Genetic approaches to the identification of the mitis group within the genus Streptococcus

Yoshiaki Kawamura1, Robert A. Whiley2, Shin-Ei Shu1, Takayuki Ezaki1 and Jeremy M. Hardie2

Department of Microbiology, Gifu University School of Medicine, 40 Tsukasa-machi, Gifu 500-8705, Japan1
Department of Oral Microbiology, St. Bartholomew’s and the Royal London School of Medicine and Dentistry, Turner Street, London E1 2AD, UK2

Author for correspondence: Yoshiaki Kawamura. Tel: +81 58 267 2240. Fax: +81 58 267 0156. e-mail: kawamura{at}cc.gifu-u.ac.jp


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The usefulness and reliability of partial sequence analysis of the manganese-dependent superoxide dismutase gene (sodA), autolysin (lytA) gene amplification and species-specific PCR based on the D-alanine:D-alanine ligase (ddl) gene for differentiating each member of the mitis group of the genus Streptococcus was investigated. On the phylogenetic tree based on sodA partial sequences (366 bp) from 96 strains, including all species currently within the mitis group isolated in different geographic areas (mainly Japan and the UK), eight well separated clusters were generated corresponding to recognized species, and all strains fell into those clusters to which they had also been assigned by DNA–DNA hybridization. The Streptococcus pneumoniae sub-cluster was located within the Streptococcus mitis cluster, but the sodA gene of S. pneumoniae was very conserved and therefore could be separated from all other species examined. Furthermore, the lytA gene amplification approach could also be used to differentiate S. pneumoniae from other species. The species-specific amplification product of the ddl gene was successfully detected in Streptococcus sanguinis and Streptococcus gordonii, but failed to be detected in some strains of Streptococcus oralis including the type strain and S. mitis. We conclude that the partial sequence analysis of the sodA gene could be applied globally as a reliable and easy method for the accurate identification of all species currently within the mitis group.


View this table:
[in this window]
[in a new window]
 
Table 1. Strains used and identification results

 
Keywords: Superoxide dismutase gene, autolysin gene, d-alanine:d-alanine ligase gene, mitis group, Streptococcus

The DDBJ accession numbers of the superoxide dismutase genes described in this paper are shown in Table 1.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Based on 16S rRNA gene sequence analysis, species of the genus Streptococcus were separated into six major clusters or species groups (Bentley et al., 1991 ; Kawamura et al., 1995 ). Within these clusters, the mitis group contained several clinically important species, such as Streptococcus pneumoniae, Streptococcus mitis, Streptococcus oralis and Streptococcus sanguinis. It is well known that members of this group are difficult to identify correctly, and show 40–60% DNA–DNA similarity values with each other. In addition, some species share greater than 99% 16S rRNA sequence homology (Kawamura et al., 1995 ). Biochemical identification schemes for oral or viridans streptococci have been proposed (Kilian et al., 1989 ; Beighton et al., 1991 ); however, these schemes are insufficient for identifying all the currently recognized species within the mitis group. Similarly, commercially available identification kits give limited identification: in a previous study, we evaluated two kinds of identification kits for streptococci (Rapid ID32 Strep. and STREPTOGRAM) and demonstrated that these kits showed less than 79% accuracy within the mitis group (Kawamura, 1996 , 1998 ). Recently, we used the BIOLOG identification system, which employed 95 different reactions and this seemed to be useful. However, even using this system, some strains of S. mitis and S. oralis were mis-identified (unpublished data). Identification of members of the mitis group remains difficult.

Recently, the epithets of some species within the mitis group have been revised for grammatical reasons: Streptococcus sanguis was corrected to S. sanguinis, similarly, Streptococcus parasanguis was corrected to Streptococcus parasanguinis and Streptococcus crista was corrected to Streptococcus cristatus (Truper & De Clari, 1997 ). We have used the revised names throughout this report. In the 1990s, four new species were described within the mitis group, S. cristatus, S. parasanguinis, S. peroris and S. infantis, giving a total of nine species classified within this group (Handley et al., 1991 ; Whiley et al., 1990 ;Kawamura et al., 1998 ). Due to the addition of these new species, we were strongly aware of the potential confusion surrounding the identification of members of the mitis group and were therefore motivated to establish a reliable, globally acceptable and easy practical identification method.

To this end, DNA probes were designed from 16S rRNA sequences. Even though some members of the mitis group share greater than 99% sequence similarities in their 16S rRNA gene, it was possible to design several DNA probes which could differentiate between the type strains of S. mitis, S. oralis, Streptococcus gordonii, S. cristatus, S. sanguinis and S. parasanguinis. When we applied these DNA probes to our clinical strains that had been identified, in advance, by DNA–DNA hybridization, all strains of S. gordonii, S. cristatus, S. sanguinis and S. parasanguinis could be correctly identified, although significantly, some clinical strains of S. oralis and S. mitis were mis-identified. From these results, we concluded that our 16S rRNA gene probe could not be applied unambiguously to the identification of S. mitis and S. oralis (unpublished data).

Identification on the basis of whole-cell protein profile comparison has also been investigated as a taxonomic tool. In a previous study, we found some clinical strains not included within the same cluster as their respective type strain (unpublished data). Similar data were published by Vandamme et al. (1998 ). According to their report, they observed that S. mitis biovar 2 strains were neither in the same cluster as S. mitis biovar 1 strains nor formed an independent cluster, and in fact migrated into several other species clusters, that included S. oralis, S. cristatus and S. parasanguinis. Furthermore, while the majority of strains of S. sanguinis formed a single cluster, this excluded the type strain of S. sanguinis, which was relatively close to the S. mitis biovar 1 cluster. From these results, it was apparent that applying protein profiles to the identification of the members of the mitis group would prove problematic.

Recently, some genetic methods for the identification of members of the mitis group at the species level have been proposed by other researchers, such as species-specific PCR primers based on the ddl gene and comparative analysis of the partial sequences of the manganese-dependent superoxide dismutase (sodA) gene (Garnier et al., 1997 ; Poyart et al., 1998 ). In this study, we have investigated the reliability and usefulness of these genetic methods by using 96 human clinical strains isolated from different geographic areas. Finally, we confirm that the sodA partial sequence analysis method is a reliable and useful method for accurate identification of members of the mitis group.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains.
The type strains used in this study are S. mitis (GTC 495T=NCTC 12261T), S. oralis (GTC 276T=NCTC 11427T), S. pneumoniae (GTC 261T=NCTC 7465T), S. gordonii (GTC 497T=ATCC 10558T), S. cristatus (GTC 631T=NCTC 12497T), S. sanguinis (GTC 217T=ATCC 10556T), S. parasanguinis (GTC 498T=ATCC 15912T), S. peroris (GTC 848T=JCM 10158T) and S. infantis (GTC 849T=JCM 10157T). We also used some reference strains and 82 human clinical strains mainly isolated in Japan and the UK (Table 1). All strains were grown on Todd–Hewitt broth (Difco) under aerobic conditions at 37 °C after confirming purity on Columbia blood (5% defibrinated sheep blood) agar plates (bioMérieux).

DNA–DNA hybridization.
DNA from all strains was prepared by a standard procedure (Marmur, 1961 ) with minor modification (Ezaki et al., 1983 ). DNA from each type strain was labelled with photobiotin (Sigma) and microplate quantitative DNA–DNA hybridization was carried out according to previously described methods (Ezaki et al., 1988 , 1989 ). Briefly, purified DNA (100 µg ml-1) of each strain was heat denatured and then diluted to 10 µg ml-1 with ice cold PBS (pH 7·4) containing 0·1 M MgCl2. The diluted DNA solution was distributed into a microplate (Maxsorp; Inter Med) at 100 µl per well, and the plate incubated at 30 °C for 12 h. The solution was discarded and the plate dried at 60 °C. The plate was prehybridized for 30 min and then hybridized in the presence of 2x SSC and 50% formamide, with biotin-labelled DNA at 31 °C for 2 h. The plate was washed three times with 1x SSC and 100 µl streptavidin–ß-d-galactosidase (diluted 1:1000 with 0·5% BSA in PBS; Gibco-BRL) was added to each well. The plate was incubated at 37 °C for 30 min and washed three times with 1x SSC. Then 100 µl of the substrate (100 µg 4-methylumbelliferyl-ß-d-galactopyranoside ml-1; Sigma) was added to each well and the fluorescence intensity was measured by Cytofluor (model 2350; Millipore). The species name of an isolate was determined if the DNA strongly hybridized and showed greater than 70% similarity value with the DNA of only one type strain.

PCR amplification of part of the sodA gene.
An internal portion of the sodA gene was amplified by PCR using our designed primer set (SOD-UP; 5'-biotin-TRCAYCATGAYAARCACCAT-3' and SOD-DOWN; 5'-ARRTARTAMGCRTGYTCCCARACRTC-3'). Each PCR reaction (100 µl) contained 1x PCR buffer (Pharmacia Biotech), 100 µM dNTPs, 0·1 µM and 0·2 µM biotin-labelled (SOD-UP) primer and non-labelled (SOD-DOWN) primer, respectively, 20 ng template DNA and 1 U Taq polymerase (Pharmacia Biotech). PCR was carried out on a thermal cycler (model 2400, Perkin Elmer) as follows: 35 cycles of 30 sec at 94 °C, 1 min at 50 °C and 1 min at 72 °C, and a final cycle of 1 min at 50 °C followed by 10 min at 72 °C.

Determination and analysis of sodA partial sequences.
After confirming the presence of a single 435 bp amplification product of the sodA gene on 1% agarose gels, the sequence was determined using a Pharmacia automatic sequencer (ALF express) with an auto load sequencing kit (Pharmacia Biotech). Briefly, the avidin-conjugate comb was soaked in the PCRed solution to capture the biotin-labelled PCR product. Then the comb was washed in 0·1 M NaOH solution to denature the amplicon and to remove other substances of the PCR reaction mixture. The sequencing reaction was carried out on the comb using T7 polymerase and Cy5-labelled primer (the sequence was the same as for the SOD-DOWN primer). The reacted comb was applied directly onto the sequencing gel and the sequencer run for 10 h.

Phylogenetic analysis of sodA partial sequences.
clustal w software originally described by Thompson et al. (1994 ) was used to align the sequences, and the phylogenetic distances were calculated using the neighbour-joining method. The phylogenetic tree was drawn using TreeView software (Page, 1996 ). The sodA partial sequence of Streptococcus agalactiae was obtained from the DNA Database of Japan (DDBJ, accession number is Z95893) and used as the outgroup.

D-Alanine:D-alanine ligase gene specific PCR.
Species-specific PCR for S. mitis, S. oralis, S. gordonii and S. sanguinis based on the d-alanine:d-alanine ligase gene (ddl gene) was carried out according to Garnier et al. (1997 ). Ten microlitres of each PCR reaction was applied to a 1% agarose gel and electrophoresed.

Autolysin gene (lytA) specific PCR.
lytA-specific PCR for the identification of S. pneumoniae was carried out according to Gillespie et al. (1994 ). PCR amplification was confirmed by electrophoresis through 1% agarose gels.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
DNA–DNA hybridization
To determine species identification, we applied our DNA–DNA hybridization method to all strains of unidentified isolates (a total of 69 isolates). All identification results are shown in Table 1.

ddl gene amplification
We confirmed the specificity of each species-specific primer set based on the ddl gene by using the type strains of members of the mitis group, except S. oralis (Table 1). We also applied this method to some of the S. mitis, S. oralis, S. gordonii and S. sanguinis strains. All strains of S. gordonii (three strains including the type strain) and S. sanguinis (five strains including the type strain) were positive. However, the S. oralis type strain and one of our clinical strains were negative. The type strain of S. mitis gave a positive result. However, one reference strain (NCTC 10712) and four of six clinical strains tested were negative (Fig. 1, Table 1).



View larger version (32K):
[in this window]
[in a new window]
 
Fig. 1. Results of ddl gene amplification using S. mitis- or S. oralis-specific primers. The molecular size marker is the 100 bp ladder (lane M). Lanes: 1, S. mitis GTC 495T; 2, S. mitis NCTC 10712; 3, S. mitis O-4; 4, S. mitis O-17; 5, S. mitis O-28; 6, S. mitis OS51; 7, S. mitis HV51; 8, S. mitis A17C1; 9, S. oralis GTC 276T; 10, S. oralis NCTC 7864; 11, S. oralis O-5; 12, S. oralis O-7; 13, S. oralis O-8; 14, S. oralis PC1467; 15, S. oralis B1493; 16, S. oralis GPD1.

 
lytA gene amplification
lytA gene amplification gave amplicons from all S. pneumoniae strains including the type strain, while all other strains, including those of closely related species S. mitis and S. oralis, gave completely negative results (data not shown, the results are summarized in Table 1)

Phylogenetic analysis of sodA partial sequences
The phylogenetic tree of partial sequences of sodA was compared with that of 16S rRNA sequences for the type strains. It can be seen that the evolutionary rate of the sodA partial gene is much faster than that of the 16S rRNA gene (Fig. 2). We took this to indicate that partial sodA sequencing would be useful for differentiating genetically closely related organisms, and decided to apply sodA partial sequencing to the main collection of isolates. We determined the partial sequences of the sodA gene from 96 strains including all species currently within the mitis group and constructed a phylogenetic tree using the neighbour-joining method. Eight clusters were clearly generated corresponding to recognized species as confirmed by DNA–DNA hybridization (Fig. 3). With respect to the Japanese and UK isolates of S. mitis and S. oralis, each species formed its own cluster with no evidence of geographic variation. S. pneumoniae strains did not form an independent cluster but formed a sub-cluster within the S. mitis cluster on the sodA phylogenetic tree.



View larger version (16K):
[in this window]
[in a new window]
 
Fig. 2. Comparative phylogenetic trees of sodA partial gene sequences and the 16S rRNA gene using the type strains of all members of the mitis group. (a) Phylogenetic tree of sodA partial sequences (366 bp); (b) phylogenetic tree of 16S rRNA sequences (1300 bp). Distances were calculated by the neighbour-joining method. The scales of each phylogenetic tree were made equivalent.

 


View larger version (39K):
[in this window]
[in a new window]
 
Fig. 3. Phylogenetic tree based on sodA partial sequences (366 bp). Distances were calculated by the neighbour-joining method. The numbers at the nodes are bootstrap values (based on 100 samplings). Each member species of the mitis group formed a separate cluster, except S. pneumoniae. All S. pneumoniae strains were located within the S. mitis cluster as a distinct sub-cluster.

 
S. pneumoniae-specific bases within sodA
S. pneumoniae strains did not cluster independently on the sodA phylogenetic tree. However, sodA partial sequences of all S. pneumoniae strains were almost identical and some species-specific base differences were detected that were helpful for differentiating S. pneumoniae from other species including S. mitis (Table 2).


View this table:
[in this window]
[in a new window]
 
Table 2. Some bases of the sodA gene helpful in differentiating S. pneumoniae from other species

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
In 1997, Garnier and co-workers developed a promising method for the identification of S. mitis, S. oralis, S. gordonii and S. sanguinis (Garnier et al., 1997 ). According to their report, they used species-specific PCR primers based on the ddl gene, and could exploit size polymorphisms between the species tested. If this method worked well, it was expected to be useful. However, no amplicon was observed here when we applied S. oralis-specific primers (the G primer pair of Garnier et al.) to the type strain of S. oralis. We could not compare this result with their report, because the previous authors did not use the type strain of S. oralis in their study. We had applied milder PCR conditions [5 °C lower annealing temperature (51 °C) than that of Garnier et al.], although even under those conditions we did not get a positive result from the type strain of S. oralis. Furthermore, we used two different laboratory versions of the type strain of S. oralis (Gifu laboratory strain GTC 276T, which was directly received from NCTC as NCTC 11427T, and another from London, strain LVG1T). Neither strain yielded a positive result. From these observations, we conclude that the reported S. oralis-specific primers based on the ddl gene could not amplify DNA of the S. oralis type strain. Similarly, one reference strain of S. mitis (NCTC 10712) did not give a positive result when using S. mitis-specific primers (the F primer pair of Garnier et al.). Finally, we applied these PCR primers to eight strains of each of S. oralis and S. mitis, and could not obtain amplification from two strains of S. oralis (the type strain and one clinical strain) and five strains of S. mitis (NCTC 10712 and four clinical strains). We therefore abandoned the idea of using these primers for species identification.

After developing lytA-specific PCR (Rudolph et al., 1993 ), many researchers have successfully detected or identified S. pneumoniae by this method using purified DNA as a template (Gillespie et al., 1994 ; Hassan-King et al., 1994 ; Ubukata et al., 1996 ). Recently, Whatmore et al. (1996 ) reported that a subset of S. mitis and S. oralis strains isolated were positive for this gene by PCR. However we have applied this method to more than 20 strains of S. pneumoniae and more than 70 strains of S. mitis or S. oralis, and did not observe any positive results from S. mitis nor S. oralis strains. From these results, we thought that the lytA-specific PCR might be helpful, in many cases, for identification of S. pneumoniae.

Recently, several identification methods using gene sequence data have been established, such as DNA gyrase B subunit (Yamamoto & Hirayama, 1995 , 1996 ), heat-shock protein 60 (Goh et al., 1996 , 1998 ) and the sodA gene (Smith & Doolittle, 1992 ; Zolg & Philippi-Schulz, 1994 ;Poyart et al., 1995 , 1998 ). Of these, the sodA partial gene sequencing seemed to be relatively useful, because less than 400 bp is enough to determine the phylogenetic position of a strain. Therefore, we selected this gene for the identification of members of the mitis group.

When the sodA phylogenetic tree from each species type strain within the mitis group was constructed, it became apparent that the evolutionary rate of the sodA gene was much faster than that of 16S rRNA sequence (Fig. 2). Thus, we suspected that this gene would be useful in differentiating genetically closely related organisms. However, it was possible that there might be too much variation within the same species, possibly due to geographic differences, for this approach to be practicable. To test the method, clinical strains from different areas, mainly from Japan and the UK, were examined but failed to reveal any obvious geographic variation (Fig. 3). For these reasons, we believe that this method could be applied globally.

Some bootstrap values of the branching point of each species on the sodA phylogenetic tree were not very high (the lowest is 66%, at the S. gordonii and S. sanguinis branching point, Fig. 3). However, before making the final phylogenetic tree (Fig. 3), we remade the tree each time we collected new sequence data, and each time the species separated in the same way as Fig. 3. From this experience, we consider that the branching of each species on the sodA tree is reproducible.

All strains of each member species of the mitis group formed a single cluster. However, many strains of the same species showed several base substitutions (maximum 31 bases) within the 366 bp region of the sodA gene examined, while the gene of S. pneumoniae, by comparison, was very conserved. Eleven of 16 S. pneumoniae strains showed completely identical sequences, and only two strains (strains 1510 and 3203) and three strains (strains 1639, 1293 and 661) showed two and one base substitutions, respectively. Similar data have been published by Poyart et al. (1998 ). These authors used eight isolates of S. pneumoniae of which only two showed a single base substitution. We cannot explain why the sodA gene of S. pneumoniae is so conserved, but are encouraged that this conservation is helpful for the identification of S. pneumoniae (Table 2).

In this study, we determined the sequences direct from PCR amplicons using a Pharmacia sequencer with an auto-load sequencing kit. By using this method, we could apply the sample onto the sequencer within a day from the bacterial culture without doing any troublesome manipulation, such as ethanol precipitation and prolonged centrifugation of the sample, and could obtain the sequence data the next morning. There are many internet home pages providing a service to enable calculation of phylogenetic relationships, e.g. clustal w could be carried out on the home page of the DNA Databank of Japan (http://www.ddbj.nig.ac.jp). By using these easy practical methods, we believe that this method could be carried out rapidly in many laboratories.

In view of the data presented above, we believe that the sodA partial sequence analysis method could be applied globally as a reliable and practical method for the accurate identification of all species currently within the mitis group.


   ACKNOWLEDGEMENTS
 
We thank Mrs D. Brigit and Dr K. Saitou for the gift of S. pneumoniae strains. This work was partially supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan, grant B(2)-09557024, and Yacult Research foundation.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Beighton, D., Hardie, J. M. & Whiley, R. A. (1991). A scheme for identification of viridans streptococci. J Med Microbiol 35, 367-372.[Abstract]

Bentley, R. W., Leigh, J. A. & Collins, M. D. (1991). Intragenetic structure of Streptococcus based on comparative analysis of small-subunit rRNA sequences. Int J Syst Bacteriol 41, 487-494.[Abstract]

Ezaki, T., Yamamoto, N., Ninomiya, K., Suzuki, S. & Yabuuchi, E. (1983). Transfer of Peptococcus indolicus, Peptococcus asaccharolyticus, Peptococcus prevotii, and Peptococcus magnus to the genus Peptostreptococcus and proposal of Peptostreptococcus tetradius sp. nov. Int J Syst Bacteriol 33, 683-698.

Ezaki, T., Hashimoto, Y., Takeuchi, N., Yamamoto, H., Liu, S., Miura, H., Matsui, K. & Yabuuchi, E. (1988). Simple genetic method to identify viridans group streptococci by colorimetric dot hybridization and fluorometric hybridization in microdilution wells. J Clin Microbiol 26, 1708-1713.[Medline]

Ezaki, T., Hashimoto, Y. & Yabuuchi, E. (1989). Fluorometric deoxyribonucleic acid-deoxyribonucleic acid hybridization in microdilution wells as an alternative to membrane filter hybridization in which radioisotopes are used to determine genetic relatedness among bacterial strains. Int J Syst Bacteriol 39, 224-229.

Garnier, F., Gerbaud, G., Courvalin, P. & Galimand, M. (1997). Identification of clinically relevant viridans group streptococci to the species level by PCR. J Clin Microbiol 35, 2337-2341.[Abstract]

Gillespie, S. H., Ullman, C., Smith, M. D. & Emery, V. (1994). Detection of Streptococcus pneumoniae in sputum sample by PCR. J Clin Microbiol 32, 1308-1311.[Abstract]

Goh, S. H., Potter, S., Wood, J. O., Hemmingsen, S. M., Reynolds, R. P. & Chow, A. W. (1996). HSP60 gene sequences as universal targets for microbial species identification: studies with coagulase-negative staphylococci. J Clin Microbiol 34, 818-823.[Abstract]

Goh, S. H., Driedger, D., Gillett, S. & 8 other authors (1998). Streptococcus iniae, a human and animal pathogen: specific identification by the chaperonin 60 gene identification method. J Clin Microbiol 36, 2164–2166.[Abstract/Free Full Text]

Handley, P., Coykendall, A., Beighton, D., Hardie, J. M. & Whiley, R. A. (1991). Streptococcus crista sp. nov. a viridans streptococcus with tufted fibrils, isolated from the human oral cavity and throat. Int J Syst Bacteriol 41, 543-547.[Abstract]

Hassan-King, M., Baldeh, I., Secka, O., Falade, A. & Greenwood, B. (1994). Detection of Streptococcus pneumoniae DNA in blood cultures by PCR. J Clin Microbiol 32, 1721-1724.[Abstract]

Kawamura, Y. (1996). Evaluation and comparison of newly available identification kits for streptococci. J Med Tech 40, 409–416 (in Japanese).

Kawamura, Y. (1998). Recent classification of the Genus Streptococcus. Jpn J Bacteriol 53, 493–507 (in Japanese).

Kawamura, Y., Hou, X. G., Sultana, F., Miura, H. & Ezaki, T. (1995). Determination of 16S rRNA sequences of Streptococcus mitis and Streptococcus gordonii and phylogenetic relationships among members of the genus Streptococcus. Int J Syst Bacteriol 45, 406-408.[Abstract]

Kawamura, Y., Hou, X. G., Todome, Y., Sultana, F., Hirose, K., Shu, S., Ezaki, T. & Ohkuni, H. (1998). Streptococcus peroris sp. nov. and Streptococcus infantis sp. nov., new members of the Streptococcus mitis group, isolated from human clinical specimens. Int J Syst Bacteriol 48, 921-927.[Abstract/Free Full Text]

Kilian, M., Mikkelsen, L. & Henrichsen, J. (1989). 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). Int J Syst Bacteriol 39, 471-484.

Marmur, J. (1961). A procedure for the isolation of deoxyribonucleic acid from micro-organisms. J Mol Biol 3, 208-218.

Page, R. D. M. (1996). TREEVIEW: an application to display phylogenetic trees on personal computers. Comp Appl Biosci 12, 357-358.[Medline]

Poyart, C., Berche, P. & Trieu-Cuot, P. (1995). Characterization of superoxide dismutase genes from gram-positive bacteria by polymerase chain reaction using degenerate primers. FEMS Microbiol Lett 131, 41-45.[Medline]

Poyart, C., Quesne, G., Coulon, S., Berche, P. & Trieu-Cuot, P. (1998). Identification of streptococci to species level by sequencing the gene encoding the manganese-dependent superoxide dismutase. J Clin Microbiol 36, 41-47.[Abstract/Free Full Text]

Rudolph, K. M., Parkinson, A. J., Black, C. M. & Mayer, L. W. (1993). Evaluation of polymerase chain reaction for diagnosis of pneumococcal pneumonia. J Clin Microbiol 31, 2661-2666.[Abstract]

Smith, M. W. & Doolittle, R. F. (1992). A comparison of evolutionary rates of the two major kinds of superoxide dismutase. J Mol Evol 34, 175-184.[Medline]

Thompson, J. D., Higgins D. G. & Gibson, T. J. (1994). clustal w: improving the sensitivity of progressive multiple sequence alignment through sequence weighing, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22, 4673–4680.[Abstract]

Trüper, H. G. & De Clari, L. (1997). Taxonomic note: necessary correction of specific epithets formed as substantives (nouns) ‘in apposition’. Int J Syst Bacteriol 47, 908-909.

Ubukata, K., Asahi, Y., Yamane, A. & Konno, M. (1996). Combinational detection of autolysin and penicillin-binding protein 2B genes of Streptococcus pneumoniae by PCR. J Clin Microbiol 34, 592-596.[Abstract]

Vandamme, P., Torck, U., Falsen, E., Pot, B., Goossens, H. & Kersters, K. (1998). Whole-cell protein electrophoretic analysis of viridans streptococci: evidence for heterogeneity among Streptococcus mitis biovars. Int J Syst Bacteriol 48, 117-125.[Abstract/Free Full Text]

Whatmore, A. M., Pickerill, A. P., Woodward, G. E. & Dowson, C. G. (1996) Allelic variation of the lytA gene of S. pneumoniae and related species. Abstracts of the Lancefield International Symposium on Streptococci and Streptococcal Diseases, Paris.

Whiley, R. A., Fraser, H. Y., Douglas, C. W. I., Hardie, J. M., Williams, A. M. & Collins, M. D. (1990). Streptococcus parasanguis sp. nov. an atypical viridans Streptococcus from human clinical specimens. FEMS Microbiol Lett 68, 115-122.

Yamamoto, S. & Hirayama, S. (1995). PCR amplification and direct sequencing of gyrB genes with universal primers and their application to the detection and taxonomic analysis of Pseudomonas putida strains. Appl Environ Microbiol 61, 1104-1109.[Abstract]

Yamamoto, S. & Hirayama, S. (1996). Phylogenetic analysis of Acinetobacter strains based on the nucleotide sequences of gyrB genes and on the amino acid sequences of their products. Int J Syst Bacteriol 46, 506-511.[Abstract]

Zolg, J. W. & Philippi-Schulz, S. (1994). The superoxide dismutase gene, a target for detection and identification of mycobacteria by PCR. J Clin Microbiol 32, 2801-2812.[Abstract]

Received 12 February 1999; revised 14 May 1999; accepted 20 May 1999.