点击显示 收起
Research Service, VA Medical Center West Los Angeles, Los Angeles, California
Infectious Diseases Section, VA Medical Center West Los Angeles, Los Angeles, California
Department of Medicine, UCLA School of Medicine, Los Angeles, California
Department of Microbiology, Immunology, and Molecular Genetics, UCLA School of Medicine, Los Angeles, California
ABSTRACT
Phenotypic and phylogenetic studies were performed on an unknown gram-negative, strictly anaerobic, rod-shaped bacterium isolated from human clinical specimens. This organism was indole negative, resistant to 20% bile, produced acetic and a lesser amount of succinic acids as the major end products of glucose metabolism, and possessed a G+C content of approximately 43 mol%. Comparative 16S rRNA gene sequencing demonstrated that the unidentified bacterium was a member of the Cytophaga-Flavobacter-Bacteroides phylum of gram-negative bacteria and formed a close association (with an average sequence similarity of 93.6%) with the second subcluster of the Porphyromonas cluster in the Bacteroides subgroup. Phylogenetically and phenotypically it resembled Bacteroides merdae; however, a 16S rRNA gene sequence divergence of approximately 5.5% between the unknown bacterium and B. merdae, as well as distinguishable biochemical characteristics, demonstrate that the unknown bacterium is genotypically and phenotypically distinct and represents a previously unknown subline within the Porphyromonas phylogenetic cluster. Furthermore, a DNA-DNA reassociation value of 17.8% between isolates WAL 12034T (the type strain of this novel taxon) and ATCC 43184T (B. merdae type strain) also documented the separateness of the unknown species and B. merdae. Based on the phenotypic and phylogenetic findings, a new species, "Bacteroides goldsteinii sp. nov," is proposed. The G+C content of the DNA is 43 mol% for Bacteroides. The type strain of "B. goldsteinii" is WAL 12034T (= CCUG 48944T = ATCC BAA-1180T).
INTRODUCTION
The taxonomy of Bacteroides has undergone significant changes in the past few years (8). Studies have shown that the genus Bacteroides contained species representing several genera. A majority of the species previously included in the genus Bacteroides have been placed in the genera Porphyromonas, Prevotella, and Bacteroides sensu stricto (9-11). Several other genera have subsequently been described for Bacteroides species which do not conform to these three major groups (e.g., Anaerorhabdus, Dichelobacter, Dialister, Fibrobacter, Megamonas, Mitsuokella, Rikenella, Sebaldella, Tannerella, Tissierella, and Alistipes). The taxonomic positions of some other species still included in the genus, such as Bacteroides distasonis and Bacteroides merdae, remain uncertain; all of these species will ultimately be transferred to other genera (6). Furthermore, several clinically important species still await formal description.
In this paper, we report on the characterization of a group of isolates which were recovered from clinical specimens of human intestinal origin. Phenotypically, the unknown bacterium is very much like B. merdae; these strains were misidentified as B. merdae previously. However, 16S rRNA sequencing reveals approximately 5.5% sequence divergence between the novel species and its phylogenetically closest species, B. merdae. A DNA-DNA hybridization study also confirmed that this unknown organism was indeed distinct from its nearest valid species, B. merdae. Based on the phenotypic and phylogenetic findings presented here, a new species, "Bacteroides goldsteinii," is proposed. In addition, we describe the phenotypic tests useful in distinguishing between this novel organism and its related taxa.
MATERIALS AND METHODS
Bacterial strains and growth conditions. The study included seven B. merdae-like strains, seven strains each of B. merdae, B. distasonis, and Tannerella forsythensis, and three strains each of Bacteroides fragilis, Bacteroides vulgatus, Bacteroides caccae, Prevotella nigrescens, Prevotella disiens, Prevotella corporis, Prevotella intermedia, Porphyromonas asaccharolytica, Porphyromonas endodontalis, and Porphyromonas gingivalis (Table 1). The novel isolates were recovered from clinical sources, such as peritoneal fluid, appendix tissue, and intra-abdominal abscess; therefore, they are likely of intestinal origin. All the clinical isolates of each species were identified by 16S DNA sequencing in our laboratory. All the strains were cultivated on Brucella blood agar (Difco, Detroit, Mich.) supplemented with 5% sheep blood and incubated anaerobically at 37°C under an N2 (86%), H2 (7%), and CO2 (7%) gas phase.
Biochemical characterization. The strains were characterized biochemically by using a combination of conventional tests as described previously in the Wadsworth manual (3), plus the API ZYM and rapid ID 32A systems (API bioMerieux, Marcy l'Etoile, France), and the RapID ANA II (Remel, Inc., Lenexa, KS) system according to the respective manufacturer's instructions. All biochemical tests were performed in duplicate. Fermentation tests were performed using prereduced, anaerobically sterilized peptone-yeast-sugar broth tubes (Anaerobe Systems, Morgan Hill, CA). The strains were grown in peptone-yeast broth and peptone-yeast-glucose broth (Anaerobe Systems, Morgan Hill, CA) for metabolic end product (short-chain volatile and nonvolatile fatty acids) analysis by gas liquid chromatography (3). The MICs of seven antimicrobial agents, which were selected either as representative of a class of compound or as drugs for which MICs for quality control strains were published, were determined by the National Committee for Clinical Laboratory Standards reference agar dilution method (5). -Lactamase production was tested using nitrocefin disks (Cefinase; BBL, Becton, Dickinson and Company, Sparks, MD) inside the anaerobic chamber.
Cellular fatty acid composition. Long-chain cellular fatty acids were detected with a Hewlett-Packard 5890 series II gas chromatograph, and cellular fatty acid profiles were determined by the Microbial Identification System software (MIDI, Newark, N.J.) as described previously (13). The corresponding library (ANAEROBE, version 5.0) was used in successive analyses. Peaks were automatically integrated, fatty acids were identified by equivalent chain length, and percentages of the total peak area were calculated. External calibration was done by using MIDI calibration mixture I.
DNA base composition. The mol% G+C content of DNA was determined by high-performance liquid chromatography according to the method of Mesbah et al. (4) except that the methanol content of the chromatographic buffer was decreased to 8% and the temperature was increased to 37°C.
16S rRNA sequencing and phylogenetic analysis. The 16S rRNA genes were amplified by PCR using universal primers 8UA (positions 8 to 28, Escherichia coli numbering) and 1485B (positions 1485 to 1507) as described previously (12). The amplified product was purified by using a QIAamp PCR purification kit (QIAGEN, Inc., Chatsworth, CA) and directly sequenced with a Biotech Diagnostic (Biotech Diagnostic, CA) BigDye sequencing kit on an ABI 377 sequencer (Applied Biosystems, Foster City, CA). The closest known relatives of the new isolates were determined by performing database searches using the BLAST software (1). Almost the full lengths of the 16S rRNA gene sequences (>1,400 nucleotides) of the unidentified bacteria and of closely related bacteria were aligned using CLUSTAL-W (http://genome.kribb.re.kr). A phylogenetic tree was reconstructed using DNA analysis software PAUP 4.0 (Sinauer Associates, Inc., Sunderland, MA). The stability of the groupings was estimated by bootstrap analysis (1,000 replications) using the same program.
DNA-DNA reassociation. DNA-DNA reassociation experiments were carried out according to the spectrophotometric method of De Ley et al. (1a), using a Gilford System model 2600 spectrophotometer equipped with a Gilford model 2527-R thermal programmer.
Nucleotide sequence accession number. The 16S rRNA sequence of strain WAL 12034T has been deposited in GenBank under accession number AY 974070.
RESULTS
The novel bacterium recovered from clinical infections of human intestinal origin was always isolated together with other anaerobes (mostly other species of the B. fragilis group) and/or aerobes. Four strains were isolated from appendix tissue of four patients with appendicitis, two were isolated from peritoneal fluid, and one was isolated from abdominal abscess. Moderate to heavy growth was obtained for most of the isolates on primary isolation blood agar plates. Gram staining showed that they were gram negative and rod shaped, and typical cells were 0.9 to 1.5 μm by 1.2 to 10 μm, observed by an optical microscope. Colonies on Brucella blood agar plates at 48 h were gray, circular, convex, entire, and opaque and attained a diameter of 1 to 2 mm. All of the isolates grew well (resistant to 20% bile) and blackened the Bacteroides bile esculin agar by hydrolyzing esculin. They all grew well anaerobically, but no growth occurred following subculture in air or in atmospheres of 2% or 6% O2. All of the strains were resistant to the kanamycin (1,000 μg), colistin sulfate (10 μg), and vancomycin (5 μg) special-potency identification disks. They were indole, urease, and nitrate negative. They all produced acid from cellobiose, glucose, rhamnose, sucrose, trehalose, and xylose but did not produce acid from arabinose and xylan. Using the API ZYM, rapid ID 32A, and RapID ANA II systems, all isolates of the same group produced the same profile. They were all identified as B. merdae with a low percentage of similarity. Positive reactions were obtained for -glucosidase, -galactosidase, -galactosidase, -N-acetyl-glucosaminidase, naphthol-AS-Bl-phosphohydrolase, acid phosphatase, alkaline phosphatase, leucine arylamidase, p-nitrophenylphosphatase, arginine arylamidase, leucyl glycine arylamidase, phenylalanine arylamidase, alanine arylamidase, glycine arylamidase, histidine arylamidase, glutamyl glutamic acid arylamidase, leucyl glycine aminopeptidase, glycine aminopeptidase, phenylalanine aminopeptidase, arginine aminopeptidase, and serine aminopeptidase. All the other tests were negative; -fucosidase was negative in all three systems. Mannose and raffinose were fermented when tested by the rapid ID 32A system. Using Rosco diagnostic tablets (Rosco, Taastrup, Denmark), -xylosidase, -glucuronidase, -glucosidase, -glucosidase, -galactosidase, -galactosidase (o-nitrophenol--D-galactopyranoside), -N-acetyl-glucosaminidase, alkaline phosphatase, and -arabinosidase were detected. Results for the production of -glucosidase and -arabinosidase differed between the different biochemical kits and the Rosco tablets. Using Rosco tablets, -arabinosidase was detected, but with the biochemical kits it was not detected. -Glucosidase was detected by rapid ID 32A and Rosco tablets but not by the API ZYM and RapID ANA II systems. In peptone yeast broth and peptone yeast glucose broth, major amounts of acetic acid and succinic acid and minor amounts of propionic acid, isovaleric acid, and formic acid were produced by all isolates. Long-chain cellular fatty acids analysis using Microbial Identification System software and the corresponding library provided a misidentification as Tannerella forsythus with a low similarity index. The principal long-chain cellular fatty acids of the isolates were anteiso-15:0 and iso-3-OH-17:0 (25 to 28% and 18 to 23% of the total, respectively). Significant amounts of 18:1 w9c and anteiso-17:0 3OH (11 to 16% and 9 to 15% of the total, respectively) were also present. Agar dilution tests showed all of the strains were susceptible to metronidazole (MIC, 2 μg/ml) and ertapenem (MIC, 1 μg/ml). Some resistance was seen with clindamycin (MIC, 8 μg/ml). Resistance was shown to penicillin G (MIC, 32 μg/ml), cefotetan (MIC, 256 μg/ml), and vancomycin (MIC 32 μg/ml) by all strains. All strains were -lactamase positive.
To assess the genealogical affinity between the unknown bacteria and their relationship with other taxa, their 16S rRNA gene sequences were determined. Pairwise analysis showed that all of the isolates of the same group were phylogenetically closely related to each other (>99.5% sequence similarity). Sequence searches of GenBank and Ribosomal Database Project libraries revealed that the unknown organism was a member of the Cytophaga-Flavobacter-Bacteroides phylum, and it represents a previously unknown subline within the Porphyromonas subgroup (6). It has a high sequence similarity (99.2%) to altered Schaedler flora strain Bacteroides sp. ASF 519, and B. merdae was the phylogenetically closest valid species. A tree, constructed by the neighbor-joining method, depicting the phylogenetic affinity of the novel bacterium as exemplified by strain CCUG 48944T, is shown in Fig. 1 and confirms the placement of the novel bacterium in the second subcluster of the Porphyromonas phylogenetic cluster. It is evident from the branching pattern in the tree that the novel species possesses a close relationship with B. merdae and B. distasonis, and bootstrap resampling showed this relationship to be statistically significant (99% recovery in 1,000 resamplings). Pairwise comparison revealed approximately 5.5% sequence divergence between the novel bacteria and the type strain of its closest valid species, B. merdae, based on almost the full length of the 16S rRNA gene sequences (1,400 nucleotides). Furthermore, a DNA-DNA reassociation value of 17.8% was observed between isolates WAL 12034T and ATCC 43184T (B. merdae type strain), thereby confirming that the unidentified bacterium represents a previously unknown subline within the second subcluster of the Porphyromonas phylogenetic cluster.
Support for the separation of the unknown bacterium from its related bacterial species also comes from the phenotypic characterization. The unknown organism can be readily distinguished from species of the genera Tannerella, Prevotella, and Porphyromonas by its resistance to 20% bile. In addition, the unknown bacterium is highly fermentative, in contrast to asaccharolytic species in the genus Porphyromonas and moderately saccharolytic species in Prevotella. It also can be readily distinguished from the genera Dysgonomonas and Tannerella by long-chain-fatty-acid analysis. The fatty-acid profile of the novel bacterium is incompatible with those of the species of the genera Dysgonomonas and Tannerella. Dysgonomonas has iso-C14:0 as one of the major fatty acids, and Tannerella has a much higher ratio (>20) of anteiso-C15:0 to iso-C15:0 than that for the novel bacterium. Within the second subcluster of the Porphyromonas phylogenetic cluster, although 16S rRNA gene sequence analysis showed the closest phylogenetic relative to the unknown bacterium is B. merdae, it can be distinguished easily from B. merdae by several biochemical characteristics, such as by strongly producing -glucosidase (tested with Rosco tablets) and -glucosidase and strongly hydrolyzing esculin. In addition, by using RapID ANA II, the unknown bacterium showed a strongly positive reaction for phenylalanine aminopeptidase, whereas B. merdae was either negative or only weakly positive. In addition, the cellular fatty acid composition of the unknown was similar to that of B. merdae except that the ratio of 17:0 iso 3OH to 17:0 anteiso 3OH in the whole-cell methanolysates of the unknown species was lower than that of B. merdae (ratio ranges from 1.4 to 2.2 for the unknown bacterium and from 6.1 to 8.3 for B. merdae). The unknown bacterium can also be differentiated from B. distasonis by producing -glucuronidase (tested by Rosco tablets), whereas B. distasonis does not. Furthermore, using RapID ANA II and API ZYM, -glucosidase activity was not detected in the unknown bacterium, in contrast to B. distasonis. The characteristics for distinguishing the unknown bacteria from their phenotypically or phlogenetically closely related species are summarized in Table 2.
DISCUSSION
In this study, we report on the characterization of a group of strains of a novel bacterium that is isolated from clinical infections and is likely of intestinal origin. The strains were isolated from seven different patients. Four patients had appendicitis; the strains were isolated from their appendix tissue. All the other three patients had peritonitis and abdominal abscess, and one of them had wound dehiscence and evisceration. Background factors were cecal inflammation of undetermined source, diverticulitis, and necrotic ileum in an incarcerated umbilical hernia. They are not commonly encountered in clinical specimens They were always found in mixed culture and were not recovered in blood cultures or in very serious infections. This group of clinical isolates was misidentified as B. merdae by routine biochemical tests. However, 16S rRNA sequencing revealed approximately 5.5% sequence divergence between the novel species and B. merdae. 16S rRNA sequence analysis showed that the novel bacterium has a close affinity with the second subcluster of the Porphyromonas cluster, which includes B. merdae and B. distasonis, with a mean sequence similarity of about 93.6%. It also displayed a lesser close phylogenetic relationship with the genera Dysgonomonas and Tannerella (with an average sequencing similarity of 88.5% and 90.5%, respectively). Other taxa displayed significantly lower levels of sequence similarity, including Porphyromonas, Bacteroides sensu stricto, Prevotella, and Bacteroides splanchnicus. It is evident from the described findings that the unidentified bacterium is not a species of Bacteroides sensu stricto affiliated with the type species (Fig. 1); it represents a hitherto-unknown line within the second subcluster of the Porphyromonas cluster, displaying a loose affinity (with an average similarity of 86%) with the first subcluster (the genus Porphyromonas). These data agree with previous findings (6); the deep branching position of the unknown bacterium suggested that a novel genus should be established to accommodate these three species. However, current phenotypic tests do not separate it from Bacteroides sensu stricto. Until the genera are revised along phylogenetic lines, the novel bacterium conforms to the present definition of the genus, and as such, this is the only appropriate home for the organism. Support for the separation of the unknown bacterium from related taxa also comes from phenotypic considerations. Table 2 summarizes the key characteristics for identification and differentiation of the novel bacterium that we are describing here from the other related taxa. Based on both phenotypic and genotypic evidence, it is clear that this group of unknown isolates recovered from infections of intestinal origin in humans represents a novel species.
We propose the unknown bacterium be classified as a new Bacteroides species, "Bacteroides goldsteinii sp. nov."
Description of "Bacteroides goldsteinii sp. nov." "Bacteroides goldsteinii" (in honor of the outstanding infectious disease clinician who has done a lot of work with anaerobes, Ellie C. Goldstein.). Cells are rod shaped, 0.9 to 1.5 μm by 1.2 to 10 μm. Colonies on Brucella blood agar plates at 48 h are gray, circular, convex, entire, and opaque and attain a diameter of 1 to 2 mm. Obligately anaerobic. Indole, urease, and nitrate negative. Esculin is hydrolyzed. Resistant to 20% bile. Acid is produced from cellobiose, glucose, rhamnose, sucrose, and xylose but not from arabinose, trehalose, and xylan. In peptone yeast broth and peptone yeast glucose broth, major amounts of acetic and succinic acids and minor amounts of isovaleric acid, propionic acid, and formic acid are produced by all isolates. Using the API ZYM, rapid ID 32A, and RapID ANA II systems, all isolates of the same group produced the same profile. Positive reactions were obtained for -glucosidase, -galactosidase, -galactosidase, -N-acetyl-glucosaminidase, naphthol-AS-Bl-phosphohydrolase, acid phosphatase, alkaline phosphatase, leucine arylamidase, p-nitrophenylphosphatase, arginine arylamidase, leucyl glycine arylamidase, leucine arylamidase, phenylalanine arylamidase, alanine arylamidase, glycine arylamidase, histidine arylamidase, glutamyl glutamic acid arylamidase, leucyl glycine aminopeptidase, glycine aminopeptidase, phenylalanine aminopeptidase, arginine aminopeptidase, and serine aminopeptidase. All the other tests were negative. -Fucosidase was negative by all three systems. Mannose and raffinose were fermented when tested by the rapid ID 32A system. Using Rosco diagnostic tablets (Rosco, Taastrup, Denmark), -xylosidase, -glucuronidase, -glucosidase, -glucosidase, -galactosidase, -galactosidase (o-nitrophenol--D-galactopyranoside), -N-acetyl-glucosaminidase, alkaline phosphatase, and -arabinosidase were detected; -arabinosidase was tested as positive only by Rosco tablets. The principal long-chain cellular fatty acids of the isolates were anteiso-15:0 and iso-3-OH-17:0 (25 to 28% and 18 to 23% of the total, respectively). Significant amount of 18:1 w9c and anteiso-17:0 3OH (11 to 16% and 9 to 15% of the total, respectively) were also present. Agar dilution tests showed that the strains were susceptible to metronidazole (MIC, 2 μg/ml), and ertapenem (MIC, 1 μg/ml). Some resistance was seen with clindamycin (MIC, 8 μg/ml). Resistance was shown to penicillin G (MIC, 32 μg/ml), cefotetan (MIC, 256 μg/ml), and vancomycin (MIC, 32 μg/ml) by all strains. All strains were -lactamase positive.
Isolated from human clinical specimens of intestinal origin. Habitat is probably the human gut. The type strain is WAL 12034T = ATCC BAA-1180T = CCUG 48944T. The G+C content of the type strain is 43 mol%.
ACKNOWLEDGMENTS
This work was carried out, in part, with financial support from Veterans Administration Merit Review funds.
REFERENCES
Benson, D. A., I. Karsch-Mizrachi, D. J. Lipman, J. Ostell, and D. L. Wheeler. 2005. GenBank. Nucleic Acids Res. 33:D34—D38.
De Ley, J., H. Cattoir, and A. Reynaerts. 1970. The quantitative measurements of DNA hybridization from renaturation rates. Eur. J. Biochem. 12:133-142.
Hofstad, T., I. Olsen, E. R. Eribe, E. Falsen, M. D. Collins, and P. A. Lawson. 2000. Dysgonomonas gen. nov. to accommodate Dysgonomonas gadei sp.nov., an organism isolated from a human gall bladder, and Dysgonomonas capnocytophagoides (formerly CDC group DF-3). Int. J. Syst. Evol. Microbiol. 50:2189-2195.
Jousimies-Somer, H. R., P. Summanen, D. Citron, E. Baron, H. M. Wexler, and S. M. Finegold. 2002. Wadsworth-KTL anaerobic bacteriology manual, 6th ed. Star Publishing, Belmont, Calif.
Mesbah, M., U. Premachandran, and W. B. Whitman. 1989. Precise measurement of the G+C content of deoxyribonucleic acid by high-performance liquid chromatography. Int. J. Syst. Bacteriol. 39:159-167.
National Committee for Clinical Laboratory Standards. 2001. Methods for antimicrobial susceptibility testing of anaerobic bacteria, 5th ed., approved standard. 2001 NCCLS document M11-A5. National Committee for Clinical Laboratory Standards, Villanova, Pa.
Paster, B. J., F. E. Dewhirst, I. Olsen, and G. J. Fraser. 1994. Phylogeny of Bacteroides, Prevotella, and Porphyromonas spp. and related bacteria. J. Bacteriol. 176:725-732.
Sakamoto, M., M. Suzuki, M. Umeda, I. Ishikawa, and Y. Benno. 2002. Reclassification of Bacteroides forsythus (Tanner et al. 1986) as Tannerella forsythensis corrig., gen.nov., comb.nov. Int. J. Syst. Evol. Microbiol. 52:841-849.
Shah, H. N. 1990. The genus Bacteroides and related taxa, p. 3593-3607. In A. Balows, H. G. Trper, M. Dworkin, W. Harder, and K. H. Schliefer (ed.), The prokaryotes, 2nd ed. Springer-Verlag, Berlin, Germany.
Shah, H. N., and D. M. Collins. 1990. Prevotella, a new genus to include Bacteroides melaninogenicus and related species formerly classified in the genus Bacteroides. Int. J. Syst. Bacteriol. 40:205-208.
Shah, H. N., and M. D. Collins. 1989. Proposal to restrict the genus Bacteroides (Castellani and Chalmers) to Bacteroides fragilis and closely related species. Int. J. Syst. Bacteriol. 39:85-87.
Shah, H. N., and M. D. Collins. 1988. Proposal for reclassification of Bacteroides asaccharolyticus, Bacteroides gingivalis, and Bacteroides endodontalis in a new genus, Porphyromonas. Int. J. Syst. Bacteriol. 38:128-131.
Song, Y., C. Liu, M. McTeague, and S. M. Finegold. 2003. 16S ribosomal DNA sequence-based analysis of clinically significant gram-positive anaerobic cocci. J. Clin. Microbiol. 41:1363-1369.
Wexler, H. M., D. Reeves, P. H. Summanen, E. Molitoris, M. McTeague, J. Duncan, K. H. Wilson, and S. M. Finegold. 1996. Sutterella wadsworthensis gen. nov., sp. nov., bile-resistant microaerophilic Campylobacter gracilis-like clinical isolates. Int. J. Syst. Bacteriol. 46:252-258.