
The numbers in parentheses are reference numbers.
VOL.
39, 1989
VIRIDANS STREPTOCOCCI
483
Proskauer reaction (acetoin). The G+C content of the DNA
is
39
to
42
mol%
(12).
Found in human oral cavities. The type
strain
is
strain NCTC
11427.
The characteristics of the type
strain are the majority reactions for
S.
oralis
given in Table
3,
with the exceptions indicated. The G+C content of the
type strain is
41
mol%.
Emended description
of
Weptucuccus
mitis
Andrewes and
Horder
1906,712.
Cells are gram-positive cocci and grow in
short or long chains in serum broth. They are nonsporing,
aerobic and facultatively anaerobic, nonmotile, fermenta-
tive, and catalase negative and show a-hemolysis on horse
blood agar and pronounced greening on chocolate agar. The
cell wall contains ribitol teichoic acid and lacks significant
amounts
of
rhamnose
(10).
The peptidoglycan type
is
Lys-
direct
(30,
41).
Biochemical, physiological, and serological
characteristics of the species and
two
biovars are shown in
Table
3.
The G+C content of the DNA is
39
to
41
mol%
(12).
Found in human mouths and pharynges. The type strain is
strain NCTC
12261.
The reactions of the type strain are the
majority reactions for
S.
mitis
biovar
1
given in Table
3.
The
G+C content of the type strain is
41
mol%.
ACKNOWLEDGMENTS
The technical help provided by Kirsten Holmgren and Ella Brandt
and the secretarial assistance of Dorthe Eggertsen are gratefully
acknowledged. Thanks are also due to Bente Nyvad for a critical
review of the manuscript, to Jesper Reinholdt for help with one of
the assays, and to several colleagues (see Table
1)
for providing
some of the strains used in the study.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
LITERATURE CITED
Andrewes, F. W., and T.
J.
Horder.
1906.
A study of the
streptococci pathogenic for man. Lancet
ii:708-713.
Bridge,
P.
D., and
P.
H.
A.
Sneath.
1982.
Streptococcus galli-
narum
sp. nov. and Streptococcus oralis sp. nov. Int.
J.
Syst.
Bacteriol.
32:410415.
Bridge,
P.
D., and
P.
H.
A.
Sneath.
1983.
Numerical taxonomy
of
Streptococcus.
J.
Gen. Microbiol.
129565-597.
Carlsson,
J.
1967.
Presence of various types
of
nonhaemolytic
streptococci in dental plaque and in other sites of the oral cavity
in man. Odontol. Revy
18:55-74.
Carlsson,
J.
1968.
A numerical taxonomic study
of
human oral
streptococci. Odontol. Revy
19:137-160.
Carlsson,
J.
1970.
Chemically defined medium for growth of
Streptococcus sanguis. Caries Res.
4:297-304.
Cole,
R.
M., G. B. Calandra, E. Huff, and
K.
M. Nugent.
1976.
Attributes of potential utility in differentiating among “group
H”
streptococci
or
Srreptococcus sanguis. J. Dent. Res.
SS(Specia1 Issue
A):A142-A153.
Colman, G.
1968.
The application of computers to the classifi-
cation of streptococci.
J.
Gen. Microbiol.
50149-158.
Colman, G., and
L.
C. Ball.
1984.
Identification
of
streptococci
in a medical laboratory. J. Appl. Bacteriol.
57:l-14.
Colman, G.,
and
R.
E.
0.
Williams.
1965.
The cell walls of
streptococci. J. Gen. Microbiol.
41:375-387.
Colman, G., and R.
E.
0.
Williams.
1972.
Taxonomy of some
human viridans streptococci, p.
282-299.
In L. W. Wannamaker
and
J.
M. Matsen (ed.), Streptococci and streptococcal dis-
eases. Academic Press, Inc., New York.
Coykendall, A. L., and
A.
J.
Munzenmaier.
1978.
Deoxyribonu-
cleic acid base sequence studies on glucan-producing and glu-
can-negative strains
of
Streptococcus mitior. Int. J. Syst. Bac-
teriol.
2851 1-515.
Coykendall,
A.
L.,
and
P.
A.
Specht.
1975.
DNA base sequence
homologies among strains
of
Streptococcus sanguis. J. Gen.
Microbiol.
91:92-98.
Coykendall,
A.
L.,
P.
M.
Wesbecher, and
K.
B. Gustafson.
1987.
“Streptococcus milleri,” Streptococcus constellatus, and
Streptococcus intermedius are later synonyms of Streptococcus
anginosus.
Int.
J.
Syst. Bacteriol.
37:222-228.
15.
Drucker,
D.
B., and T. H. Melville.
1971.
The classification of
some oral streptococci
of human and rat origin. Arch. Oral Biol.
16:845-853.
16.
Edwardsson,
S.
1968.
Characteristics of canes-inducing strepto-
cocci resembling
Streptococcus mutans. Arch. Oral Biol.
13:
637-646.
17.
Facklam,
R. R.
1977.
Physiological differentiation of viridans
streptococci. J. Clin. Microbiol.
5184201.
18.
Facklam,
R. R.
1984.
The major differences in the American and
British streptococcus taxonomy schemes with special reference
to
Streptococcus milleri. Eur. J. Clin. Microbiol.
3:91-93.
19.
Gibbons,
R.
J., and
S.
B. Banghart.
1967.
Synthesis of extracel-
lular dextran by cariogenic bacteria and its presence in human
dental plaque. Arch. Oral Biol.
12:ll-24.
20.
Gilmour, M.
N.,
T.
S.
Whittam, M. Kilian, and R.
K.
Selander.
1987.
Genetic relationships among the oral streptococci.
J.
Bac teriol.
16952474257.
21.
Gordon, M. H.
1905.
A ready method of differentiating strepto-
cocci and some results already obtained by its application.
Lancet
i:
1400-1403.
22.
Guggenheim, B.
1968.
Streptococci
of
dental plaques. Caries
Res.
2:147-163.
23.
Hamada,
S.,
M. Torii,
Y.
Tsuchitani, and
S.
Kotani.
1980.
Isolation and immunobiological classification of Streptococcus
sanguis
from human tooth surfaces. J. Clin. Microbiol.
12:
809-817.
24.
Hardie, J.
M.
1986.
Oral streptococci,
p.
1054-1063.
In
P. H.
A.
Sneath, N.
S.
Mair, M.
E.
Sharpe, and
J.
G. Holt (ed.),
Bergey’s manual of systematic bacteriology, vol.
2.
The
Williams
&
Wilkins
Co.,
Baltimore.
25.
Hardie,
J.
M.,
R.
A.
Whiley, and
M.
J.
Sakin.
1982.
A numerical
taxonomic study of oral streptococci, p.
59-60.
In
S.
E.
Holm
and
P.
Christensen (ed.), Basic concepts of streptococci and
streptococcal diseases. Reedbooks, Ltd., Surrey, United King-
dom.
26.
Hehre, E.
J.,
and
J.
M. Neill.
1946.
Formation of serologically
reactive dextran by streptococci from subacute bacterial en-
docarditis. J. Exp. Med.
83:147-162.
27.
Kilian, M., and
K.
Holmgren.
1981.
Ecology and nature of
immunoglobulin
A1
protease-producing streptococci in the hu-
man oral cavity and pharynx. Infect. Immun.
31:868-873.
28.
Kilian, M., L. Mikkelsen, and
J.
Henrichsen.
1989.
Replacement
of the type strain of
Streptococcus mitis. Request for an Opinion
under Rule
18g.
Int. J. Syst. Bacteriol.
39:498-499.
29.
Kilian,
M.,
B.
Nyvad, and L. Mikkelsen.
1986.
Taxonomic and
ecological aspects of some oral streptococci, p.
391-400.
In
S.
Hamada,
S.
M. Michalek,
H.
Kiyono, L. Menaker, and J. R.
McGhee (ed.), Molecular microbiology and immunobiology of
Streptococcus mutans. Elsevier Science Publishing Co., Inc.,
New York.
30.
Kilpper-Biilz,
R.,
P. Wenzig, and
K.
H. Schleifer.
1985.
Molec-
ular relationships and classification
of some viridans strepto-
cocci as
Streptococcus oralis and emended description of Strep-
tococcus oralis
(Bridge and Sneath
1982).
Int.
J.
Syst. Bacteriol.
35482488.
31.
Kilpper-Balz,
R.,
B.
L.
Williams,
R.
Lutticken, and
K.
H.
Schleifer.
1984.
Relatedness of “Streptococcus milleri” with
Streptococcus anginosus and Streptococcus constellatus. Syst.
Appl. Microbiol.
5494500.
32.
Lautrop, H.
1960.
Laboratory diagnosis of whooping-cough
or
Bordetella infections. Bull. W .H.O.
23:15-31.
33.
Mikkelsen, L.,
S.
B. Jensen, and
J.
Jakobsen.
1981.
Microbial
studies on plaque from carious and caries-free proximal tooth
surfaces in a population with high caries experience. Caries Res.
15428435.
34.
Mizuno, J.,
M.
Torii, T. Yamamoto, and
S.
Hamada.
1983.
Detection of amphipathic antigen unique to biotype B Strepto-
coccus sanguis.
FEMS Microbiol. Lett.
16:209-212.
35.
Mfiller,
V.
1955.
Simplified tests for some amino acid decarbox-
ylases and for the arginine dihydrolase system. Acta Pathol.
Microbiol. Scand.
36:158-172.
484 KILIAN ET
AL.
INT.
J.
SYST.
BACTERIOL.
36.
Muller,
H.
E.
1984. ABTS perioxidase medium as a highly
sensitive plate assay
for
detection
of
hydrogen peroxide produc-
tion in bacteria.
J.
Microbiol. Methods 2:lOl-102.
37.
Nyvad, B., and
M.
Kilian.
1987. Microbiology
of
the early
colonization
of
human enamel and root surfaces in vivo. Scand.
J.
Dent. Res. 95:369-380.
38.
Ottens, H., and
K.
R.
Winkler.
1962. Indifferent and hemolytic
streptococci possessing group antigen
F.
J.
Gen. Microbiol.
39.
Rosan, B.
1978. Absence of glycerol teichoic acids in certain
oral streptococci. Science 201:918-920.
40.
Rosan,
B.
1983. Antigens
of
Streptococcus sanguis. Infect.
Immun. 7:205-211.
41.
Schmidhuber,
S.,
R.
Kilpper-Bdz, and
K.
H. Schleifer.
1987. A
taxonomic study of
Streptococcus mitis,
S.
oralis,
and
S.
sanguis. Syst. Appl. Microbiol. 10:74-77.
42.
Sherman, J. M.,
C.
F.
Niven, Jr., and
K.
L.
Smiley.
1943.
Streptococcus salivarius and other non-hemolytic streptococci
of the human throat.
J.
Bacteriol. 45249-263.
43.
Skerman,
V.
B.
D.,
V.
McGowan, and
P.
H.
A.
Sneath
(ed.).
1980. Approved lists
of
bacterial names. Int.
J.
Syst. Bacteriol.
30:225420.
44.
Varki,
A.,
and
S.
Diaz.
1983. A neuraminidase from Streptococ-
28:181-191.
cus sanguis that can release 0-acetylated sialic acids.
J.
Biol.
Chem. 258:12465-12471.
45.
Waitkins,
S.
A.,
L.
C.
Ball,
and
C.
A.
M. Fraser.
1980. Use
of
the API-ZYM system in rapid identification of
a
and non-
haemolytic streptococci.
J.
Clin. Pathol. 3353-57.
46.
Washburn,
M.
R.,
J.
C.
White, and
C.
F. Niven, Jr.
1946.
Streptococcus s.b.e.: immunological characteristics.
J.
Bacte-
rial.
51:723-729.
47.
Welborn,
P.
P.,
W.
K.
Hadley,
E.
Newbrun, and D.
M.
Yajko.
1983. Characterization
of
strains of viridans streptococci by
deoxyribonucleic acid hybridization and physiological tests. Int.
J.
Syst. Bacteriol. 33:293-299.
48.
Whiley,
R.
A.,
and
J.
M. Hardie.
1988. Streptococcus vestibu-
laris
sp. nov. from the human oral cavity. Int.
J.
Syst. Bacteriol.
49.
Whiley,
R.
A.,
J.
M. Hardie, and
P.
J.
H. Jachman.
1982.
SDS-polyacrylamide gel electrophoresis of oral streptococci, p.
61-62.
In
S.
E.
Holm and
P.
Christensen (ed.), Basic concepts
of
streptococci and streptococcal diseases. Reedbooks, Ltd.,
Surrey, United Kingdom.
50.
White, J.
C.,
and
C.
F. Niven, Jr.
1946. Streptococcus S.B.E.:
a
streptococcus associated with subacute bacterial endocarditis.
J.
Bacteriol. 51:717-722.
38:
3 35-339.
