INTERNATIONAL
JOURNAL
OF SYSTEMATIC
BACTERIOLOGY,
July 1987, p. 222-228
0020-7713/87/030222-07$02.OO/O
Copyright 0 1987, International Union of Microbiological Societies
Vol. 37, No. 3
“Streptococcus milleri,” Streptococcus constellutus, and
Streptococcus intermedius Are Later Synonyms of
Streptococcus anginosus
ALAN L. COYKENDALL,” PAMELA M. WESBECHER, AND KATHLEEN B. GUSTAFSON
Department of Oral Diagnosis, School of Dental Medicine, University of Connecticut Health Center,
Farmington, Connecticut 06032
Streptococci that produced acetoin and alkaline phosphatase, hydrolyzed arginine, and fermented trehalose
but did not produce pyrrolidonylarylamidase or P-glucuronidase, split hippurate, or ferment ribose or
glycogen were collected and compared, These streptococciwere considered members of the unapproved species
‘LStreptococcus millen’,” but most of them would fit the description of one of three approved species:
Streptococcus anginosus, Streptococcus constellatus, or Streptococcus intermedius. Most were recent clinical
isolates. Some hydrolyzed esculin and fermented lactose, while others did pot. Some fermented mannitol and
rafknose. Many were beta-hemolytic, and several reacted to antiserum of Lancefield group A, C, F, or G. From
a total of 111 strains, 40 were selected for comparison of their deoxyribbnucleic acid (DNA) base sequence
similarities by DNA-DNA hybridization on membrane Filters. All biotypes, hemolytic types, and serotypes were
included, as well as the type strains of S. anginosus, S. constellatus, and S . intermedius, and Lancefield group
F Streptococcus sp., plus two strains derived from GuthoPs ‘ ( S . millen’” isolates. The results showed
considerable genetic similarity within the group. DNA from most strains hybridized at a level of 70% or more
of the homologous control, even under very stringent conditions. There was somewhat less homology between
DNAs of some of the least reactive strains (lactose, mannitol, and esculin negative) and the most reactive strains
(lactose, mannitol, and esculin positive). The Lancefield F strain and the type strains of S. anginosus and S.
constellatus were genetically similar. Traits such as hemolysis and lactose fermentation could not be ascribed
to plasmids. The results support the unification of these streptococci into a single species.
Although it has no standing in taxofiomy at the present
time, the name “Streptococcus milleri” is used for a large
group of medically important streptococci that produce
acetoin and hydrolyze arginine but do not ferment ribdse and
are resistant to bacitracin. Many are hemolytic, and many
react with antiserum to Lancefield group A, C, F, or G.
Although originally described and named by Guthof (16), the
current understanding of the chdracteristics and importance
of this taxon derives from the studies of Colman (Ph.D.
thesis, University of London, London, England, 1970) and
Colman and Williams (6), who found that the Guthof
streptococci, the nonhemolytic streptococci of Lancefield
groups A, C, and G, and the hemolytic streptococci of
Lancefield groups F and G which produced minute colonies
were all more similar than different. In their extensive study
of the streptococci of human infections, Parker and Ball (34)
adopted this classification and also included all hemolytic
strains that were otherwise similar to “S. milleri” and all
group F isolates. Ball and Parker (3) elaborated on the
characteristics and medical importance of “ S . milleri” and
brought into the species strains that fermented mzinnitol and
raffinose. Liitticken et al. (24) agreed that streptococci with
the characteristics set forth by Colman should be considered
one species, “ S . milleri,” regardless of hemolysis.
Thus, “ S . milleri” became a species that included hemolytic and nonhemolytic members which could react with one
of four Lancefield antisera, or none. Facklam (12) also noted
biochemical similarities among several tyges of streptococci
which had been previously described. He found that the
so-called Streptococcus sp. strain MG (30) and the recognized species Streptococcus intermedius (18) are identical.
* Corresponding author.
Also, his reference Strains of Streptococcus anginosus (1, 9)
and Streptococcus constellatus (18) could not be differentiated from each other by phenotypic tests, and group F
streptococci were separable from the latter two only by their
hemolytic traits. Finally, most hemolytic streptococci that
had no Lancefield antigens were biochemically like S .
constellatus. Although Facklam acknowledged that Colman’s “ S . milleri” would embrace all these streptococci, he
recommended two “species” based on lactose fermentation:
“ S . anginosus-constelfatus” for those that did not ferment
lactose, and “ S . MG-intermedius” for those that did. Conversely, Moore et al. (31) recognized that S. intermedius was
similar to S . anginosus and considered S . intermedius a later
synonym of S . anginosus. The confusion created by the use
of so many names has been ohly slightly relieved by an
attempt to resolve the differences between use of “ S .
milleri” (mainly in Europe) and the Facklam scheme used in
North America (13). In a review of the history and pathogenic potential of group F and related streptococci, Shlaes et
al. (38) used S . milleri to denote lactose fermenters, and S.
anginosus for lactose negative and all group F strains.
However, both the current (9) and original (1) descriptions of
S. angiriosus state that this streptococcus ferments lactose.
The name “S. milleri” has been used in recent publications
(20, 37) that have focused on the identification and importance of these streptococci in the United States. Both Ruoff
et al. (37) and Lawrence et al. (22) agreed that this species
accounted for a large proportion of hemolytic streptococci of
groups C and G and all of group F.
Welborn et al. (41) tested the genetic similarity of some of
these bacteria and found a high degree of DNA base sequerice similarity among the type strains of S . intermedius
and S . constellatus, two group F strains, ATCC 9895 (Streptococcus sp. biotype MG), and the one “S. MG-intermedi-
222
Downloaded from www.microbiologyresearch.org
by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 14:23:54
SYNONYMS OF STREPTOCOCCUS ANGINOSUS
VOL. 37, 1987
us” strain tested. Farrow and Collins (15) found DNA
homology among S. anginosus, S. constellatus, S . intermedius, “S. milleri,” and group F. Recently, Ezaki et al.
(11) showed strong genetic relationships among the minute
hemolytic streptococci and S . anginosus. Kilpper-Balz et al.
(20) compared the DNAs of several strains that resembled
“S. milleri.” They found two genetic groups. DNA from one
group hybridized with DNA from the type strain of S.
constellatus, and DNA from members of the other group
hybridized with DNA from the type strain of S . anginosus.
Hybridization between the two groups was about 60% and
decreased under very stringent conditions. These results
supported the continued acceptance of these two species.
However, like Facklam, Kilpper-Balz et al. (20) could find
no reliable phenotypic traits by which S . constellatus could
be distinguished from S . anginosus. DNA from the type
strain of S . intermedius did not hybridize above 70% with
any other DNA.
Thus, the taxonomic relationships among these streptococci remain uncertain. In an effort to measure the level of
kinship within this group, we collected a wide variety of
streptococci resembling “S. miller?’ and assessed their
DNA base sequence similarities to each other and to the type
strains of S. anginosus, S. constellatus, and S. intermedius
by ONA-DNA hybridization.
MATERIALS AND METHODS
Bacteria. Most of the streptococci used in this study were
recent clinical isolates obtained from the clinical laboratory
of the University of Connecticut Health Center. They were
selected from streptococcal cultures that had been determined to be other than S . pyogenes, enterococci, pneumococci, or group B. These isolates were identified using the
tests described below, and those resembling “ S . miller?’
were included in the investigation. Such strains are designated by numbers preceded by UC. Other strains of “S.
milleri” were received from the Massachusetts General
Hospital (MGH), Boston, the Streptococcal Reference Laboratory of the Central Public Health Laboratory (COL),
London, England, and the Centers for Disease Control,
Atlanta, Ga. (CDC). Type strains and other reference strains
were purchased from the American Type Culture Collection
(ATCC), Rockville, Md., and the National Collection of
Type Cultures (NCTC), London, England. All cultures were
tested for purity by streaking on blood agar plates and by
Gram stain. Ail subsequent cultures that were to be used for
a series of biochemical tests or for extraction of DNA were
similarly examined. Stock cultures wete maintained i r ~ToddHewitt broth with 1% extra glucose and CaC03. Duplicates
were frozen and stored at -30°C.
Phenotypic characterization. The API Rapid Strep system
was used for most biochemical tests. This system has been
described and evaluated before (2, 14, 36, 40). In our
laboratory, this system was found to identify reliably
streptococci whose identity had been authenticated previously by DNA-DNA hybridization. In addition, the following conventional tests were used: hemolysis on casein soy
agar with 5% sheep blood, esculin hydrolysis (4), fermentation of salicin, production of hydrogen peroxide (42), and
colony morphology on mitis salivarius agar (5). Most strains
were also tested for growth in 6.5% NaCI, growth on
bile-esculin agar, and production of urease (27). To confirm
the reliability of the Rapid Strep tests, a group of 26 strains
was tested for acid production in cysteine Trypticase agar
(CTA; BBL Microbiology Systems, Cockeysville, Md .) with
223
TABLE 1. Distribution of serotypes arid biotypes of “S. milleri””
Lancefield
group’
A
C
F
G
None
Total
No. of strains‘ in biotype:
1
2
P
-P
P
-P
2(2)1”
0
0
0
0
4(0)0
0
0
6(3)1
15(1)1
17(5)3
1(1)1 13(4)4 l(0)O
0
0
4(4)2
2(2)2
l(1)O
0
7(2)2 2(1)1
9(5)4
0
0
31(10)6 2(1)1 18(4)4 28(12)9 5f4)2
3
-p
Total
0
0
5(3)1
0
8(3)3
13(614
2
25
37
7
26
97
The biotypes were as determined by the API Rapid Strep system. p means
beta-hemolytic, and -p means not beta-hemolytic.
Lancefield group as determined by Streptex.
The first number is the total number of strains in the collection with that
set of characteristics. The number of strains whose DNAs were hybridized is
in parentheses, and the third number represents strains that were examined
for plasmids.
ti
’
1% trehalose, raffinose, mannitol, lactose, and ribose plus
CTA with no carbohydrate as a control; the production of
acetoin (27); and the production of ammonia from arginine
(32). Except for the H202 tests and the CTA cultures, all
incubations were in candle jars or in a C 0 2 incubator, For
stock cultures, and to generate the lawns for the API tests,
cultures were incubated in anaerobic jars. The API fermentation tests were under oil.
The presence of Lancefield antigens A, C, D, F, and G
was determined by a commercial latex bead agglutination
test (Streptex; Wellcome Reagents Ltd., Research Triangle
Park, N.C.).
Preparation of DNA. Bacteria were grown and lysed, and
the DNA was isolated as described previously (8). Briefly,
the cells were harvested from cultures grown in 1 liter of
Todd-Hewitt broth with 1% extra glucose. For tritiumlabeled DNA, 1 mCi of [methyl-3H]thymidine was added.
Penicillin was added to the cultures while they were still in
the log phase of growth. Washed, harvested cells were
treated with lysozyme and finally lysed with sodium dodecyl
sulfate. The DNA was purified by deproteinizations with
phenol and chloroform, treatment with ribonuclease, and
precipitations with ethanol and isopropanol (28).
DNA hybridizatiods. Base sequence similarities among the
DNAs of 40 representative strains were assessed by hybridization on membrane filters (type HAHY; Millipore Corp.,
Bedford, Mass.) in 0.3 M NaC1-0.03 M sodium citrate in 30%
(vol/vol) dimethyl sulfoxide (23) as we have described earlier
(8). Hybridizations were carried out at two temperatures, 48
and 57°C. For DNA of about 40 mol% guanine plus cytosine
(G+C), 48°C is the optimum temperature according to the
formula of DeLey et al. (10). They showed that hybrid
duplexes formed at the optimum temperature are stable, and
that higher incubation temperatures did not result ih lower
relative hybridization among various DNAs (10). Nevertheless, we wished tb test the fidelity of hybrids formed at 48°C
by repeatirig some experiments at 57°C.
Base composition of DNA. The DNA base compositions of
47 strains were determined by thermal denaturation in 0.15
M NaC1-0.015 M sodium citrate (29). The temperature of the
samples was raised O.S”C/min while the absorbance was
recorded continually. One to three determinations were
made for each straid. DNA from Escherichia coli K-12 with
a base composition of 51.5 mol% G + C was used as a
standard.
Plasmids. To determine whether characteristics such as
hemolysis and lactose fermentation were conferred by plas-
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 14:23:54
:224
INT. J . SYST.BACTERIOL.
COYKENDALL ET AL.
mids, representative strains were examined for the presence
(ofcovalently closed circular DNA (Table 1). Cells grown in
10 ml of Todd-Hewitt broth were lysed as described above.
Plasmid DNA was concentrated by removing chromosomal
DNA with 5 M potassium acetate and precipitation of
circular DNA by ethanol after phenol-chloroform
deproteinization (27). Electrophoresis of this material was
performed through 0.7% agarose gel in 40 mM Tris acetate
buffer pH 8.1 for 16 h at 2 V/cm. Plasmids from E. coli V517
>wereused as molecular weight standards (26). The gels were
:stainedwith ethidium bromide and photographed.
RESULTS
From 256 clinical isolates, 76 were identified as having
characteristics of “S. milleri.” An additional 35 strains were
received from the laboratories identified above. These
istreptococci produced acetoin, hydrolyzed arginine, and
]fermented trehalose. They did not hydrolyze hippurate or
ferment ribose, sorbitol, or inulin. They produced alkaline
phosphatase, but not p-glucuronidase or pyrrolidonylarylamidase. On mitis salivarius agar (9,colonies were small
and peaked or convex without visible signs of extracellular
polysaccharide production. The great majority of our strains
could be assigned to one of the three “ S . milleri” biotypes
recognized by the Rapid Strep scheme which identifies
istreptococci to species and biotypes based on the numerical
:system of Lapage et al. (21). Typical biotype 1 strains do not
:split esculin or ferment lactose and are usually hemolytic.
‘Variantsthat ferment lactose are included. Biotype 2 strains
Jferment lactose, split esculin, and often acidify starch.
Hemolysis is variable. Biotype 3 strains are like those of
biotype 2 with the addition of mannitol and raffinose fermen1,ation. Many variations can exist within a biotype, and the
borders between biotypes are indistinct. Table 1 shows the
numbers of each biotype, serotype, and hemolytic type in
our collection. This table includes 97 strains that were fully
characterized; 14 others either were not fully characterized
(mainly lacking serotyping) or had doubtful identities. Most
biotype 1 strains were hemolytic and bore a Lancefield
antigen, usually F. Of these strains, 41% had the typical
biotype 1 characteristics. The most common departure was
fermentation of lactose in 22%. Less than half of our biotype
2 strains were hemolytic, but the majority had a Lancefield
antigen. Nonhemolytic isolates tended to be group F, and
the hemolytic ones often were group C. These streptococci
could be called S. intermedius. The typical biochemical
pattern with hemolysis was found in 28% of these strains,
and another 16% were not hemolytic. Strains that were both
nonhemolytic and starch negative accounted for 22% of
biotype 2 strains. Most biotype 3 strains (72%) had the
typical pattern and were not hemolytic. Lack of mannitol
fermentation together with hemolysis was the only variation
seen in more than one strain (3 or 17%). Four were hemolytic
and group G . Several nonhemolytic strains were group F.
13iotype 3 corresponds to the “S. milleri” isolates with the
“wide fermentation patterns” noted by Ball and Parker (3)
and the “urine isolates” described by Ruoff et al. (33).
(herall, there were 26 biochemical profiles observed within
the 111 “S. miller?’ strains; 11 profiles occurred only once.
In the original description of S. constellatus and S .
intermedius (18), aod in other descriptions (12, 38), the
fermentation of lactose is an important differentiating test.
Also, Facklam’s recent scheme (13) for the identification of
the “S. milleri group” implies that all of these streptococci
hydrolyze esculin. However, all combinations of lactose and
esculin reactions were represented in our collection. The
most common pattern was lactose and esculin positive (57
strains, of which 25 were hemolytic), but there were 19
strains that reacted with neither lactose nor esculin (16 were
hemolytic). An additional 18 isolates were esculin negative
and lactose positive. Of 37 esculin-negative strains, 30 were
hemolytic.
There was excellent agreement between the Rapid Strep
test results and the results of conventional tests, except for
the hydrolysis of esculin. While there was agreement in most
cases, our esculin test was sometimes positive after four
days incubation when the commercial product gave a negative result. (In the Rapid Strep test the 24 h result is
recorded). In the results shown in the tables and text the
Rapid Strep result is used.
Only six isolates produced H202. Three were biotype 2
and later proved to be related to other “ S . milleri” strains. A
Lancefield group F strain which was identified as S . sanguis
also produced H202, which is typical for the species (17).
The remaining two (strains UCAbb and NCTC 11062) were
not closely related to other strains (see below).
Half of the strains were beta-hemolytic; alpha-hemolysis
was rare. The hemolytic colonies produced zones of clearing
that were very large relative to the size of the colonies. It
was not unusual to see hemolysis before the colony was
visible (9). S. intermedius 27335Twas biochemically like “S.
milleri” biotype 2 and was not hemolytic. However, this
strain repeatedly gave weak reactions with both group F and
D antisera in Streptex latex bead agglutination tests. This
streptococcus was examined by K. Ruoff (MGH, Boston).
Antigens were extracted by pronase, albus enzyme, and hot
HC1. No antigens were detected in precipitin tests, but an F
reaction and no D reaction were observed in latex bead tests.
S. constellatus 27823T split esculin but did not ferment
lactose or starch. It was not hemolytic and had no group
antigen. S. anginosus 33397T was hemolytic and reacted
with the group G antibody. It hydrolyzed esculin and fermented lactose and raffinose, but not starch or mannitol.
This was the only strain in our collection with this reaction
pattern. However, it does fit the original description of S.
anginosus by Andrewes and Horder (1). The group F reference strain ATCC 12393 (Lancefield no. C628) was a typical
biotype 1strain, but the F antigen could not be demonstrated
by the Streptex agglutination test. Ottens and Winkler (33)
also failed to detect the F antigen on this strain. At the
ATCC, the F antigen could be demonstrated by counterimmunoelectrophoresis, although not by Streptex (J. Scully,
ATCC, personal communication).
DNA hybridization. The results of hybridization experiments are listed in Table 2. They are arranged in order of
decreasing hybridization in experiment 1, in which the
radioactively labeled DNA was from the biotype 1 hemolytic
group F strain UC4483. This DNA shared greater than 74%
base sequence homology with DNA from the eight other
biotype 1 strains tested. In general, the biotype 2 and 3
strains tested showed a lower level of hybridization (41 to
79%). DNA from three of eight biotype 3 strains tested
hybridized at less than 60%. These results were consistent in
the small experiment 2. These levels of hybridization were
nearly the same when DNA of 11 “S. milleri” strains were
hybridized at 57°C (data not shown). In contrast, hybridization of the group F S . sanguis strain UC4989 was halved
under these very stringent conditions (from 31 to 15%).
Results of experiments 3 and 4 show that labeled DNA
from the biotype 2 strain UC2314 hybridized well with DNA
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 14:23:54
SYNONYMS OF STREPTOCOCCUS ANGINOSUS
VOL. 37, 1987
225
TABLE 2. Hybridization among strains resembling “ S . rnillevi””
% Hybridization with labeled DNA from strains in expt no.:
Source of unlabeled
DNA
Biotypeb
UC4483
1
UC4483
UC8499
Group F(?) reference
ATCC 12393
UC2953
UC0398
UC5246
COLR85/2096
MG H-A 1784
MGH-G14
COLR85/2 190
MGH-NGR5
S . mginosus
ATCC 33397
MGH-NGR3
MG H 144
MGH611
UC2314
S . constellatus
ATCC 27823T
MGH580
UC2326
UC2305
UC7793B
U C0921B
S . intivtnedius
ATCC 27335T
MGH485
UC3133
UC7895
MGH-G17
UC6042
UC2013
UC7793A
UC4912
CO LR85/2210
UC9017
MGH455
UC3370
“ S . tnilleri” reference
NCTC 10708
NCTC 10709
NCTC 11062
UC2232
UCAbb
S.sctnguis
UC4989
E. coli
UC2314
MGH611
2
3
4
5
S . constellutus
ATCC 27823”
6
lPF--
100
100
64
71
58
102
1pF- lpF+1pF--
92
105
91
108
69
93
71
99
99
111
85
l-pn--
lpF+lpA-+
lpA+1PG+2pc++
1pn- +
95
91
90
79
79
77
74
70
67
114
68
60
86
86
67
61
73
100
103
100
81
100
72
2-pn-+
3-PF++
2pG+ +
2-PF-+
69
76
69
77
96
80
68
68
83
91
52
75
2-pc++
3-PF++
2-pn++
3PG++
3-pn++
3-pF++
3-pn++
3pG++
2-pn++
2-pn++
2-pn++
68
67
65
64
64
41
31
-PF++
31
6
79
74
75
79
98
66
69
83
65
89
74
57
57
?PC+-
?PF+ +
79
83
67
73
76
61
78
87
87
88
85
91
98
32
27
9
72
74
73
62
2-@F/D+ +
2-@F++
3PG++
2pF++
3pG++
62
70
77
66
74
73
2PG++
103
82
3-pn++
2-pc++
2-pc++
8
114
75
74
2-PF-+
7
70
lpn- -
?PG++
S. intermedius
ATCC 27355T
7
65
100
86
100
67
69
65
68
78
78
83
75
100
58
74
120
112
56
34
7
Data are expressed as percentages of the homologous control value.
The first number is the biotype as determined by the API Rapid Strep system (see the text). The next figure indicates the strain was beta-hemolytic (p) or not
beta-hemolytic C-p). The third item is the Lancefield group (n means no group). The fourth and fifth symbols represent ability to ferment lactose and hydrolyze esculin, respectively, in the Rapid Strep system.
”
”
of the seven other biotype 2 strains tested, although S.
intermedius ATCC 27335T DNA hybridized at only 66%.
Labeled index DNA from biotype 3 strain MGH611 hybridized extensively with DNA from all biotype 2 strains, and
the four biotype 3 strains tested (experiment 5 ) . When these
hybridizations were repeated at 57”C, the results were similar.
In these experiments, DNA from the type strain of S.
anginosus hybridized with DNA from the index strains of all
three biotypes (67 to 75%). The S. constellatus type strain
formed hybrid duplexes particularly well with labeled DNAs
of biotype 2 and 3 index strains, but the S. intermedius type
strain DNA never hybridized above 69%.
In other experiments, labeled DNA from S . constellatus
hybridized strongly to DNA of other type strains, although
S . intermedim DNA was, as usual, less reactive (experiment
6). S. constellatus DNA was also used to test the similarity
of DNA extracted from two strains (NCTC 10708 and 10709)
originally isolated by Guthof, and one other NCTC “S.
mdleri” strain (NCTC 11062). In this experiment (experiment 7), the labeled DNA bound more to the DNA of the
Guthof strains than to the unlabeled S. constellatus DNA,
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 14:23:54
226
INT. J.
COYKENDALL ET AL.
but the hybridization to NCTC 11062 DNA was low. When
S. intermedius DNA was labeled, it did not hybridize
strongly to other DNA (experiment 8).
Three strains that resembled “ S . milleri“ did not hybridize appreciably with any of the reference DNAs tested.
Strain UCAbb was a hemolytic group F streptococcus but
differed from other “ S . milleri” isolates by failure to ferment
trehalose and by the production of H202. It fermented
raffinose but not mannitol. The Rapid Strep system would
not accept this pattern as representing “ S . milleri”. Strain
UC2232 also failed to ferment trehalose and could not be
identified by Rapid Strep. Except for the indifference to
trehalose and lack of hemolysis, this strain had the biotype 1
pattern, Strain NCTC 11062 was biochemically typical of
biotype 2 “ S . milleri” but it produced H202.
G+C content of DNA. The DNA base composition was
determined for 42 strains. All but seven strains had between
38.0 and 40.0 mol% G+C in their DNAs. The lowest base
ratio was 37.6 mol% G+C ( S . intermedius 27335=) and the
highest was 40.5 mol% G+C (strain UC2953).
Plasmid DNA. A total of 27 strains representing 14 phenotypes and all the type strains were examined for plasmid
DNA (Table 1;there are 26 strains noted, and 1other strain,
S . anginosus ATCC 33397T, was not included in the table
because of its unusual biochemical pattern). Only two plasmids were found. Strain UC0921, a hemolytic group G
biotype 2 strain carried a 6.9-kilobase plasmid, and strain
UC8948 (hemolytic, group C, biotype 2) bore a 4.5-kilobase
plasmid.
DISCUSSION
Our results support the contention of Colman and Williams (6) that these streptococci are more alike than different
and are consistent with the data of Welborn et al. (41), which
indicated that streptococci that resemble “ S . miZler?’ share
considerable base sequences. DNAs from representatives of
a wide variety of biotypes, serotypes, and hemolytic and
nonhemolytic strains were at least moderately homologous.
Within this group were streptococci resembling S . anginosus, S . constellatus, and S . intermedius, as well as the type
strains of both S . anginosus and S . constellatus. The type
strain of S . intermedius may represent a separate species
since it never hybridized as well as other “ S . milleri”
iisolates and the other type strains; yet DNA from other
:strains that were biochemically like S . intermedius showed
good hybridization with other “ S . milleri” DNA. There
were no genetic distinctions between hemolytic and nonhemolytic strains or between different serotypes. There was
;a tendency for the least reactive strains (esculin, lactose,
raffinose, and mannitol negative) to be less related to the
imost adroit strains (esculin, lactose, raffinose, and mannitol
positive). Nevertheless, we consider these two biotypes
(which represent the extremes of a continuum) insufficiently
disparate to be considered different species. There are two
reasons for this. First, as pointed out by Ball and Parker (1)
these organisms have many traits in common. More importantly, the hybrid DNA duplexes formed in our hybridization
experiments with strains of all biotypes persisted under very
stringent temperatures, If the “ S . milleri” DNA that hybridized at 60 to 70% of controls had actually represented
specious, poorly base-paired hybrid duplexes at 48”C, then
the levels of hybridization with these DNAs would have
been lower at 57°C (7, 19). However, hybridization levels at
!57”C were almost exactly the same as those at 48°C. This
indicated that the hybrids were formed from truly compli-
SYST.
BACTERIOL.
mentary DNA sequences. The effect of the higher temperature on infidelitous hybrids was evident with the S . sanguis
DNA. Therefore, we conclude that these strains, with the
exceptions noted above, have at least 60% base sequence
homology (and usually more) and should be considered
members of a single species. This contention is consistent
with previously published DNA hybridization data generated from studies of smaller collections (11, 15, 41). We did
not see the large difference between the S . anginosus and S .
constellatus type strains that Kilpper-Balz et al. (20) observed. Perhaps the hot formamide method that they used is
so very stringent that DNA base sequence heterogeneities
were revealed, which neither our method nor the S1 nuclease method (11, 41) could detect. It is unlikely that traits
such as lactose fermentation or hemolysis are conferred by
plasmids.
The oldest accepted name amongst these bacteria is S .
anginosus, which was described by Andrewes and Horder in
1906 (1). They isolated hemolytic, lactose-positive streptococci which often acidified raf€inose, a sugar rarely fermented by hemolytic streptococci. Smith and Sherman (39)
noted the similarity between hemolytic group G isolates that
fermented raffinose, and the description of S . anginosus (1).
The present type strain of S . anginosus has these characteristics. Since it is genetically related to the strains we called
“S.milleri,” this large, phenotypically diverse group should
be called S . anginosus. S . constellatus and “ S . milleri”
become later synonyms of S . anginosus. S . intermedius is
also a later synonym of S . anginosus as noted by Moore et
al. (31).
Emended description of Streptococcus anginosus. Streptococcus anginosus Andrewes and Horder 1906 (1).Small (0.5
to 1.0 pm) gram-positive cocci in short chains, which form
small or minute (less than 1 mm), white, convex or pulvinate, entire colonies on the surface of blood agar. Growth is
feeble in aerobic conditions and is enhanced by Cot. Some
strains require an anaerobic atmosphere for optimum
growth. Strains may be beta-hemolytic or not hemolytic,
rarely alpha-hemolytic. Acetoin and alkaline phosphatase
are produced, arginine is hydrolyzed, and trehalose is fermented. Hippurate is not split, and neither ribose nor
glycogen is acidifed. Isolates may react with antibody produced against Lancefield antigens A, C, F, or G. Others give
no Lancefield reactions. Other reactions are variable, but
some correlate with others. For example, strains that are
hemolytic often fail to split esculin or ferment lactose,
especially among hemolytic group F strains. Nonhemolytic
strains usually split esculin and ferment lactose. Strains that
are Lancefield group G are nearly always beta-hemolytic.
Some strains ferment raffinose and mannitol in addition to
lactose.
The type strain is ATCC 33397 (NCTC 10713). It is a
TABLE 3. Characteristics that separate S. anginosus from
nonhemolytic acetoin-producing streptococciu
Reaction
Organism
ALKP
production
Ribose
fermentation
*-
-
S.
S.
S.
S.
anginosus
salivarius
cremoris
lactis
S. uberis
+
-
2
Arg
hydrolysis
+
+
a ALKP, Alkaline phosphatase. S . mutans, S. equinus, and S . bovis are
negative for all three tests.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 14:23:54
VOL. 37. 1987
SYNONYMS OF STREPTOCOCCUS ANGZNOSUS
beta-hemolytic, Lancefield group G strain that ferments
lactose and raffinose, but not mannitol, and corresponds to
the original description of S . anginosus (1).
The combination of acetoin production with the failure to
ferment ribose serves to distinguish S . anginosus from other
hemolytic species. Among the nonhemolytic streptococci
that produce acetoin, S . anginosus can be separated by
considering its production of alkaline phosphatase, failure to
ferment ribose, and hydrolysis of arginine (Table 3).
ACKNOWLEDGEMENTS
We are grateful to G. Colman, R. R. Facklam, K. L. Ruoff, and
the staff of the Clinical Microbiology Laboratory of the Uconn
Health Center for giving us strains and isolates, and to L. J. Kunz
for his encouragement. We thank M. B. Latham and K. Edmiston
for preparing the manuscript.
This work was supported by a Public Health Service grant
DE07126 from the National Institute for Dental Research.
LITERATURE CITED
1. Andrew-, F. W., and T. J. Horder. 1906. A study of the
streptococci pathogenic for man. Lancet ii:703-713.
2. Appelbaum, P. C., P. S. Chaurushiya, M. R. Jacobs, and A.
Duffett. 1984. Evaluation of the Rapid Strep system for species
identification of streptococci. J. Clin. Microbiol. 19588-591.
3. Ball, L. C., and M. T. Parker. 1979. The cultural and biochemical characters of Streptococcus milleri strains isolated from
human sources. J. Hyg. 82:63-78.
4. Carbon, J. 1967. Presence of various types of non-haemolytic
streptococci in dental plaque and in other sites of the oral cavity
in man. Odontol. Revy 1855-74.
5. Chapman, G. H. 1946. The isolation and testing of fecal
streptococci. Am. J. Digest. Dis. 13:105-107.
6. 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 diseases. Academic Press, Inc., New York.
7. Coykendall, A. L. 1971. Genetic heterogeneity in Streptococcus
mutans. J. Bacteriol. 106:192-1%.
8. Coykendall, A. L., and K. B. Gustafson. 1985. Deoxyribonucleic
acid hybridizations among strains Streptococcus salivarius and
Streptococcus bovis. Int. J. Syst. Bacteriol. 35274280.
9. Deibel, R. H., and H. W. Seeley, Jr. 1974. Genus I. Streptococcus Rosenbach 1884, p, 490-509. In R. E. Buchanan and N. E.
Gibbons (ed.), Bergey’s manual of determinative bacteriology,
8th ed. The Williams and Wilkins Co., Baltimore.
10. DeLey, J., and R. Tijtgat. 1970. Evaluation of membrane filters
for DNA-DNA hybridization. Antonie van Leeuwenhoek J.
Microbiol. Serol. 36:461474.
11. Ezaki, T., R. Facklam, N. Takeuchi, and E. Yabuuchi. 1986.
Genetic relatedness between the type strain of Streptococcus
anginosus and minute-colony-forming beta-hemolytic streptococci carrying different Lancefield grouping antigens. Int. J.
Syst. Bacteriol. 36:345-347.
12. Facklam, R. R. 1977. Physiological differentiation of viridans
streptococci. J. Clin. Microbiol. 5:184-201.
13. 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.
14. Facklam, R. R., D. L. Rhoden, and P. B. Smith. 1984. Evaluation of the Rapid Strep system for the identification of clinical
isolates of Streptococcus species. J. Clin. Microbiol. 20:894898.
15. Farrow, J. A. E., and M. D. Collins. 1984. Taxonomic studies on
streptococci of serological groups C, G and L and possibly
related taxa. Syst. Appl. Microbiol. 5483493.
16. Guthof, 0. 1956. Ueber Pathogene “Vergriinende Streptokokken” Streptokokken-Befunde bei dentogenen abszessen und
227
infiltraten in bereich der Mundhohle. Zentralbl. Bakteriol.
Parasitenkd. Infektionskr. Hyg. Abt. 1 Orig. 166553-564.
17. Hardie, J. M., and G. H. Bowden. 1976. Physiological classification of oral viridans streptococci. J. Dent. Res. 55:A16&
A176.
18. Holdeman, L. V., and W. E. C. Moore. 1974. New genus,
Coprococcus, twelve new species, and emended descriptions of
four previously described species of bacteria from human feces.
Int. J. Syst. Bacteriol. 24:260-277.
19. Johnson, J. L., and E. J. Ordal. 1968. Deoxyribonucleic acid
homology in bacterial taxonomy: effect of incubation temperature on reaction specificity. J. Bacteriol. 95893-900.
20. Kilpper-Balz, R., B. L. Williams, R. Liitticken, and K. H.
Schleifer. 1984. Relatedness of “Streptococcus milleri” with
Streptococcus anginosus and Streptococcus constellatus. Syst .
Appl. Microbiol. 5:494500.
21. Lapage, S. P., S. Bascomb, W. R. Wilcox, and M. A. Curtis.
1973. Identification of bacteria by computer: general aspects
and perspectives. J. Gen. Microbiol. 77:273-290.
22. Lawrence, J., D. M. Yajko, and W. K. Hadley. 1985. Incidence
and characterization of beta-hemolytic Streptococcus rnilleri
and differentation form S . pyogenes (group A), S . equisimilis
(group C ) , and large-colony group G streptococci. J. Clin.
Microbiol. 22:772-777.
23. Legault-Dkmare, J., B. Desseaux, T. Heyman, S. Seror, and
G. P. Ress. 1967. Studies on hybrid molecules of nucleic acids.
1. DNA-DNA hybrids on nitrocellulose filters. Biochem.
Biophys. Res. Commun. 28550-557.
24. Lutticken, R., U. WendorE, D. Liitticken, E. A. Johnson, and
L. W. Wannamaker. 1978. Studies on streptococci resembling
Streptococcus milleri and on an associated surface-protein
antigen. J. Med. Microbiol. 11:419-431,
25. MacFaddin, J. R. 1977. Biochemical tests for the identification
of medical bacteria. The Williams and Wilkins Co., Baltimore.
26. Macrina, F. L., D. J. Kopecko, K. R. Jones, D. J. Ayers, and
S. M. McCowen. 1978. A multiple plasmid-containing Escherichia coli strain: convenient source of size reference plasmid
molecules. Plasmid 1:417420.
27. Macrina, F. L., P. H. Wood, and K. R. Jones. 1980. Simple
method for demonstrating small plasmid deoxyribonucleic acid
molecules in oral streptococci. Appl. Environ. Microbiol.
39:1070-1073.
28. Marmur, J. 1961. A procedure for the isolation of deoxyribonucleic acid from microorganisms. J. Mol. Biol. 3:208-218.
29. Marmur, J., and P. Doty. 1962. Determination of the base
composition of deoxyribonucleic acid from its thermal .denaturation temperature. J. Mol. Biol. 5109-118.
30. Mirik, G. S., L. Thomas, E. C. Curnen, and F. L. Horsfall, Jr.
1944. Studies on a non-hemolytic streptococcus isolated from
the respiratory tract of human beings. J. Exp. Med. 80:391407.
31. Moore, W. E. C., L. V. Holdeman, R. M. Smibert, I. J. Good,
J. A. Burmeister, K. G. Palcanis, and R. R. Raney. 1982.
Bacteriology of experimental gingivitis in young adult humans.
Infect. Immun. 38:651467.
32. Niven, C. F., K. L. Smiley, and J. M. Sherman. 1942. The
hydrolysis of arginine by streptococci. J. Bacteriol. 43:65146.0.
33. Ottens, H., and K. C. Winkler. 1962. Indifferent and haemolytic
streptococci possessing group-antigen F. J. Gen. Microbiol.
28: 181-191.
34. Parker, M. T., and L. C. Ball. 1976. Streptococci and aerococci
associated with systemic infection in man. J. Med. Microbiol.
9:275-302.
35. Ruoff, K. L., and L. J. Kunz. 1982. Identification of viridans
streptococci isolated from clinical specimens. J. Clin. Microbiol. 15:920-925.
36. Ruoff, K. L., and L. J. Kunz. 1983. Use of the Rapid Strep
system for identification of viridans streptococcal species. J.
Clin. Microbiol. 18:1138-1 140.
37. Ruoff, K. L., L. J. Kunz, and M. J. Ferraro. 1985. Occurrence
of Streptococcus milleri among beta-hemolytic streptococci
isolated from clinical specimens. J. Clin. Microbiol. 22:149151.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 14:23:54
228
INT.J. SYST.BACTERIOL.
COYKENDALL ET AL.
38. Shlaes, D. M., P. I. Lerner, E. Wolinsky, and K. V. Gopalakrishna. 1981. Infections due to Lancefield group F and related
streptococci (S. milleri,S.anginosus). Medicine 60:197-207.
39. Smith, F. R., and J. M. Sherman. 1938. The hemolytic
streptococci of human feces. J. Infect. Dis. 62:189-192.
40. Titlotson, G. S. 1982. An evaluation of the API-20 Strep system.
J. Clin. Pathol. 35468472.
41. Welborn, P. P., W. K. Hadley, E. Newbrun, and D. M. Yajko.
1983. Characterization of strains of viridans streptococci by
deoxyribonucleic acid hybridization and physiologic tests. Int.
J. Syst. Bacteriol. 33:293-299.
42. Whittenbury, R. 1964. Hydrogen peroxide formation and
catalase activity in the lactic acid bacteria. J. Gen. Microbiol.
35: 13-26.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sun, 18 Jun 2017 14:23:54