Abstract
Six strains of a hitherto undescribed gram-positive, catalase-negative, facultatively anaerobic coccus isolated from human sources were characterized by phenotypic and molecular taxonomic methods. Comparative 16S rRNA gene sequencing studies demonstrated that the unknown strains were genealogically identical and constitute a new subline within the genus Gemella. The unknown bacterium was readily distinguished from Gemella haemolysans, Gemella bergeriae, and Gemella morbillorum by biochemical tests and electrophoretic analysis of whole-cell proteins. Based on phylogenetic and phenotypic evidence, it is proposed that the unknown bacterium be classified as Gemella sanguinis sp. nov. The type strain is CCUG 37820T.
The genus Gemella consists of catalase-negative, facultatively anaerobic, gram-positive coccoid organisms, which occur in pairs, tetrads, or short chains, and which have DNA with a low G+C content (1). Until recently, two species, Gemella haemolysans, the type species of the genus, and Gemella morbillorum, were recognized (1). G. haemolysans and G. morbillorum are considered commensal with humans, although each has been shown to cause severe localized and generalized infection, particularly in immunocompromised patients (e.g., references 1, 4, 8, 10, and 11). During a study of Gemella-like organisms from human sources, we recently described a third species, G. bergeriae (2). In the course of this ongoing investigation, we have performed a polyphasic taxonomic study of six Gemella-like isolates from human blood which biochemically did not appear to correspond to any of the recognized species of this genus. In this article, we report the phenotypic properties of these organisms and the results of a phylogenetic analysis. Based on the findings presented here, a new species, Gemella sanguinis sp. nov., is described.
Six isolates originating from human clinical specimens were examined. Four of the strains were received by the Centers for Disease Control and Prevention (CDC [Atlanta, Ga.]). Strain 2029-87 (Culture Collection of the University of Göteborg, CCUG 37970) was received from the Virginia State Health Department and was isolated from blood cultures of a 69-year-old male with subacute bacterial endocarditis. This strain was initially identified as a Gemella species by the CDC laboratory. Strain 522-94 (CCUG 37821) was received from the New York State Health Department; the patient was a resident of Schenectady, N.Y. The strain was isolated from a blood culture of a 54-year-old male, and no clinical diagnosis was given. This culture was initially identified as G. haemolysans by the CDC laboratory. Strain 2045-94 (CCUG 37820) was received from the Georgia State Health Department. The patient was a resident of Augusta, Ga. The strain was isolated from a blood culture of a 41-year-old male, and no clinical diagnosis was given. This culture was initially identified as a possible Gemella species by the CDC laboratory. Strain 1519-95 (CCUG 37822) was received from Susan Novak, Kaiser Permanente Laboratory, North Hollywood, Calif. The isolate was obtained from a blood culture of an adult male; no other clinical information was given. This strain was initially identified as G. morbillorum by the CDC laboratory. The two remaining strains (CCUG 24073 and CCUG 33602) were also isolated from human blood but were referred to the Culture Collection of the University of Göteborg, Göteborg, Sweden, for identification. Strain CCUG 24073 was received from the Public Health Laboratory, Göteborg, and was isolated from a 46-year-old woman. Strain CCUG 33602 was received from the Public Health Laboratory, Vaxjo, Sweden, and originated from a 69-year-old male. No clinical information about these strains is available.
All strains were cultured on Columbia agar (Difco, Detroit, Mich.), supplemented with 5% horse blood at 37°C, in air plus 5% CO2. The four strains from the CDC were tested with conventional tests as previously described (5). All of the strains were biochemically characterized by using the API rapid ID32 strep and API ZYM systems according to the manufacturer’s instructions (API bioMérieux, Marcy l’Etoile, France). Polyacrylamide gel electrophoretic (PAGE) analysis of whole-cell proteins was performed as described by Pot et al. (12). For densitometric analysis, normalization, and interpretation of protein patterns, the GelCompar GCW 3.0 software package (Applied Maths, Kortrijk, Belgium) was used. DNA was prepared by the method of Marmur (9), and DNA-DNA hybridization experiments were performed by the S1 nuclease method (7). The G+C content of DNA was determined by thermal denaturation. A large fragment of the 16S rRNA gene (corresponding to positions 30 to 1521 of the Escherichia coli 16S rRNA gene) was amplified by PCR and directly sequenced with a dye-deoxy terminator cycle sequencing kit (Applied Biosystems, Foster City, Calif.) and an automatic DNA sequencer (model 373A; Applied Biosystems). The sequences of the closest relatives of the unknown strains were retrieved from the GenBank or Ribosomal Database Project (RDP) data library and aligned with the newly determined sequences by using the program PILEUP (3). The resulting multiple-sequence alignment was corrected manually, and a phylogenetic tree was constructed according to the neighbor-joining method with the program NEIGHBOR (6).
The six clinical isolates were ovoid in shape and formed single cells, pairs, or short chains. All of the strains were gram-positive, non-spore-forming catalase-negative, oxidase-negative facultative anaerobes which produced small pinhead colonies on blood agar. The four strains tested at the CDC with conventional tests were identified as Gemella species according to the criteria described in reference 5. All strains were vancomycin sensitive, failed to produce gas from glucose in Mann-Rogosa-Sharpe (MRS) broth, were pyrrolidonyl arylamidase and leucine aminopeptidase disk test positive, failed to grow in broth containing 6.5% NaCl or at 10°C, and were nonmotile. All strains produced acid in maltose, mannitol, and sorbitol broths. Variable reactions were observed in lactose and sucrose broths. The strains did not hydrolyze esculin, hippurate, or starch. Arginine deamination was not observed. Urease was not produced. Pyruvate was not utilized, and 0.04% tellurite was not tolerated. None of the four strains produced acid from arabinose, glycerol, inulin, melibiose, raffinose, ribose, sorbose, or trehalose broths. According to the commercial API systems, all of the isolates were similar in producing acid from glucose, mannitol, sorbitol, and sucrose. Five of the six strains also produced acid from maltose. None of the isolates produced acid from d-arabitol, l-arabinose, cyclodextrin, glycogen, lactose, melibiose, melezitose, methyl-β-d-glucopyranoside, ribose, raffinose, trehalose, or d-xylose. According to the API systems, all six isolates produced acid phosphatase, alkaline phosphatase, and ester lipase C8. Most of the strains produced alanine-phenylalanine-proline arylamidase (four of six weak positive), pyrrolidonyl arylamidase (four of six weak positive), pyrazinamidase (four of six weak positive), and esterase C-4 (five of six weak positive). None of the isolates showed arginine dihydrolase, N-acetyl-β-glucosaminidase, chymotrypsin, cystine arylamidase, α-fucosidase, α-galactosidase, β-galactosidase, β-galacturonidase, β-glucuronidase, glycyl-tryptophan arylamidase, leucine arylamidase, lipase C4, β-mannosidase, trypsin, valine arylamidase, or urease activity, and none hydrolyzed esculin, hippurate, or gelatin. Acetoin production was variable. Although the isolates resembled gemellae in their appearance, they differed from G. haemolysans in producing acid from mannitol and sorbitol. Similarly, the isolates were readily distinguished from G. bergeriae and G. morbillorum in producing acid and alkaline phosphatase. The isolates also differed from G. bergeriae in producing acid from sorbitol and sucrose. The close phenotypic affinity between the isolates and their separateness from G. haemolysans, G. bergeriae, and G. morbillorum were also demonstrated by PAGE analysis of whole-cell proteins. The unknown clinical isolates formed a robust and tight cluster which was quite separate from the three currently recognized Gemella species and all other gram-positive catalase-negative reference organisms examined (Fig. 1).
FIG. 1.
Similarity dendrogram based on whole-cell protein patterns of G. sanguinis sp. nov. and related species. Levels of correlation are expressed as percentages of similarity for convenience.
To assess the genealogical affinity between the unknown isolates and their relationship with other gram-positive catalase-negative taxa, comparative 16S rRNA gene sequence analyses were performed. Almost complete (>1,400 nucleotides) 16S rRNA gene sequences were determined for each isolate, and pairwise analysis showed no differences between the six strains (i.e., 100% sequence similarity), thereby demonstrating their genealogical homogeneity. Sequence searches of GenBank and RDP data libraries revealed that the unknown organism was phylogenetically most closely associated with species of the genus Gemella. A tree depicting the phylogenetic affinity of the unknown coccus with closely related taxa is shown in Fig. 2. The unknown coccus clustered within the Gemella clade, where it formed a distinct subline.
FIG. 2.
Unrooted tree showing the phylogenetic relationships of G. sanguinis sp. nov. and some other low-G+C-content gram-positive bacteria. The tree constructed by the neighbor-joining method was based on a comparison of approximately 1,320 nucleotides.
The six isolates from human clinical sources clearly constitute a hitherto unknown species within the genus Gemella. Biochemically the unknown coccus does not correspond to any of the three recognized species, and PAGE analysis of whole-cell proteins confirmed the phenotypic separateness of the bacterium. G. haemolysans and G. morbillorum are phylogenetically closely related species exhibiting 25 base differences (24 mismatches, 1 unmatched) in their 16S rRNA genes. Despite this high genealogical affinity, earlier DNA-DNA pairing studies have shown they are distinct species (14). The unknown coccus reported here is of a comparable genealogical distance, showing 23 (21 mismatches, 2 unmatched) base differences with G. haemolysans and 31 (29 mismatches, 2 unmatched) differences with G. morbillorum. DNA-DNA hybridization experiments conducted in this study with 3H-labelled DNA from isolate 2045-94 showed reassociation values of 38% and 30% with the type strains of G. haemolysans and G. morbillorum, respectively. These findings confirm the genotypic separateness of the unknown coccus from these species. Based on the high phenotypic and genealogical homogeneity of the isolates and their distinctiveness from G. haemolysans, G. bergeriae, and G. morbillorum, we consider they represent a hitherto unrecognized species within the genus Gemella, for which the name Gemella sanguinis sp. nov. is proposed.
G. haemolysans and G. morbillorum are residents of the mucous membranes of humans. Although considered commensal with humans, both species have been shown to cause endocarditis and a variety of other infections (e.g., references 1, 4, 8, 10, 11, and 13). Of the novel isolates examined in this study, strain 2029-87 (= CCUG 37970) was isolated from blood cultures of a male patient diagnosed with subacute bacterial endocarditis. The five other isolates also originated from human blood, but no clinical information about these strains is available. It nevertheless seems highly likely that the newly described coccus, like other gemellae, may represent another gram-positive, catalase-negative opportunistic human pathogen. It is pertinent to note that the unknown Gemella species is biochemically distinct. The tests outlined in Table 1 readily distinguish the unknown coccus from G. haemolysans, G. bergeriae, and G. morbillorum and in the future should help to clarify the range of infections caused by this hitherto unknown bacterium.
TABLE 1.
Characteristics useful in differentiating G. sanguinis sp. nov. from G. haemolysans, G. bergeriae, and G. morbillorum
Test | Result fora:
|
|||
---|---|---|---|---|
G. sanguinis sp. nov. | G. haemo- lysans | G. ber- geriae | G. morbil- lorum | |
Acid from: | ||||
Mannitol | + | − | V | V |
Sorbitol | + | − | − | −(+) |
Sucrose | + | V | − | + |
Production of: | ||||
Alkaline phosphatase | + | + | − | − |
Acid phosphatase | + | + | − | − |
Alanine-phenylalanine-proline arylamidase | +(−) | − | − | V |
Voges-Proskauer | V | − | − | − |
+, positive; −, negative; +(−), a few strains negative; −(+), a few strains positive; V, variable.
Description of Gemella sanguinis sp. nov.
Gemella sanguinis (sanguinis, L. genitive noun, of the blood) cells are gram-positive, non-spore-forming cocci that occur singly, in pairs, or in short chains. Cocci are sometimes elongate. Colonies on blood agar plates after 48 h are small, circular, entire, low-convex, nonpigmented, translucent to opaque, and smooth. Some strains are hemolytic. They are facultatively anaerobic and catalase and oxidase negative. According to the API systems, acid is produced from glucose, mannitol, sorbitol, and sucrose. Most strains produce acid from maltose. Acid is not produced from d-arabitol, l-arabinose, cyclodextrin, glycogen, lactose, melibiose, melezitose, methyl-β-d-glucopyranoside, pullulan, raffinose, ribose, tagatose, trehalose, and d-xylose. Alkaline phosphatase, acid phosphatase, and ester lipase C8 are produced. Some strains produce alanine-phenylalanine-proline arylamidase, pyrrolidonyl arylamidase, pyrazinamidase, esterase C-4, and pyroglutamic acid arylamidase. Activity of arginine dihydrolase, N-acetyl-β-glucosaminidase, chymotrypsin, cystine arylamidase, α-fucosidase, α-galactosidase, β-galactosidase, β-galacturonidase, β-glucuronidase, glycyl-tryptophan arylamidase, leucine arylamidase, lipase C4, β-mannosidase, trypsin, valine arylamidase, and urease is not detected. Esculin, hippurate, and gelatin are not hydrolyzed. The Voges-Proskauer test is variable. Nitrate is not reduced. The G+C content of DNA is 31 mol%. The type strain of G. sanguinis is 2045-94 (=CCUG 37820T). The type strain produces acid from maltose and is Voges-Proskauer, pyrazinamidase, and pyrrolidonyl arylamidase positive but esterase C-4 negative.
Nucleotide sequence accession number.
The 16S rRNA gene sequence of strain 2045-94 (CCUG 37820T) has been deposited in GenBank under accession no. Y13364.
Acknowledgments
This work was supported in part by a grant from the European Union (BI02-CT94-3098).
We are grateful to colleagues for kindly supplying cultures.
REFERENCES
- 1.Berger U. The genus Gemella. In: Balows A, Trüper H G, Dworkin M, Harder W, Schleifer K-H, editors. The prokaryotes. 2nd ed. New York, N.Y: Springer-Verlag; 1992. pp. 1643–1653. [Google Scholar]
- 2.Collins M D, Hutson R A, Falsen E, Sjöden B, Facklam R R. Gemella bergeriae sp. nov., isolated from human clinical specimens. J Clin Microbiol. 1998;36:1290–1293. doi: 10.1128/jcm.36.5.1290-1293.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Devereux J, Haeberli P, Smithies D. A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 1984;12:387–395. doi: 10.1093/nar/12.1part1.387. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Eggelmeier F, Petit P, Dijkmans B A C. Total knee arthroplasty infection due to Gemella haemolysans. Br J Rheumatol. 1992;31:67–69. doi: 10.1093/rheumatology/31.1.67. [DOI] [PubMed] [Google Scholar]
- 5.Facklam R, Elliot J A. Identification, classification, and clinical relevance of catalase-negative, gram-positive cocci, excluding streptococci and enterococci. Clin Microbiol Rev. 1995;8:479–495. doi: 10.1128/cmr.8.4.479. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Felsenstein J. PHYLIP-phylogeny inference package (version 3.2) Cladistics. 1989;5:164–166. [Google Scholar]
- 7.Grimont P A D, Popoff M Y, Grimont F, Coynault C, Lemelin M. Reproducibility and correction study of three deoxyribonucleic acid hybridization procedures. Curr Microbiol. 1980;4:325–330. [Google Scholar]
- 8.Kaufhold A, Franzen D, Lütticken R. Endocarditis caused by Gemella haemolysans. Infection. 1989;17:385–387. doi: 10.1007/BF01645552. [DOI] [PubMed] [Google Scholar]
- 9.Marmur J. A procedure for the isolation of deoxyribonucleic acid from microorganisms. J Mol Biol. 1961;3:208–218. [Google Scholar]
- 10.Mitchell R G, Teddy P L. Meningitis due to Gemella haemolysans after radiofrequency trigeminal rhizotomy. J Clin Pathol. 1986;38:558–560. doi: 10.1136/jcp.38.5.558. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Petit Y J, Layre J C, Lamaury I, Perez C, Jonquet Q, Janbon F. Purulent-meningitis caused by Gemella haemolysans. Presse Med. 1993;22:444. [PubMed] [Google Scholar]
- 12.Pot B, Vandamme P, Kersters K. Analysis of electrophoretic whole-organism protein fingerprints. In: Goodfellow M, O’Donnell A G, editors. Modern microbial methods. Chemical methods in prokaryotic systematics. Chichester, United Kingdom: J. Wiley and Sons, Ltd.; 1994. pp. 493–521. [Google Scholar]
- 13.Pradeep R, Ali M, Encarnacion C F. Retropharyngeal abscess due to Gemella morbillorum. Clin Infect Dis. 1997;24:284–285. doi: 10.1093/clinids/24.2.284. [DOI] [PubMed] [Google Scholar]
- 14.Whitney A M, O’Connor S P. Phylogenetic relationship of Gemella morbillorum to Gemella haemolysans. Int J Syst Bacteriol. 1993;43:832–838. doi: 10.1099/00207713-43-4-832. [DOI] [PubMed] [Google Scholar]